You are on page 1of 660

2.

003ProblemArchive
David Trumper
Last Updated April 19, 2005
Contents
Problems 12
1 1st Order Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.1 First Order Time Constant . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.2 Rise and Settling Times . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.3 First Order System Response . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2 2nd Order Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.1 Second-Order System Response . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.2 Second Order Responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.3 Second Order Derivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.4 Second Order Derivation Continued . . . . . . . . . . . . . . . . . . . . . . . 20
3 Higher Order Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.1 Structure of Higher Order System Solutions . . . . . . . . . . . . . . . . . . . 21
3.2 Structure of Higher Order System Solutions . . . . . . . . . . . . . . . . . . . 22
4 Mechanical Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.1 Balloon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.2 Bungee Jumper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.3 Elevator Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.4 Hoisting for Engineers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.5 Blocked Springs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.6 Engine Vibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.7 Garage Door . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.8 Rotating Damped . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
4.9 Car Suspension 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.10 Disk Drive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
4.11 Crashworthiness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.12 Nonlinear Rotational System . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4.13 Developing Dierential Equations . . . . . . . . . . . . . . . . . . . . . . . . . 41
4.14 Mass Spring System Frequency Response . . . . . . . . . . . . . . . . . . . . 42
4.15 Modeling Practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
4.16 Small Motion Transfer Function . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.17 Mass Spring Damper System Frequency Response 1 . . . . . . . . . . . . . . 45
4.18 Mass Spring Damper System Frequency Response 2 . . . . . . . . . . . . . . 46
4.19 Propeller Shaft Vibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
4.20 Safe Packaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
4.21 Sliding Damped . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.22 Toy Flywheel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.23 Truck Shocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
2
4.24 Car Suspension 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4.25 Kid-Skid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.26 Rolling Machine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
4.27 Sunday Bike Ride . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.28 Compound Mass Spring Damper System 1 . . . . . . . . . . . . . . . . . . . 57
4.29 Compound Mass Spring Damper System 2 . . . . . . . . . . . . . . . . . . . 59
4.30 Wind Induced Building Vibrations . . . . . . . . . . . . . . . . . . . . . . . . 60
5 Electrical Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
5.1 Camera Flash Circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
5.2 Batteries and Capacitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.3 Loaded Motor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
6 Fluid Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
6.1 Rotational Fluid System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
6.2 Fluid Leak . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
6.3 Whered the water go? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
7 Thermal Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
7.1 Thermal Block Question . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
7.2 The Hot Copper Block 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
7.3 The Hot Copper Block 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
7.4 Transistor on heat sink . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
8 Circuits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
8.1 RC Transfer Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
8.2 RLC Transfer Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
8.3 Bandpass Filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
8.4 LC Circuit Dierential Equations . . . . . . . . . . . . . . . . . . . . . . . . . 82
8.5 Equivalent Impedance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
8.6 Circuit Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
8.7 LRC Circuit 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
8.8 LRC Circuit 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
8.9 LR Circuit Step . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
9 Op Amps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
9.1 Opamp Arithmetic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
9.2 Inverting Op-Amp . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
9.3 Opamp Transfer Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
9.4 Op-Amp Dierentiator 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
9.5 Op-Amp Proportional plus Integral Controller . . . . . . . . . . . . . . . . . 94
9.6 Op-Amp Circuit Design 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
9.7 Op-Amp Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
9.8 Op-Amp Circuit Design 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
9.9 Op-Amp Dierentiator 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
9.10 RC Op-Amp Frequency Response . . . . . . . . . . . . . . . . . . . . . . . . . 99
10 Dierential Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
10.1 Dierential Equations 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
10.2 Dierential Equations 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
10.3 Dierential Equations 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
10.4 Eulers Theorem 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
10.5 Eulers Theorem 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
10.6 Force Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
3
10.7 Input for Zero Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
10.8 Zero Input Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
10.9 Zero Step Response 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
10.10 Zero Step Response 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
10.11 Nonlinear String System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
10.12 Homogenous and Particular Solutions . . . . . . . . . . . . . . . . . . . . . . 111
11 Frequency Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
11.1 Circuit Bode Plots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
11.2 Sketch Bode Plots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
11.3 LC Circuit Frequency Response . . . . . . . . . . . . . . . . . . . . . . . . . . 115
11.4 LRC Circuit Frequency Response . . . . . . . . . . . . . . . . . . . . . . . . . 116
11.5 Pole-Zero Plots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
12 Signals, Transforms, and Transfer Functions . . . . . . . . . . . . . . . . . . . . . . . 118
12.1 First Order Zeros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
12.2 Laplace Practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
12.3 Step-Zero . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
12.4 Laplace to Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
12.5 Time to Laplace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
12.6 Laplace Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
12.7 Signal in Time and Frequency Domain . . . . . . . . . . . . . . . . . . . . . . 124
12.8 Transfer Function and Time Constants . . . . . . . . . . . . . . . . . . . . . . 125
12.9 Fourier Series Input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
12.10 Spring Mass System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
12.11 Transfer Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
13 Controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
13.1 Rooftop Antenna . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
13.2 Lead Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
13.3 Third Order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
13.4 Error Constants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
13.5 System Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
13.6 Unity Feedback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
13.7 P and PI Controllers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
14 Motors and Transformers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
14.1 Locked Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
14.2 Motoring with a Capacitor . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
14.3 Leadscrew with Translating Stage . . . . . . . . . . . . . . . . . . . . . . . . 141
14.4 Mass Pulley System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
14.5 DC Motors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
14.6 Gear Train . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
14.7 Non-Ideal Transformer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
14.8 Equivalent Mass and Inertia . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
14.9 Inertia in Geartrain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
14.10 Motor Driving Inertia Through Gear Train . . . . . . . . . . . . . . . . . . . 148
15 MATLAB and Simulink . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
15.1 Matlab Spirograph . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
15.2 MATLAB Plotting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
15.3 Simulink Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
15.4 Wackygraph . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
4
16 Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
16.1 CD Player 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
16.2 CD Player 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
16.3 CD Player 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
16.4 Pinewood Derby 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
16.5 Pinewood Derby 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
16.6 Pinewood Derby 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
16.7 Pinewood Derby 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
16.8 Pinewood Derby 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
16.9 Engine Block Vibration 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
16.10 Engine Block Vibration 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
16.11 Engine Block Vibration 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
16.12 Plate On Springs 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
16.13 Plate On Springs 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
16.14 Mousetrap Dynamics 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
16.15 Mousetrap Dynamics 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
16.16 Mousetrap Dynamics 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
16.17 Mousetrap Dynamics 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
16.18 Hydraulic Elevator Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
16.19 Servo Position Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
16.20 Servo Frequency Compensation . . . . . . . . . . . . . . . . . . . . . . . . . . 175
16.21 Servo Torque Disturbance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
16.22 Plate On Springs Damped 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
16.23 Plate On Springs Damped 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
16.24 Plate On Springs Damped 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
16.25 Plate On Springs Damped 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
16.26 Plate On Springs Damped 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
16.27 Plate On Springs Damped 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
16.28 Shipping Crate 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
16.29 Shipping Crate 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
17 Quiz Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
17.1 Fun with Block Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
17.2 Complex Translation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
17.3 Cylinder Step Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
17.4 JKC Frequency Response 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
17.5 Mass Spring Damper Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . 194
17.6 Op-Amp Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
17.7 RLC Circuit Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
17.8 Sailing for Engineers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
17.9 Second Order Step Response . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
17.10 Spring Damper Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
17.11 Derive Blocks for Op-amp Circuit . . . . . . . . . . . . . . . . . . . . . . . . 200
17.12 Automobile Fender Spring/Damper System . . . . . . . . . . . . . . . . . . . 202
17.13 Cu Flywheel with Eddy Current Damper . . . . . . . . . . . . . . . . . . . . 203
17.14 Current Driven RC Circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
17.15 Driven Mass Spring System . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
17.16 Flywheel Driven by Hanging Mass . . . . . . . . . . . . . . . . . . . . . . . . 206
17.17 Homogeneous Second Order DE . . . . . . . . . . . . . . . . . . . . . . . . . 207
5
17.18 Match Pole/Zero Plots with Step Response . . . . . . . . . . . . . . . . . . . 208
17.19 Opamp Block Transfer Functions . . . . . . . . . . . . . . . . . . . . . . . . . 210
17.20 Piston with 2nd Order Translation . . . . . . . . . . . . . . . . . . . . . . . . 212
17.21 Pole Zero Bode Matching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
17.22 Power Semiconductor Thermal Problem . . . . . . . . . . . . . . . . . . . . . 215
17.23 Submersible Capsule Hoist System . . . . . . . . . . . . . . . . . . . . . . . . 216
17.24 Transfer Function from Pole Zero Plot . . . . . . . . . . . . . . . . . . . . . . 218
17.25 Vaccine Cooler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
17.26 Voltage Driven RRC Circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
17.27 Linear Mechanical System with Position Input . . . . . . . . . . . . . . . . . 222
17.28 Tank with Pump Inlet Lower than Outlet . . . . . . . . . . . . . . . . . . . . 223
17.29 Thermal Power Chip Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
18 Math Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
18.1 Complex Expression Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . 226
18.2 Complex Expressions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
18.3 Matrix Operation Practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
19 Recitation Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
19.1 Recitation 1 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
19.2 Recitation 2 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
19.3 Recitation 3 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
19.4 Recitation 4 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
19.5 Recitation 5 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
19.6 Recitation 6 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
19.7 Recitation 7 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
19.8 Recitation 8 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
19.9 Recitation 9 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
19.10 Recitation 10 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
20 Recitation Quizzes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
20.1 Recitation 1 Quiz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
20.2 Recitation 2 Quiz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
20.3 Recitation 3 Quiz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
20.4 Recitation 4 Quiz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
20.5 Recitation 5 Quiz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
20.6 Recitation 6 Quiz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246
20.7 Recitation 7 Quiz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
20.8 Recitation 8 Quiz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
20.9 Recitation 9 Quiz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
20.10 Recitation 10 Quiz Spring 2005 . . . . . . . . . . . . . . . . . . . . . . . . . . 250
Solutions 251
1 1st Order Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
1.1 First Order Time Constant . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
1.2 Rise and Settling Times . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
1.3 First Order System Response . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
2 2nd Order Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
2.1 Second-Order System Response . . . . . . . . . . . . . . . . . . . . . . . . . . 256
2.2 Second Order Responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
2.3 Second Order Derivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
6
2.4 Second Order Derivation Continued . . . . . . . . . . . . . . . . . . . . . . . 264
3 Higher Order Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
3.1 Structure of Higher Order System Solutions . . . . . . . . . . . . . . . . . . . 265
3.2 Structure of Higher Order System Solutions . . . . . . . . . . . . . . . . . . . 266
4 Mechanical Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
4.1 Balloon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
4.2 Bungee Jumper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
4.3 Elevator Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
4.4 Hoisting for Engineers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
4.5 Blocked Springs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
4.6 Engine Vibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
4.7 Garage Door . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284
4.8 Rotating Damped . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
4.9 Car Suspension 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
4.10 Disk Drive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
4.11 Crashworthiness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
4.12 Nonlinear Rotational System . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
4.13 Developing Dierential Equations . . . . . . . . . . . . . . . . . . . . . . . . . 301
4.14 Mass Spring System Frequency Response . . . . . . . . . . . . . . . . . . . . 304
4.15 Modeling Practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
4.16 Small Motion Transfer Function . . . . . . . . . . . . . . . . . . . . . . . . . 306
4.17 Mass Spring Damper System Frequency Response 1 . . . . . . . . . . . . . . 307
4.18 Mass Spring Damper System Frequency Response 2 . . . . . . . . . . . . . . 309
4.19 Propeller Shaft Vibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
4.20 Safe Packaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
4.21 Sliding Damped . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
4.22 Toy Flywheel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
4.23 Truck Shocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
4.24 Car Suspension 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
4.25 Kid-Skid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
4.26 Rolling Machine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
4.27 Sunday Bike Ride . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332
4.28 Compound Mass Spring Damper System 1 . . . . . . . . . . . . . . . . . . . 335
4.29 Compound Mass Spring Damper System 2 . . . . . . . . . . . . . . . . . . . 340
4.30 Wind Induced Building Vibrations . . . . . . . . . . . . . . . . . . . . . . . . 342
5 Electrical Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
5.1 Camera Flash Circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
5.2 Batteries and Capacitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
5.3 Loaded Motor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
6 Fluid Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
6.1 Rotational Fluid System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
6.2 Fluid Leak . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354
6.3 Whered the water go? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
7 Thermal Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
7.1 Thermal Block Question . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
7.2 The Hot Copper Block 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
7.3 The Hot Copper Block 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366
7.4 Transistor on heat sink . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370
7
8 Circuits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
8.1 RC Transfer Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
8.2 RLC Transfer Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
8.3 Bandpass Filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379
8.4 LC Circuit Dierential Equations . . . . . . . . . . . . . . . . . . . . . . . . . 382
8.5 Equivalent Impedance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
8.6 Circuit Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
8.7 LRC Circuit 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388
8.8 LRC Circuit 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
8.9 LR Circuit Step . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392
9 Op Amps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
9.1 Opamp Arithmetic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
9.2 Inverting Op-Amp . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
9.3 Opamp Transfer Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
9.4 Op-Amp Dierentiator 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
9.5 Op-Amp Proportional plus Integral Controller . . . . . . . . . . . . . . . . . 401
9.6 Op-Amp Circuit Design 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
9.7 Op-Amp Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405
9.8 Op-Amp Circuit Design 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406
9.9 Op-Amp Dierentiator 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408
9.10 RC Op-Amp Frequency Response . . . . . . . . . . . . . . . . . . . . . . . . . 413
10 Dierential Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416
10.1 Dierential Equations 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416
10.2 Dierential Equations 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419
10.3 Dierential Equations 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422
10.4 Eulers Theorem 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
10.5 Eulers Theorem 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
10.6 Force Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
10.7 Input for Zero Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429
10.8 Zero Input Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
10.9 Zero Step Response 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432
10.10 Zero Step Response 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
10.11 Nonlinear String System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
10.12 Homogenous and Particular Solutions . . . . . . . . . . . . . . . . . . . . . . 437
11 Frequency Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441
11.1 Circuit Bode Plots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441
11.2 Sketch Bode Plots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442
11.3 LC Circuit Frequency Response . . . . . . . . . . . . . . . . . . . . . . . . . . 444
11.4 LRC Circuit Frequency Response . . . . . . . . . . . . . . . . . . . . . . . . . 447
11.5 Pole-Zero Plots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
12 Signals, Transforms, and Transfer Functions . . . . . . . . . . . . . . . . . . . . . . . 452
12.1 First Order Zeros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452
12.2 Laplace Practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456
12.3 Step-Zero . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
12.4 Laplace to Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459
12.5 Time to Laplace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 460
12.6 Laplace Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461
12.7 Signal in Time and Frequency Domain . . . . . . . . . . . . . . . . . . . . . . 462
8
12.8 Transfer Function and Time Constants . . . . . . . . . . . . . . . . . . . . . . 463
12.9 Fourier Series Input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467
12.10 Spring Mass System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
12.11 Transfer Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478
13 Controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480
13.1 Rooftop Antenna . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480
13.2 Lead Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489
13.3 Third Order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 490
13.4 Error Constants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 493
13.5 System Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 494
13.6 Unity Feedback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499
13.7 P and PI Controllers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501
14 Motors and Transformers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 502
14.1 Locked Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 502
14.2 Motoring with a Capacitor . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505
14.3 Leadscrew with Translating Stage . . . . . . . . . . . . . . . . . . . . . . . . 507
14.4 Mass Pulley System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 508
14.5 DC Motors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 510
14.6 Gear Train . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 511
14.7 Non-Ideal Transformer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513
14.8 Equivalent Mass and Inertia . . . . . . . . . . . . . . . . . . . . . . . . . . . . 516
14.9 Inertia in Geartrain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 518
14.10 Motor Driving Inertia Through Gear Train . . . . . . . . . . . . . . . . . . . 519
15 MATLAB and Simulink . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 520
15.1 Matlab Spirograph . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 520
15.2 MATLAB Plotting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 524
15.3 Simulink Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 526
15.4 Wackygraph . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527
16 Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 531
16.1 CD Player 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 531
16.2 CD Player 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533
16.3 CD Player 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 536
16.4 Pinewood Derby 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 539
16.5 Pinewood Derby 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 541
16.6 Pinewood Derby 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542
16.7 Pinewood Derby 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 546
16.8 Pinewood Derby 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 548
16.9 Engine Block Vibration 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 552
16.10 Engine Block Vibration 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556
16.11 Engine Block Vibration 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 559
16.12 Plate On Springs 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563
16.13 Plate On Springs 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 565
16.14 Mousetrap Dynamics 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 566
16.15 Mousetrap Dynamics 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567
16.16 Mousetrap Dynamics 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 568
16.17 Mousetrap Dynamics 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 569
16.18 Hydraulic Elevator Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 570
16.19 Servo Position Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 576
9
16.20 Servo Frequency Compensation . . . . . . . . . . . . . . . . . . . . . . . . . . 577
16.21 Servo Torque Disturbance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 578
16.22 Plate On Springs Damped 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579
16.23 Plate On Springs Damped 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 586
16.24 Plate On Springs Damped 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 588
16.25 Plate On Springs Damped 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 590
16.26 Plate On Springs Damped 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
16.27 Plate On Springs Damped 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592
16.28 Shipping Crate 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596
16.29 Shipping Crate 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597
17 Quiz Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 599
17.1 Fun with Block Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 599
17.2 Complex Translation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 600
17.3 Cylinder Step Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 601
17.4 JKC Frequency Response 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602
17.5 Mass Spring Damper Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . 603
17.6 Op-Amp Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604
17.7 RLC Circuit Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 606
17.8 Sailing for Engineers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608
17.9 Second Order Step Response . . . . . . . . . . . . . . . . . . . . . . . . . . . 609
17.10 Spring Damper Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 610
17.11 Derive Blocks for Op-amp Circuit . . . . . . . . . . . . . . . . . . . . . . . . 611
17.12 Automobile Fender Spring/Damper System . . . . . . . . . . . . . . . . . . . 612
17.13 Cu Flywheel with Eddy Current Damper . . . . . . . . . . . . . . . . . . . . 613
17.14 Current Driven RC Circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . 614
17.15 Driven Mass Spring System . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615
17.16 Flywheel Driven by Hanging Mass . . . . . . . . . . . . . . . . . . . . . . . . 616
17.17 Homogeneous Second Order DE . . . . . . . . . . . . . . . . . . . . . . . . . 617
17.18 Match Pole/Zero Plots with Step Response . . . . . . . . . . . . . . . . . . . 618
17.19 Opamp Block Transfer Functions . . . . . . . . . . . . . . . . . . . . . . . . . 619
17.20 Piston with 2nd Order Translation . . . . . . . . . . . . . . . . . . . . . . . . 620
17.21 Pole Zero Bode Matching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 621
17.22 Power Semiconductor Thermal Problem . . . . . . . . . . . . . . . . . . . . . 622
17.23 Submersible Capsule Hoist System . . . . . . . . . . . . . . . . . . . . . . . . 623
17.24 Transfer Function from Pole Zero Plot . . . . . . . . . . . . . . . . . . . . . . 624
17.25 Vaccine Cooler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 625
17.26 Voltage Driven RRC Circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626
17.27 Linear Mechanical System with Position Input . . . . . . . . . . . . . . . . . 627
17.28 Tank with Pump Inlet Lower than Outlet . . . . . . . . . . . . . . . . . . . . 628
17.29 Thermal Power Chip Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 629
18 Math Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 630
18.1 Complex Expression Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . 630
18.2 Complex Expressions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 631
18.3 Matrix Operation Practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 632
19 Recitation Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 633
19.1 Recitation 1 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 633
19.2 Recitation 2 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 634
19.3 Recitation 3 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 640
10
19.4 Recitation 4 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 642
19.5 Recitation 5 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 643
19.6 Recitation 6 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645
19.7 Recitation 7 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 647
19.8 Recitation 8 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 648
19.9 Recitation 9 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 649
19.10 Recitation 10 Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650
20 Recitation Quizzes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651
20.1 Recitation 1 Quiz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651
20.2 Recitation 2 Quiz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 652
20.3 Recitation 3 Quiz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 653
20.4 Recitation 4 Quiz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654
20.5 Recitation 5 Quiz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655
20.6 Recitation 6 Quiz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656
20.7 Recitation 7 Quiz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 657
20.8 Recitation 8 Quiz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 658
20.9 Recitation 9 Quiz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
20.10 Recitation 10 Quiz Spring 2005 . . . . . . . . . . . . . . . . . . . . . . . . . . 660
11
Chapter
Problems
12
1 1st Order Systems
1.1 First Order Time Constant
This problem concerns the rst-order response shown in Figure 1.1.1.
A
m
p
l
i
t
u
d
e

0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
0 0.1 0.2 0.3 0.4 0.5 0.6
Time (sec.)
Figure 1.1.1: Plot of Response
(a) Whatisthetimeconstantofthisresponse? (Youcanworkdirectlyonthegiventimeresponse
plot and turn this in as part of your homework.)
(b) Show us a mechanical system which will give this response, under the assumption that the
indicated response is angle in radians. What are possible numerical values of the system
parametersandinitialconditionsthatgowiththisresponse? Besuretoshowyourreasoning.
13
1.2 Rise and Settling Times
Consider the rst-order system
y + y=u
driven with a unit step from zero initial conditions. The input to this system is u and the output
is y. Derive expressions for the 1090% rise time t
r
and the settling time t
s
, where the settling is
towithinanerrorfromthenalvalueof1. Howmanytimeconstantsarerequiredinorderto
settle to within an error of =10
6
?
14
1.3 First Order System Response
You are given an equation of motion of the form:
y + 5y= 10u
(a) What is the time constant for this system?
(b) Ifu=10,whatisthenalorsteady-statevaluefory(t)? Nowassumethatu=0(noinput)
and the system was started at some initial position y
o
, which you do not know. But, you do
know that 0.5 seconds later it was at a position y(0.5)=2.
(c) What was the initial condition y(0) that would lead to this result?
(d) Sketch as accurately as you can, the time response for case (c) starting from t= 0 until the
response is nearly zero, and indicate where the one data point at t= 0.5 would lie.
(e) When does the response reach 2% of the initial value?
(f) When does the response reach a value of 0.02?
15
2 2nd Order Systems
2.1 Second-Order System Response
For a system described by the homogenous equation:
20 y+160 y+8000y= 0
Determine the solution y (t) for three dierent initial conditions:
(a) y(0)=0
y(0)=10
(b) y(0)=10
y(0)=0
(c) y(0)=1
y(0)=10
(d) Forallthreecases,createaseparateplotofy(t)usingMATLAB,butbesuretousethesame
scale on all three plots.
(e) Compare the time for each of the three plots to reach steady-state. Comment on how you
can make this comparison and why the three answers do or do not dier.
16
2.2 Second Order Responses
This problem concerns the second-order response shown in Figure 2.2.1.
A
m
p
l
i
t
u
d
e

-200
-100
0
100
200
300
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time (sec.)
Figure 2.2.1: Second Order Response
A
m
p
l
i
t
u
d
e

-200
-100
0
100
200
300
0 0.05 0.1
Time (sec.)
Figure 2.2.2: Zoomed Second Order Response
(a) What are the natural frequency and damping ratio of this response? (You can work directly
17
on the given time response plot and turn this in as part of your homework paper.)
(b) Show us a mechanical system which will give this response, under the assumption that the
indicatedresponseispositioninmillimeters. Whatarepossiblenumericalvaluesofthesystem
parametersandinitialconditionsthatgowiththisresponse? Besuretoshowyourreasoning.
18
2.3 Second Order Derivation
Thesupplementalnoteshandedoutinclassgiveresultsforthepositionresponseofasecond-order
system to an initial condition of x
0
= 0 and v
0
= 1 for the underdamped, critically damped, and
overdampedcases,inEquations(15),(16),and(17),respectively. However,theseresultsaregiven
without showing the derivation. Carry out the detailed calculations to verify these homogeneous
responses.
19
2.4 Second Order Derivation Continued
Repeat the results of Problem 2.3 for the dierent initial condition x
0
= 1 and v
0
= 0, showing
thedetailsofyourderivation,aswellasanexpressionforthepositionasafunctionoftimeforthe
underdamped,criticallydamped,andoverdampedcases. UseMatlabtoplottheresultingposition
as a function of time for each of the parameter sets of values used in Figures 15, 16, and 17.
20
3 Higher Order Systems
3.1 Structure of Higher Order System Solutions
As we will see later in the term, the ship roll stabilizer problem can lead to an equation of motion
that is 4th order:
d
4
x d
3
x d
2
x d
2
x
a
4
+a
3
+a
2
+a
2
+a
0
x=bu(t)
dt
4
dt
3
dt
2
dt
2
Where x would be the roll angle and u the desired stabilizer n angle. Right now all we want
to do is understand the basic nature and range of possible homogeneous solutions for this type of
equation. ForthisproblemwewillnotuseLaPlacetransforms,butinsteadrelyonwhatwealready
know about homogeneous solutions for linear ordinary dierential equations. Assuming that u=0,
consider three dierent sets of parameters . a = a4 a3 a2 a1 a0 [ ]: Case I: . a = 1 10 35 50 24 [
] Case II: . a = 1 8 42 85 50 [ ] Case III: . a = 1 7 135 550 2500 [ ] Assuming that only x(0) is
non-zero:
(a) For each case, write out the basic form of the homogeneous solution where m is the number
of complex conjugate root pairs and n-2m is the number of real distinct roots, leaving any
real coecients as undetermined variables.
(b) Describe (Briey in words!) the resulting characteristic response.
(c) Sketchanapproximationofwhatitwilllooklike,butdonotsolvefortheciorcjcoecients.
HINT: You might nd it useful to use the MATLAB command roots(a), which will nd the roots
of a polynomial whose coecients are in a vector a.
21
3.2 Structure of Higher Order System Solutions
As we will see later in the term, the ship roll stabilizer problem can lead to an equation of motion
that is 4th order:
d
4
x d
3
x d
2
x d
2
x
a
4
+a
3
+a
2
+a
2
+a
0
x=bu(t)
dt
4
dt
3
dt
2
dt
2
where x would be the roll angle and u the desired stabilizer n angle. Right now all we want
to do is understand the basic nature and range of possible homogeneous solutions for this type of
equation. ForthisproblemwewillnotuseLaPlacetransforms,butinsteadrelyonwhatwealready
know about homogeneous solutions for linear ordinary dierential equations. Assuming thatu=0,
consider three dierent sets of parameters. a= [a4, a3, a2, a1, a0]:
Case I: a =[1 10 35 50 24]
Case II: a = [1 8 42 85 50]
Case III: a = [1 7 135 550 2500]
Assuming that only x(0) is non-zero:
(a) For each case, write out the basic form of the homogeneous solution where m is the number
of complex conjugate root pairs and n-2m is the number of real distinct roots, leaving any
real coecients as undetermined variables.
(b) Describe (briey in words!) the resulting characteristic response.
(c) Sketchanapproximationofwhatitwilllooklike,butdonotsolveforthec
i
orc
j
coecients.
HINT: You might nd it useful to use the MATLAB command roots(a), which will nd the roots
of a polynomial whose coecients are in a vector a.
22
4 Mechanical Systems
4.1 Balloon
Astringdanglingfromahelium-lledballoonhasitsfreeendrestingontheoorasshowninFig.
4.1.1. As a result the balloon hovers at a xed height o the oor and when it is deected a little
from that height it oscillates up and down for a while, eventually returning to the same height.
Developthesimplestmathematicalmodelcompetenttodescribetheverticalmotionoftheballoon.
You may take all the elements of your model to be linear. Show that your model is competent to
describe the observed hovering behavior.
Figure 4.1.1: Hovering Balloon
(a) Show that your model has a steady state.
(b) Showthatdisplacementfromthesteadystateevokesarestoringforce.Whatistheequivalent
stiness parameter?
(c) Show that your model could predict oscillation about the steady state.
(d) Show that your model predicts that the motion will eventually approach the steady state.
The balloon is released from rest at a small distance a above its steady-state height of 3.0 feet.
Afterthersthalf-cycleofoscillation,whichtakes1.5seconds,theballoonisatapoint0.2abelow
the steady-state height,with an instantaneous velocity of zero. The string is known to weigh 0.5
ounces per foot. Use these data to estimate the following behavioral parameters:
(e) the damping ratio ;
(f) the undamped natural frequency
o
;
and the following model element parameters:
(g) the eective mass m;
23
(h) the eective damping coecient b;
(i) the eective stiness k.
24
4.2 Bungee Jumper
A bungee jumper weighs 150 pounds. Bungee cords attached to her ankles have a slack length of
100feet. Shedivesoahightower,theelasticcordsextendandinstantaneouslyarresthermotion
at a lowest point A (well above the ground). The cords then retract and she bounces through
several cycles before nally coming to rest at a point B, where her ankles are 120 feet below the
upper attachment point of the bungee cord.
(a) Estimate the location of the low-point A.
(b) Duringtheportionofherjumpfromtheinitialtake-otothepointA,estimatehermaximum
downward acceleration?
(c) Duringtheportionofherjumpfromtheinitialtake-otothepointA,estimatehermaximum
upward acceleration.
(d) Briey explain the assumptions made in obtaining the previous estimates.
Seeking an ever-greater thrill the jumper doubles the slack length of the bungee cords(the new
lowest point A of her jump is still comfortably above the ground). For this second jump
(e) Estimate the location of the low-point A.
(f) Estimate the location of B where she nally comes to rest.
(g) Duringtheportionofherjumpfromtheinitialtake-otothepointA,estimatehermaximum
downward acceleration?
(h) Duringtheportionofherjumpfromtheinitialtake-otothepointA,estimatehermaximum
upward acceleration?
(i) Make a careful sketch of the time histories of vertical position during the initial downward
phase (from initial take-o to point A , or A) for the two jumps.Use the same time and
position axes for both curves.
25
4.3 Elevator Model
T
T
v
m
T
mg
O
3 ft/sec
1 sec Time
Velocity
Figure 4.3.1: Elevator Dynamics
Anelevator(Fig.4.3.1)isdesignedtomovebetweenoorsasfollows: startingatrestitaccelerates
to a speed of 3 feet/second in 1 second, then moves at a constant speed until it decelerates to rest
in1second. Woulditbereasonabletodesignthestructuresupportingthewinchmotor(andhence
theelevator)withoutconsideringdynamicforces? Thatis,doesdynamicsmatterinthissituation?
Provide a quantitative justication for your answer. Work in SI units.
26
4.4 Hoisting for Engineers
M
X

r
1
r
2
R

r
Combined inertia of gear and drums J
Motor
Pinion
Drums
Gear
Rigid base fixed to ground
,
Neglect Friction
Number of teeth on gear N , on pinion N
g
g p
Pulley P
Figure 4.4.1: Hoisting Mechanism
Figure4.4.1showsamechanismforhoistingamassM upaplaneinclinedatanangle tothehor-
izontal. ThemasslesshoistingcablerollswithnoslipoverapulleyP andovertwodrums(ofradii
r
1
andr
2
)attachedrigidlytoagearofradiusR.ThecombinedinertiaofthegearandthedrumsisJ.
A motor drives the pinion of radius r with an input torque . The transmission ratio of the gear
driveisN
g
/N
p
, where N
g
andN
p
arethenumberofteethonthegearandthepinionrespectively.
Note that this ratio is the same as R/r.
The system is mounted on a rigid base xed to the ground. To simplify our model, we neglect (i)
friction, (ii) inertia of the motor and the pinion and (iii) mass and inertia of the pulley P. The
cable remains under positive tension at all times. X is positive when measured upward along the
inclined plane as shown in the gure. Gravity acts on the system in the downward direction.
(a) Ifthepinionrotatesbyanangleintheclockwisedirectionasshowninthegure,showthat
the mass M moves up the inclined plane by a distance X given by
X =
1
2
(r
1
+r
2
)
r
R
(1)
Clearly show your reasoning.
(b) Derivetheequationofmotionforthesystem. Useandshowappropriatefreebodydiagrams
showing all the forces and torques acting on
(i) the pinion
(ii) the gear and drum unit
27
(iii) the pulley P and (iv) the mass M.
Your answer must be a dierential equation in X alone, with and g as inputs.
(c) Given the values R = 400 mm, r
2
= 300 mm, r
1
= 200 mm, r = 50 mm, M = 50 kg, g =
9.8 m/s
2
, = 30

, and J = 5 kgm
2
, what value of torque should be applied to the pinion
by the motor so that the mass M has an acceleration of 0.1g up the inclined plane?
(d) Now consider the system in Part (d) without the gear transmission drive. The drums are
driven directly by the motor in this new system. J, in this case, is the combined inertia of
themotorandthedrums. IfthevalueofJ remainsunchanged,calculatethevalueoftorque
thatshouldbeappliedtothedrumsbythemotortoachievethesameaccelerationof 0.1g
for the mass M.
(e) Assuming that the cost of a motor is proportional to its torque output rating, does your
answer in Part (d) support the use of the gear transmission drive? Explain.
28
coi l spri ng
shock absorber
wheel hub
t i r e
poi nt of at t achment
to car body
4.5 Blocked Springs
Figure 4.5.1 depicts the main components of the suspension for one wheel of an automobile. To
change the ride and handling qualities, automobile enthusiasts sometimes insert blocks between
some of the coils of the spring to prevent that part of it from deecting. Consider the case where
blocks are added to immobilize exactly half of the coils of each spring. Assume that:
z
Figure 4.5.1: Automobile Suspension
All four wheels are identical and have identical suspensions.
The car moves vertically as a rigid body.
The tire deections are negligibly small compared to the spring deections.
The shock absorbers exhibit linear viscous behavior.
(In practice these assumptions are not especially accurate but they will keep the analysis simple
and provide insight to the behavior of the suspension.)
It is known that the vehicle weighs 2,500 pounds, and that before the blocks were added the
suspensionwascriticallydamped( =1). Aftertheblockswereadded,therideheightwaschanged
by 2.5 inches. Use these data to estimate the following parameters. Work in SI units.
(a) The suspension stiness before, and after, the blocks were added.
(b) The undamped natural frequency of oscillation before, and after, the blocks were added.
(c) The suspension damping coecient before the blocks were added.
29
(d) The suspension damping ratio after the blocks were added.
Writestateequationsandoutputequationstocomputetheresponsetoanabruptlyappliedvertical
load of 1,000 pounds. Adapt the MATLAB scripts below to provide plots of the vertical displace-
ment of the vehicle from its resting height vs. time, for (i) the suspension without blocks, and for
(ii) the suspension with blocks added.
eqpos.m
Download
% eqpos.m Provides equation of motion for plate on springs
% to be integrated by script POS.m
function Xdot = eqpos(t,X)
global m k b fa
Xdot = [ 0 1 ; -k/m -b/m ]*X + [ 0 ; fa/m ];
pos.m
Download
% POS.m A MATLAB script for Plate On Springs. Produces plots of
% (i) position vs. time
% (ii) velocity vs. time
% (iii) velocity vs. position
% for the response of a steel plate on springs, with mass m, stiffness k,
% and damping parameter b, when the plate starts from initial conditions
% y = y0 and v = v0 under the action of a suddenly applied force fa at t = 0.
clear variables
global m k b fa
% Input parameters
m = input(Enter the mass "m" in kilograms );
k = input(Enter the stiffness "k" in Newtons/meter );
b = input(Enter the damping constant "b" in kilograms/sec );
fa= input(Enter the magnitude "fa" of the suddenly applied force in Newtons );
% Input initial conditions.
y0= input(Enter the initial displacement from equilibrium, in meters );
v0= input(Enter the initial velocity, in meters/second );
tspan = input(Enter the duration "T" of the desired time history, in seconds );
X0 = [ y0 ; v0 ];
% Integrate equations of motion
[t,X] = ode45(eqpos, tspan, X0);
% Plot results
plot(t,X(:,1)), title(Time History of Displacement),
xlabel(Time [Seconds]), ylabel(Displacement [meters]), pause
plot(t,X(:,2)), title(Time History of Velocity),
xlabel(Time [Seconds]), ylabel(Velcity [meters/second]), pause
plot(X(:,1), X(:,2)), title(Velocity vs. Displacement),
xlabel(Displacement [meters]), ylabel(Velcity [meters/second])
30
4.6 Engine Vibration
Asinglepistonweighs1poundandoscillatesatfrequencywithatotalstroke(toptobottomof
piston travel) of 2 inches.
(a) Assumingthatthepistonmotionissinusoidal,calculatetheamplitudeofthesinusoidalforce
that is required to move the piston at the following crankshaft speeds in revolutions per
minute (rpm):
N = 1500 rpm
N = 3000 rpm
N = 6000 rpm
The piston in (a) is assumed to be part of an engine, which weighs 200 pounds,and which is sup-
ported on a xed frame by mounts which have an eective stiness (for vertical motion) of 18,000
pounds/inch,andaneectivelineardampingcoecientof2pounds/inch/sec. Theoscillatingforces
consideredin(a)areforcesactingonthepiston. ByNewtonslawofactionandreaction,equaland
opposite forces act on the engine block whenever the crankshaft rotates at the speeds considered.
The rest of the problem is devoted to estimating how much vibration of the engine block results
from the action of one piston.
(b) Formulateanequationofmotionfortheengineblockinwhichtheoutputisthedisplacement
of the engine block, and the input is the reaction force from the motion of one piston.
(c) Derive expressions for the amplitude and phase of the steady-state displacement response to
piston reaction forces of the form f(t) = f
a
sint.
(d) Estimate the amplitude of the engine block displacement, due to the motion of one piston,
when the engine operates at
N = 1600rpm
N = 1800rpm
N = 2000rpm
(e) If these three speeds are generated by starting with the engine block at rest,in equilibrium,
and then immediately rotating the crankshaft at the full indicated speed,estimate how long
it would take the engine block, in each case, to reach steady state vibration.
31
4.7 Garage Door
Side View of Garage Door support
cable attached to fixed support
pulley
pulley
coil spring
door in four
hinged segments
cable attached to
lowest door segment
track to guide
door segments
chain drive to raise
door not shown
Figure 4.7.1: Schematic of Garage-Door-Support System
Figure 4.7.1 shows a side view of one side of the support mechanism used in many garage-door
openers. On each side of the garage door a portion of the weight of the door is supported by a
longcoilspringattachedtothedoorthroughacableandpulleysystem. Itwasobserved,afterthe
mechanism was recently serviced and freshly lubricated, that the suspended door had a tendency
to oscillate up-and-down when disturbed.

k

g m/2
Figure 4.7.2: Simplied Schematic of Garage-Door-Support System
32
Figure 4.7.2 shows a simplied schematic of one-half of the door support system. Half of the total
doorinertiaiscoupledtooneoftheoverheadspringsbythepulleysystem. Takethetotalweight
of the door to be 200 pounds and the stiness of each spring to be 5 pounds/inch.
(a) Formulate a model to analyze the oscillations of the door.
(b) Estimate the frequency, in Hz, of the oscillations.
(c) List the main assumptions underlying your model.
33
4.8 Rotating Damped
ConsidertherotorwithmomentofinertiaI rotatingundertheinuenceofanappliedtorqueT and
the frictional torques from two bearings, each of which can be approximated by a linear frictional
element with coecient B (Fig. 4.8.1).
T
I
B B
t
Figure 4.8.1: Rotating Inertia
(a) Formulate the state-determined equation of motion for the angular velocity as output and
the torque T as input.
(b) Consider the case where:
I = 0.001 kg-m
2
B = 0.005 N-m/r/s
Whatisthesteady-statevelocity
ss
whentheinputisaconstanttorqueof10Newton-meters?
(c) When the torque T varies sinusoidally at a frequency , the steady-state angular velocity
alsovariessinusoidallyatfrequency. Deriveaformulaforthesteady-stateangularvelocity
when T = T
a
sint. Evaluate the steady-state angular velocity response for the following
cases:
(i) T = 10 sin(50t) Newton-meters
(ii) T = 10 sin(5t) Newton-meters
(iii) T = 10 sin(0.5t) Newton-meters
(d) ConsidertheamplitudesA()ofthesteady-stateangularvelocityresponseasthefrequency
approacheszero. WhatisthelimitingamplitudeA
o
as0? EvaluatetheratioA()/A
o
for =50 rad/sec, =5 rad/sec, and =0.5 rad/sec.
(e) What is the break frequency of this system?
(f) Makeanaccurate, labeled,sketchoftheBodeplotsfortheamplituderatioA()/A
o
andthe
phase angle () of this system.
34
coi l spri ng
shockabsorber
wheel hub
t i r e
poi nt of at t achment
tocarbody
4.9 Car Suspension 1
For the second-order car suspension model shown in Figure 4.9.1, perform the following tasks:
(a) Solve for the unit step response from zero initial conditions. Write an expression for the
responseasafunctionoftime,anduseMatlabtographthisresponse. Forthisproblem,use
the parameters: m=500 kg, k= 5 10
4
N/m, and c= 2 10
3
Ns/m.
(b) For these parameters, the system will be underdamped. What are the values of
n
, ,
d
,
and ?
(c) Comment on the quality of the response to the step. Will the passengers like this ride? For
what value of c would the system have critical damping? Recalculate the step response for
this critical value of damping, and compare the maximum resulting acceleration experienced
by the passengers to the underdamped case.
(d) Compare the 5% settling time for both cases.
(e) UseaLaplacetransformapproachtosolveforthetimeresponse. Ofcourse,thisresultshould
be the same as what you found in part (a).
z
Figure 4.9.1: Car Suspension
35
4.10 Disk Drive
Read Head
Arm Assembly
Voice Coil Motor
Spindle
Disk
Figure 4.10.1: Picture of Hard Disk drive
A hard disk drive as shown in Figure 4.10.1 has two main components:
1. The head/arm assembly which moves the read/write heads laterally over the disk surface to
the desired tracks. The arm is driven by a voice coil motor.
2. The spindle/disk assembly which is driven by a permanent magnet spindle motor which
rotates at near constant speed.
This problem studies the spin-up/spin-down transients of the spindle assembly. Assume that the
spindleassemblyhasarotationalinertiaJ,andthatthemotoractsasasourceoftorque
m
, which
is constant independent of speed. Further assume that we can model the air drag acting on the
spindleaslinearlydependentupontheangularvelocity,witharotationaldampingcoecientb.
(a) ForthissystemdrawafreebodydiagramforthespindleinertiaJ showingthetorquesacting
on it.
(b) Use this free body diagram to derive a dierential equation in terms of the spindle angular
speed , an input torque
m
, and using the parameters given above.
(c) Now assume that the spindle inertia is 10
3
kgm
2
. (This is a big disk drive from an older
computer!) Weexperimentallyobservethatthediskspinsdownwithatimeconstant = 0.5
sec. What is the numerical value of the associated damping term b?
(d) Finally, assume that the disk is initially at rest at t = 0, is spun up by the motor using a
constanttorqueof0.5Nm. Whatistheresultingtransient(t)? Makeadimensionedgraph
of this response.
36
(e) After a long time, at t = t
1
, the motor is turned o and exerts no torque on the spindle.
Write an expression for (t) for t > t
1
, and make a dimensioned graph of the response.
37
4.11 Crashworthiness
v
o
k
m
b
Figure 4.11.1: Model of Vehicle Impacting a Barrier
A vehicle weighing 1 ton is driven into a xed concrete barrier at 10mph. The vehicles fender,
which strikes the barrier rst, is of the type that can deform under this load and return to its
original shape (undamaged) when it is unloaded. If the fender were unable to dissipate energy, its
maximum deection would be 6 inches.
(a) Estimate the eective stiness of the fender.
(b) Estimate the peak deceleration of the vehicle in SI units.
To absorb collision energy, linear dampers are added to the fender. Write state equations and
outputequationstopredictthedynamicsofthevehicle-fendersystemwhilethefenderisincontact
with the barrier. The outputs should include:
(i) The deection of the fender.
(ii) The deceleration of the vehicle.
(iii) The total force exerted on the barrier.
AdapttheMATLABscriptpos2.mandeqpos2shownbelowtointegratethesestateequations
tondtheresponseoftheseoutputsstartingfromthemomentthefenderrstcontactsthebarrier.
Makeplotsofthetimehistoriesofthesethreeoutputsforthefollowingvaluesofthedampingratio
:
(c) = 0.25
(d) = 0.50
(e) = 0.75
(f) = 1.00
(g) Inwhich,ifany,ofthecases(c)through(f)doesthefenderremainincontactwiththebarrier
after the impact is over?
(h) Can the peak deceleration of a vehicle with a fender with damping ever be greater than the
peak deceleration of a vehicle with an undamped fender? Give a brief physical explanation
for your answer.
(i) In which of the cases (c) through (f) is the peak deceleration the greatest?
(j) Estimate the value of the damping ratio which would minimize the peak deceleration.
38
eqpos2.m
Download
% eqpos2.m Provides equation of motion for plate on springs
% to be integrated by script POS.m
function Xdot = eqpos2(t,X)
global m k b fa
Xdot = [ 0 1 ; -k/m -b/m ]*X + [ 0 ; fa/m ];
pos2.m
Download
% POS2.m A MATLAB script for Plate On Springs. Produces plots of
% (i) position vs. time
% (ii) velocity vs. time
% (iii) velocity vs. position
% for the response of a steel plate on springs, with mass m, stiffness k,
% and damping parameter b, when the plate starts from initial conditions
% y = y0 and v = v0 under the action of a suddenly applied force fa at t = 0.
clear variables
global m k b fa
% Input parameters
m = input(Enter the mass "m" in kilograms );
k = input(Enter the stiffness "k" in Newtons/meter );
b = input(Enter the damping constant "b" in kilograms/sec );
fa= input(Enter the magnitude "fa" of the suddenly applied force in Newtons );
% Input initial conditions.
y0= input(Enter the initial displacement from equilibrium, in meters );
v0= input(Enter the initial velocity, in meters/second );
tspan = input(Enter the duration "T" of the desired time history, in seconds );
X0 = [ y0 ; v0 ];
% Integrate equations of motion
[t,X] = ode45(eqpos2, tspan, X0);
% Plot results
plot(t,X(:,1)), title(Time History of Displacement),
xlabel(Time [Seconds]), ylabel(Displacement [meters]), pause
plot(t,X(:,2)), title(Time History of Velocity),
xlabel(Time [Seconds]), ylabel(Velcity [meters/second]), pause
plot(X(:,1), X(:,2)), title(Velocity vs. Displacement),
xlabel(Displacement [meters]), ylabel(Velcity [meters/second])
39
4.12 Nonlinear Rotational System
Figure 4.12.1 shows a rotational inertia and damper system. The moment inertial of the rotor is
I. The damping torque from the damper is a nonlinear function of angular velocity,T
c
() = a
3
.
For each of three inputs (i)
1
= 1, (ii)
2
=8 and (iii)
3
=64
(a) Find the equilibrium point
(b) Derive a linearized model about the equilibrium point
(c) Solve for roots and plot them on the complex place.
(d) Plot the response of for step input =
i
(1+0.01u
s
(t)).
, input
I
, output
C
Figure 4.12.1: Nonlinear Rotational System
40
4.13 Developing Dierential Equations
C C
J
, output , input
(i): Rotational system
C C
M
X, output
F, input
K K
(ii): Translational system
Figure 4.13.1: System Figures
For each of the systems shown in Figure 4.13.1:
(a) Separate the system at a node or nodes into a free body diagram to show the forces acting
on each element.
(b) Use the free body diagram to develop a dierential equation describing the system in terms
of the indicated input and output. For each system, what is the system order?
41
4.14 Mass Spring System Frequency Response
F
m
x
k
Figure 4.14.1: Mass Spring System
In this problem, m=1 [kg], k=100 [N/m] and F(t) = sin t[N].
(a) Calculate an expression for the steady-state response x(t) = Msin(t+), with expressions
for M() and ().
(b) Make hand sketches of M() versus on log-log coordinates and () versus on semi-log
coordinates (linear in phase and log in frequency).
42
4.15 Modeling Practice
This problem concerns the spring-mass-damper system shown in Figure 4.15.1.
c
1 k
1
c
2 k
2
M
F
Position input w(t)
Mass position x(t)

g
Figure 4.15.1: Sprin-Mass-Damper System
In this gure, gravity acts on the mass in the downward direction as shown. The position of the
mass in the downward direction is x(t). A force F acts on the mass in the upward direction. The
upper spring and damper are connected to a position source w(t). That is, the position w(t) is
specied as in independent input. The position x is dened to be zero when gravity is not acting
on the mass, and when the applied inputs are zero.
(a) Draw a free-body diagram for the mass which shows all the forces and associated reference
directionsactingonthemass. Besuretolabeltheforceswiththeirdependenceonthesystem
position/velocity variables (if any).
(b) Use this free-body diagram to derive a dierential equation in terms of x and the system
parameters and inputs which describes the dynamics of this system.
43
4.16 Small Motion Transfer Function
For the system shown in Figure 4.16.1, calculate the transfer function X(s)/F(s), under the as-
sumption of small motions. Clearly show the steps in your derivation. The massless linkage has
lever arms l
1
and l
2
as shown.
Figure 4.16.1: Small Motion System
44
4.17 Mass Spring Damper System Frequency Response 1
F
k b
m
1
x
Figure 4.17.1: Mass Spring Damper System
In this problem, m=1 [kg], k=100 [N/m], b=1 [Ns/m] and F(t) = sin t[N].
(a) Calculate an expression for the steady-state response x(t) = Msin(t+), with expressions
for M() and ().
(b) Make hand sketches of M() versus on log-log coordinates and () versus on semi-log
coordinates (linear in phase and log in frequency).
45
4.18 Mass Spring Damper System Frequency Response 2
Consider the mechanical system shown in Figure 4.18.1. Note that F acts on m
1
in the direction
of x
1
.
k
1
b
1
m
1
F
k
2
b
2
m
2
x
1
x
2
Figure 4.18.1: Mass Spring Damper System
(a) Calculate the transfer functions H
1
(s) = X
1
(s)/F(s) and H
2
(s) = X
2
(s)/F(s) in terms of
the given parameters.
(b) Now let m
1
= 25 [kg], m
2
= 1 [kg], k
1
= 100 [N/m], k
2
= 10
4
[N/m] and b
1
= b
2
= 1
[Ns/m]. Use MATLAB to plot the Bode plots for H
1
and H
2
. Also plot the poles/zeros for
bothtransferfunctions. RelatefeaturesontheBodeplotstothepole/zerolocationsandthe
damping ratio of the poles and zeros. For a unit sinusoidal input F(t), at what frequency
is the motion on m
1
a relative maximum? What is the magnitude of motion on m
2
at this
frequency? ForaunitsinusoidalinputF(t),atwhatfrequencyisthemotiononm
1
arelative
minimum? What is the magnitude of motion on m
2
at this frequency?
46
4.19 Propeller Shaft Vibration
Thisisadierentkindofvibrationproblemforthelightaircraftenginewehavebeenconsidering.
Previously, we considered uniaxial translation of the engine block due to the inertia loading from
accelerating pistons. We now consider oscillations in the rotational speed of the propeller due to
torsionalvibrationoftheshortelasticcouplershaftconnectingthepropellertothecrankshaft. The
source of the oscillation is the uctuating speed generated by the reciprocating engine.
Periodicringofthecylindersinaninternalcombustionenginecausesitsrotationalspeedtovary
periodically. One stroke of a piston is one move from top dead center to bottom dead center (or
from bottom to top). In a four-stroke engine, three strokes are used to clear out the products of
combustion from the previous ring, let in fresh air and fuel, and compress the mixture prior to
ring. Itisonlyinthefourthstrokethattheexplosionoccursandaverylargeforceonthepiston
exertstorquearoundtheaxisofthecrankshaft. Inafour-cylinderenginetheresultingtorqueonthe
crankshaft is smoothed out considerably by arranging it so that one cylinder res on every stroke.
The remaining uctuation in torque, when applied to the inertia of the crankshaft, results in a
uctuation in the output speed of the engine, which varies in an approximately sinusoidal manner
at the ring frequency, which is half the rotational frequency of the crankshaft. It is this periodic
engine speed uctuation which excites the torsional vibration.
Consider the case of a four-cylinder 150 horsepower engine which operates between 500 and 2700
rpm. Themomentofinertiaofthetwo-bladedpropellercanbeestimatedtobethesameasthatof
auniformsolidrodofaluminum,sixfeetlongandtwoinchesindiameter(thedensityofaluminum
is 2.72grams/cc). It is observed that the steady-state oscillations of propeller speed at the ring
frequency reach a peak amplitude when the engine runs at2200 rpm. Furthermore the magnitude
of the oscillation at 2200 rpm is four times larger than the magnitude at 500 rpm.
(a) Develop a model to describe the steady-state uctuations in propeller speed(output) in re-
sponse to the uctuations in engine speed (input). To keep the analysis simple, assume that
theamplitudeoftheengine-speeductuationsdeliveredtothecouplershaftareindependent
of the engine speed, so that in the steady state the angular position
eng
and the angular
speed
eng
of the engine can be assumed to take the form

eng
= t+sin t
2

eng
= (1 + cos t)
2 2
(b) Use your model to estimate the torsional stiness K of the elastic coupler shaft.
47
4.20 Safe Packaging
A packing crate was designed to protect a fragile instrument during shipment. Assuming that the
packing material can be modeled as an ideal linear spring of stiness, k , in parallel with an ideal
linear damper, b , and that the instrument and crate are of mass, m
1
and m
2
, respectively, the
system can be modeled as shown in Figure 4.20.1A.
Instrument
Crate
Packing
Material
b k
g
m
1
m
2
Lo
h
b k
m
1
L(0)=Lo
v(0)=Vo
g
A B C
Figure 4.20.1: Shows actual system and two models of the situation.
Thepackingcrate(withinstrumentinside)isdroppedfromaheight,h,asshowninFigure4.20.1B.
Theheightissucientlylargethatbythetimethecratehitstheground,thespringisfullyextended
to its unloaded length, L
o
, as shown in Figure 4.20.1C. Note that the crate hits the ground with
velocity, V
o
, and in the presence of gravity.
(a) Derive the dierential equation for the system. Clearly indicate the initial conditions, and
any inputs present.
(b) For what values of, b , will the instrument oscillate?
(c) Assumingthattheinstrumentdoesoscillate,deriveananalyticalexpressionforthecomplete
solution.
48
4.21 Sliding Damped
Consider the mass m sliding horizontally under the inuence of the applied force f and a friction
force which can be approximated by a linear friction element with coecient b (Fig. 4.21.1.
m
v
f
Friction, b
Figure 4.21.1: Sliding Mass
(a) Formulate the state-determined equationof motion for the velocityv as output and the force
f as input.
(b) Consider the case where:
m = 1000 kg
b = 100 N/m/s
What is the steady-state velocity v
ss
when the input is a constant force of 10Newtons?
(c) When the force f varies sinusoidally at a frequency , the steady-state velocity also varies
sinusoidallyatfrequency. Deriveaformulaforthesteady-statevelocitywhenf =f
a
sint.
Evaluate the steady-state velocity response for the following cases:
(i) f = 10 sin(0.5t) Newtons
(ii) f = 10 sin(0.05t) Newtons
(iii) f = 10 sin(0.005t) Newtons
(d) Consider the amplitudes A() of the steady-state velocity response as the frequency ap-
proacheszero. WhatisthelimitingamplitudeA
o
as0? EvaluatetheratioA()/A
o
for
= 0.5 rad/sec, =0.05 rad/sec, and =0.005 rad/sec.
(e) What is the break frequency of this system?
(f) Make an accurate, labeled sketch of the Bode plots for the amplitude ratioA()/A
o
and the
phase angle () of this system.
49
4.22 Toy Flywheel
A toy consists of a rotating ywheel supported on a pair of bearings as shown in Figure 4.22.1.
The ywheel is connected to a pulley , around which is wrapped a exible but inextensible cable
connected to a spring. In operation, the ywheel is initially at rest, the string made taut, and at
t= 0, the input x
s
(t) undergoes a step change in position of magnitude x
0
.
x(t)
x (t)
s
t=0
t=0
r
J
c
t
flywheel
bearings
pulley
t
x (t)
s
x
o
Figure 4.22.1: Flywheel Diagrams
Assumethattheywheel-shaft-pulleyunithasrotationalinertia,J.Thebearingscanbemodelled
as a viscous rotational damper of coecient, c
t
. the pulley is of radius, r. The spring is an ideal
linear spring with spring constant, k.
(a) Writethesystemequationasadierentialequationinx(t), the length of cable unwound from
the pulley, as well as system parameters, J, r, and k. [Note that x(0)=0]
(b) For what range of values of c
t
(expressed in terms of system parameters J, r, and k) will the
cable never go slack?
(c) Assumingthatc
t
hassomenon-zerovaluesuchthatthecabledoesgoslack,writeanexpres-
sion (in terms of system parameters J,r, and k) for the response x(t), i.e. the length of cable
unwound from the pulley. Sketch the response, x(t). Carefully indicate the time over which
the expression and the sketch are valid.
(d) Assume that c
t
is zero. Write an expression (in terms of system parameters J, r, and k) for
t, the time at which the cable rst goes slack.
50
4.23 Truck Shocks
At a tailgate party, it was observed that when Uncle Massive, a sprightly 250-pounder,hopped up
onthebackofhispick-uptruck,thetrucksuspensionbouncedformorethan5clearlydenedcycles
(it has been some time since Uncle Massive has replaced his shock-absorbers) at a frequency of 2
Hz. When the vibration stopped it was noted that the back of the truck was one inch lower than
it was before Uncle Massive sat down.
(a) Estimate the eective mass of the back end of the truck.
(b) Briey explain any assumptions you have made in arriving at your estimate.
51
4.24 Car Suspension 2
A simplied model for a car suspension is shown in Figure 4.24.1. Here, the road position is r(t),
and the car body position is x(t). A force F is applied to the body mass. Assume r(t) = 0
throughout this problem. We also let the force be a step function F(t)=500u
s
(t) [N].
(a) Solvefortheresultingstepresponsex(t)fromzeroinitialconditions. Writeanexpressionfor
theresponseasafunctionoftime,anduseMatlabtographthisresponse. Forthisproblem,
use the parameters: m=500 kg, k= 5 10
4
N/m, and b= 2 10
3
Ns/m.
(b) For these parameters, the system will be underdamped. What are the values of
n
, ,
d
,
and ?
(c) Comment on the quality of the response to the step. Will the passengers like this ride? For
what value of b would the system have critical damping? Recalculate the step response for
this critical value of damping, and compare the maximum resulting acceleration experienced
by the passengers to the underdamped case.
(d) Compare the 5% settling time for position x(t) for both cases.
[[
U
P
N E
)
Figure 4.24.1: Car suspension quarter model
52
4.25 Kid-Skid
You have a great new business idea! Sick and tired of seeing kids injured when their friends jump
o the see-saw (teeter-totter, etc), you invent the Kid-SkidR _ is simply a spring _. The Kid-SkidR
dampermechanismthatinsuresasoftlandingifthefriendontheotherendjumpso. Aschematic
is shown in Figure 4.25.1:
Kid-Skid
h
b k
Figure 4.25.1: Kid-SkidR _ Schematic
Assuming that the see-saw has an length L, and a mass m
s
, and that the kid on the right hand
side is of mass m
k
:
(a) Howwouldyoumodelthissystem,forthecasewherethenastykidontheleftjumpsowhile
thepoorkidontherightisatthefullheighthabovetheground. Sincewedonotknowhow
to model impact(wait until 2.004 for that), assume that the Kid-SkidR _ is massless.
(b) Ifyouwanttowritetheequationofmotionforthekid-seesawKid-SkidR _systemoncecontact
has been made, what would the equation be, and what would be the initial conditions. (Be
specic about the initial conditions, and express them in terms of the system parameters
given.)
(c) If the Kid-SkidR = y
max
, how would you ensure that it was not _ has a full range of motion
exceeded given the parameters above?
53
4.26 Rolling Machine
Figure 4.26.1: Rolling Machine Schematic
A model of a commercial rolling machine (for metals processing) has been created and is shown
above. Themodelcomprisesacylinderwithamass,m,witharadius,r= 0.5m,thatspinsabout
an axle. The cylinder rolls without slip on the lower surface. Attached to the axle housing are a
damper, b, a spring, k=200 N/m, and a force source, f. In testing the machine, to impose a step
ontheforcesource f, andmeasurethefollowingresponse for the horizontalmotion ofthe axle,x.
Figure 4.26.2: System Step Response
Using the given parameters and the measured step response in Figure 4.26.2,
(a) Showtheafreebodydiagramofthemass/inertia,anddeterminetheequationofmotionfor
x(t)
54
(b) From the data determine the damping constant, b, and the equivalent mass, m
eq
, where m
eq
is mass equivalent of the combined rotational inertia and linear mass.
(c) If the cylinder has a m=3 kg, determine the inertia of the cylinder.
55
4.27 Sunday Bike Ride
AcoupleisoutforaSundaybikeride. Theguy(weighing 200lbs)isona30yrold,allsteel(very
heavy)10speed,andhiswife(weighing 100lbs)isonalightweightracingbikewithhigh-pressure
tires, aluminum rims; etc. They are riding alongside each other when they reach a signicant
downhill. They agree to coast and not pedal or brake on the hill. At the bottom, the guy ends up
veryfaraheadofhiswife. Theyarebothconfounded;whyspend$1000onherfancybikeifthisis
going to happen?
You must explain it using system dynamics.
(a) Create a model for the bike that will lead to a rst order equation of motion (even though
it has two rotating wheels with inertia + the inertia of the frame and rider
1
). Consider all
sources of friction or drag on the bike as it is pulled downhill.
(b) Draw the free body diagram for the bike that will show all forces acting on it.
(c) Write the equation of motion and justify the assumption that the system is linear.
(d) Based on the description of the bikes and the riders, use the model to show why the hefty
guy on a clunky bike ends up going faster at the bottom of the hill.
Remember we said we could only deal with xed axes of rotation using 2.003
56
1
4.28 Compound Mass Spring Damper System 1
m
n
1
n
2
b
1
b
2
k
1
k
2
b
3
k
3
x
g
F
1
F
2
Figure 4.28.1: Mass Spring Damper System
Refering to Figure 4.28.1 and neglecting gravity, please do the following:
(a) Draw free body diagrams for the mass m, and each of the nodes n
1
and n
2
.
(b) Derive the 2
nd
order dierential equation for the position x(t) of the mass m.
(c) Using the following parameters
m= 1[kg]
k
1
=k
2
=10000 [N/m]
k
3
=5000 [N/m]
b
1
=b
2
=10 [Ns/m]
b
3
= 5 [Ns/m]
give values for the damped natural frequency
d
and the damping ratio . Plot the poles of
this system on the complex s-plane.
(d) Using MATLAB, plot the response of this system for the following inputs and initial condi-
57
tions:
x(0)=10
2
[m]
x (0)=0 [m/s]
F
1
=F
2
= 0 [N]
58
4.29 Compound Mass Spring Damper System 2
m
n
1
n
2
b
1
b
2
k
1
k
2
b
3
k
3
x
g
F
1
F
2
Figure 4.29.1: Mass Spring Damper System
This problem is a continuation of Problem 4.28. You should use the solutions for that problem to
aid you in this problem. Use the same parameters for the mass, damping coecients and spring
constants. Referring to Figure 4.29.1 and neglecting gravity, solve x(t) and sketch the solution by
hand using the following parameters:
F
1
= 10 [N]
F
2
= 0 [N]
x(0)=x (0) =0
59
4.30 Wind Induced Building Vibrations
Tallbuildingsareoftensubjecttonoticeableswayorhorizontalmotionattheupperoors. This
swayisdrivenbywindpressurevariationsonthebuilding. Thebuildingitselfisaninterconnected
set of columns and beams, all with continuously distributed mass and elasticity. Assume that we
want to create a simple, lumped parameter model of this building to capture the sway motion
dynamics.
A measurement was made of this motion along with wind velocity, and the data is shown in
Figures 4.30.1 and 4.30.2.
(a) Propose a mechanical model to describe the motion that would lead to the equation:
x + ax + bx=cu
wherexistheswaymotionofthebuildinganduistheforcefromthewind. Besuretoprovide
a simple sketch of your equivalent mechanical model and the resulting free body diagram.
(b) From the data, which of the parameters of the equation (a, b and c) could you determine?
(Or in other words, why cant you determine all three?)
(c) For this data, determine:
n
, , , and
d
.
(d) Plot the roots of the characteristic equation for this system on an s-plane. A highly paid
consultant has said that she can double the damping of the building with a feedback control
system.
(e) On the s-plane plot from d) show where the new roots would end up if she succeeds.
(f) Describe how the building would have reacted to the increased damping, and sketch the
responseonaplotwithscalessimilartoFig. 2. Pleasemakeyourdescriptionintermsof2nd
order system transient characteristics.
Figure 4.30.1: Wind pressure versus time. (Note the wind stops suddenly at t=50.)
60
0 10 20 30 40 50 60 70 80 90 100
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 10 20 30 40 50 60 70 80 90 100
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
Figure 4.30.2: Building Motion (x(t)) vs. time.
61
5 Electrical Systems
5.1 Camera Flash Circuit
This problem considers further the ash circuit described in class. Suppose that the capacitor has
a value C =120F, and the battery voltage is V
bat
= 1.2V. Also suppose that the largest inductor
you can use is L=1 mH, and that this inductor has a maximum current limit of 0.1 A.
(a) How much energy can be transferred from the inductor to the capacitor in one cycle?
(b) How many cycles would be required to charge the capacitor from 0 to 300 volts? What
frequency of charging (in inductor charge/discharge cycles per second) would be required to
do this in one second?
(c) What is the maximum charging frequency in cycles per second which is possible given the
parametersforL,C, and V
bat
above? Explainyourreasoning. Whatchangescouldyoumake
to the components to allow faster charging?
62
5.2 Batteries and Capacitors
With terminals shorted, a battery has a short circuit current dened as I
sc
. With its terminals
open, it has an open-circuit voltage dened as V
oc
. For example, a 1.5V AA cell as shown in the
gure 5.2.1
Figure 5.2.1: AA Battery
can be modeled internally as a 1.5 V voltage source in series with an internal resistance R
b
as
shown in the gure 5.2.2
63
Figure 5.2.2: AA Battery Schematic
If we eliminate the packaging details, this circuit appears more simply as gure 5.2.3, where V
b
is
the voltage appearing at the battery terminals.
-
+
+
-
Figure 5.2.3: Simplied battery circuit
(a) With a short briey applied to such a 1.5 V battery, I measured a current out of the +
terminal and back into the - terminal of I
sc
= 5.7A. What is the value of R
b
?
(b) In the experiment above, how much power is the battery voltage source supplying? Where
is it being dissipated? (Batteries are really not happy having ashort-circuit applied to them
for any signicant time, as they can dangerously overheat due to the internal dissipation.).
An electrolytic capacitor, as shown in the picture 5.2.4, allows a relatively large amount of
capacitanceinasmallpackageascomparedwithothertypesofcapacitors. Thesecapacitors
64
arepolarized,inthesensethattheyshouldonlyhavevoltageappliedoftheindicatedpolarity.
Reversal of voltage can be hazardous, and cause failure or explosion.
Figure 5.2.4: Electrolitic capacitor of a single use ash camera
Thecapacitorshownisusedtoprovideapulseofcurrentusedtoreaashlampinasingle
use camera.
Afterthecapacitorischarged,weobservethatthecapacitorvoltagebleedsawaytozeroover
the course of about 30 minutes. This internal loss can be modeled by including a leakage
resistor R
l
in parallel with the capacitor as shown in gure 5.2.5.
+
Figure 5.2.5: Electrolitic capacitor model
If we eliminate the packaging details, this circuit appears more simply as in gure 5.2.6.
65
+
+
-
Figure 5.2.6: Electrolitic capacitor circuit
The + sign is usual to indicate that this is an electrolytic capacitor, as well as its associated
allowed polarity.
(c) SupposethatC =120uF,andthatweobservethatfromaninitialchargeittakes10minfor
thecapacitorvoltagetofallto37percentofitsinitialvalue. Whatisthevalueoftheleakage
resistor R
l
?
Now, we want to study using the battery to charge the capacitor. At t = 0, the battery is
connected to the initially uncharged capacitor.
(d) Draw a circuit diagram showing this connection.
(e) Write a dierential equation which discribes the behavior of the circuit in this connection.
The dierential equation should be written in terms of the capacitor voltage v
c
.
(f) From the initially uncharged state at t = 0, write a solution v
c
(t) for this charging process.
Whatistheassociatedtimeconstant? Makeadimensionedgraphofv
c
(t) and show the time
constant on this graph. Where is the system pole located? Make an s-plane sketch showing
this pole location.
After charging for a long time, the capacitor voltage reaches a nal, or steady state value.
Then, at time t
1
, the battery is disconnected from the capacitor.
(g) Write a dierential equation which describes the behavior of the capacitor voltage v
c
(t) in
this disconnected condition. Solve the dierential equation for v
c
(t), t > t
1
, and make a
dimensioned sketch of v
c
(t). Where is the system pole located? Make an s-plane sketch
showing this pole location.
66
5.3 Loaded Motor
This problem considers the motor system shown below.
Figure 5.3.1: Loaded Motor
The motor in Figure 5.3.1 is driven by a voltage source V through a switch, which can be opened
and closed. The motor internal coil resistance is modeled by the resistor R as shown. The motor
itself is ideal and described by the relations = Ki, and e = K, where is the torque applied
to the load, and is the load speed. The load consists of a rotary inertia J, and a rotational
damper b.
(a) Suppose the system is initially at rest. At t=0 the switch is closed. Solve for the resulting
rotor speed as a function of time, and write an expression for (t) in terms of the system
parameters. Clearlyshowthestepsinyoursolution. Makealabeledanddimensionedsketch
of (t).
(b) Aftertheswitchhasbeenclosedforalongtime,theswitchisopened. Letscallthisopening
time t
1
. Solve for the resulting rotor speed as a function of time for t > t
1
, and write an
expressionfor(t)intermsofthesystemparameters. Clearlyshowthestepsinyoursolution.
Make a labeled and dimensioned sketch of (t).
(c) Solve for the value of the coil current i as a function of time throughout the motions deter-
mined above, and write an expression for i(t) in terms of the system parameters. Clearly
show the steps in your solution. Make a labeled and dimensioned sketch of i(t).
67
6 Fluid Systems
6.1 Rotational Fluid System
Cross-section
(t)
1
(t)
2
c
1
c
2
Inertia, J
fluid-filled
cup
(t)
2
(t)
1
c
2
J
c
1
Top View
Figure 6.1.1: Cross-Sectional and Top view
This problem considers the rotational mechanical system shown in Figure 6.1.1.
The setup is similar to the one used in Lab 2. The central shaft rotates at an arbitrary velocity

1
(t) in a uid lled cup, which is xed. Unlike Lab 2, there is an intermediate ring, of inertia J.
Thisringissupportedonbearingswhicharenotshowninthegure. Theringrotateswithangular
velocity
2
(t) as shown.
Theuid-lledannulicreateadamperc
1
betweentheshaftandthering,andadamperc2 between
the ring and the cup. Assume that any other damping is negligible.
(a) Draw a free-body diagram for the ring showing the torques acting on the ring.
(b) Usethisfree-bodydiagramtoderiveadierentialequationintermsof
1
(t) and
2
(t) which
describes this system. Note that
1
(t) is an arbitrary velocity which is externally specied.
(c) Assume that
1
(t) is a step, i.e.,
1
(t) = u
s
(t), and that
2
(0) = 0. solve for the resulting
motion
2
(t) for t0.
(d) In steady-state, what torque must be exerted on the input shaft? Why?
68
6.2 Fluid Leak
Figure 6.2.1: Diagram of Fluid Problem
YoumustreadtheFluidExampleonthewebpagebeforestartingthisproblem. Imagine
that in the system in Figure 6.2.1 the valve is shut so that there is no ow at t=0 and that P
s
(t)
leaks down from 3 10
5
Paveryrapidlyafterthevalveisclosed,ie. P
s
(0

) = 3 10
5
andP
s
(0
+
) = 0.
(a) Solve for P
c
(t).
(b) At what time is P
c
(t) at its minimum value?
(c) How long does it take this system to settle withen P
c
(t)=300 Pa?
69
6.3 Whered the water go?
This problem extends the plumbing example presented in the supplementary notes. The gure
from the notes is repeated below.
Figure 6.3.1: Shower Figure
Youareupstairstakingashowerwhileyour9yearolddaughterisdownstairsskatingmadlyaround
in the basement on her in-line skates with a hockey stick. (Why indoors? Why in the basement?
Why with a hockey stick? Why didnt she ask if this was a good idea? If these questions havent
occurred to you, you arent a parent yet.)
Anyway,takinganespeciallysharpturn,sheslidesoutskatesrstintothepipeswheretheyconnect
to the pressure tank which breaks them o essentially instantaneously. With calm engineering
detachment, we will model this event as taking the pressure rapidly to zero, i.e., as P
s
(t) =
310
5
(1u
s
(t)) [Pa].
P
s
(t)[Pa]
3
.
10
5
[Pa]
t[s]
Figure 6.3.2: Pressure function
(a) Usethemodelandparametersfromtheexampleproblem. Developananalysiswhichpredicts
theresultingpressureandowattheshowernozzle. Sketchthesetwofunctions. Abouthow
long does it take for the shower ow to go to zero?
70
(b) So now youre standing there naked in the shower with soap in your hair and no water ow
wondering what happened, when you hear your kid screaming that the basement is starting
to ll up with water. (The water pump is valiantly trying to restore pressure to P
s
= 3 10
5
Pa). We leave the next steps to your imagination...
(c) Ifallofthisseemsabitrigorous,perhapsyoulackthesternconstitutionrequiredforparent-
hood?
(d) (Adapted from a true story).
71
7 Thermal Systems
7.1 Thermal Block Question
A block of copper is brought out of an oven at an initial temperature of 1000 K, and allowed to
cooltotheambienttemperatureof300K.Thegraphbelowshowsdatawhichmightrepresentthis
experiment. The copper block has a mass of 0.1 kg. Given that copper has a specic heat of 385
J
, propose a thermal model of this experiment, and write the equation of motion. Given the
kg
o
C
experimental data, determine the value of thermal resistance to ambient. Clearly show the steps
in your derivation.
200
300
400
500
600
700
800
900
1000
T
b
l
o
c
k
(
K
)

T
amb
0 500 1000 1500 2000 2500 3000
time(s)
Figure7.1.1: Theblocktemperature(T
block
)isplottedagainsttime(t). Theambienttemperature
is given by T
amb
.
72
7.2 The Hot Copper Block 1
Thisproblemconsidersthethermaldynamicsofwaterandacopperblockinsideavacuumbottle.
A diagram for the system is shown below in Figure 7.2.1.
Figure 7.2.1: Copper block and water thermal system
Initially, a 100 gm copper block is heated outside the bottle in an oven to a temperature of 95

C.
Wearenotconcernedwiththeheatingprocess. Thewaterinthevacuumbottlehasbeentherefor
a long time and is equilibrated to the ambient temperature T
a
= 25

C.
At t = 0, the stopper is removed, the copper block is lowered into the bottle, and the stopper is
replaced. Now we would like to develop a model which can describe the temperature dynamics as
a function of time.
Thesystemparametersareasfollows: Thecopperblockhasamassof100gm;copperhasaspecic
heat of 0.385 J/gm

C=385J/kg

C. The bottle contains 1 liter of water; water has a specic heat


of 4.2 J/gm

C=4200J/kg

C.
ThetemperatureoftheblockisdenedasT
b
. Thewaterisassumeduniformintemperature,with
a value of T
w
.
The heat ow from the block to the water is dened as q
bw
[W]. The heat ow from the water
to ambient is dened as q
wa
[W]. These ows are modeled as passing through thermal resistances
R
bw
= 0.1

C/W and R
wa
= 17

C/W, respectively. Note: Take a look at the 2


nd
order ther-
mal example on the course web page before attempting this problem.
73
(a) Write the governing dierential equations for the system in state-space form as
dT
w
=f
1
(T
w
, T
b
, T
a
)
dt
dT
b
=f
2
(T
w
, T
b
)
dt
where f
1
and f
2
are linear functions.
(b) Convert this representation to a 2
nd
order dierential equation in T
w
(t). You may assume
that the ambient temperature is constant, and thus
dTa
= 0.
dt
(c) Solve for the response T
w
(t) for t 0. What are the system natural frequencies? Plot the
system poles on an s-plane plot.
(d) Use Matlab to plot T
w
(t), and its two constituent modes, over an interesting time range.
You may need to use two plots with dierent time scales to capture the dynamics. Explain
qualitatively why the plot looks as it does. Can you show us the eect of the system time
constants?
(e) Supposetheblockwereinitiallyatonly60

C. Make a sketch on your time response plot of the


resulting T
w
(t). Explain your reasoning. (Hint: If you exploit linearity, this should require
no new calculations.)
74
7.3 The Hot Copper Block 2
This problem is a continuation of Problem 7.2. Reconsider the system of Problem 7.2, but with
an electrical heater added to the inside of the copper block. The heater provides an input power
q
in
= 4 W. This new system is shown in Figure 7.3.1. The system starts at t = 0 in a rest state
with T
b
=T
w
=T
a
= 25

C. At t=0, the heater is turned on.


Q
IN
Figure 7.3.1: Copper block and water thermal system with heater
(a) Write the governing dierential equation in terms of water temperature T
w
.
(b) Solve the dierential equation for T
w
(t), with the numerical values given in Problem 7.2.
(c) Use MATLAB to plot T
w
(t) over both a short and long time interval to show the interesting
partsofthegraphs. Whatisthesteady-statevalueofT
w
? Whatisthe10%to90%risetime
t
r
? How long does the response take to settle within 1% of the nal value?
75
7.4 Transistor on heat sink
Figure 7.4.1 is a schematic cross-section of a power transistor mounted on a heat sink. The tran-
sistor itself is fabricated on a thin piece of silicon, perhaps 5mm x 5mm x 0.5mm. This device
is mounted within a package (case) which provides mechanical protection as well as electrical and
thermal connections to the outside world. The case is mounted on a nned heat sink in order to
transfer heat dissipation in the transistor into the ambient air.
Figure 7.4.1: Transistor Schematic
The case of most transistors needs to be electrically insulated from the heat sink. This is ac-
complished with the indicated thermal washer, which has the conicting requirements of thermal
conduction and electrical insulation.
WedenethedevicetemperatureasT
d
,thecasetemperatureasT
c
,andtheheatsinktemperature
asT
s
. TheambienttemperatureisT
a
= 25

C.Thethermalresistancebetweenthedeviceandcase
is R
dc
, between the case and heat sink is R
cs
, and between the heat sink and air is R
sa
.
Thetransistorisusedasanelectricalswitchoramplier. Intheprocesselectricalpowerisconverted
to heat. The heat ow needs to be managed without allowing the device to become too hot
(generally the device temperature should stay below 100

C).
(a) Assume R
dc
= 0.1

C/W, R
cs
= 0.2

C/W, and R
sa
= 0.5

C/W. Further assume that in


steady-state the transistor is dissipating 50W. What are the steady-state values of T
d
, T
c
,
and T
s
?
(b) Nowassumetheheatsinkisconstructedwith500gmofaluminum,whichhasaspecicheat
of0.90J/g

K.Thesystemisoperatinginthesteady-stateconditionabovewhen,att= 0 the
transistor is pulsed with 500W of power dissipation for 1 second, and returns to dissipating
76
50W.WriteexpressionsforT
d
(t),T
c
(t),andT
s
(t) for t0,andmakeplotsofthesefunctions.
Assume that the device and case have zero thermal capacitance. What is the peak device
temperature in this linear model. Will it likely survive the transient? Would a bigger heat
sink help?
77
8 Circuits
8.1 RC Transfer Function
This problem considers the circuit shown in Figure 8.1.1.
Figure 8.1.1: RC Circuit
(a) Calculate the transfer function V
o
(s)/V
i
(s).
(b) For the values R
1
= 3k, R
2
= 12k, and C = 10 F, solve for the pole and zero of the
system, and plot these on the s-plane.
(c) For the values given in part (b), calculate the transfer function magnitude and phase as a
function of frequency for a sinusoidal input. Make an accurate hand-sketch of the Bode plot
forthistransferfunction,i.e.,themagnitudeandphaseasafunctionoftheinputfrequency.
78
8.2 RLC Transfer Function
This problem considers the circuit shown in Figure 8.2.1.
Figure 8.2.1: RLC Circuit
(a) Calculate the transfer function V
o
(s)/V
i
(s).
(b) Let L = 0.01 H. Choose the remaining parameters of the circuit such that
n
= 10
5
and
= 0.05. Plotthepolesofthesystem,indicatingthepolepositionsinrectangularcoordinates
and polar coordinates.
(c) For these parameters, make a carefully-dimensioned sketch of the response to a unit step in
inputvoltage. Specically,showtheovershootvalue,thetimetothepeak,andthetime-scale
for each of the next 3 peaks in the response. What is the time required to settle to 5%?
(d) For these parameters, make an accurate sketch of the Bode plot for the transfer function in
part (a).
79
8.3 Bandpass Filter
The circuit shown in Figure 8.3.1 acts as a bandpass lter, in that input frequencies will only sig-
nicantlypropagatetotheoutputiftheyarenearthelterresonantfrequency. Inputcomponents
away from the resonant frequency will be attenuated; the degree of attenuation will depend upon
the separation between the driving and resonant frequencies.
I
s
C L
i
C
i
L
+
_
v
L
+
_
v
C
R
+
_
v
R
i
R
Figure 8.3.1: Bandpass Filter Circuit
An example of the use of such a lter would be at the front end of a radio. Here the lter is used
to select a particular radio station carrier frequency for listening, and to attempt to reject other
radio station carrier frequencies. In this situation, the current i would represent current from the
antennaattachedtotheradio,andtheoutputsignalv
R
wouldrepresentavoltageusedintheradio
for further processing.
In this problem, the numerical values have been selected to tune the lter to a Boston-area FM
radio station. We ask you to work out the frequency response function for the lter and then
consider how much an adjacent stations signal is attenuated. For the purposes of this problem,
assumethatradiostationsemitapuresinusoidalsignalattheircarrierfrequency. So,forexample,
astationsuchasWBCNoperatesat104.1MHz(104.110
6
Hz). Althoughtherealsignalconsists
of a band in the vicinity of this frequency, we will assume that the station puts out a pure sine
wave at 104.1 MHz, and that this is what the lter is processing. In this problem, be careful to
keep track of what frequencies are in Hz and what frequencies are in rad/sec.
(a) SolveforthedierentialequationwhichrelatestheinputcurrentI
s
totheoutputvoltagev
R
.
Show the steps in your derivation.
(b) Use the results of a) to derive the transfer function from input I(s) to output V
R
(s).
(c) In terms of the circuit parameters, what are the values of the natural frequency
n
and the
damping ratio ? What are the locations of the system poles and zeros?
(d) In the following parts of the problem assume the components take the values:R = 50 k,
C = 10
11
F=10pF,andL= 3.1481510
7
H.(Caution: Thevalueschosenmustbeused
withallgivendigitstoguaranteeasucientlyaccurateanswer.) Fortheseparametervalues,
what are the numerical values of
n
and ?
80
(e) Use the MATLAB bode command to plot the frequency response magnitude and phase for
this lter. Can you see why it is called a bandpass lter? What is the center frequency of
the lter in MHz? What Boston-area radio station operates at this frequency?
(f) In a given area, radio stations are generally spaced in frequency by at least 400 kHz. Use
thefrequencyresponsefunctiontodeterminehowmuchattenuationwillbeexperiencedbya
signal at 400 kHz above the lter resonant frequency. Show this frequency and the resulting
attenuation on a Bode plot zoomed in on the relevant frequency region.
81
8.4 LC Circuit Dierential Equations
Consider the circuit in Figure 8.4.1
L
C
+
_
v
i
v
o
+
_
Figure 8.4.1: LC Circuit
(a) Write the governing dierential equation for this circuit.
(b) The input is a unit step v
i
(t) = u
s
(t). Calculate v
o
(t), t > 0, assuming initial rest. Write an
expression for v
o
(t), and make a properly labeled sketch of v
o
(t).
82
8.5 Equivalent Impedance
Calculate the equivalent impedance, Z
eq
, at the indicated terminal pair in Figure 8.5.1.
L
R
1
C
2 Z
C
1
eq
R
2
Figure 8.5.1: LRC Circuit
83
8.6 Circuit Response
For each of the circuits below, derive the dierential equation which relates input to output, and
solve for the unit step response for the listed initial conditions. Note that some of the circuits use
a current source input. In each case, make a sketch of the output vs. time.
a) Solve for i
c
as output
-
+
+
-
Figure 8.6.1: Circuit for Part (a)
(b) Solve for v
L
as output
-
+
+
-
Figure 8.6.2: Circuit for Part (b)
(c) Solve for v
L
as output
i
in
R
L
i
L
+
v
L
i)i (0)=0A
L
-
Figure 8.6.3: Circuit for Part (c)
(d) Solve for v
c
as output
84
R
+
i
in
v
C
i)v
C
(0)=0V
C
-
Figure 8.6.4: Circuit for Part (d)
(e) For parts (a)(d), use impedance techniques to directly nd the appropriate transfer func-
tions.
85
8.7 LRC Circuit 1
This problem studies the circuit shown below.
-
+

Figure 8.7.1: LRC Circuit


(a) For t<0, what are the values of v
1
, v
2
, i
L
, and i
c
(b) What are the initial conditions at t= 0
+
, i.e., the values of v
1
(0
+
) and
dv
1
(0
+
)? Be sure to
dt
show how you derived these results.
(c) For t0 a dierential equation describing this circuit in terms of v
1
is
d
2
v
1
dv
1
+ 2 + 2v
1
= 0.
dt
2
dt
Write the solution of this equation as a function of time for T 0. Show carefully how you
derivedthesolutionandsatisedtheinitialconditionsfromtheprevioussection. Ifyouwere
not able to derive the initial conditions, you can write your solution in terms of v
1
(0
+
) and
dv
1
(0
+
).
dt
86
8.8 LRC Circuit 2
Thisproblemstudiesthecircuitshownbelow. Notethatthecircuitisdrivenbyanunstepfunc-
tion i
s
(t) = 1 u
s
(t), and thus the input is 1 [A] for t < 0 and 0 [A] for t > 0.
v
1
1
v
2
i =1-u
s
(t)
s
1 1F
1H
i
c i
L
Figure 8.8.1: LRC Circuit
(a) Writenodeequationsatv
1
andv
2
toderiveadierentialequationdescribingthecircuit. The
dierential equation should be in terms of v
1
and i
s
(that is v
2
should be eliminated).
(b) For t < 0, what are the values of i
s
, v
1
, v
2
, i
L
, and i
c
?
(c) Use Laplace transform techniques to develop a solution v
1
(t) for this system for t 0.
(d) Make a hand sketch of this solution.
L
(e) Calculate
V
Is
1
(
(
s
s
)
)
and
I
Is(
(
s
s
)
)
87
8.9 LR Circuit Step
V
o
V
i
L R
1
R
2
+
-
Figure 8.9.1: Circuit Diagram
(a) Derive a dierential equation which describes the circuit behavior in terms of the variables
V
o
(t) and V
i
(t). Plot the system pole on the complex plane.
(b) For this section of the problem, let R
2
= 1k. The output V
o
(t) in response to a unit step
input, from zero initial conditions is shown below. Determine the values of L and R
1
that
correspond to this response. Be sure to show your reasoning.
Step Response
A
m
p
l
i
t
u
d
e

0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0 1 2 3 4 5 6
Time (sec.) x 10
-4
Figure 8.9.2: Step Input Response from zero initial conditions
(c) Now let the input the to circuit be a sinusoid V
i
(t) = sin t. In steady-state, the output will
have the form V
o
(t) = Msin(t+). Using the numerical values of the circuit elements you
derived above, calculate expressions for M and as a function of .
88
(d) Make plots by hand of M() and (). For the plots, use logarithmic axes for , log axes
for M, and linear axes for . That is, the magnitude is displayed on a log-log plot, and the
phase is displayed on a semi-log plot. What is the DC gain of this circuit? What is the
high-frequency asymptotic gain?
89
9 Op Amps
9.1 Opamp Arithmetic
Design an opamp circuit that realizes the following input-output relationship
v
o
= 5v
1
+ 3v
2
10v
3
where v
1
, v
2
, and v
3
are three indepdent output voltages. Use the minimum number of opamps
that you nd possible, and assume that the opamps follow the ideal innite gain model. Explain
your thinking.
90
9.2 Inverting Op-Amp
ThecircuitshowninFigure9.2.1isaninvertingopampconnection,withacapacitorC

connected
between the inverting input and ground. As we saw in lecture, adding a capacitor in this location
canhaveasignicanteectonthestabilityoftheopampconnection. Thisproblemexploressuch
an eect.
Figure 9.2.1: Inverting Op-Amp Circuit
(a) For this connection, draw a block diagram representing the circuit.
(b) Let the components take the values R
i
= R
f
= 10 k, C = 10 F, and a(s) = 210
6
/s.
Calculate the transfer function V
o
/V
i
. What are the closed-loop damping ratio and natural
frequency?
91
9.3 Opamp Transfer Function
Design an opamp circuit that realizes the following input-output transfer function
V
o
(s) 10
3
s+ 1
=100 .
V
i
(s) s(10
4
s+ 1)
Clearly show your thinking. You can design this circuit using only resistors, capacitors, and op
amps. Assume that the opamps follow the ideal innite gain model. Explain your thinking.
92
9.4 Op-Amp Dierentiator 1
The circuit in Figure 9.4.1 acts approximately as a dierentiator.
R
R
1
C
+
_
V
i
+
_
V
1
V
o
R
2
R
2
+
_
Figure 9.4.1: Op-Amp Dierentiator
The second amplier just acts as an inverter to cancel out the inverter associated with the rst
stage. ForthevaluesR
1
=R
2
=100k,R=10k,andC = 20 F,andassumingtheop-ampis
modeled as ideal with innite gain,
(a) Calculate the pole and zero locations.
(b) CalculatethemagnitudeandphaseofthetransferfunctionV
o
(j)/V
i
(j)asafunctionof.
(c) Use MATLAB to plot the step response of this circuit. Does it act as a dierentiator? How
is the response non-ideal?
(d) Use MATLAB to generate the Bode plot for this transfer function. Indicate the breakpoint
frequency on the Bode plot. How is this related to the pole location?
93
9.5 Op-Amp Proportional plus Integral Controller
The circuit in Figure 9.5.1 can be used as a proportional plus integral controller.
R
f
C
R
i
+
_
V
i
+
_
V
o
Figure 9.5.1: Op-Amp PI Circuit
For the values R
i
= 10 k, R
f
=100k,andC = 10 F,andassumingtheop-ampismodeledas
ideal with innite gain,
(a) Calculate the pole and zero locations.
(b) CalculatethemagnitudeandphaseofthetransferfunctionV
o
(j)/V
i
(j)asafunctionof.
(c) Use MATLAB to plot the step response of this circuit. Does it act as a proportional plus
integral (PI) lter? How so?
(d) Use MATLAB to generate the Bode plot for this transfer function. Indicate the breakpoint
frequency on the Bode plot. How is this related to the zero location?
94
9.6 Op-Amp Circuit Design 1
Design an op-amp circuit that realizes the following input-output relationship
v
o
= 5v
1
+ 3v
2
10v
3
wherev
1
,v
2
, and v
3
arethreeindependentoutputvoltages. Usetheminimumnumberofop-amps
that you nd possible, and assume that the op-amps follow the ideal innite gain model. Explain
your thinking.
95
9.7 Op-Amp Block Diagram
In the op-amp circuit shown in Figure 9.7.1, the op-amp has a transfer function a(s) from the
dierential input to the output voltage.
R
3
R
2
C
+
-
a(s) +
R
1
V
-
+
o
V
in
-
Figure 9.7.1: Op-Amp Circuit
(a) Draw a block diagram representing the circuit. Your block diagram should show the vari-
ables V
in
, V
o
, v
+
, and v

. Be sure to show how the block diagram was derived.


(b) Let a(s) = g/s. What is the transfer function from input V
in
to output V
o
, in terms of the
given system parameters?
96
9.8 Op-Amp Circuit Design 2
Design an opamp circuit that realizes the following input-output transfer function
V
o
(s) 10
3
s+ 1
=100 . (1)
V
i
(s) s(10
4
s+ 1)
Clearly show your thinking. You can design this circuit using only resistors, capacitors, and op
amps. Assume that the opamps follow the ideal innite gain model. Explain your thinking.
97
9.9 Op-Amp Dierentiator 2
The circuit in Figure 9.9.1 is the circuit from Problem 9.4.
R
R
1
C
+
_
V
i
+
_
V
1
V
o
R
2
R
2
+
_
Figure 9.9.1: Op-Amp Dierentiator
The second amplier just acts as an inverter to cancel out the inverter associated with the rst
stage. ForthevaluesR
1
=R
2
=100k,R=10k,andC = 20 F,andassumingtheop-ampis
modeled as ideal with innite gain,
(a) DrawablockdiagramforthecircuitofFigure9.9.1inwhichtheopampismodeledashaving
dynamics. That is, the op-amp output voltage is V
out
=A(s)(v
+
v

).
(b) Solve for the input-output transfer function for this circuit conguration in the case where
A(s) = g/sand g= 210
7
, and using the numerical values of the parameters given above.
(c) Plot the poles and zeros of the transfer function from part (b). What dynamics can you
estimate will dominate the step response? Make a sketch of the step response that would
result from these dominant dynamics.
(d) Enter the exact transfer function into MATLAB, and thereby plot the exact step response.
How does this dier from the response calculated in Problem 9.4? How are the dominant
dynamics of part (c) evident in this step response?
98
9.10 RC Op-Amp Frequency Response
(a) ForthecircuitshownbelowcalculatethetransferfunctionV
o
/V
in
undertheassumptionthat
the op amp has innite dierential gain as shown.
+
_

C
2
C
1
R
2
R
3
V
o
R
1
V
in
Figure 9.10.1: RC Op-Amp Circuit
(b) NowlettheresistorR
1
=10[k]. Choosetheothercomponentssuchthattheinput/output
transfer function Bode plot has the asymptotic magnitude shown in Figure 9.10.2. Explain
your calculations. With the indicated slopes, it is possible to determine the value of A as
shown on the plot. What must this value be?
V
V
in
o
(j)
A
-1 +1
5
200 2000 5000
Figure 9.10.2: Asymptotic Magnitude of Bode plot
99
10 Dierential Equations
10.1 Dierential Equations 1
Here we consider a system described by the dierential equation
500 y+1000y= 0.
(a) Plot the system pole on the complex plane.
(b) Supposethaty(3)=10. Writeanexpressionfory(t)validforalltime. Showyourreasoning.
Sketch this response.
(c) At what time does the response fall below a value of 10
6
?
(d) Use Matlab to plot the response in the range 4t1.
(e) Design a mechanical system which will have the dierential equation given above. Clearly
indicate the physical variable that y represents, as well as the numerical values of the me-
chanical system parameters. Note that there are many possible systems which will have the
same dierential equation.
100
10.2 Dierential Equations 2
Here we consider a system described by the dierential equation
y +10 y+10000y= 0.
(a) Whatarethevaluesofthenaturalfrequency
n
,thedampingratio,andthedampednatural
frequency
d
? Plot the system poles on the complex plane.
(b) Supposethaty(0.1) = -3, and that y(0.1)=0. Writeanexpressionfory(t)validforalltime.
Show your reasoning. Sketch this response.
(c) Use Matlab to plot this response in the range 0.5 t1. Indicate on your plot that the
auxilary conditions are met.
(c) Atwhattimeisthevalueoftheresponseguaranteedtobebelow10
3
? Donotcalculatethe
exacttimevaluewhenthisoccurs,butprovideaboundusingtheexponentialenvelopeofthe
waveform.
(d) Designamechanicalsystemwhichwillhavethedierentialequationgivenabove. Clearlyin-
dicatethephysicalvariablethatyrepresents, as well as the numerical values of the mechanical
system parameters.
101
10.3 Dierential Equations 3
Here we consider a system described by the dierential equation
y +200 y+10000y= 0.
(a) Whatarethevaluesofthenaturalfrequency
n
,thedampingratio,andthedampednatural
frequency
d
? Plot the system poles on the complex plane.
(b) Suppose that y(0) = 1, and that y(0) = 1. Write an expression for y(t) valid for all time.
Show your reasoning. Sketch this response.
(c) Use Matlab to plot this response over an interesting range of time values. Indicate on your
plot that the auxilary conditions are met.
102
10.4 Eulers Theorem 1
(a) Use Eulers theorem to demonstrate the following equivalence
ce
jt
+c

e
jt
=Acost+Bsint
wherec=+jisacomplexnumberandAandBarereal. Whatistheresultingrelationship
betweenc and A and B?
(b) Given A= 7, B = 2, and =100, what is the corresponding value of c?
(c) UseMatlabtoplotthreecyclesofthevaveformdenedinpartb),andhandintheresulting
plot.
103
10.5 Eulers Theorem 2
(a) Use Eulers theorem to demonstrate the following equivalence
Acost+Bsint=Mcos(t+)
where A, B, M, and are real. What is the resulting relationship between A,B and M, ?
(b) What is the relationship between M, , and c as dened in Problem 10.4? How are these
related to the amplitude of the waveform you plotted in Problem 10.4?
104
_
10.6 Force Steps
Assume that we apply a unit step in force separately to a mass m, a dashpot c, and a spring k.
Themassmovesininertialspace. Thespringanddashpothaveoneendconnectedtoinertialspace
(reference velocity = 0), and the force is applied to the other end. Solve for and plot the resulting
velocity and position of the mass, and the velocity and position of the free ends of the spring and
damper under this step in force.
Recall that the unit step function u
S
(t) is dened as u
S
(t) = 0; t < 0 and u
S
(t) = 1; t 0. We
will also nd it useful to introduce the unit impulse function (t) which can be dened via
t
u
S
(t) = () d

This means that we can also view the unit impulse as the derivative of the unit step:
du
S
(t)
(t) =
dt
105
10.7 Input for Zero Output
A system with an input u and an output y is described by
d
2
u du
y= + 2 15u (1)
dt
2
dt
What is the most general input u(t) applied for all t which yields y(t)=0? Show your reasoning,
and verify by direct substitution into (1).
106
10.8 Zero Input Response
For each of the systems shown in Figure 10.8.1, suppose that the indicated input is set equal to
zero. That is, we will solve for the zero input response(ZIR). Dont use Laplace techniques to nd
the requested solution.
C C
J
, output , input
(i): Rotational system
C C
M
X, output
F, input
K K
(ii): Translational system
Figure 10.8.1: System Diagrams
(a) For the rotational system, solve for (t) with the initial conditions (0)=
0
and

(0)=
0
.
Write expressions for the system time constants.
(b) Forthetranslationalsystem,solveforx(t)withtheinitialconditionsx(0)=x
0
and x(0)=v
0
,
under the assumption that the system is under-damped. Write expressions for the system
natural frequency, damped natural frequency and damping ratio.
107
10.9 Zero Step Response 1
C C
J
, output , input
(i): Rotational system
C C
M
X, output
F, input
K K
(ii): Translational system
Figure 10.9.1: System Diagrams
For each of the systems shown in Figure 10.9.1, suppose that the initial state is equal to zero, and
that the input is active. That is, solve for the zero state response(ZSR). Specically assume that
the input is a step, i.e.,
(i) (t) = u
s
(t); (0)=

(0)=0.
(ii) F(t) = u
s
(t); x(0)=x (0)=0.
Now solve for the resulting output variable as a function of time. Again, in the case of the trans-
lational system, you may assume that it is under-damped.
108
10.10 Zero Step Response 2
Now assume the following numerical parameters for the systems in Problem 10.9
(i) J = 0.1 kg-m
2
; C
1
= 10
3
N-m-s/rad; C
2
= 3 10
3
N-m-s/rad.
(ii) M =10kg; K
1
= 10
5
N/m; K
2
= 8 10
5
N/m; C
1
=C
2
=300N-s/m.
(a) Fortheseparameters,writeexpressionsfor(t) and x(t)forthezerostateresponseofProblem
10.9, and make hand-sketches of these responses.
(b) Use Matlab to plot the responses of Part a.
109
10.11 Nonlinear String System
Figure10.11.1showsamassMsupportedbytheconstantforcestring,whichhasconstanttension
force no matter how much it is stretched. The mass is attached to the center point of the string.
L= 0.5m. T=100N.Foreachofthreecases(i)M
1
=10kg,(ii)M
2
= 14.5kgand(iii)M
3
= 17.7
kg.
(a) Find the equilibrium point X.
(b) Derive a linearized model about the equilibrium point.
(c) Solve for roots and plot them on the complex place.
(d) Plot the zero input response from zero initial velocity and the initial position is 1cm above
the equilibrium point, i.e. X(0)=X0.01m.
2L

X
gravity

Figure 10.11.1: Nonlinear String System
110
10.12 Homogenous and Particular Solutions
Forthisproblem,weusetherotationalsystemofLab2,seeninFigures10.12.1and10.12.2. Refer
toPrelab2formoredetails. Intheprelab,youareaskedtocalculatethemomentofinertiaforthe
shaftwithandwithouttheywheel,andtousethevideoshownonthecoursewebpagetoextract
an estimate of the damping coecient c resulting from the honey in the cup at the bottom of the
shaft. Notethatinthevideotheywheelisattachedtotheshaft,andsoforthisproblemyouwill
use the inertia value with the ywheel present. Use the following values in this problem:
K = 10
2
Nm , I = 0.000317kgm
2
and c= 0.0011N-m-sec/rad
Flywheel
Airbearings
Cup with
viscous uid
Shaft
Figure 10.12.1: Rotational System
(t)
c
I
Figure 10.12.2: Lumped System Model
(a) Assume that the Lab 2 system is driven by an input torqueT
in
acting on the rotating shaft,
and has a resulting angular velocity . Write a dierential equation which describes the
system with the given torque input.
111
(b) NowassumethatthetorqueisastepofmagnitudeK,i.e.,T
in
=Ku
s
(t),andthattheshaft
hasaninitialangularvelocity(0)=
0
. Findtheparticularandhomogeneoussolutionsand
combine these to write an expression for the total solution. Given the input magnitude K,
for what initial velocity will the solution reach nal value in zero time, i.e., there will be no
transient?
(c) AccuratelysketchthesolutionsforK = 10
2
Nm,andforeachofthefollowinginitialangular
velocities:
0
=0 rad/sec;
0
=40 rad/sec.
(d) NowwechangethetorquesourcetoasinusoidalinputT
in
=u
s
(t)Kcost. Againthesystem
has an initial angular velocity (0)=
0
. Note that the step function u
s
(t) which multiplies
the cosine function turns on the input at t = 0.The homogeneous solution will have the
same form as in part b), so you will not need to repeat this calculation. We ask you to nd
the particular solution in two ways. The rst is by using only trigonometric functions, and
nocomplexvariablerepresentations. Thesecondapproachisbywritingtheinputasthereal
part of a complex exponential.
(i) Solve for the particular solution
p
(t) to the cosinusoidal drive without making use of
complex notation. Youll need to work through some trigonometric identities to work
out the solution.
(ii) Solve for the particular solution
p
(t) to the cosinusoidal drive by using complex nota-
tion. That is, we recognize that the input can be expressed as T
in
= Reu
s
(t)Ke
jt
.
Then because the system is linear, we can rst solve for the particular solution for a
complex-valued input T
in
=u
s
(t)Ke
jt
, taking the real part of the solution at the end
tondtheparticularsolutionfortheoriginalreal-valuedinput. Thesolutionshouldbe
the same as you found in part i). As youll see, the math is simpler via this path than
forthesolutiontakenini)byusingtrigonometricidentities. Asthesystemordergrows
larger than this rst-order example, this simplication becomes more signicant.
(e) Assume that the initial shaft velocity is zero, i.e.,
0
=0. For the particular solution to the
sinusoidal drive that you found above, what homogeneous solution is required to satisfy this
initial condition? Use Matlab to plot the total solution for K = 10
2
, and for the following
two values of input frequency: =0.5 rad/sec; =20 rad/sec. How do these two solutions
dier in appearance? Specically, how signicant is the homogeneous solution in these two?
Also,whateectdoesthesystemhaveontheoutputsinusoidalmagnitudeandphaserelative
to the input?
112
11 Frequency Response
11.1 Circuit Bode Plots
For each of the circuits shown in Figure 11.1.1
(a) Determine the associated transfer function.
(b) Accurately sketch the associated Bode plots (magnitude and phase).
Figure 11.1.1: Circuits for analysis
113
11.2 Sketch Bode Plots
Using the appropriate graph paper, sketch the Bode plots for the following transfer functions:
5s+1
(a) G(s) =
s+10
5s+1
(b) G(s) =
s
2
+3s+1
s+10
(c) G(s) =
(s+2)(s
2
+10s+100)
114
11.3 LC Circuit Frequency Response
I
s
C L
i
C
i
L
+
_
v
L
+
_
v
C
Figure 11.3.1: LC Circuit
(a) DeriveadierentialequationwhichdescribesthecircuitbehaviorintermsoftheinputI
s
(t)
andtakingtheoutputasV
c
(t). Whatarethevaluesof
n
,
d
,, and forthiscircuit? Plot
the poles/zero on the complex plane.
(b) Calculate and make a carefully labeled graph of the response to a unit step current input
I
s
(t) = u
s
(t), from non-zero initial conditions v
c
(0) = 0 [V] and i
L
(0)=1 [A].
(c) Now let the input the to circuit be a sinusoid I
s
(t) = sin t. In steady-state, the output will
have the form V
o
(t) = Msin(t+). Calculate expressions for M and as a function of .
(d) MakeplotsbyhandofM() and (). Fortheplots,uselogarithmicaxesfor,logaxesfor
M,andlinearaxesfor. Thatis,themagnitudeisdisplayedonalog-logplot,andthephase
isdisplayedonasemi-logplot. Youshouldndthatmagnituderesponsehasaninnitepeak

at = 1/ LC. Thus this circuit can be used as a band-select lter, predominantly passing
a single frequency. Such a lter is useful for instance in a radio tuner to select a particular
station.
115
_ _
11.4 LRC Circuit Frequency Response
i
C
i
L
i
R
+
+ +
I
s
C
v
C
L
v
L
R
v
R
_
Figure 11.4.1: LRC Circuit
(a) DeriveadierentialequationwhichdescribesthecircuitbehaviorintermsoftheinputI
s
(t)
andtakingtheoutputasV
c
(t). Whatarethevaluesof
n
,
d
,, and forthiscircuit? Plot
the poles/zero on the complex plane.
(b) NowletthecircuitparametershavethevaluesL=5[mH],C = 0.02[F],andR= 2.5[k].
Calculate and make a carefully labeled graph of the response to a unit step current input
I
s
(t) = u
s
(t), from zero initial conditions v
c
(0) = 0 [V] and i
L
(0) = 0 [A].
(c) Now let the input the to circuit be a sinusoid I
s
(t) = sin t. In steady-state, the output will
have the form V
o
(t) = Msin(t+). Calculate expressions for M and as a function of .
(d) MakeplotsbyhandofM() and (). Fortheplots,uselogarithmicaxesfor,logaxesfor
M,andlinearaxesfor. Thatis,themagnitudeisdisplayedonalog-logplot,andthephase
is displayed on a semi-log plot. You should nd that magnitude response has a nite peak.
What is the peak frequency? How is the peak magnitude related to the circuit parameters?
116
11.5 Pole-Zero Plots
For the pole-zero plots shown in Figure 11.5.1:
(a) Determine the associated transfer function, under the assumption that the magnitude of the
Bode plot passes through 1 at = 1.
(b) Accurately sketch the associated Bode plots (magnitude and phase), under the assumption
from (a).
Figure 11.5.1: Pole Zero Plots
117
12 Signals, Transforms, and Transfer Functions
12.1 First Order Zeros
This problem considers the eect of zeros on the step response of a rst-order system. Speci-
cally,consider a system with input u and output y described by the dierential equation
y + y=au + u
Note that this system has a pole at s=1/ and a zero at s=1/a.
(a) Write an analytical solution for the response of this system to a unit step, assuming zero
initial conditions. Use impulse matching to work through the transient att=0, and use the
approach of particular and homogeneous solutions to solve the problem. Do not use Laplace
transform techniques.
(b) We now want to study the step response for a number of values of the parametera. Let =
0.1sec,andsketchthestepresponseforeachthefollowingvalues: a= 0.05,0.09,0.11,1,1,.1.
You should be able to sketch the response by hand, without any need for computational as-
sistance. Along with the step response, include a pole-zero plot showing the locations of the
pole and zero on thes-plane.
(c) Collocated poles and zeros in some sense cancel each other. Argue that if we let a = =
0.1 sec, then the system reverts a simple gain of 1 from input to output, that is, from a
condition of initial rest, the output at all times equals the input. What is the step response
for this choice of parameters? Show that your response in part b) above approaches this
response in the limit a = . Can you see this asymptotic behavior in the step responses
youve plotted in part b)?
(d) Anotherwaytothinkaboutzerosisthattheyallowproperlychosenexponentialinputswith
nocorrespondingoutput. Specically,assumethattheinputtakestheformu(t) = e
s
1
t
. Show
that by the proper choice of s
1
, the driven output is identically zero. How is this value of s
1
related to the zero location?
(e) Solve for the step response from a zero initial condition via Laplace techniques. This should
yield the same result as you found in part (a).
118
12.2 Laplace Practice
A signal w(t) is dened as
w(t) = u
s
(t)u
s
(tT)
whereT isaxedtimeinsecondsandu
s
(t)istheunitstep. ComputetheLaplacetransformW(s)
of w(t). Show your work.
119
12.3 Step-Zero
This problem studies the eect of a real-axis zero on the step response of a second-order system.
Specically, consider the transfer function
s/+ 1
H(s) = .
s
2
+ 2s+ 1
Note that we have normalized by choosing
n
=1. This is the same as allowing a non-unity value
of
n
and making all plots with respect to
n
t. For this transfer function:
(a) Plot on the s-plane the poles and zero as a function of for the values = 0.1, 1, 10,and
for = 0.01, 0.1,0.7. Make a separate plot for each value of , showing the eect of on
the zero location. Clearly label the pole and zero locations on the real and imaginary axes.
Note that you cannot use a linearly scaled pole/zero plot if you want to show the pole and
zero locations with adequate resolution.
(b) UseLaplacetechniquestoderiveananalyticalexpressionfortheresponseofthissystemy(t)
to a unit step u(t) = u
S
(t) from initial condition y(0)=y(0)=0. Carefully show all steps.
Be sure to include an expression which gives the output transform Y(s) in terms of the
input transform U(s). Expand this expression in terms of given entries in a table of Laplace
Transforms, performing inversion from the Laplace Transform to a function of time.
(c) UseMatlabtoplottheresponsefunctionyouhavederivedversustimeandforthevaluesof
and listedinparta). Includeseparateplotsforeachvalueof,whichoverlaythevariation
with. Howdoesthezerolocationaectthestepresponse? Canyouseethattheterms/
contributes a derivative eect in the step response? Explain.
120
12.4 Laplace to Time
For each Laplace Transform Y(s), nd function y(t):
(a)
1
Y(s) =
(s+a)(s+b)
(b)
s
Y(s) =
2
s
+
2
s+ 1

2
n
n
121
12.5 Time to Laplace
For each of the functions y(t), nd the Laplace Transform Y(s):
(a) y(t) = e
at
(b) y(t) = e
t
sin
d
t
(c) y(t) = e
t
cos
d
t
122
12.6 Laplace Transform
A signal has a Laplace transform
a
X(s) = b +
s(s +a)
where a, b > 0, and with a region of convergence of [s[> 0. Find x(t), t > 0.
123
12.7 Signal in Time and Frequency Domain
A signal x(t) is given by
x(t) = (e
t
e
1
)
_
u
s
(t)u
s
(t1)
_
(a) Sketch this signal as a function of time.
(b) Calculate X(s). Clearly show the steps in your calculation.
124
12.8 Transfer Function and Time Constants
(a) Calculate the transfer function v
o
(s)/v
i
(s) for the circuit shown in Figure 12.8.1
C
2
C
1
+
_
v
i
v
o
+
_
R
1
R
2
Figure 12.8.1: Electrical Circuit
Write this transfer function in the form

1
s+ 1
K

2
s+ 1
Writeexpressions forK,
1
, and
2
. UseLaplace initialandnal value theorems to calculate
v
o
(0
+
) and v(). Plotthepole andzeroofthissystemonthe s-plane. Whathappenswhen
R
1
C
1
=R
2
C
2
?
(b) Letv
i
(t) = u
s
(t),aunitstep. Assuminginitialrestconditions,calculateanexpressionforthe
step response v
o
(t). Make sketches of v
o
(t) for the following three sets of parameter values:
(i) R
1
= 1k, R
2
= 9k, C
1
= 1F, C
2
= 9F
(ii) R
1
= 1k, R
2
= 9k, C
1
= 9F, C
2
= 1F
(iii) R
1
= 9k, R
2
= 1k, C
1
= 9F, C
2
= 1F
Also plot the pole/zero diagram for each set of parameters.
125

_
12.9 Fourier Series Input
Acommonperiodicsignalisatrianglewavesimilartothatshownbelow. Itcanbethoughtofasa
seriesoframpsrepeatingatafrequencyof
o
. Thepurposeofthisproblemistoseehowdierent
dynamicsystems,eachwithadierenttransferfunctionG(s),willreacttothatsignal. Todothis,
wecanmakeuseoftheFourierseries,thatstatesthatanyperiodicfunctionf(t)canbedescribed
by an innite series:
1

f(t) = a
o
+ [a
n
cos(n
o
t) + b
n
sin(n
o
t))]
2
n=1
where:
2
t
1
+T
a
n
= f(t)cos(n
o
t)dt
T
t
1
2
_
t
1
+T
b
n
= f(t)sin(n
o
t)dt
T
t
1
where
o
is the fundamental frequency of the waveform.
(a) Usingreferencesyoumighthavefrom18.01/2/3,determinetheappropriateFourierSeriesfor
the triangle wave shown below
(b) For each of the transfer functions below, sketch (not using MATLAB) the Bode diagram.
G
1
(s) =
K
+1
with K =10, =20.
s
K
G
2
(s) =
(
1
+1)(
2
+1)
with K = 1,
1
= 5, and
2
=10.
s
2
n
G
3
(s) =
s
2
+2
n
s+
n
2
with = 0.5 and
n
=50.
K
s
2
+2
z

zn
s+
2
zn
G
4
(s) =
s+1 s
2
+2
n
s+
2
with K = 1, = 0.01, = 1.0,
n
= 100,
z
= 0.05, and
n

z
n=10.
(c) Using the diagram only, write the steady-state output y(t) that would result if the triangle
wave u(t) of Figure 12.9.1 were applied to the input of each transfer function. For this
equation, consider only the rst 5 terms of the series in your answer.
(d) Using MATLAB, plot 2 full cycles of the following:
(i) u(t)
(ii) u(t) with only 3 term in the series included
(iii) y(t) for G1 AND G2 based on the series solutions from part C)
(e) Nowconsiderthatwewouldliketohavetheoutputofthetransferfunction,y(t),beapure
sinewaveoffrequency3
o
. (Inotherwordswewanttotuneinthefrequency3
o
and
tune out all the others.) Propose a transfer function that will come close to this and then
check how it does using the above procedure.
126
0.6
0.4
0.2
0
-0.2
-0.4
-0.6
T
Figure 12.9.1: Triangle wave u(t) with T = 0.628 sec.
127

12.10 Spring Mass System
Consider the simple spring mass system shown in Figure 12.10.1 subjected to a force F(t) that is
a pulse as shown in Figure 12.10.2.
k
m
b
F(t)
Figure 12.10.1: Spring Mass System
2 10 time
F(t)
1.0
Figure 12.10.2: Force Pulse
If m= 1; k = 1 and b = 1:
(a) Show how you would model the F(t) using Laplace transforms. (i.e. nd F(s).)
(b) Using Laplace Transforms, nd the solution of x(t) in response to F(t) for 0t20 sec.
(c) Using MATLAB, plot x(t).
(d) Describe why the response look as it does.
(e) How would the response change if the pulse were 2 seconds wide instead of 8? Just describe
qualitatively and make a sketch of what youd expect. Do not recompute the response.
128
12.11 Transfer Function
Given the transfer function
X(s) s
3
+110s
2
+1100s+1000
G(s) = = (1)
U(s) s
4
+ 2s
3
+ 52s
2
+ 51s+ 50
(a) Find the poles and zeros (you can use the MATLAB roots() command for this).
(b) Plot these poles on a s-plane.
(c) Based on the location of these poles, describe the expected response to a step input. Please
explain your reasoning.
(d) UsingtheMATLABstepcommand,createaplotofxvs. ttoseeifyourpredictionwasOK,
and compare.
129
13 Controls
13.1 Rooftop Antenna
(The following problem is adapted from Roberge, J.K., Operational Ampliers, Wiley and Sons,
1975, problem P2.9.) A negative-feedback system used to rotate a rooftop antenna is shown in
Figure 13.1.1(a). The total inertia of the output member (antenna, motor armature, and poten-
tiometerwiper)is2kg-m
2
. Themotorcanbemodeledasaresistorinserieswithaspeed-dependent
voltage generator as shown in Figure 13.1.1(b).
Figure 13.1.1: Antenna Diagram
The torque provided by the motor that accelerates the total output-member inertia is 10 Nm per
ampere of motor current I
a
. The polarity of the motor dependent generator is such that it tends
to reduce the value of I
a
as the motor speed increases. The scale factor is such that I
a
becomes
zero for a motor shaft velocity equal to V
m
/10 rad/sec.
(Note: There are a number of repeated analyses in this problem. You will want to use Matlab to
automate these. The problem shows some of the key tradeos associated with feedback control,
so please do think carefully about what you observe as the controller gain is changed through the
values 10, 100, and 1000.)
(a) Drawablockdiagramthatrelates
o
to
i
. Youmayincludeasmanyintermediatevariables
130
as you wish, but be sure to include V
m
and I
a
in your block diagram. Calculate the transfer
function
o
/
i
.
AlsoincludeinyourblockdiagramatorquedisturbanceT
d
appliedtotheantennashaft. For
instance,thistorquedisturbancemightrepresenttheeectofwindontheantenna. Calculate
the transfer function
o
/T
d
which represents the response to such a disturbance.
(b) Plot the system poles in the s-plane. Calculate the response in
o
to a 1 rad step input
i
.
Plot this response showing relevant time scales.
(c) Generate the Bode plot for
o
/
i
.
(d) Calculate the unit step response and Bode plots for the disturbance transfer function
o
/T
d
.
(e) Calculatethetransferfunctionofthenegativeofthelooptransmissionforthisfeedbackloop.
Plot a Bode plot for this transfer function. What are the crossover frequency and phase
margin for this loop? (These terms have been discussed in lecture.)
(f) Nowchangethegainofthedierentialamplierintheloopfrom10to100. (Thiswillincrease
the loop transmission magnitude by a factor of 10.) Repeat parts (b), (c), (d), and (e) with
this new value of loop gain. Comment on the eect of this larger loop gain on the system
closed-loop dynamics exhibited in the speed and damping of the step response, Bode plots,
command following ability, and the ability to reject a disturbance torque. How do the poles
move as the gain is increased?
(g) Repeat the analysis of part (f) for an amplier gain of 1000.
131
13.2 Lead Controller
We now extend the laboratory system to consider position control, and the design of a lead con-
troller. Thesystemtransferfunctionsusedinthelaballrefertothemotoroutputshaftangle
m
.
Wewilldenethepotentiometershaftangleas
p
. Thepotentiometerisgeareddownfromthemo-
torby6:1. Thatis,n
d
= 6n
p
, and thus 6
p
=
m
. Further,thepotentiometeroutputsV
pot
=10V
for a potentiometer shaft rotation of
p
= rad.
(a) ShowthattherelationshipsaboveimplythatV
pot
=K
pot

p
, where K
pot
= 10/ V/rad. Note
that we have dened K
pot
asdependinglinearlyupon
p
.
(b) Modify the block diagram of Fig. 2 of the prelab to include the motor angle
m
, the po-
tentiometer angle
p
, and the potentiometer gain K
pot
. Use this modied block diagram to
calculate the transfer function V
pot
/V
in
. (Recall that V
in
is the amplier input voltage as
shown in the prelab.) Sketch the poles of this transfer function as well as its Bode plot.
(c) Now assume that the motor position controller is a lead controller of the form
V
in
(s) 10s+ 1
=K .
E(s) s+ 1
Here E(s) = R(s)V
pot
(s), where R(s) is a position reference voltage, and E(s) is thus
the position error. The controller is embedded in the position feedback loop as shown in
Figure 13.2.1.
Figure 13.2.1: Block Diagram
Choose the controller parameters K and to achieve a loop crossover frequency
c
=
100 rad/sec, and phase margin 45

. Explain how you developed a design satisfying


these design values. Include a Bode plot of the negative of the loop transmission which
showsthatyouhaveachievedthesedesignvalues. Alsoincludeaplotoftheclosed-looppole
locations.
132
13.3 Third Order
Consider the mechanical drive of Problem 14.1 with the load free to move. The model parameters
have the following values:
I
rotor
= 5e-5 kg-m
2
B
motor
= 1e-4 N-m-s/r
B
coupler
= 2e-5 N-m-s/r
K
coupler
= 1.24e-2 N-m/r
I
load
= 5e-5 kg-m
2
B
load
= 1e-4 N-m-s/r
This is a system with fairly special symmetry. Note that I
motor
= I
load
= I,and that B
motor
=
B
load
= B. For this special case, develop a model to predict the angular velocity of the load in
response to an applied motor torqueT
m
.
(a) Use the variables:
, relative angular displacement of motor with respect to load,

m
, motor angular velocity, and

l
, load angular velocity,
as the state variables for this model in a state-determined system. Construct the state
equations for this model.
(b) Consider a change of variables to the following three variables:
, =
m

l
, and
avg
=

m
+
l
2
Express the variables
m
and
l
in terms of the variables, and
avg
.
(c) Re-write the state-determined system using ,, and
avg
as state variables. Show that
the dynamic behavior of the system may be viewed as a combination of a rst-order system
with state variable
avg
and a second-order system with state variables , and . Note
that both sub-systems are excited by the motor torque T
m
but are otherwise independent of
one another.
(d) Evaluate the rst-order behavioral parameter: the decay time-constant. Estimate the
steady-state speed
ss
when T
m
= 6.0e-3 N-m.
(e) Evaluate the second-order behavioral parameters: the undamped natural frequency
o
, and
the damping ratio .
133
13.4 Error Constants
Figure 13.4.1:
134
13.5 System Type
KG(s)
R(s)
Y(s) E(s)
WhereG(s)isthelooptransferfunctionandKistheDCloopgain.
Usingthesamecontrollerformasabove,youaregivenatransferfunction
400
G(s)
s(s
2
10s 400)
Bytheabovedefinitions,thisisatype1system,andshouldhavezeroerrortoastep(or
displacement)input. However,thevelocityerror(thatresultingfromarampinput)
willnotbezero. Indeedourgoalistoachieveavelocityerrorof<2%.
Forthiswecanexploretwodifferentapproaches.
a)FirstplotabodediagramofG(s)
b)Fromthatdiagram,whatwillbethevelocityerroriftheclosed-loopsystemhasa
phasemarginof50? (Thatis,afteryouhavefoundaloopgaintoachievethephase
margin)
c)Sincetheerroristoolarge,aformer2.003studenttellsyousimplytoincreasethe
systemtypefrom1to2andallyourproblemswillbeover: justlookattheerrortable!
Good advice or not? - Please explain by looking at the resulting open loop bode
diagram.
Nowlookatthealternativeofusingalagcompensator.
d)Giventhatwewantavelocityerror<2%,designalagcompensatorthatwillachieve
thatresultwithoutchangingthephasemarginatthecrossoverfrequencyyoufoundin
partb). ShowthefinalBodeplottoconfirmtheresult
e)Nowplottheresultingstepresponseofthesystemdesignedinpartd) Howdoesit
comparetotheexpectedvalues?
f)Lookingattheclosed-loopsystemresponsetoarampinput,confirmthatyouhave
achieveasteady-statevelocityerrorof1%
Figure 13.5.1:
135
13.6 Unity Feedback
Consider the feedback system of Figure 13.6.1 in block diagram form where G(s) = 1/s.
K
c
G(s)
-
R(s) Y(s)
Figure 13.6.1: System with unity feedback
(a) Determine the closed-loop transfer function T(s) = Y(s)/R(s) for this system.
(b) Find the gain K
c
that will result in a closed-loop step response with a settling time of 1
millisecond.
(c) On a single Bode magnitude plot, show the following:
G(s)
T
1
(s) for K = 1
T
2
(s) for K = 10
T
3
(s) for K =100
Be sure to label each plot clearly.
(d) If we dene bandwidth as the range of frequencies over which the magnitude is >0.5, what
are the corresponding bandwidths of the three T(s) plots in part (c)? Recall that we said a
true model of an op amp is: G
a
(s) = K/swhere K 10
6
(e) Given this information, what is the expected step response of a unity-gain op-amp?
(f) What is the corresponding bandwidth of that op-amp?
136
13.7 P and PI Controllers
In the lab this week, you are designing P and PI controllers for the motor velocity. This problem
refers to the designs youve developed for the lab. Please use the results from your Prelab 9
calculations to work this problem.
(a) Fortheproportionalvelocityloop,writeanexpressionforthetransferfunctionofthenegative
ofthelooptransmission. PlottheBodeplotforthistransferfunction. Whatarethecrossover
frequency and phase margin for this loop? What is the closed-loop pole location? How does
this compare with the crossover frequency?
(b) For the proportional-integral velocity loop, write an expression for the transfer function of
the negative of the loop transmission. Plot the Bode plot for this transfer function. What
are the crossover frequency and phase margin for this loop? Also plot the closed-loop poles.
(c) For the proportional-integral velocity loop, the controller has the form
s + 1
K .
s
In the prelab, you have calculated values of K and to locate the poles with = 0.5 and

n
=30. In part b) above, you have looked at this design from a loop transmission perspec-
tive. Now we want to consider the design as it is aected by the loop phase margin.
In this section, modify the PI controller so that you achieve a crossover frequency of 200
rad/sec, with a phase margin of 15 degrees. What values of K and result? Be sure to
explain how you arrived at these parameter values. Plot the closed-loop poles that result.
Whatarethevaluesof and
n
associatedwiththesenewrootlocations? Howdotheseroot
locations compare with the values in part b)? What root parameter seems to be associated
with the crossover frequency, and what parameter seems to be associated with the phase
margin?
137
14 Motors and Transformers
14.1 Locked Load
The sketch (Fig. 14.1.1) depicts a common mechanical drive conguration discussed in class. To
identify parameters of the motor and coupler it is often useful to conduct a test of the locked-load
dynamic response. The load is immobilized; a step change in motor torque is applied; and the
resulting motion of the motor is observed.
flexible
motor load
coupler
bearings
Figure 14.1.1: Mechanical Drive
(a) Assuming that all model elements have linear constitutive equations, develop a model to
predicttheangularpositionofthemotorinresponsetoanappliedmotortorqueT
m
. Express
this model as a state-determined system.
(b) Calculate
(i) the damped natural period, and
(ii) the decrement ratio for the oscillatory response, when the model parameters have the
following values:
Irotor= 5e-5 kg-m
2
B
motor
= 1e-4 N-m-s/r
B
coupler
= 2e-5 N-m-s/r
K
coupler
= 1.24e-2 N-m/r
T
m
= 4e-3 N-m abruptly applied constant torque
(0)=0 initial angular velocity
(0)=0 initial displacement
(c) A proportional velocity feedback controller is added to the uncontrolled system described in
(b) above. The motor angular speed is sensed and used in the control algorithm
T
m
=G
1
(r
1

sensed
) (1)
138
where r
1
is a reference speed parameter, and G
1
is a gain parameter. In what units should
G
1
be expressed? Analyze the closed-loop system and construct a state-determined system
which describes its dynamic behavior. Calculate the controller gain, G
1
, required to yield a
critically-damped locked-load response.
(d) A proportional position feedback controller is added to the uncontrolled system described in
(b) above. The motor angular position is sensed and used in the control algorithm
T
m
=G
2
(r
2

sensed
) (2)
wherer
2
isareferencepositionparameter,andG
2
isagainparameter.InwhatunitsshouldG
2
beexpressed? Analyzetheclosed-loopsystemandconstructastate-determinedsystemwhich
describes its dynamic behavior. Derive algebraic expressions which show how the following
closed-loop behavioral parameters depend on the feedback gain, G
2
:
(i) The locked-load undamped natural frequency
O
.
(ii) The locked-load damping ratio .
(e) Theusualimplementationofthecontrolalgorithms(1)and(2)iswithelectronicsignalpro-
cessing and amplication. It is however theoretically possible to implement these algorithms
mechanically by interpreting the control algorithms as constitutive equations for elements
connected between the motor rotor and a reference rotor which is driven with the desired
motion. Can you identify the type of mechanical element which is represented by:
(i) The algorithm (1)?
(ii) The algorithm (2)?
139

14.2 Motoring with a Capacitor
This problem concerns the motor connection shown in Figure 14.2.1. In this system, we assume
that the motor torque is related linearly to current
m
=Ki
m
, and thus that the motor voltage is
given by v
m
= K
m
. We further assume that the motor coil has zero resistance R = 0 and zero
inductanceL=0. ThecapacitorC isconnectedtothemotorterminalsasshown. Themotorrotor
has a moment of inertia J.
Assumethatthemotorisdrivenwithatorquesource,i.e.,that
in
isspeciedasanexternalinput
independent of the shaft speed. The shaft of the torque source is rigidly connected to the motor
output shaft as shown in the gure. The torque source is massless.
Note that we have dened the reference direction for motor speed
m
to be the same as that for
the source speed
in
. Further note that with these denitions the motor torque is applied to the
rotor in the same direction as the reference for
m
.
,
in
i
m
in
Motor,
with rotor
Inertia J
v
m
+
-
C
m
Torque
Source
Figure 14.2.1: Motor Diagram
(a) Draw a free body diagram for the motor rotor showing the torques acting on it. Write
Newtons second law for this free body. Note that continuity requires that
in
=
m
.
(b) How are the motor voltage and current related in light of the capacitor connection?
(c) Use the information gathered above to write a dierential equation for this system with
in
as the input and the resulting speed
in
as the output. Be sure to show the steps you used
to develop this dierential equation. The electrical variables should be eliminated in your
dierential equation.
(d) What is the equivalent mechanical model seen by the input torque source? Why?
(e) Let the torque take a step from 0 to A Nm at t=0 with zero initial rotor velocity. What is
the resulting velocity
in
as a function of time?
140
14.3 Leadscrew with Translating Stage
The system shown in Figure 14.3.1 has a translational stage of mass M which moves on friction-
less bearings. The stage is coupled to a lead-screw via the nut shown; the lead screw has a pitch
of 10
2
m/revolution, and is xed in the x-direction by the bearings shown, but is free to rotate
about its axis. The lead screw has a moment of inertia J about this axis. The mass of the nut is
included in M. The stage is connected to translational spring K
2
and damper c
2
. The lead screw
is connected to rotational spring K
1
and rotational damper c
1
.
Write the transfer function X(s)/F(s) where F is an input force driving the stage, and x is the
stage position.
Figure 14.3.1: Leadscrew
141
14.4 Mass Pulley System
This problem concerns the mechanical system shown below.
Figure 14.4.1: Pulley System
As shown in the gure, the system has a force input F which acts on the pulley carrier. The
position of the carrier is x
2
. The pulley has a radius R, moment of inertia J, and mass M
2
. The
carrier is included in M
2
. The angular rotation of the pulley is , as shown.
AmassM
1
hangsonastringwhichiswrappedaroundthepulleyasshown,andanchoredtoground.
ThepositionofM
1
isx
1
. Notealsothatgravityactsonthesysteminthedownwarddirection. We
assume that the motions of the system are such that the string remains under positive tension.
(a) What is the relationship between x
2
and ? Show your reasoning.
(b) What is the relationship between x
2
and x
1
? Show your reasoning.
(c) NowweaskyoutodevelopadynamicalrelationshipbetweentheappliedforceF,thegravity
load(s),andtheresultingmotionx
2
. Thatis,writeadierentialequationthatdescribesthe
motion of the system with F and g as inputs, and with only x
2
as the dependent variable.
The variables x
1
and should not appear in your equation. Clearly show your reasoning in
developingthismodel. Besuretoincluderelevantfree-bodydiagrams,andotherguresthat
explain the steps in your development.
142
14.5 DC Motors
Consider the DC motor shown below. It is connected to a rudimentary circuit with only a single
resistor, which follow Ohms Law : e=Ri where R is the circuit resistance.
Figure 14.5.1: DC Motor with resistor
(a) Given the motor constant K
m
and the circuit as shown nd the relationship between T and
as seen at the motor shaft. Be careful with signs and look at your answer critically to be
sure you got it right.
Figure 14.5.2: DC Motors connected together
(b) Now consider the system shown in Figure 14.5.2, where two motors are connected together
asshown. Assumingnegligiblemechanicalfrictionandinertiainthemotors, ndtheoverall
transformer relationship
1
/
2
and T
1
/T
2
as a function of K
m
1
, K
m
2
and R.
(c) What would be a purely mechanical equivalent of this?
143
14.6 Gear Train
For the system shown below with k=10 [Nm/rad], J = 1 [Nm s
2
/rad] and b= 1.25 [Nms/rad]:
Figure 14.6.1: System Diagram
(a) Find the basic dynamic properties and
n
as seen at
2
.
(b) Nowreectthesepropertiesbacktotheinputshaft. Whatarethebasicproperties and
n
as seen at
1
.
(c) Sketch the response of
1
and
2
to a unit step in T
1
in the same plot.
(d) What eect does the gear train have on the characteristic dynamics of this system? Please
explain your answer.
144
14.7 Non-Ideal Transformer
For the system shown in the gure below, we cannot ignore the dynamic properties of the trans-
former and must try to account for them in our model.
Bearings
Figure 14.7.1: Diagram of Non-Ideal Transformer
Assuming that k
1
is very large, the load inertia is large but of the same order of magnitude as J
1
and J
2
, and that the bearings exert a viscous drag on the shaft.
(a) Write the equation of motion for each important mass in this system.
(b) Express each of those equations in terms of appropriate input/output transfer functions.
(c) Using a block diagram, show how these transfer functions combine to give us the overall
transfer function G(s) =
2
/T
1
.
(d) Nowusingblockdiagramalgebra,derivetheoveralltransferfunctionforthissystemexplicitly.
145
14.8 Equivalent Mass and Inertia
(a) Derive an expression for the equivalent mass at the point of application of the force in Fig-
ure 14.8.1.
f
m v
a b
Figure 14.8.1: Mass on cantilever.
(b) Derive an expression for the equivalent mass at the point of application of the force in Fig-
ure 14.8.2. Assume that each wheel has a mass, m , and a rotational inertia of J . Assume
there is no energy loss due to bearings and that the wheels do not slip.
m,J
f
v
M
Figure 14.8.2: Cart of mass M and wheels each of inertia J.
(c) Derive an expression for the equivalent rotational inertia in Figure 14.8.3. Assume that the
two pulleys have negligible inertia, there are no losses due to bearings and that the ywheel
has inertia, J .
r1
r2
J
T,
Figure 14.8.3: Pulley System
146
14.9 Inertia in Geartrain
This problem considers the system shown in Figure 14.9.1 in which two rotational inertias J
1
and
J
2
are connected by a gear train.
Figure 14.9.1: Rotational inertias connected through a gear train.
TherotationalspeedofJ
1
is
1
andtherotationalspeedofJ
2
is
2
. Aninputsourceoftorque
1
is applied to J
1
in the direction of
1
. As shown in Figure 14.9.1, the gear train has a ratio of
N : 1; that is,
2
=N
1
.
(a) Assume that the input torque source has a constant value of =
0
. What value of N will
maximize the acceleration of the load
2
?
(b) For this acceleration-optimum gear ratio, what is the equivalent inertia of J
2
as seen by J
1
looking through the gear train? That is, what is the reected inertia of J
2
on the J
1
side?
How does this compare with J
1
?
(c) The power input to the system is P
in
=
1

1
. For the optimum gear ratio calculated above,
make a plot of P
in
(t) assuming that the load starts at rest at t= 0.
147
14.10 Motor Driving Inertia Through Gear Train
This problem focuses on the motor connected to a load as shown in Figure 14.10.1. The motor is
driven by an input voltage V in series with a coil resistance R. The motor is assumed ideal, with
no energy storage or losses inside the motor. The motor is connected to a load inertia J > 0 and
rotationaldamperb0. Themotorshaftandloadrotateatangularvelocity. The motor applies
a torque to the load =Kiin the direction of ; correspondingly, the back emf is e=K.
Figure 14.10.1: Motor driving a rotational inertia and damper.
(a) WriteadierentialequationdescribingthesystemintermsoftheinputvoltageV(t) and the
outputspeed(t). WriteanequivalentdierentialequationwithinputV(t) and output i(t).
(b) Assume that the system is initially at rest, and that at t=0 the input voltage takes a step
V(t) = V
0
u
s
(t). Solve for the resulting transient in i(t) and (t), and make a plot of these
two quantities as a function of time.
(c) What are the steady-state values of i and ? In steady-state, write an expression for the
powerbeingdissipatedonthemechanicalsideintheloaddamperb,andontheelectricalside
intheresistorR. Howmuchpowerisbeingsuppliedinsteady-statebythevoltagesource? Is
this in balance with the dissipation? In this steady-state, how much kinetic energy is stored
in the load inertia J?
(d) Makeaplotofsteady-stateloadpowerdissipationasafunctionofloaddampingbforb0.
What value of b results in maximum power dissipation in the load? How does this compare
withtheelectricalequivalentdampingtermK
2
/R? Forthismaximumpowervalueofb, how
much power is being dissipated in the resistor R?
(e) Finally, supposeweallownegativevaluesoftheloaddamping. Notethatanegativedamper
will supply power to the load. For what range of b < 0 will the system be stable? For what
range of b < 0 will the system be unstable?
148
15 MATLAB and Simulink
15.1 Matlab Spirograph
Your instructor has written a small Matlab m-le called spirograph.m which he claims makes
spirograph plots. This le is listed below, and is available for download from the course web site.
Most of you have probably played with a simple spirograph made from a sheet of plastic. In such
a device, an inner gear of some radius rolls inside an internal gear of larger radius. A pencil is
insertedintheinnergearinoneofanumberofpossiblelocations. Astheinnergearrolls,thepencil
thus traces out a trajectory on a piece of paper lying under the spirograph. The spirograph has a
number of choices for the sizes of the inner and outer gears, which result in interesting variations
in the generated patterns.
Unfortunately,theinstructorwastoolazytoaddmuchinthewayofcommentswhenhewrotethe
le. Yourjobistolookthroughthisletounderstandtheunderlyinggeometryandcomputations.
Specically, we ask you to:
(a) Develop a diagram which explains the geometry which is being computed. Be sure this
diagram includes all three of the radii r
0
, r
2
, and r
3
, as well as the angles
1
and
3
.
(b) Draw a diagram of the gear system, showing how this relates to the geometry above.
(c) Explain the operation of the program and the function of each line of code, by annotating a
listing of the m-le.
(d) Run the m-le in Matlab by choosing some values for the parameters in the edit window,
saving the le, then typing the name spirograph at the Matlab command line. Find some
interestingvaluesforthechoiceofradii andincludetheresultingplots. Canyouexplainthe
essential features of the plots on the basis of the chosen radii?
(e) Rewrite the le as a Matlab function which can be called with the three radii as arguments.
Include a listing of your function. To nd out more about Matlab m-les and functions,
reviewthematerialonMatlabinCh.1ofthetext. Youcanalsousetheon-lineMatlabhelp
system.
spirograph.m
Download
% spirograph
% DL Trumper 9/13/01
% revised 9/17/01 to fix minus sign error in theta3
r0 = 2;
r2 = .84;
r3 = .7;
theta1 = [0:pi/1000:100*pi];
theta3 = -theta1*((r0/r2) - 1);
position = (r0-r2)*exp(j*theta1) + r3*exp(j*theta3);
figure(1)
plot(position)
axis equal
149
15.2 MATLAB Plotting
Use MATLAB to plot the following signals. Choose the appropriate time range and sampling
points.
(a) 10sin100t
(b) e
t
cos100t
(c) Ree
st
; s=1 + 10i
(d) 5e
10t
(e) 5e
10t
(f) 8e
100t
5e
1000t
150

15.3 Simulink Introduction
Fun and games with SIMULINK
The purpose of this short but sweeeeet assignment is to learn about the block diagram language within
MATLAB know as SIMULINK. SIMULINK lets you create block diagrams of dynamic systems, invoke
feedback control designs and then run the system to see the predicted performance.
To get started, simply type Simulink in the MATLAB command window. The GUI will then include a
menu of various block options, along with a model window in which you build the system. For example,
the simple velocity servo we have discussed in class would look like this in SIMULINK:
1
s +as
2
Transfer Fcn
output
To Workspace
Sum
Step
Scope
K
Gain
time
Clock
To Workspace1
And if I set K=5 and a = 1 in the command window, the simulation (using default simulation parameters)
becomes:
Page 1 of 2
Figure 15.3.1:
151
For this assignment I encourage you to look at the various blocks in SIMULNIK and to run some of the
DEMOS.
In addition, please look at the following:
Consider the problem we have done in class: The PI control of the velocity servo. However, here we
address the problem of noise in the velocity measurement. This is typically done with a tachometer (really
a generator) and leads to high frequency noise from the brushes in the generator. As a result, we often
place a low pass filter in the feedback loop. Thus:
K
f
H (s)
W
f
s 1
Lets look at this effect by using both analysis and SIMULINK simulation:
a) First create the model of the PI controller for the velocity servo:
Sum
Step
Kp
s+a
Plant
s+z
s
PI Controller
Kc
Gain
b) Calculate analytically the values for K
c
and z (for the PI controller) that will give a critically damped
response with a settling time of 0.5 second. (Assume a=Kp=1)
c) Now implement those gains in SIMULINK and see if it agrees with expectations. If there is any
discrepancy, please explain it.
d) Now add the feedback filter H(s) to the SIMULINK diagram. Assume that K
f
= 1 and that the filter time
constant (W
f
= 0.5 sec). Without changing K
c
or z, run the simulation again, and plot the result. What
happened and why?
e) Now go ahead and vary the gain K
c
until you get critically damped response again as seen by the
simulation response. (Try to keep track of how many simulations you have to do until you get the right
answer.) What is the settling time for this case?
Figure 15.3.2:
152
15.4 Wackygraph
Your instructor has written a small Matlab m-le called wackygraph.m which he claims makes
interesting plots. This le is listed below, and is available for download from the course web site.
% wackygraph
% DL Trumper 9/9/02
r0 = 10;
r2 = 1;
N = 10;
g1 = 40;
g2 = 0.2*r0;
theta1 = [0:pi/1000:100*pi];
theta3 = -g1*theta1;
r1 = r0 + g2*sin(N*theta1);
position = (r0+r1).*exp(j*theta1) + r2*exp(j*theta3);
figure(1)
plot(position)
axis equal
Unfortunately,theinstructorwastoolazytoaddmuchinthewayofcommentswhenhewrotethe
le. Yourjobistolookthroughthisletounderstandtheunderlyinggeometryandcomputations.
Specically, we ask you to:
(a) Develop a diagram which explains the geometry which is being computed. Be sure this
diagram includes all three of the radii r
0
, r
1
, and r
2
, as well as the angles
1
and
3
. What
are the eects of the parameters g
1
, g
2
, and N?
(b) Write expressions for the real part of the position variable, and the imaginary part of the
position variable, i.e., Reposition and Imposition, respectively.
(c) Explain the operation of the program and the function of each line of code, by annotating
a listing of the m-le. Note the use of the .* operator which does element by element
multiplication. Why did the program use this. What happens if the regular * operator is
used?
(d) Run the m-le in Matlab by chosing some values for the parameters in the edit window,
saving the le, then typing the name wackygraph at the Matlab command line. Find some
interestingvaluesforthechoiceofparametersandincludetheresultingplots. Canyouexplain
the essential features of the plots on the basis of the chosen parameters?
(e) Rewrite the le as a Matlab function which can be called with the three radii and three
parametersasarguments(6totalarguments). Includealistingofyourfunction. Tondout
more about Matlab m-les and functions you can use the on-line Matlab help system.
153
16 Case Studies
16.1 CD Player 1
In a typical CD player the disc is mounted directly on the shaft of an electric motor. The rotor of
the motor has a moment of inertia I
r
and experiences a frictional retarding torque which can be
modeled by a linear damper with damping coecient B
m
.
(a) Themassofacompactdiscis0.028kg,anditsdiameteris0.120meters. Thediameterofthe
hole in the center is 0.015 meters. Calculate the moment of inertia I
l
of the CD.
(b) Assume that the electric motor, when it is powered, produces a constant torque T
m
inde-
pendently of the rotational speed
m
. It is observed that,after an initial starting transient,
the steady-state rotor speed
ss
is the same, whether a CD is mounted or not. The start-up
transients are, however, dierent. With no disc mounted, the rotor spindle-speed reaches
95%of
ss
in0.30seconds,butwhenadiscismountedittakes2.0secondstoreachthatsame
speed. Assuming that this behavior can described by a simple linear model, estimate the
value of the damping coecient B
m
,
(i) when there is no disc mounted, and
(ii) when a CD is mounted.
(c) Estimate the moment of inertia of the motor rotor I
r
.
(d) Constructastate-determinedrepresentationofyourlinearmodelforthesystemwiththeCD
mounted. Take T
m
to be the input and
m
to be the output.
154
16.2 CD Player 2
A common design for a speed controller uses simple proportional velocity feedback. The actual
speed of the disc is sensed from information written onto the disc. The motor torque T
m
is then
generated according to the following control algorithm
T
m
=G(r
sensed
)
where r is a constant reference speed, and G is a constant gain.
(a) WhatarethedimensionsofthegainparameterG? WhatwouldbetheunitsforGintheSI
system of units?
(b) Consider that this controller is applied to the CD player modeled in Problem 16.1 (d). Con-
structastate-determinedrepresentationofalinearmodelforthecontrolledsystemwithr as
the input and
m
as the output.
(c) Compare the inputs required to achieve a steady-state disc rotation speed of 4.0 revolutions
per second for the uncontrolled open-loop system of Problem1 and for the controlled closed-
loop system of the present Problem:
(i) What is the magnitude of the constant, suddenly applied, motor torqueT
m
required to
reach the desired steady-state speed in the uncontrolled system.
(ii) Obtain a formula (involving the gain parameter G) for the magnitude of the constant,
suddenly applied, reference speed r required to reach the desired steady-state speed in
the controlled system.
(d) Designthecontroller;i.e.,choosethevalueofG,suchthatthetimetoreach95%ofthesteady-
state speed for the controlled system is 1/5 of thecorresponding time for the uncontrolled
system.
(e) WriteaMATLABscriptwhichcanintegratetheequation(orequations)ofPart(b)andplot
the time history of the motor torque T
m
during the startingtransient of the controlled sys-
tem. Compare the magnitude of the maximum motortorque in the controlled case with the
magnitude of the constant torquefound in (ci) for the uncontrolled case.
155
16.3 CD Player 3
In a CD player the disc with moment of inertia I
d
is mounted directly on the shaft of an electric
motor, whose rotor has a moment of inertia I
r
. The unit experiences a frictional retarding torque
that can be modeled by a linear damper with damping coecient B
m
. The parameter values for
the system are estimated to have the following values:
I
d
= 5e-5 kg-m
2
I
r
= 9e-6 kg-m
2
B
m
= 9e-5 N-m/rad/sec
Formulatethestate-determinedequationofmotionforrotationwiththeangularvelocity
m
asthe
output and the motor torque T
m
as the input.
(a) Consider the steady-state sinusoidal oscillation of
m
in response to an input torque of the
formT
m
=T
a
sint. Calculatetheresponseamplitudeandphaseangleforthreefrequencies:
= 1rad/sec
= 5rad/sec
= 15rad/sec
for each of the following cases:
(i) No disc mounted on the motor spindle
(ii) A disc is mounted on the spindle
(b) Evaluate the break frequency for each of the following cases:
(i) No disc mounted on the motor spindle
(ii) A disc is mounted on the spindle
(c) Make an accurate, labeled, sketch of the Bode plots for amplitude ratio and phase angle for
each of the following cases:
(i) No disc mounted on the motor spindle
(ii) A disc is mounted on the spindle
(d) Iftheinputtorquesareappliedtotherotorinitiallyatrest,estimatehowlongittakestoreach
the steady state (for most practical purposes, steady state may be assumed to be reached at
the end of 4 decay time-constants, when 98% of the transienthas decayed). Do this for each
of the three values of , for each of the followingcases:
(i) No disc mounted on the motor spindle
(ii) A disc is mounted on the spindle
156
16.4 Pinewood Derby 1
1
6
f
e
e
t
6
in
c
h
e
s

3

f
e
e
t

alp
h
a
Figure 16.4.1: Pinewood Derby
Referring to the Pinewood Derby (Fig. 16.4.1), answer the following questions:
(a) In SI units, how much energy is available to move the race car?
(b) A typical Derby car is 6 inches long. If there were no friction of any kind,what would be the
maximumspeed(inSIunits)aracecarcouldattainifallitsweightwereconcentratedatthe
front end of the car?
(c) Whatwouldbethemaximumspeedifallitsweightwereconcentratedattherearendofthe
car?
(d) In case (c) how long does it take the car to reach the bottom of the inclined section of track?
157
16.5 Pinewood Derby 2
Consideraddingaviscousfrictionretardingforcebvtothepreviousmodel. Withsucharetarding
forcewhichincreaseswithvelocity,thespeedofthecar,evenonaninclinedtrackofinnitelength,
is limited by the speed at which the retarding force just balances the accelerating force. If this
limiting speed is twice the velocity obtained in part (c) of Problem 16.4, what is the value (in SI
units) of the viscous friction parameter b ?
158
16.6 Pinewood Derby 3
ThetwoMATLABscriptswhichfollowpermityoutointegratetheequationofmotionforaracecar
on a long incline with angle alpha, subjected to a viscous friction force. Study the programs and
experiment with them , using the Pinewood Derby parameters and the friction parameter b
obtained in Problem 16.5.
(a) Plot a graph of the racecar speed vs. time for a car starting at rest, over the time period
obtained in Problem 16.5, part (d).
(b) With a long incline the speed of the car is asymptotic to the limiting speed described in
Problem16.5. Usethescriptstodeterminehowlongittakesforthespeedtoreach99.9%of
the limiting speed.
car.m
Download
% car.m
% This program calculates the velocity of a car rolling down an incline
% under the influence of viscous friction
% The differential equation used to model the car is:
% m dv/dt = m g sin(alpha) - b v
% This equation is stored in a separate m-file car_visc called by this
% program (a tedious MATLAB detail).
% Declare global parameters (a MATLAB detail)
global g alpha b m
% Input parameter values.
g= 9.81
m = input(Enter the mass of the car in SI units: )
alpha = input(Enter the angle of the incline in degrees: )
b = input(Enter the value of the friction coefficient "b" in SI units: );
% Input initial conditions
v0 = input(Enter the initial velocity in SI units: );
% Input integration time.
tspan = input(Enter time interval of integration in seconds: );
% Call a numerical integration algorithm
[t,v] = ode45(carvisc, tspan, v0);
% Plot the results.
plot(t,v)
xlabel(Time [seconds]
ylabel(Speed, [meters/second]
title(Speed of car on incline
grid
% Print out speed at final instant.
N = max(size(v))
Finalspeed = v(N)
carvisc.m
Download
159
% carvisc.m
% This function, which is called by the program car.m, contains the
% equation of motion of a car on an incline subjected to viscous friction
% Introduce V_dot, the time derivative of v.
function V_dot = carvisc(t,v)
% Declare global parameters (a MATLAB detail).
global g alpha b m
% The mathematical model of the car provides an equation for v_dot.
V_dot = g * sin(alpha * pi/180) - (b/m) * v;
% Thats all there is to it!
160
16.7 Pinewood Derby 4
In Problem 16.4 a Pinewood Derby race car was analyzed under the assumption that its motion
was opposed by linear viscous friction with a constitutive relation
f
viscous
=bv
(a) Formulatearevisedmodelinwhichtheviscousfrictionelementisreplacedbyamodelofair
resistance (sometimes called form drag) with a constitutive equation
f
drag
=AC
d
v[v[=AC
d
v
2
sgn(v)
where A is the frontal area of the car and C
d
is a drag coecient that depends on the cars
geometry; i.e., on its form.
(b) What are the dimensions, in SI units, of the drag coecient C
d
in the above constitutive
relation?
(c) Deriveamathematicalexpressionforthetime-historyofthespeedofthecar,onahorizontal
track, if it starts at t=0 with initial velocity v
o
.
161
16.8 Pinewood Derby 5
Formulate a dierential equation for the velocity of a race car in a model in which the car of mass
m descends a very long inclined track with angle (alpha), under the inuence of gravity and a
formdrag, which has the constitutive equation described in Problem 12.8.
(a) Show that this model predicts a terminal velocity v
ss
on an innitely long track.Derive a
formula for v
ss
in terms of the parameters m, (alpha), g, A, and C
d
.
(b) If, for the same parameter values considered in Assignment 1 (m = 0.1418 kg, (alpha) =
10.81degrees),theterminalvelocityisthesameasitwasforviscousfriction(v
ss
= 8.60m/s),
what is (are) the parameter value(s) in the form-drag model?
(c) Study the MATLAB scripts car2.m and car form.m which can be used to explore the
model described in this Problem. Determine the time T it takes for the car starting at rest
to accelerate to 99.9% of the terminal velocity.
(d) Compare the result of 2(c) with the corresponding result in 4(b) of Problem 16.7. Which
friction model has a quicker approach to the terminal velocity? Give a physical and/or
graphical explanation for this result.
162
16.9 Engine Block Vibration 1
Consideruniaxialvibrationofanengineblockthatweighs200poundsandissupportedonmounts
thathaveaneectivestinessof18,000poundsperinch. andaneectivelineardampingcoecient
of 2 pounds per inch per second, and is subjected to an oscillating force of the form
f(t) = f
a
sint
where the amplitude f
a
equals 2 pounds. Derive an equation of motion for the displacement y(t),
of the engine block from its equilibrium position.
(a) Write an expression for the complex amplitude A of the steady-state displacement response.
(b) Obtain:
(i) the low-frequency asymptote for the complex amplitude (a);
(ii) the high-frequency asymptote for the complex amplitude (a);
(iii) the break frequency
break
where the magnitudes of (i)and (ii) are equal.
(c) Evaluate the input frequency
peak
at which the engine displacement amplitude has the
greatest magnitude.
(d) Evaluate the ratio [A(
peak
)/A(0)[ where A(0) is the limit of A() as 0.
(e) Use MATLAB to make Bode plots for:
(i) the magnitude of the ratio A()/A(0);
(ii) the phase of the complex amplitude A().
Use a logarithmic scale for frequency which extends at least one decade below, and one decade
above,thebreakfrequency. Expressthemagnitudescaleindecibels;i.e.,plot10log
10
(magnitude)
2
.
163
16.10 Engine Block Vibration 2
Reconsider Problem 16.9 for the same engine block and mounts, but instead of the input force
with constant amplitude f
a
= 2 pounds, consider the amplitude of the input force f(t) = f
a
sint
to increase with the square of the frequency, as it does when the force is a reaction to the force
requiredtoaccelerateamassm=w/gwhichoscillatesatfrequencywithdisplacementamplitude
r. Take f
a
= (w/g)r
2
, where (w/g)r= (1.93/386)(1)=0.005 pound-sec
2
.
(a) Write an expression for the complex amplitude A of the steady-state displacement response.
(b) Obtain:
(i) the low-frequency asymptote for the complex amplitude (a);
(ii) the high-frequency asymptote for the complex amplitude (a);
(iii) the break frequency
break
where the magnitudes of (i) and (ii) are equal.
(c) Evaluate the input frequency
peak
at which the engine displacement amplitude has the
greatest magnitude.
(d) Evaluate the ratio [A(
peak
)/A()[ where A()is the limit of A() as .
(e) Use MATLAB to make Bode plots for:
(i) the magnitude of the ratio A()/A();
(ii) the phase of the complex amplitude A().
Use a logarithmic scale for frequency which extends at least one decade below, and one decade
above,thebreakfrequency. Expressthemagnitudescaleindecibels;i.e.,plot10log
10
(magnitude)
2
.
164
16.11 Engine Block Vibration 3
Reconsider Problem 16.9, but now with emphasis on the start-up transient. Cast the equations of
motion in the standard form for state-determined systems:
d
x = Ax+Bu
dt
y = Cx+Du
withtheforcef(t)asinputandthedisplacementy(t)astheoutput. Considerthattheengineisat
restinitsequilibriumpositionuntilt=0. Taketheforcef(t)tobeanabruptlyinitiatedsinusoidal
force of the form _ _
_
0, t < 0
_
f(t) =
_
f
a
sint, t >0
_
with f
a
=2 pounds.
(a) WriteaMATLABscripttointegratethestateequationsfromthegiveninitialcondition. The
script should have the capability of accepting an arbitrary value of the forcing frequency .
(b) The MATLAB solution for x is a two-column matrix consisting of the functions y(t) and
v(t) = dy/dt(t) evaluated at the many intermediate t-values listed in the one-column t-
matrix. WritesomeadditionalscripttomakeMATLABconstructthreeone-columnmatrices
of the following power quantities evaluated at the same set of intermediatet-values:
(i) P
in
(t) = f(t)v(t), the power transmitted to the vibratory system by the input force
f(t).
(ii) P
diss
(t) = f
fric
(t)v(t) = bv
2
(t), the power dissipated by the damping element.
(iii) P
stored
(t) = P
in
(t)P
diss
(t), the power transmitted to the storage elements where it is
continually shifted back and forth from kinetic energy of the engine block to potential
energy of the suspension springs.
(c) Have MATLAB make the following four plots:
Engine block displacement y(t) vs. t;
Power input P
in
(t) vs. t;
Power dissipated P
diss
(t) vs. t;
Power stored P
stored
(t) vs. t;
for each of three input frequencies:
(i) = 1/3
break
;
(ii) =
break
;
(iii) = 3
break
.
In order to see both the transient response and the steady state response in the plots, the time
span in the plots should be at least ten times longer than the system decay time-constant.
165
16.12 Plate On Springs 1
m
y
k
k k
c c
Figure 16.12.1: Steel Plate on Springs
In class a model of the vertical motion of a steel plate mounted on four springs (Fig. 16.12.1)
analyzed. Consider the limiting case where all friction is absent. Derive an equation for the dis-
placementhistoryoftheplatewhenitstartsatt=0fromitsequilibriumpositionwiththeinitial
velocity v
o
.
166
16.13 Plate On Springs 2
Consider the experiment demonstrated in class, in which a textbook is dropped on the spring-
supported plate. Again, for simplicity, consider the limiting case where all friction is absent. Take
the mass of the textbook to be one half the mass of the plate, and assume that during the motion
the textbook remains in contact with the plate. There are two cases to be considered. In Case I
there is a zero height dropin which the book is released from rest when just in contact with the
plate. InCaseIIthedropisfromanon-zeroheightsuchthatthebookstrikesthemotionlessplate
with a downward velocity of v
o
. In both cases the motion begins at t = 0, at the instant when the
book contacts the plate.
(a) What is the ratio of the frequency of oscillation in Case I to the frequency of free oscillation
of the unloaded plate?
(b) WhatistheratioofthefrequencyofoscillationinCaseIItothefrequencyoffreeoscillation
of the unloaded plate?
(c) Locatetheequilibriumpositionofthe(book+plate)systemwithrespecttotheequilibrium
position of the unloaded plate .
(d) Determine the initial conditions for y
k
and v
k
for Case I.
(e) Determine the initial conditions for y
k
and v
k
for Case II.
(f) For each case write a mathematical expression for the vertical motion of the book and plate
for t > 0.
167
16.14 Mousetrap Dynamics 1
Thisproblemmakesuseofthemousetrapswhichwerehandedoutinclass. Theproblemisinten-
tionally open-ended; we havent told you what methods to use to nd the requested information.
Think physically about the problem, and develop the simplest methods which will allow you to
answer the questions which we pose with reasonable accuracy. In your answers, we expect you to
provide a good description of the experiments you conducted and the assumptions made in devel-
oping answers. Relevant and clear sketches, gures, and experimental data clearly presented will
greatly help in telling us what you did. We will come back to this problem over the course of the
next few weeks, so please retain copies of your data and calculations for later reference. We will
not provide solutions for this problem, so it is up to you to develop a solid understanding now so
that you will be able to eectively answer later modeling questions.
You are welcome to work in groups of your own creation for solving the experimental parts of this
problem. However, it is up to you to be sure whatever collaboration you undertake is eective
for increasing your own learning. The mousetraps are yours to keep, so feel free to drill holes, tie
on strings, etc., if that will facilitate your experiments. Just be sure that your trap can remain
functional. If you are working in a group, you can take one of your traps completely apart if that
helps. Werecommendputtingsomesuperglueonthetwostaplesinthebasewheretheseenterthe
baseinordertogluethemintothebase. Otherwisethestapleshaveatendencytopulloutinabout
20-30triggersofthetrap. Safety note: Parts of the trap attain signicant velocities when
thetrapistriggered.Thus,youmustweartheprovidedsafetyglassesatalltimeswhen
you are conducting experiments on the trap! Your vision is a precious commodity!
Please return the safety glasses to us when requested at the close of this experiment.
To have a common vocabulary, we will usethe following terminology: Thetrap is built on a wood
piece we will call the base. We will refer to the main pivoting wire portion of the trap as the bail.
The bail is driven by the main spring wrapped around its pivot point. When cocked, the bail is
retainedbyawirethatwewillcallthetriggerbar.Thebaitisplacedonapieceofmetalwewillcall
the trigger plate. When cocked, the trigger bar is captured by the trigger plate, and then released
whenthetriggerplateistouchedwithsucientforce.Pleaseanswerthefollowingquestions. Think
carefully about how to develop these answers in a accurate,ecient manner. Use fundamentals to
guide your thinking!
(a) Dene the cross-bar as the portion of the wire bail at the largest radius from the pivot, i.e.,
the portion of the wire that hits a pencil when the trap is triggered. What is the force
acting on the crossbar when the trap is cocked, i.e., how much force is required to keep the
trap in the cocked position? What element supplies this force to the cross-bar? How is
this force transferred to the base?About what fraction of the force is carried by the trigger
plate? Includeadiagramtoexplainyourthinking.Deviseameanstomeasurethishold-open
force to an accuracy on the order of a fraction of a Newton. Tell us how you made your
measurements,andwhytheyshouldbevalid. Whatprovidestheweightcalibration? Include
relevant diagrams and experimental data.
(b) Whenthetrapisintheclosedposition(i.e.,initsrestingstate),thereisstillatorqueacting
on the bail. For example, when the cross-bar is resting on a pencil, a force is exerted on the
pencil,forcing it against the base. We will call this the resting-state force. Devise a means to
measurethisresting-stateforcetoanaccuracyontheorderofafractionofaNewton. Tellus
168
howyoumadeyourmeasurements,andwhytheyshouldbevalid. Whatprovidestheweight
calibration? Include relevant diagrams and experimental data.
(c) Use the force measurements you made above to develop a model for the spring torque as a
function of opening angle, under the assumption that the force, and thus the torque, varies
linearlywithopeningangle. Asaconvention,assumethattheresting-stateangleis0radians,
and thus that the hold-open angle (i.e., when the trap is cocked) is radians. Express your
torquemodelinunitsofNm/rad. Whatistheangleatwhichzerotorqueisexerted(itsnot
0 rad,since there is still a signicant torque at this angle)? Make a sketch of the torque as a
function of angle.
(d) Onthecoursewebsitethereareseveralhigh-speedvideoswhichshowthetrapbeingtriggered.
The video labeled mtrap1 shows a pencil triggering the trap. The frame rate of this video is
1000frames/sec. On the basis of your observation of the video, about how long does it take
forthetraptocloseafteritistriggered? Whatisapproximatelythehighestvelocityinm/sec
attainedbyanypartofthetrap? Abouthowfastisthecross-bargoinginm/secwhenithits
thebaseafterbeingtriggered?Explainhowyoudevelopedtheseestimates. Togiveasenseof
scale,anarrowshotfromamodernbowgoesabout100m/sec,andabulletfromariegoes
about 1000 m/sec.
169
16.15 Mousetrap Dynamics 2
Thisproblemfurtherconsidersthemousetrapsystem. Nowwedevelopamodelwhichwillpredict
the closing time of the trap.
(a) From the geometry of the bail, and using the density of steel of about 810
3
/m
3
, calculate
the moment of inertia of the bail about the pivot point, in units of kg-m
2
.
(b) Last week you measured the force of the bail in the open and closed positions, and thereby
developed a model of the torque acting on the bail as a function of angle. For the rest of
thisweeksproblem,wewillassumethatthetorquedoesnotvaryasafunctionofangle,but
rather has a constant value equal to the open position torque. This will allow us to simplify
the analytical calculation of the trap closing time. What is the value of this constant torque
in N-m?
(c) Usethesimpliedconstantspringtorquemodeldevelopedaboveaspartofananalysiswhich
will predict the time required for the trap to close when triggered. Clearly describe your
thinking in developing this model. Please provide plots of the angular velocity and position
asafunctionoftimeaswellasthemagnitudeofthevelocityofthecross-barasafunctionof
time. Whatisthevelocityofthecross-barwhenitreachestheclosedposition(=0)? How
long does it take the trap to close from when it is triggered?
(d) Compare your predicted transition time and velocity with data that you extract from the
high-speed video which we have placed on the web. Tell us how you made this comparison.
How might your model need to be modied in order to match the measured data on the
video?
170
16.16 Mousetrap Dynamics 3
Thisproblemrevisitsthecontinuingsagaofmousetrapdynamics. Inourlastexcitinginstallment,
youcalculatedthetimetocloseundertheassumptionthatthetorqueisconstantwithangle. This
allowed a fairly simple analysis to predict the closing time with reasonable accuracy. This week,
we will allow the torque on the bail to vary linearly with angle. With this linear variation, and
assuming no damping, the mousetrap is really a second-order oscillator, with a pair of poles on
the imaginary axis during the time that the bail is closing. Nonlinearity manifests itself in a very
strong fashion when the bail hits the wood part of the trap. However, the closing motion can be
well-predicted by this linear model. Specically, we ask you to:
(a) Recall your earlier results to write a model for the spring torque as a function of angle in the
form
=K(+
0
).
Here is the torque in Nm applied to the bail from an external source, K is the spring
constant in Nm/rad, is the rotational angle of the mousetrap, and =
0
is the angle at
which the torque applied to the bail is zero.
(b) Usethisspringmodelalongwiththevalueofmomentofinertiathatyoucomputedlastweek
towriteanundampedsecond-orderdierentialequationdescribingthemousetrapmotionin
the form
J

=K(+
0
).
Intheopenposition,wehavedened=. Fromthisinitialcondition,solveforthemotion
assuming that the bail is released at t=0. Be clear as to how you handle the
0
term.
How long does it take for the bail to reach the closed position = 0? You may assume for
thepurposesofthisanalysisthatthemotioncomestoaninstantaneousstopwhentheangle
reaches 0. Solve for and write an expression for the force required to arrest this motion in
zero time.
(c) Use Matlab to make a plot of (t) during the closing event based upon your solution above.
Also plot on the same axes the function (t) based upon last weeks analysis in which the
torque was constant as a function of angle. How do these two solutions dier? Why?
(d) Use the high-speed video of the mousetrap to develop a measure of (t) for all the frames
during the closing event. Overlay this experimental data on your plot from part c). How
does this data dier from the analytical results? Why? How might you rene your analysis
to better model the event? You do not need to carry out this analysis, just tell us what
additional elements or analysis might be added.
171
16.17 Mousetrap Dynamics 4
Triggeryourmousetrapwithawoodpencil,andnoticethedentthatappearsinthepencil. Inthis
last installment of the mousetrap problem, we ask you to develop a calculation which will predict
the force experienced by the pencil, and the approximate time-scale in which this force is applied.
(a) Take a look at the high-speed video on the web of the mousetrap being triggered with a
pencil. On the basis of the 1000 frame per second rate of the video, what can you say about
an upper bound on the time in which the pencil is dented?
(b) Observe and estimate the depth of the dent in the pencil. In the simplest of models, we
might assume that this dent is made at constant force. Under this assumption, you can
readily calculate the force and duration of the denting process. You will need to rely on
the values of angular velocity and moment of inertia that you developed earlier. Show your
assumptionsandcalculations,andmakeagraphofyourestimatedforceasafunctionoftime.
How does this estimate compare with the upper bound that you established in part a)?
(c) During the denting process, what is the instantaneous power ow as a function of time?
That is, what is the time rate of energy transfer from the kinetic energy of the bail to the
mechanical deformation of the pencil? Express this result in watts. Also, convert this power
to units of horsepower. How does this power ow compare to typical devices such as light
bulbs, lawn mower engines, etc.?
172
16.18 Hydraulic Elevator Design
Forbuildingsofmoderateheight,passengerelevatorsarefrequentlydrivenbyhydraulicactuators.
Such elevators also include some form of spring/damper assembly in the bottom of the elevator
shafttocatchtheelevatorintheeventofafailure inthehydraulicdriveorintheelevatorcontrol
system. This problem studies an idealized model of the design of this crash stop. Specically,
suppose we model the elevator as a lumped mass m = 2000 kg. This mass (the car) rides in a
vertical shaft, driven from the bottom by a hydraulic cylinder. At the bottom of the shaft are a
springanddamperwhichformthecrashstop. Denemotionintheverticaldirectionasx. Assume
that at t=0, with the bottom of the elevator at a height of 10 m above the crash stop, and with
the elevator at zero velocity, a valve fails in the hydraulic cylinder. The elevator begins falling
under the inuence of gravity. During the fall, the oil remaining in the cylinder sprays out of the
failed valve, and acts as a damper with a damping coecient c=2000 Ns/m.
(a) Develop a model to predict the motion of the elevator during the fall. Make plots of the
elevator position and velocity as a function of time during the fall. What is the total force
exerted on a passenger during the fall? For this purpose, we model the passenger as a rigid
mass of 100 kg (already included in the total m=2000 kg).
(b) Afterfalling10m,thebottomoftheelevatorcontactsthecrashstop. Yourjobistodesignthe
crashstopsothattheelevatorcomestorestinamannerwhichissafeforthepassengersand
which also doesnt use up too much space in the building.Develop a model for the behavior
oftheelevatorsystemafterithitsthecrashstop. Inyourmodel,assumethatthedampingc
from the oil cylinder is present throughout the motion. You may also assume that the crash
stop mass is insignicant.Use this model to choose the values of the crash spring k
r
and
damper c
r
to meet the following specications:
(i) The maximum acceleration d
2
x/dt
2
on the car must be less than 20 m/s
2
at all times.
(ii) The maximum deection of the crash stop is to be minimized.
(iii) The settling time will be dened as the time from initial contact with the crash stop
until the time that the elevator velocity remains below 0.1 m/sec. The settling time
should be made small.
Developamodelandanalyseswhichallowyoutomeetthespecicationsabove. Clearlyshow
the key steps, assumptions, and results. Note that this is a typical design problem in that
thereisnoonerightanswer,onlydegreesofgoodnessinsatisfyingthespecications. Tellus
how you attempted to solve this problem. Be sure to include plots of the car position and
velocity as a function of time, the deection of the crash stop, and the force exerted on a
100 kg passenger. Also show on your plots the peak acceleration of the car and the time at
which it occurs, the maximum deection of the crash stop and the associated time, and the
time at which the car velocity meets the settling specication.
173
16.19 Servo Position Control
Figure 16.19.1:
174
16.20 Servo Frequency Compensation
This problem is a continuation of Problem 16.19.
Figure 16.20.1:
175
16.21 Servo Torque Disturbance
This problem is a continuation of Problems 16.19 and 16.20.
Now we encounter a true disturbance: the application of some unknown and
undesirable torque on the motor shaft. Our design goal will be to completely reject this
disturbance, and to do so we must make use of more complex controller forms.
For the moment keep K
p
at the value used in Problem 1b), but now assume T
r
= 0 and T
d
is a unit step.
T
a) What is the disturbance transfer function for this system?
T
d
b) Sketch the expected response to the step in T
d
based on this transfer function
.
.
c) What will be the steady-state error to this disturbance?
Now, to completely eliminate this error, well try an integral controller of the form
G
c
= K
p
/s.
With this new controller:
d) Explain why this controller form will not yield an acceptable control system by
reference to the bode plot for this new system.
As a remedy to this problem, we could replace the integral controller with a proportional
integral (PI) controller of the form:
G
c
(s) = K
p
K
I
s
that can be re-written as:
G
c
(s) =
K
c
(s D)
s
Figure 16.21.1:
176
16.22 Plate On Springs Damped 1
f(t)
m
k
c
k
c
b
c
bc
y
k
Figure 16.22.1: Plate On Springs With Damper
Consider the vertical oscillations of a 5-pound steel plate supported on springs, as demonstrated
in class (Fig. 16.22.1). Begin by formulating a simple linear model for the motion of the plate in
response to an applied force f(t). Your job is to use the MATLAB scriptsMassSprgDmpr1, 2 &
3 based on closed form analytical solutions, of the type developed in class, to obtain twelve plots
of the time history of response for twelve dierent sets of model parameters. In every case the
plate is at rest in its equilibrium position, when, att=0 a constant force of 5 pounds is suddenly
applied in the direction of positive displacement. In each case you must select the model stiness
and damping parameters to achieve the specied behavioral parameter values.
(a) In each of the rst four plots the undamped natural frequency is 5 Hz(cycles/second), and
the damping ratios have the following values: (i) = 0.1; (ii) = 0.5; (iii) = 1.0; (iv)
= 1.5. You must choose the model parameters to insert in the MATLAB scripts to get
the time histories of response corresponding to these combinations of
o
and . The script
MassSprgDmpr1.mmustbeusedforCases(i)and(ii). ThescriptMassSprgDmpr2.mmust
be used for Case (iii), and the script MassSprgDmpr3.m must be used for Case (iv). For
Cases (i), (ii), and (iii) take the total duration of the time history to be equal to 5 times the
decaytimeconstant. InCase(iv)takethetotaldurationofthetimehistorytobe5timesthe
longest time constant.
177
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

Case (i) Case (ii)
0.02 0.012
0.01
0.015
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]
0.008
0.006
0.004
0.002
0.01
0.005
0 0
0 0.5 1 1.5 2 0 0.1 0.2 0.3 0.4
Time [seconds] Time [seconds]
Case (iii) Case (iv)
0.01 0.01
0.008
0.006
0.004
0.002
0.008
0.006
0.004
0.002
0 0
0 0.05 0.1 0.15 0.2 0 0.1 0.2 0.3 0.4 0.5
Time [seconds] Time [seconds]
Figure 16.22.2: Figure 1
178
Time History of Displacement Response
0
0.002
0.004
0.006
0.008
0.01
0.012
0.014
0.016
0.018
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

Case (i)
Case (ii)
Case (iii)
Case (iv)
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Time [seconds]
Figure 16.22.3: Figure 2
(b) Ineachofthenextfourplots,thedampednaturalfrequencyis5Hz,andthedampingratios
have the following values: (i) = 0.1; (ii) = 0.3; (iii) = 0.5; (iv) = 0.7. Again choose
model parameters to get the time histories of response corresponding to these combinations
of
d
and .Here all cases can be run with MassSprgDmpr1.m. In each case take the total
duration of the time history to be equal to 5 times the decay time constant.
179
Case (i) Case (ii)
0.02 0.014
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]
0.012
0.015
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

0.002
0 0
0 0.5 1 1.5 2 0 0.2 0.4 0.6 0.8
Time [seconds] Time [seconds]
Case (iii)
x 10
-3
Case (iv)
0.01 6
0.01
0.008
0.006
0.004
0.01
0.005
0
0 0.1 0.2 0.3 0.4
0
Time [seconds]
0 0.05 0.1 0.15 0.2
Time [seconds]
Figure 16.22.4: Figure 3
5
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]
0.008
4
0.006
3
0.004
2
0.002
1
180
Time History of Displacement Response
0
0.002
0.004
0.006
0.008
0.01
0.012
0.014
0.016
0.018
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

Case (i)
Case (ii)
Case (iii)
Case (iv)
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Time [seconds]
Figure 16.22.5: Figure 4
(c) In the nal four plots the decay time constant is xed at = 0.10seconds, and the damping
ratios have the following values: (i) = 0.3; (ii) = 0.5; (iii) = 0.7; (iv) = 0.9. Choose
model parameters to get the time histories of response corresponding to these combinations
of decay time constant and . In each case take the total duration of the time history to be
equal to 5 times the decay time constant.
181
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

Case (i)
Case (ii)
Case (iii)
Case (iv)
Time History of Displacement Response
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45
Time [seconds]
Figure 16.22.6: Figure 5
182
0.5
16.23 Plate On Springs Damped 2
Makeacarefulsketchofthecomplexplaneshowingthelocationoftheeigenvalues,corresponding
to the twelve parameter sets used to produce the time histories in Problem 16.22.
183
16.24 Plate On Springs Damped 3
Reconsiderthespring-supportedplateofProblem16.22,butnowtakethemotiontobecausedby
a2-poundrubberballwhichstrikestheplateandbouncesaway. Assumethattheplateisatrestin
itsequilibriumposition, whenitis struckbythe ballatt=0. Assumetheballis fallingvertically
withavelocityof10feet/secondandreboundswithanupwardvelocityof3feet/second,andthat
it does not impact the plate again.
(a) Determine the initial velocity of the plate, immediately after the impact.
(b) Select the model stiness and damping parameters so that the damped natural frequency is
5 Hz, and the damping ratio is 0.3.
(c) Insert the model parameters in MassSprgDmpr1.m and plot the displacement response of
the plate for an interval equal to 5 times the decay time constant.
(d) What is the relationship between the time history in Problem 16.23 Part (c) and the time
history in the plot for b(ii) in Problem 16.22?
184
16.25 Plate On Springs Damped 4
A rigid plate supported on four identical springs, like the system in Problem 16.22, has unknown
model parameters, m, k, and b. A dynamic test is performed in which a constant force of 1.0
Newton is suddenly applied at t = 0. The measured displacement response is displayed in Fig.
16.25.1
x 10
-4
Time History of Displacement Response
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
0
1
2
3
4
5
6
7
8
9
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

Time [seconds]
Figure 16.25.1: Response to 1.0 Newton Step-Force
Your job is to use the data revealed in this Figure to estimate:
(a) The eective stiness k of the four springs;
(b) The eective mass m of the plate;
(c) The eective damping coecient b of the system.
185
16.26 Plate On Springs Damped 5
Each, of two identical plate-on-springs assemblies, like those of problem 16.22,is accurately rep-
resented by a model with m = 2.0 kg, k = 2000 N/m, and b = 20N/m/s. Consider the system
obtained by placing the two individual units face-to-face, with the rst plate resting on a rm
table with its springs extending upward, and the second plate reversed, so that its springs extend
downwards and are connected to the springs of the rst plate. Your job is to derive a dierential
equation for free vertical motion of the second plate, and:
(a) Evaluate the undamped natural frequency
o
for the model of the combined system;
(b) Evaluate the damping ratio for the model of the combined system;
(c) Evaluate the decay time constant for the model of the combined system.
186
16.27 Plate On Springs Damped 6
The steel plate on springs shown in Problem 16.22 has been modeled by the dierential equation
d
2
y dy
m +b +ky=f(t)
dt
2
dt
Remodelthisasastate-determinedsystemwiththeforcef(t)astheinputandtheverticalposition
and velocity of the plate as the outputs.
(a) The eective stiness of the springs is known to be 3000 N/m. Choose the mass parameter
m and the damping parameter b such that the steady state deection of the plate under its
own weight is =7 millimeters and the damping ratio is = 0.5.
(b) Write a MATLAB script to integrate the state equations for the case where,at t = 0, an
abruptly applied force equal to the weight of the plate is applied in the upward direction.
(c) WriteaMATLABscripttoproduceplotsofthersttwosecondsoftransientresponsein(b).
Produce plots of the following forms:
(i) position vs. time
(ii) velocity vs. time
(iii) velocity vs. position
187
16.28 Shipping Crate 1
Sensitive machinery and instruments are typically packed with shock-absorbing material for ship-
ping. Fig. 16.28.1showsaplanviewsketchofa500-poundengineinashippingcratewithpacking
materialateachend. Inrealitypackingmaterialisprovidedonallsides,buttokeepthingssimple
we onlyconsidertheend-to-endhorizontaltranslationoftheengine withrespecttothecrate. As-
sume that the packing material at each end can be adequately modeled as a linear elastic element
with stiness k combined with a linear friction element with damping parameter b.
Crate
Engine
Packing Material
Figure 16.28.1: Engine in Shipping Crate
(a) Derive a dierential equation for the end-to-end motion of the engine with respect to a
stationary crate.
(b) Specify the values of k and b required to provide a damped natural frequency of 1 Hz and a
damping ratio of 0.707.
188
16.29 Shipping Crate 2
Reconsider the 500-pound engine with shock-absorbing packaging in a shipping crate of Problem
16.28. Take the packing material at each end to have an eective stiness k of 100 pounds/inch,
andaneectivedampingparameterbof150pounds/inch/second. Nowsupposethecrateisplaced
on a at-bed truck so that the end-to-end orientation of the crate is parallel to the direction of
motion of the truck. The crate is secured to the truck so that there is no relative motion of the
crate with respect to the truck. However, because of the shock-absorbing packaging, the engine
may move with respect to the crate when the truck accelerates and decelerates.
(a) Formulateamathematicalmodeltodescribethelongitudinalmotionoftheenginewithrespect
to the crate in response to longitudinal motion of truck.
(b) Derive a single ordinary dierential equation relating the (input) absolute velocity of the
truck v
truck
(t) to the the (output) relative velocity of the engine to the crate v
engine
(t).
(c) Derive a state-determined system in which v
truck
(t) is the input andv
engine
(t) is the output.
189
17 Quiz Problems
17.1 Fun with Block Diagrams
(a) Reduce the block diagram in Figure 17.1.1 to derive the transfer function for the system.
Find the value of K that will result in a critically damped response.
K
s(s+20)
+
-
X(s) Y(s)
Figure 17.1.1: Block diagram for part (a)
(b) Derivethetransferfunctions
Vout(s)
and
Vout(s)
fortheblockdiagramshowninFigure17.1.2.
V
1
(s) V
2
(s)
(Hint: youcanalwayslabelindividualpositionsalongtheblockdiagramwithvariablenames
and write out and solve the relevant algebraic equations.)
V
2
6s+1
14s+8
s+3
+
-
+
+
+
+
V
1
V
out
Figure 17.1.2: Block diagram for part (b)
(c) Write the complete dierential equation for V
out
(t) in terms of V
1
(t) and V
2
(t).
190
17.2 Complex Translation
This problem concerns the translational mechanical system shown in Figure 17.2.1
c
2
F
in
M
2
M
1
c
1
k
1 k
2
x
1 x
2
x
in
Figure 17.2.1: Translational Mechanical System
The system is driven by two inputs, a force input F
in
, and a position input x
in
. The damper c
1
shown acts between the masses M
1
and M
2
. The mass M
1
rolls on the ground surface without
friction.
(a) Draw free body diagrams for the two masses, clearly labeling the forces and associated ref-
erence directions which act on the mass. The functional dependence of these forces on the
system variables should be indicated.
(b) Usethefreebodydiagramsdevelopedabovetowritethegoverningsystemdierentialequa-
tions in the form:
M
1
x
1
=
M
2
x
2
=
where you supply the right hand side of the equations. You do not need to further reduce
the form of the system equations.
191
17.3 Cylinder Step Response
x
k
r
c
f
m

Figure 17.3.1: System Model
YouhavebeengiventhesystemillustratedinFigure17.3.1. Thesystemconsistofacylinderwith
amassmwitharadiusr= 0.5mwhichspinsaboutanaxel. Thecylinderrollswithoutsliponthe
ground. Attachedtotheaxelhousingareadamperc, a spring k=200N/m,andaforcesourcef.
You measure the following response x(t) to a step input of the force source.
Step Response
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Time (sec)
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
Figure 17.3.2: Step Response
A
m
p
l
i
t
u
d
e

(a) Usingtheprovidedparametersandthestepresponse,determinethedampingconstantcand
the equivalent mass m
eq
, where m
eq
is mass equivalent of the combined inertia and mass.
(b) If the cylinder has a m=3 kg, determine the inertia of the cylinder.
192
17.4 JKC Frequency Response 1
(t)
k
c
(t)
J
Figure 17.4.1: Rotational System Model
Bode Diagram
0 1
10 10
Frequency (rad/sec)
Figure 17.4.2: Bode Plot
You perform a frequency analysis of the system shown in Figure 17.4.1 and obtain the bode plot
shown in Figure 17.4.2. The inertia J =15 N/m
2
.
(a) Using the data in the bode plot determine, and
n
for this system.
(b) Using J and the values determined in part (a), determine c and k for the system.
(c) Determine an expression for the system output (t) when (t) = sin (t) and = 1.1, 10,
and 20 rad/s.
193
P
h
a
s
e

(
d
e
g
)

M
a
g
n
i
t
u
d
e

(
d
B
)

180
135
90
45
0
20
15
10
5
0
5
17.5 Mass Spring Damper Dynamics
A second-order system is driven with a position source x
in
.
x
in
c
M
x
k
Figure 17.5.1: Mass Spring Damper System
The mass position is x(t). There is no gravity acting in this problem.
(a) Writethegoverningdierentialequationforthissystemwithaninputx
in
,andanoutputx.
Clearly show the steps in your solution.
(b) Now let the parameters take the values M = 0.1 kg, k =10 N/m, and c= 0.2 Ns/m. What
are the values of and
n
?
(c) Assume the system is in a condition of initial rest. The input position is given to be a unit
step, x
in
(t) = u
s
(t) m. Solve for the mass position x(t) as a function of time. Clearly show
the steps in your solution.
(d) Make a carefully labeled and dimensioned sketch of x(t).
194
17.6 Op-Amp Analysis
C
2
R
2
R
1
C
1
+
+
-
+
v
i
v
o
- -
Figure 17.6.1: Op-Amp Circuit, R
1
= 1M, R
f
=100k, C
1
= 1F, C
2
= 1nF
(a) Derive the Transfer function H(s) relating V
o
to V
i
(assume the Op-Amp acts as an innite
gain).
(b) Derive the expressions for the magnitude and phase as a function of frequency.
(c) Make a bode plot of the system indicating the major points.
(d) Determine the pole and zero locations(s) and plot them on the s-plane.
195
17.7 RLC Circuit Analysis
R L
-
v
i
v
o
+
+
-
C
Figure 17.7.1: RLC Circuit
(a) Write the transfer function,
Vout(s)
, for the circuit shown above.
V
in
(s)
(b) Given C = 1 10
6
F, nd the values of R and L such that = 0.707 and the undamped
natural frequency is 5 kHz. (Dont forget to convert to rad/sec!!!)
(c) Using the same values of L and C from part(b), nd the locations of any system pole(s)
and zero(s) given R=1000. Sketch the unit step response, clearly indicating the time and
magnitude scaling. (Hint: use a dominant pole approximation.) Use the IVT and FVT to
show that your response starts and ends at the appropriate values.
(d) SketchthelogMagnitudevslogfrequency,andlinearphasevslogfrequency(Bodeplot)for
this system based on your calculated poles from part 3. Show that your plot approaches the
correct values of magnitude both as 0 and .
(e) Again using C = 1 10
6
F and R = 1000, now let L = 0 H. (i.e. Remove the inductor
from the circuit) Calculate the location of the pole and compare this to the dominant pole
found in part 2.
(f) Write the transfer function,
Vout(s)
, for the two circuits shown below.
V
in
(s)
(i) (ii)
-
v
i
v
o
+
+
-
R L
C R
2
1
-
v
i
v
o
+
+
-
R L
C
R
2
1
Figure 17.7.2: RLC Circuit
196
17.8 Sailing for Engineers
A sailing ship of mass m is initially at rest, i.e. v(0) = 0. At time t = 0, a strong wind arises of
magnitude V
o
=10 m/s.
9R
YW
P
Figure 17.8.1: Sailboat Diagram
Assume that the force of the wind on the sails in the direction of travel is given by F
w
(t) =
B
w
[V
o
v(t)]. Assume that the viscous drag of the water on the ship is given by F
b
(t) = B
d
v(t)
(a) Formulate a dierential equation that describes the ships velocity v(t).
(b) Solve the dierential equation from (a) and write an expression for the ships velocity v(t).
(c) Sketch the response v(t).
(d) Write an expression for the steady-state velocity v
ss
in terms of system parameters.
197
17.9 Second Order Step Response
Thefollowingplotrepresentsthestepresponseofasecondordersystem,startingfromzeroinitial
conditions.
2.5
2
1.5
0 1 2 3 4 5 6 7 8 9 10
x
(
t
)

1
0.5
0
t
Figure 17.9.1: Step Response Plot
(a) Fromtheplot,ndthe10%-90%risingtimet
r
, the peak time t
p
,the1%settlingtimet
s
and
the overshoot M
p
.
(b) Estimate the natural frequency
n
and the damping coecient .
(c) Write down the ODE corresponding to the plot.
198
17.10 Spring Damper Dynamics
This problem studies the spring-damper arrangement shown in Figure 17.10.1. The spring and
damper are joined by a massless link, which has a vertical position x. A force F acts on this link.
c
k
F
x
Figure 17.10.1: Spring Damper System
(a) Make a sketch of the pole location in the s-plane.
(b) The input force has the following form, valid for all time
F(t) = 10 + 3(t1),
where(t)istheunitimpulse. Notethatthedrivingimpulse(t1)isdelayedby1second.
Be sure you see that this input is for all time, not just after t= 0.
Calculate the link position x as a function of time, and give an expression for this position
valid for all time. Clearly show your thinking and the steps in your calculation. Your result
should be in terms of the parameters of the system, as we have not given numerical values
for these parameters.
(c) Makeacarefullylabeledanddimensionedsketchofx(t). Again,thissketchshouldbedimen-
sioned in terms of the system parameters.
199
17.11 Derive Blocks for Op-amp Circuit
Consider the Op-amp circuit shown below.
V
r
+
_

R
a
C
a
R
b
V
o
+
_
+
_

R
1
C
1
R
2
V
f
V
a
Figure 17.11.1: Op-Amp Circuit.
(a) Anincompleteblockdiagramforthiscircuitisshownonthenextpage. Developexpressions
for each of the transfer functions in the blocks of this block diagram, and ll them in on the
blockdiagram. Carefullynotethesystemvariablesalreadyshownontheblockdiagram,and
ll in the transfer functions appropriately.
(b) Set V
r
to 1 [V]. What are the steady state values of V
o
and V
a
?
200
V

+
r

V
f

Figure 17.11.2:
201
17.12 Automobile Fender Spring/Damper System
A fender is mounted on an automobile though dampers (to absorb collision energy) and springs
(so that the fender can recover after low-speed collisions). During a crash-test, the automobile is
moving at 2 m/s when its fender strikes a concrete barrier. The vehicle mass, m, is 1,000 kg. (In
comparison, the fender itself is essentially massless.) The springs that mount the fender have a
stiness, k, of 1,000,000 N/m.
m
c
k
y
barrier
Figure 17.12.1:
(a) Write a dierential equation for the deection of the springs when the fender is in contact
with the barrier. (We dene y = 0 at the point of rst contact.)
(b) If the damping coecient, c, is 30,000 N-s/m, what is the damping ratio of the mass-spring-
damper system when the fender is in contact with the barrier?
(c) Is there any value of the damping coecient, c, that would yield no rebound of the vehicle
from the barrier? If so, what it is? If not, why not?
202
17.13 Cu Flywheel with Eddy Current Damper
The gure below shows a stylized picture of the copper ywheel, eddy current damper, and mass
used in Laboratory 2.
t
J
M
x
g
b
R
2
R
1
PULLEY
Figure 17.13.1: Stylized version of Lab 2 setup.
(a) Sketch two free body diagrams for the ywheel and hanging weight respectively, detailing
all the forces or torques acting on these elements. In developing these free body diagrams,
express the tension in the string as T. Note that T ,= mg in general.
(b) Write the equations of motion for each of your free body diagrams. Eliminate T and x
to combine these into one rst-order dierential equation in terms of the ywheel angular
velocity . To eliminate x, note that x=R
1
.
(c) Write the characteristic equation for the system in terms of the given system parameters.
What is an expression for the time constant?
(d) At t = 0, with the system initially at rest, the mass is released and begins to accelerate
downward. Make a plot of the resulting ywheel angular velocity (t). What is the nal
angular velocity as t?
203
17.14 Current Driven RC Circuit
This problem considers the circuit shown below
I
c

I
r
C
V
c
R
V
r
I
s
Figure 17.14.1: RC circuit driven by current source.
(a) Note that the circuit is driven with a current source I
s
(t). Write a dierential equation for
this circuit in terms of the capacitor voltage V
c
(t), and with input I
s
(t). Clearly show the
steps in your development.
(b) Show that with the values R = 1 k, and C = 1 F, the dierential equation becomes
10
3
V

c
(t) + V
c
(t) = 10
3
I
s
(t)
Weassumethattheinputcurrenttakestheformofa0.1AstepI
s
(t) = 0.1u
s
(t)[A],andthat
the system is initially at rest. Write an expression for the resulting output V
c
(t) for t > 0,
and make a properly-dimensioned plot of this response.
204
17.15 Driven Mass Spring System
Consider the mechanical system shown below.
k
m
F
x
Figure 17.15.1: 3.0in
(a) Write the governing dierential equation in terms of F and x.
(b) The input is a unit step F(t) = u
s
(t). Calculate x(t) for t > 0, assuming initial rest. Write
an expression for x(t), and make a properly dimensioned sketch of x(t).
205
17.16 Flywheel Driven by Hanging Mass
Consider the mechanical system shown below.
Figure 17.16.1:
In this system, a ywheel of inertia J [kgm
2
] is constrained to rotate about the indicated axis. A
torsional spring k [Nm/rad] is attached to the ywheel. A weight of mass m hangs on a massless
stringwhichiswrappedaroundahubofradiusRontheywheel. Theforceofgravitymg actson
the weight in the downward direction. In the absence of any torque on the spring, the weight sits
at the position x = 0, and x increases as the weight moves down from this position. There is no
damping in the system.
(a) Writeadierentialequationdescribingthedynamicsofthissystemintermsofthemotionof
the weight x, with input force mg. The variable should be eliminated from this equation.
(b) Assume that the weight is initially supported at the position x= 0 by a shelf. At t= 0 this
shelf is removed. Write an expression for the resulting motion x(t) for t > 0, and make a
dimensioned graph of this motion as a function of time.
206
17.17 Homogeneous Second Order DE
Consider the homogeneous dierential equation
d
2
y
+ 2
dy
+(1+100
2
)y= 0.
dt
2
dt
This has poles at s
1
=1 + j10 and s
2
=1j10.
(a) What are the values of the damped natural frequency w
d
, the natural frequency w
n
, the
attenuation , and the damping ratio ?
(b) The homogeneous solution is of the form
y(t) = c
1
e
s
1
t
+c
2
e
s
2
t
.
For the initial values of y(0)=y(0)=1, what are the numerical values of c
1
and c
2
?
(c) The solution can be written in the form
y(t) = Me
t
cos(
d
t+).
What are the numerical values of M and ?
(d) Make a reasonably accurate plot of y(t), indicating key values on the time and magnitude
axes.
207
17.18 Match Pole/Zero Plots with Step Response
Six pole/zero plots labeled A-F are provided below. The following page has eight step response
plotslabeled1-8. Matcheachofthesixpole/zerotoitscorrespondingstepresponse. Yoursolution
should include a list of A through F with the corresponding number directly next to it. Note the
time and amplitude scales are not necessarily the same in each gure. No partial credit will be
given.
30 30
D A
20 20
10 10
I
m

I
m
I
m

I
m

I
m
I
m0 0
-10
-20
-30
-30 -20 -10 0 10 20
Re
E
100
50
-10
-20
-30
B
-30 -20 -10 0 10 20
Re
100
50
0
0
-50
-50
-100
-100
-300 -250 -200 -150 -100 -50 0
-300 -250 -200 -150 -100 -50
Re
Re
C F
30 30
20 20
10 10
-30 -20 -10 0 10 20
Re
0
-10
-20
-30
-30 -20 -10 0 10 20
Re
0
-10
-20
-30
Figure 17.18.1:
208
0
0.5
2
3
4
1
5 2
1
1.5
0.8
0.6
A
m
p
l
i
t
u
d
e

A
m
p
l
i
t
u
d
e

A
m
p
l
i
t
u
d
e

A
m
p
l
i
t
u
d
e

A
m
p
l
i
t
u
d
e

A
m
p
l
i
t
u
d
e

A
m
p
l
i
t
u
d
e

A
m
p
l
i
t
u
d
e
1
0.4
0.2
0 0
0 0.1 0.2 0.3 0.4 0.5 0 0.5 1 1.5
Time (s) Time (s)
6
10
10
8
6
8
4
2
0
6
4
-2
-4
2
-6
0
0 0.005 0.01 0.015 0.02 0 1 2 3 4 5
Time (s) Time (s)
2 7
1
1.5
0.5
0
1
-0.5
0.5
-1
0
0 1 2 3 4 5 0 0.05 0.1 0.15 0.2 0.25
Time (s) Time (s)
8
1.2
1
1
0.8
0.8
0.6
0.4
0.2
0.6
0.4
0
0.2 -0.2
-0.4
0
0 0.05 0.1 0.15 0.2 0 0.5 1 1.5 2
Time (s) Time (s)
209
Figure 17.18.2:
17.19 Opamp Block Transfer Functions
This problem considers the op amp circuit shown below
Figure 17.19.1: Op amp circuit.
(a) The circuit can be represented by the block diagram given on the next page. You are to ll
ineachoftheblockswiththecorrecttransferfunctiononthatpage,andreturnthepagestill
attached to the quiz. Show the relevant calculations in your quiz book.
(b) Suppose that the input is V
i
=1 V, and that this has been applied for a long time. What is
the value of V
o
? Explain your reasoning.
210
211
Figure 17.19.2:
17.20 Piston with 2nd Order Translation
HydraulicPiston
P,Q
A
M
x
K
B
Rollers
Figure 17.20.1: Mass Spring Damper System with Hydraulic Piston
ThegureaboveshowsalargestagewithmassM(kg)drivenbyahydraulicpiston. Thehydraulic
pistonisdrivenbyapressuresourceP(t)(N/m
2
)withavolumeowQ(t) (m
3
/s)(theowresistance
ofthepipingisignored). Thepistonhasaforce/pressurerelationshipofF=P*A,andthusQ=Ax .
(a) FindthedierentialequationforthissystemintermsofaninputP(t)andanoutputx. Note:
Ignore the roller dynamics.
(b) GivenM=5000kg,K=250,000N/m,B=75,000N*s/m,andA=0.1m
2
showthatthetransfer
function from the pressure input P to ow Q (the uidic admittance) is given by
Q 0.01s
(s) =
P 5000s
2
+ 75,000s+250,000
(c) ThesystemhasbeenheldinequilibriumwithP=500kPa=500,000N/m
2
whenthepressure
is removed suddenly at t=0. That is, P(t)=P
0
(1-u
s
(t)). Find and make a dimensioned plot
of Q(t). Use the transfer function from part b) for this calculation even if you were not able
to derive it.
212

17.21 Pole Zero Bode Matching
Six pole/zero plots labeled A-F are shown below. The following page has eight Bode plots labeled 1-8.
Match each of the six pole/zero plots to its corresponding Bode plot. Your solution should consist of
a list of letters A through F with the corresponding Bode plot number directly next to it.
No partial credit will be given.
50
A
100
B
50
0
I
m0
I
m

-50
-100
-100 -80 -60
-50
-40
Re
-20 0 -100 -80 -60 -40 -20 0
Re
100
C
100
D
50
50
0
I
m0
I
m

-50
-50
-100
-300 -250 -200
-100
-150
Re
-100 -50 0 -50 -40 -30 -20 -10 0 10
Re
50
E
40
60
F
20
0
I
m0
I
m

-20
-40
-150 -100
-50
Re
-50 0 -50 -40 -30 -20 -10 0 10
-60
Re
Figure 17.21.1:
213
1
-1
0
2 0
-2
L
o
g

M
a
g
n
i
t
u
d
e
-1
L
o
g

M
a
g
n
i
t
u
d
e

P
h
a
s
e

o

-3
-180
-90
0
10
1
10
1
10
2
10
2
Frequency r/s
10
3
10
3
P
h
a
s
e

o

10
0
-2
10
0
-180
-90
0
10
1
10
1
Frequency r/s
10
2
10
2
3
-1
4
1
L
o
g

M
a
g
n
i
t
u
d
e

P
h
a
s
e

o

10
0
-2
L
o
g

M
a
g
n
i
t
u
d
e

0
10
1
10
2
10
3
10
0
0
90
P
h
a
s
e

o

10
1
10
2
10
3
10
0
-90
10
1
10
2
Frequency r/s
10
3
10
0
0
10
1
10
2
Frequency r/s
10
3
P
h
a
s
e

o

5
10
0
-1
0
1
L
o
g

M
a
g
n
i
t
u
d
e

10
0
-180
-90
0
10
1
10
2
10
1
10
2
Frequency r/s
10
3
10
3
P
h
a
s
e

o

6
-4
-3
-2
-1
0
L
o
g

M
a
g
n
i
t
u
d
e

-360
-270
-180
-90
0
10
1
10
1
Frequency r/s
10
2
10
2
7
-1
0
L
o
g

M
a
g
n
i
t
u
d
e
8
0
L
o
g

M
a
g
n
i
t
u
d
e
-2
-1
P
h
a
s
e

o

-3
-180
-90
0
10
1
10
1
10
2
10
2
10
3
10
3
P
h
a
s
e

o

-90
0
10
1
10
2
10
1
10
2
10
3
10
3
Frequency r/s Frequency r/s
Figure 17.21.2:
214
17.22 Power Semiconductor Thermal Problem
The thermal system shown below represents large power semiconductor device which is captured
between two plates. The device has a thermalcapacitance C
1
[J/K], and its temperature is T
1
[K].
Thermal resistances R
1
[K/W] and R
2
[K/W] connect the device to the upper and lower plates
respectively. TheseplatesaremaintainedatambienttemperatureT
A
[K]. Powerdissipationinthe
deviceismodeledasheatowq
in
[W]intothedevice. Assumenoheatowoccursthroughtheside
walls of the device.
Figure 17.22.1: Schematic of thermal conguration.
(a) Write the governing dierential equation in terms of T
1
and q
in
.
(b) Assume that in steady state, q
in
= q
o
, a constant. What is the steady state temperature
T
1ss
?
215
17.23 Submersible Capsule Hoist System
A submersible capsule requires in case of malfunction a rescue hoist system, as shown in Figure
17.23.1. The capsule mass is M, and the hoist cable stiness is assumed constant and of value K.
The drum has inertia J, and radius r. The drum is driven by a velocity source w(t). The capsule
is subject to a drag force equal to F
b
= bv(t) and to a otation force F
w
. Note: The system is
designed in a way that the hoist cable will be always in tension, under normal operation.
Figure 17.23.1: Capsule hoist system
(a) Whenthecapsuleisatrestthecableisextendedbyanamounte. Find easafunctionofM
and F
w
.
(b) Find the transfer function V(s)/W(s).
(c) Given M = 10000 [kg] and the bode plot shown on Figure 17.23.2, try to nd the values
of J, r, K, and B. Use the correct units for all values, and indicate what value cannot be
determined.
216
Figure 17.23.2: Bode Plot of V(s)/W(s)
217
17.24 Transfer Function from Pole Zero Plot
Consider the following pole-zero plot:
Figure 17.24.1: System Bode Plot.
The DC gain for the system is 10.
(a) Derive the systems transfer function.
(b) Sketch the systems Bode plots.
218
_ _
17.25 Vaccine Cooler
Youre a 2.009 student trying to verify the feasibility of a portable vaccine cooler. The vaccine
chamberofthecoolerisintheshapeofacubeandisshownbelowinFigure17.25.1. Thevaccines
within the chamber have mass m and specic heat c. The combined eective resistance of all six
walls is given by R. T
vac
is the temperature of the vaccines within the cooler, and T
amb
is the
temperatureoutsidethecooler. Theheatowintothevaccinechamberdueto the coolingsystem
is given by q
in
. The width of each surrounding wall is L.
Q
IN
4
AMB
M C
2
4
VAC
Q
OUT
Figure 17.25.1: Cross-section of a cubical portable vaccine cooler
(a) Derive the ordinary dierential equation that describes this system.
(b) You are told that the mass of the vaccines is 1 kg, their specic heat is 4.2 kJ/(kgK), and
the total eective thermal resistance resistance around the vaccine chamber is given by the
following equation:
K
R= 40 L
Wm
Ifthewidthofthesurroundingwallsis2.5cm,theambienttemperatureis25

C,thevaccine
chamber is initially 2

C, and the cooling system is o, how long will it take for the vaccine
chamber to reach a temperature of 10

C?
(c) Youvebeenswampedwithworkfor2.003,andyour2.009teammakesanincredibleamount
ofprogresswithoutyou. Youcomeback2weekslaterandndanalmostworkingprototype
cooler. The cooler uses a non-linear on/o controller with a dead band. On a workbench
youseethefollowingplots. Whatwallthicknessdidyourteammateschooseforyourcooler?
Whenthecoolingsystemisturnedon,whatisthevalueforq
in
? Pleaseuseappropriateunits.
219
Control System Off with Cooling System On
0 2 4 6 8 10
Time (s)
x 10
4
Control System On
0 2 4 6 8 10
Time (s)
x 10
4
Figure17.25.2: PrototypeCoolerTests. Thetopplotshowsthecoolingsystembeingruncontinu-
ously. The bottom plot shows the cooling system being run with the non-linear controller.
-20
-10
0
10
20
30
T
v
a
c

(

C
)

0
5
10
15
20
25
T
v
a
c

(

C
)

220
17.26 Voltage Driven RRC Circuit
Consider the circuit shown below.
R
R
C
1
+
_
V
in
+
V
o
Figure 17.26.1: Voltage divider with capacitor in parallel
(a) Derive the dierential equation describing V
o
(t) in terms of V
i
(t).
(b) The output V
o
(t) is given by
V
o
(t) = 1; t <= 0
V
o
(t) = 2e
10t
1;t > 0
This waveform is sketched below
V
o
(t)
t(sec)
t=-ln(1/2)/10=0.0693sec
1
-1
Figure 17.26.2: Sketch of V
o
What is the input v
i
(t) for all t? Explain your reasoning, and make a sketch of v
i
(t).
221
17.27 Linear Mechanical System with Position Input
Consider the mechanical system shown in Figure 17.27.1, where w(t) is an input position source,
and x(t) is the mass position.
w(t)
m
k
1
k
2
b
1
b
2
x(t)
Figure 17.27.1: Linear mechanical system with two springs, two dampers, and a position source
input, w(t).
(a) Draw a free body diagram for the mass m, showing forces acting on the mass, as a function
of the displacements w(t) and x(t) and in terms of the system parameters.
(b) Write a dierential equation for the system, in terms of input w(t) and output x(t).
(c) Letm = 10 kg, k
1
= 30 N/m,k
2
= 10 N/m,b
1
= 2 Ns/m,b
2
= 6 Ns/m,withw(t) = 1u
s
(t) [m]
(a unit step). Plot the system poles in the s-plane. Solve for x(t), t > 0, from rest initial
conditions.
(d) Plot x(t), using accurate dimensions, and units. Also indicate the overshoot value, with its
associated peak time, as well as the 1090 % rise time.
222
17.28 Tank with Pump Inlet Lower than Outlet
This problem considers the system shown in Figure 17.28.1.
Figure 17.28.1: Diagram of tank conguration.
Asshowninthegure,apumpactsasasourceofowq
p
(t)intothecylinder;thatisthepumpow
isaspeciedconstantindependentofloadpressure. Thecylinderisofcross-sectionalareaA[m
2
].
At a height h
o
[m] from the bottom of the cylinder, we connect a uid resistance R [Pasec/m
3
].
The entire system is exposed to atmospheric pressure P
atm
.
At t = 0, the cylinder is empty (h = 0), and the pump is turned on at constant ow rate q
o
=
10
6
m
3
/sec,andsothecylinderbeginstollup. Theuidheightabovethebottomofthecylinder
is dened as h(t) [m]. Note that no ow enters the resistance R until h(t)h
o
. The liquid in the
systemiswaterwith = 10
3
kg/m
3
. Theheightoftheuidisrecordedasafunctionoftime,and
the data is plotted as shown below.
223
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
]
s
r
e
t
e
m
[

)
t
(
h
h
f
h
o
t
1
0 10 20 30 40 50 60 70 80 90 100
t [seconds]
Figure 17.28.2: Fluid height as a function of time.
Note that the uid height reaches the height of the resistance at t = t
1
= 30 sec; that is h(t
1
) =
h
o
= 0.1 m. For t > t
1
, the height exponentially approaches a nal value of h
f
= 0.2 m.
(a) Using the data for 0t t
1
, what is the value of A in [m
2
].
(b) Using the data for t > t
1
, what is the value of R in [Pasec/m
3
]? (This can be answered
relativelyeasilyifyouthinkaboutitcorrectly.) Whatisthetimeconstant [sec]associated
with the response for t > t
1
? (It is acceptable to determine this either graphically or via an
appropriate calculation.)
(c) A long time later at t =t
2
>> t
1
, the pump ow is set to zero, q
p
=0. Make a dimensioned
sketch of h(t), t > t
2
.
224
17.29 Thermal Power Chip Analysis
ApowerelectronicchipisattachedtoasubstrateandalsoexposedtoairasshowninFigure17.29.1.
HeatislosttotheairthroughathermalresistanceR
1
[

K/W]andtotheboardthroughathermal
resistance R
2
[

K/W]. The air and circuit board are at ambient temperature T


A
[

K]. The chip


has a thermal capacitance of C [J/

K] and is at temperature T
c
[

K]. Power dissipation in the


chip is modeled as heat ow in, q
in
[W].
Figure 17.29.1:
(a) Write the governing dierential equation for this system in terms of T
c
and q
in
.
(b) For R
1
= 90

K/W, R
2
= 10

K/W, and C = 1 J/

K, solve for T
c
(t) and plot T
c
as a
function of time for q
in
= 10 W from rest initial conditions where T
c
(0) = T
A
. Be sure to
label and dimension your time and temperature scales.
(c) If the maximum allowable steady-state temperature of the chip is 100

K above ambient,
what is the maximum allowable power the chip can dissipate?
225
18 Math Techniques
18.1 Complex Expression Reduction
(a) Reduce the following expressions to a single complex number in rectangular form:
1. (2 + 3i)(45i)
2. (2 + 3i)/(45i)
3. (6 + 7i)(3+7i)
4. (6 + 7i)/(3+7i)
(b) Expressallabovecomplexnumbersinpolarform,thencalculatethenalresultinthepolar
formats.
226
18.2 Complex Expressions
(a) Reduce the following expressions to a single complex number in rectangular form:
(i) (1 + 2i)(34i)
(ii) (1 + 2i)/(34i)
(iii) (6 + 4i)(5+7i)
(iv) (6 + 4i)/(5+7i)
(b) Expressallabovecomplexnumbersinpolarform,thencalculatethenalresultinthepolar
formats.
227
18.3 Matrix Operation Practice
AssumeX = [1 2 3 ]

,Y = [9 8 0]

andZ = [s
2
s1]

, where

denotestransposeoperation. Calculate
the following expressions:
(a) 5X;
(b) X

Y;
(c) XY

;
(d) Find the roots of the equation X

Z = 0
228
19 Recitation Problems
19.1 Recitation 1 Problem
Just how eective are snow banks at stopping a sliding car? We can model the system and study
its dynamic and static behavior.
x
b
m
F
x
0
Figure 19.1.1: Sliding Car Model
Consider the model shown in Figure 19.1.1. A mass m, initially at rest, is subject to a constant
applied force F. After the mass has travelled a distance x
o
it impacts a damper with damping
coecient b. Assume the surface is frictionless.
(a) Formulatetherstorderdierentialequationdescribingthevelocityofthecarasafunction
of time, before the collision.
(b) What is the velocity at the collision?
(c) Formulate the rst order dierential equation describing the system after the collision.
(d) What is the time constant for the system? How long will it take to stop the car?
(e) What is the peak force? As a function of x
o
? What if x
o
is 0? What if b is ?
(f) Solve the problem again using energy methods.
229
_
19.2 Recitation 2 Problem
b
m
k
x
Figure 19.2.1: Automobile Suspension Model
Thesuspensionforanautomobilecanbemodeledbythespring-mass-dampersystemshowninthe
gurewithmassm=500[kg],springconstantk= 4 10
3
[N/m],anddamperb= 2 10
3
[Ns/m].
(a) Determine the characteristic equation for the system. What is the form of the homogeneous
response to an initial condition x
0
= 0 and x
0
= 1? To an initial condition x
0
= 1 and
x
0
= 0?
(b) Find the damping ratio , undamped natural frequency
n
, damped natural frequency
d
,
attenuation , and the time constant of the exponential envelope .
Hint:
b k
= and
n
=
2 km
m
(c) Is the system undamped, underdamped, critically-damped, overdamped (or just right)?
(d) From the homogeneous response determine if the system is stable, marginally stable, or
unstable. Does this match your physical intuition?
(e) What if the shock absorber is leaky? Discuss what happens to the system as b changes.
(f) What happens when you drive down Massachusetts Avenue and hit a pothole? Think about
the forced response.
(g) If you were designing the suspension for a car, which damping behavior (undamped, under-
damped, critically-damped, overdamped) would you choose and why? For a sports car? For
grandmas Cadillac? For a pickup truck?
230
19.3 Recitation 3 Problem
Consideraswingingdoor, suchasyoumightndattheentrancetothekitchenina restaurantor
atthefrontdoortoasaloonintheWildWest. Thesetypesofdoorsdontslam;theycanovershoot
and oscillate. The schematic representation of the door is shown in the Figure 19.3.1.
b
J
k
T
T(t)
Figure 19.3.1: Swinging door model
Thesystemrotatesaboutacentralshaftwithangulardisplacement, moment of inertia J, spring
constant k, and damping constant b. When someone pushes on the door, they apply a torque T.
(a) Determine the characteristic equation for the system.
(b) Find the damping ratio , undamped natural frequency
n
, and damped natural frequency

d
. Remember the standard form:
f(t) =

1
n
2

2
n

+
(c) Sketch the response of the system to a step torque of the form
T(t) = T
0
u
s
(t)
with initial conditions
0
= 0,

0
= 0. Use
n
=1 [rad/sec].
(d) Dene the maximum value of the angular displacement . Add it to the sketch.
(e) Dene rise time. Show rise time on the sketch.
(f) Dene settling time. Show settling time on the sketch.
(g) How does the above forced response of a second-order system compare to the homogeneous
response?
231
19.4 Recitation 4 Problem
Two identical tanks with cross sectional area A are partially lled with a volume V of incom-
pressible uid of density . The uid ows between the two tanks through a short tube with a
small diameter which can be modeled as a lumped resistance, R. The top of tank 2 is open to the
atmosphere (with pressure P
a
).
Initially, the top of tank 1 is closed and the pressure at the top of tank 1 is set so that the level of
the uid in the rst tank, h
1
is substantially higher than the level of the uid in the second tank,
h
2
. At time t = 0, the top of tank 1 is opened to the atmosphere. 19.4.1 shows the system just
after the top is opened.
P
h
1
h
2 R
A,U
A,U
P
a
a
Tank 1 Tank 2
Figure 19.4.1: Tank Setup
(a) How many independent variables as a function of time do you need to completely describe
the state of the system?
(b) Whatdoyouexpectthesystemtodophysically? Beforemakinganycalculations,sketchthe
height of the uid in tank 1 and the height of the uid in tank 2 as functions of time.
(c) Findarelationshipbetweentheheightoftheuidintank1totheheightoftheuidintank
2.
(d) What is the pressure P
1
at the bottom of tank 1? What is the pressure P
2
at the bottom of
tank 2?
(e) Find a constitutive relationship for the lumped resistance in the tube.
(f) Find the dierential equation in h
1
for the system.
(g) Sketch the height of the uid in the tank 1 and tank 2 as a function of time, given initial
conditions h
1
(0) and h
2
(0). Will this system oscillate?
(h) Does the time constant of the system depend on the amount of uid? What could you do
to make the system respond twice as quickly?
(i) This system is rst-order, but the demonstration we saw in class on Wednesday of water
owing through a tube was a second-order system. Why are these systems dierent?
232
19.5 Recitation 5 Problem
T(t)
b
t
b
Z
J
R
x
m
k
t
T
Figure 19.5.1: Rack and Pinion
Arackandpinionconvertsrotationalmotiontotranslationalmotionandvice-versa. Thetracking
mechanism in the CD player passed around in class contained a rack and pinion.
Thegureshowsalumpedmodelwheretheangularpositionofthepiniongearisgivenby. The
combined shaft and gear inertia is given by J, the rotational damping on the shaft is given by b

andtheappliedtorqueT(t)isafunctionoftime. Therackpositionisx,themassoftherackism,
the linear damping acting on the rack is given by b
t
, and the linear spring constant is k
t
. Assume
that there is adequate room for the rack to roll back and forth.
(a) What is the relationship between x and ?
(b) Foreachlumpedelementinthesystemndtheenergystoredorpowerdissipated. Compare
it to the constitutive relationship. For example, the potential energy stored in the spring is:
PE=.5k
t
x
2
The constitutive relationship for the spring is: F
k
=k
t
x
(c) Determine the total kinetic energy, potential energy and power dissipated for the system.
Express them as a function of (Use part (a) to eliminate x).
(d) Use the energy equations to nd the second-order dierential equation for the system as a
function of .
(e) Whatistheequivalentinertia,equivalentdampingcoecient,andequivalentspringconstant
for the system? Is the equivalent inertia for the system larger or smaller than J?
(f) Findthedampingratio,undampednaturalfrequency
n
asfunctionsofthesystemparam-
eters. Remember the standard form:
f(t) =

1
n
2

2
n

+
(g) IftheradiusofthegearRisincreased,whathappenstothedampingratio andundamped
natural frequency
n
?
233
(h) Rewrite the equations as a function of x instead of . What happens to the damping ratio
and undamped natural frequency
n
?
(i) Assumethesystemisunderdampedandsketchtheresponseoftoasteptorqueoftheform:
T(t) = T
0
u
s
(t)
with zero initial conditions (0)=

(0)=0.
234
19.6 Recitation 6 Problem
The circuit in the gure consists of a voltage source, a capacitor and three resistors.
R
1
+
Q
R
R
2
C
+
- V
in
-
R
3
Figure 19.6.1: RC Circuit
(a) Find the dierential equation for this circuit in terms of v
o
.
(b) Find v
o
as a function of time, if V
in
is a step input and v
o
(0)=0.
(c) What is the equivalent resistance for the circuit?
235
19.7 Recitation 7 Problem
R
L
J
Motor
K
b
+
i
+
e
V
-
-
T Z
Figure 19.7.1: Motor Circuit
Figure 19.7.1 shows an ideal motor connected to a load rotating with angular velocity . The
motors resistance is lumped into the element R. This motor also has non-negligible inductance
given by L. The motor is driven by an input voltage V. The motor applies a torque to the load
T =Ki. Themotorsbackemfise=K. ThecombinedmotorandloadinertiaisgivenbyJ and
the load is also subject to a rotational damper b.
(a) Find a dierential equation for the rotational load in terms of T and .
(b) Findadierentialequationforthecircuitintermsofe,V andi. Rememberthattheelemental
=L
di
equation for an inductor is V
L
dt
.
(c) Combineparts(a)and(b)tondadierentialequationforthesystemintermsof andV
in
(Eliminate T and e). What is the steady state behavior of this system?
(d) WhencantheinductanceLbeneglected? WhathappensasLgoestozero? Doesthesteady
state behavior of the system change?
(e) WithL=0, thinkaboutthe torqueinputT totherotationalload. IftheinputvoltageV is
a step, would you also expect the torque T to be a step?
236
19.8 Recitation 8 Problem
b
J
T
T
Figure 19.8.1: Rotational System
Figure 19.8.1 shows a rotational system whose angular position is and angular velocity is =

.
The load has inertia J and rotates on bearings which are modeled as an ideal rotational damper
with coecient b. A torque T is applied to the load. The initial velocity is (0)=
0
.
(a) Write the dierential equation for the system in terms of T and .
(b) Find the Laplace transform for the dierential equation.
(c) IftheinputtorqueisaunitstepT(t) = T
0
u
s
(t)(withconstantmagnitudeT
0
),whatis(s)?
(d) How would you solve for the angular velocity as a function of time (t)?
(e) What is the transfer function
(
T(s
s
)
)
?
237
19.9 Recitation 9 Problem
R
v
+
-
i
J
T W
+
-
e
Motor
k
K
m
b
Figure 19.9.1: Motor with Rotational Load
Figure 19.9.1 shows a motor connected to a rotational load. The motors resistance is given by R
andthemotorsinductanceisnegligible. Themotorisdrivenbyaninputvoltagevwhichgenerates
a current iinthecircuit. ThecombinedmotorandloadinertiaisgivenbyJ. Theangularposition
oftheloadis. Flexibilityintheloadismodeledasrotationalspringwithconstantk. Rotational
dampingfromthemotorandtheloadislumpedintotheelementb. Thesystemisinitiallyatrest.
The motor applies a torque to the load: =K
m
i
The motors back emf is: e=K
m

(a) Find the ordinary dierential equation (ODE) for the motor circuit in terms of v, and .
Laplace transform this equation and put it in the form:
V(s) = H
1
(s)T(s) + H
2
(s)(s)
(b) Find the ODE for the load in terms of and . Find the transfer function:
(s)
H
3
(s) =
T(s)
(c) Combine the results of parts (a) and (b) to nd the system transfer function from v to :
H

(s) =
V
(
(
s
s
)
)
. Howwouldyoundthetransferfunctionfromvto? Determine H
T
(s) =
V
T(
(
s
s
)
)
(d) If v(t) is a unit step voltage with magnitude V
0
, use the nal value theorem to solve for the
steadystatevalue(). Whatisthesteadystateangularvelocity

()? Whatisthesteady
state torque ()?
(e) Find the power dissipated by the rotational damper at steady state. Is this surprising? How
would you determine when the power reached its maximum? Where is the power coming
from?
(f) Repeatpart(d)forarampinput. Whatisthepowerdissipatedbytherotationaldamperat
steady state for this case?
(g) How would you redesign this system to get better behavior?
238
(h) If H(s) can be written in the form
0.7
H

(s) =
s
2
+ 62 + 25
sketchthepolesandzerosofH(s)inthes-plane. Basedonyourpole-zerosketch,howwould
you expect the system to respond to a step input?
holy crap you are a big pain in the ass what the hell is happening to the red sox they are totally
sucking some major ass and i hate it.
239
19.10 Recitation 10 Problem
Construct Bode plots for the frequency response function
2
G(j) =
j(1+j/2)(1+j/5)
240
20 Recitation Quizzes
20.1 Recitation 1 Quiz
The snow from last weeks storm on the at roof of Building 3 has melted all at once, and now
physical plant wants to know how long it will take for all that water to drain. We can model this
as a rst order system and calculate the time constant.
h
R
q
p
A,
Figure 20.1.1: Fluid Model
Consider the system shown in Figure 20.1.1, which consists of a uid tank lled to a height h.
Water is draining out of the tank at a rate q through an opening with resistance R. The volume
in the tank, V, is equal to the area A times the height V =Ah. The change in volume is thus:
dV dh
=A
dt dt
Fromconservationofmass,weknowthatthechangeinvolumeofliquidinthetankisequaltothe
outow (note the sign is negative because the water is owing out):
dV
q=
dt
The pressure p at the bottom of the tank is p = gh where is the (incompressible) uid mass
density and g is acceleration due to gravity. The linear uid resistance can be written as p=qR.
(a) Formulatetherstorderdierentialequationdescribingtheheightofthewaterasafunction
of time.
(b) What is the time constant for the system?
(c) If the initial height of the water is h
0
, what will the height of the water be after one time
constant? After 2.3 time constants?
241
20.2 Recitation 2 Quiz
0 2 4 6 8 10 12 14 16 18
-40
-30
-20
-10
0
10
20
30
40
a
m
p
l
i
t
u
d
e

(
m
i
l
l
i
m
e
t
e
r
s
)

time (seconds)
Figure 20.2.1: Second Order Response
Consider the plot shown above of the homogeneous response of a second-order system:
x + 2
n
x +
n
2
x= 0
Estimate the undamped natural frequency
n
, the damping ratio , and the initial position x(0).
Is this system undamped, underdamped, critically-damped, or overdamped?
242
20.3 Recitation 3 Quiz
b
m
x
F
Figure 20.3.1: Cart System
The system shown in the gure has the rst-order dierential equation:
mv + bv=F
where m is the mass, b is the damping coecient, x is the displacement of the mass, v =x is the
velocity of the mass, and the force F is given by a step function:
F(t) = F
0
u
s
(t)
withmagnitudeF
0
constant. Assumezeroinitialposition(x
0
=0)andzeroinitialvelocity(v
0
= 0).
(a) Sketch the velocity of the mass as a function of time.
(b) What is the velocity of the mass at steady state?
243
20.4 Recitation 4 Quiz
T
k b
x
y
m
g
R
Figure 20.4.1: Pulley System
ThemechanicalsystemshowninFigure20.4.1consistsofapulleyofradiusRsuspendedfromthe
ceiling by a spring and a damper. A cable runs over the pulley and is attached to ground on the
right and a hanging mass m on the left. The vertical position of the center of the pulley is x, the
verticalpositionofthemassisy,andtherotationalangleofthepulleyis. Notealsothatgravity
acts in a downward direction.
Assume that the mass of the pulley, cable and support is negligible, the pulley rotates without
friction, and that there is no slip between the pulley and the cable.
(a) Draw a free body diagram for the pulley.
(b) What is the relationship between x and y?
244
20.5 Recitation 5 Quiz
T(t)
b
t
b
Z
J
R
x
m
k
t
T
Figure 20.5.1: Rack and Pinion System
Arackandpinionconvertsrotationalmotiontotranslationalmotionandvice-versa. Thetracking
mechanism in the CD player passed around in class contained a rack and pinion.
Figure 20.5.1 shows a lumped model where the angular position of the pinion gear is given by .
The combined shaft and gear inertia is given by J, the rotational damping on the shaft is given
by b

and the applied torque T(t) is a function of time. The rack position is x, the mass of the
rack is m, the linear damping acting on the rack is given by b
t
, and the linear spring constant is
k
t
. Assume that there is adequate room for the rack to roll back and forth.
(a) Draw free body diagrams for the rack and for the pinion.
(b) What is the relationship between x and ?
(c) Findthesecond-orderdierentialequationforthesystemasafunctionof (Usepart(b)to
eliminate x).
245
20.6 Recitation 6 Quiz
R
1
+
+
R
2
C
Q
R
V
in
-
-
Figure 20.6.1: RC Circuit
The circuit in the gure consists of a voltage source connected to two resistors in parallel and a
capacitor. Find the dierential equation for this circuit in terms of v
o
.
246
_ _
20.7 Recitation 7 Quiz
R
V
+
-
i
b
J
Z T
+
-
e
Motor
K
Figure 20.7.1: RC Circuit
Figure20.7.1showsanidealmotorconnectedtoaloadrotatingatangularvelocity. The motors
resistance is lumped into the element R. The motor is driven by an input voltage V. The motor
applies a torque to the load T =Ki. The motors back emf is e=K. The combined motor and
load inertia is given by J and the load is also subject to a rotational damper b.
Without working through the mathematics, select the dierential equation which best represents
this system:
(a)
K
_
K
2
_
V =J b+
R R
(b)
K
_
K
2
_
V =J + b+
R R
(c)
K K
2
V =J + b+
R R
Justifyyouranswer(thinkoftransientandsteadystateresponse,thephysicsofthesystem,etc.)
247
20.8 Recitation 8 Quiz
R
x
i
+
+
e V
in
-
-
Voice
Coil
m
Figure 20.8.1: Voice Coil Circuit
ThevoicecoilshownaboveisdrivenwithaninputvoltageV
in
. Weareinterestedinthemotionof
the load m, which moves with velocity v=x .
Consider the voice coil to be ideal, with lumped resistance R. Neglect the voice coil inductance,
voice coil inertia, and any translational damping. The voice coil equations are:
F =Ki
e=K
e
v
wherethedirectionoftheappliedforceF onthemassisthesameasx. (Pleasekeepthenotation
for the back emf constant K
e
dierent than the voice coil constant K.)
Find the dierential equation for the system in terms of V
in
and v (eliminate e and i).
248
20.9 Recitation 9 Quiz
2(s+ 2)
H(s) =
s
2
+ 7s+ 12
ThetransferfunctionH(s)givenaboveistheratiooftwopolynomials. Usepartialfractionexpan-
sion to rewrite the transfer function as the sum of the ratio of two smaller polynomials.
249
20.10 Recitation 10 Quiz Spring 2005
F
m
x
b
k
Figure 20.10.1: Mass Spring Damper System
The input force F to the mass spring damper system shown above is F(t) = sin t. The output
position steady state response is x(t) = Msint+. Let m= 1, b= 1, and k= 1.
Find M and for = 1.
250
Chapter
Solutions
251
1 1st Order Systems
1.1 First Order Time Constant
The solution for this problem is not available.
252
1.2 Rise and Settling Times
We are given the rst-order transfer function
1
H(s) = (1)
s + 1
Theresponsetoaunitstepwithzeroinitialconditionswillbey(t) = 1 e
t/
. To determine the
amount of time it take y to settle to within of its nal value, we want to nd the time t
s
such
that y(t
s
) = 1 . Thus, we obtain
= e
ts/
(2)
t
s
=ln (3)
The10-90%risetimet
r
maybethoughtofasthedierencebetweenthe90%settlingtime(=0.1)
and the 10% settling time (=0.9).
t
r
=t
=0.1
t
=0.9
(4)
Therefore, we nd t
r
= 2.2 . For = 10
6
, t
s
= 13.82.
253
1.3 First Order System Response
Figure 1.3.1:
254
Figure 1.3.2:
255
2 2nd Order Systems
2.1 Second-Order System Response
Figure 2.1.1:
256
Figure 2.1.2:
257
Figure 2.1.3:
258
Figure 2.1.4:
259
Figure 2.1.5:
260
Figure 2.1.6:
261
_
2.2 Second Order Responses
(a) From the gures in the problem set, it can be estimated that the period T = 0.1sec and the
time constant = 0.35 sec.
d
= 2/T = 62.8. Since the system is slightly damped, the
estimated damping ratio
1 1 1
= = = 0.045 (1)

d
0.3563
Further the natural frequency
n
=
d
/ 1
2
= 62.8
(b) A mechanical mass-spring-damper system will give this response. K = M
n
2
, C = 2
n
M.
For example, M=1kg, K= 62.8
2
= 3944 N/m, C= 20.04562.81 = 5.6520 N-sec/m.
The initial position x(0) = 0 mm, v(0) A
n
= 300mm62.8rad/sec= 18850mm/sec =
18.85m/sec, where A is the initial amplitude of response.
262
2.3 Second Order Derivation
The solution for this problem is not available.
263
2.4 Second Order Derivation Continued
The solution for this problem is not available.
264
3 Higher Order Systems
3.1 Structure of Higher Order System Solutions
The solution for this problem is not available.
265
3.2 Structure of Higher Order System Solutions
Individual components of x(t)
1
0.8
0.6
0.4
0.2
0
Total Response x(t)
4
3.5
3
2.5
2
1.5
1
0.5
0
0 2 4 6 0 2 4 6
Time (s) Time (s)
Figure 3.2.1: Sketch Solution for Case I
Individual components of x(t)
1.2
1
0.8
0.6
0.4
0.2
0
0. 2
0. 4
Total Response x(t)
3.5
3
2.5
2
1.5
1
0.5
0
0 2 4 6 0 2 4 6
Time (s) Time (s)
Figure 3.2.2: Sketch Solution for Case II
266
Individual components of x(t) Total Response x(t)
0 2 4 6
1
0. 5
0
0.5
1
1.5
0 2 4 6
1
0. 5
0
0.5
1
1.5
2
2.5
Time (s) Time (s)
Figure 3.2.3: Sketch Solution for Case III
(a) Case I
s
1
=4, s
2
=3, s
3
=2, and s
4
=1
x(t) = C
1
e
4t
+C
2
e
3t
+C
3
e
2t
+C
4
e
t
Case II
s
1
=1, s
2
=2, s
3,4
=2.54.33j
n
= 5,
1
x(t) = C
1
e
t
+C
2
e
2t
+C
3
e
2.5t
cos(4.33t+
3
)
10.5
2
Case III
s
1,2
=19.95j
n
=10, s
3,4
=2.54.33j
n
= 5
1 1
x(t) = C
1
e
t
cos(9.95t+
1
) + C
2
e
2.5t
cos(4.33t+
2
)
10.1
2
10.5
2
(b) Case A - This is a summation of four 1
s
t order responses. Thus the system does not have
an oscillation. Because all the terms are rst order, the system will not overshoot the nal
value.
Case B - This is a summation of 2 rst order response and a decaying sinusoid. Since the
decay constant of the sinusoid is faster than the slower rst order terms, the sinusoid will
appear in the initial response but decay to a simple rst order response. Depending on the
magnitude of the sinusoid it is possible for the system to overshoot.
Case C - This is the summation of two decaying sinusoids. Since the time constant of the
decay envelope of the 4.33 r/s sinusoid is faster than the 9.95 r/s sinusoid, the response will
initially be a combination of the two sinusoids decaying to just the 9.95 r/s sinusoid. This
system can overshoot the nal value.
(c) To sketch the responses for each of these systems, we need to make a couple of assumptions.
One assumption I made was that all of the coecients were equal to 1 (Reasonable for this
267
case). The 2nd assumption I made was that
n
=0. Figures 3.2.1-3.2.3 show the responses
forthethreecases. Theleftplotshowstheindividualcomponents,whiletherightplotshows
thetotalresponse(x
1
+x
2
...+x
n
). Normally,weareinterestedinthesystemresponsetoaset
of initial conditions or an input force, in which case the
n
,=0 and c
1
,=c
2
.... For reference,
I have included the system response for the initial conditions x(0) = 1, x(0) = x(0) =
...
x (0)
(Note: thenumberofinitialconditionsforthesystemisequaltotheorderofthesystemthus
arstordersystemneedsoneinitialcondition,a2ndordersystemneeds2IC,ect.). Figures
3.2.4-3.2.6 show the results.
LinearSimulationResults
0 1 2 3 4 5 6
Time(sec)
Figure 3.2.4: Solution for Case I
LinearSimulationResults
0 1 2 3 4 5 6
Time(sec)
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
A
m
p
lit
u
d
e

0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
A
m
p
lit
u
d
e

Figure 3.2.5: Solution for Case II


268
LinearSimulationResults
0 1 2 3 4 5 6
0.4
0.2
0
0.2
0.4
0.6
0.8
1
A
m
p
lit
u
d
e

Time(sec)
Figure 3.2.6: Solution for Case III
structurehighersol.m
Download
clear all; close all;
A1=[1 10 35 50 24];
A2=[1 8 42 85 50];
A3=[1 7 135 550 2500];
S1=tf(1,A1); S2=tf(1,A2); S3=tf(1,A3);
%stuff to make plots
%Case I
t=[0:0.01:10];
x1=exp(-4*t);x2=exp(-3*t);x3=exp(-2*t);x4=exp(-t);
figure(1)
subplot(1,2,1)
plot(t,x1,t,x2,t,x3,t,x4)
title(Individual components of x(t))
xlabel(Time (s))
axis square;
subplot(1,2,2)
plot(t,x1+x2+x3+x4)
title(Total Response x(t))
xlabel(Time (s))
axis square;
omegan=1;
t=[0:0.01:20];
xt=(t-sin(omegan*t)/omegan)/omegan^2;
figure(4)
plot(t,xt)
269
4 Mechanical Systems
4.1 Balloon
Assume hovering balloon has mass m and is subject to a buoyancy force B which tends to lift the
balloon. Consider only vertical motion of the balloon. Let the length of the string between the
balloon and the oor be y. Assume the string has mass/unit length so that a string mass of y
must move whenever the balloon moves. Both the balloon mass and the string mass experience
gravitational attraction pulling them down. Assume that there is no tension in the string at the
pointwherethestringmeetstheoor. Finallyassumetherearedampingmechanisms(airdragand
losses associated with coiling and uncoiling the string on the oor) which can be approximately
modeled by a linear damping force f
d
= bdy/dt opposing the velocity dy/dyt. These forces are
displayed in the free body diagram in Fig. 4.1.1.
B
y
mg
yg
m
f
d
Figure 4.1.1: Free-Body Diagram of Hovering Balloon
To construct a model we consider (i) geometric compatibility requirements, (ii) constitutive equa-
tions, and (iii) force-balance requirements.
(i) The displacement of the balloon is y(t). The velocity and acceleration are
dy dv d
2
y
v= a= =
dt dt dt
2
(ii) The constitutive equations are
d
f
m
= [(m+y)v] f
d
=bv f
gravity
= (m+y)g
dt
270
(iii) The force-balance requirement is
f
m
=Bf
gravity
f
d
Anequationofmotionforthemodelisobtainedbysubstitutingtheconstitutiveequationsintothe
force-balance requirement to get
d dy
[(m+y)v] + b + (m+y)g=B (1)
dt dt
(a) In a steady state y has a xed value y
o
and all time derivatives. When we set the time
derivatives equal to zero in (1) we nd that the equation is satised if
Bmg
y=y
o
=
g
The variable mass term in (1) makes the equation nonlinear. It can be approximated by
a linear equation if the varying mass of the string, y is replaced by the xed mass y
o
of
the string in its equilibrium position. This is a reasonable approximation for studying small
oscillations about the equilibrium position. The dynamics of such oscillations are claried if
we introduce the displacement z=yy
o
of the balloon from its equilibrium position. After
inserting y=y
o
+z in (1) we nd
dv dz
(m+y
o
) +b +gz=B(m+y
o
)g= 0
dt dt
Finally, introducing the total mass M =m+y
o
and putting v=dy/dt=dz/dt, we obtain
d
2
z dz
M +b +gz= 0 (2)
dt
2
dt
This is the standard form for free oscillation of a linear second-order system.
(b) Physically,whentheballoonrisesaboveitsequilibriumpositionitliftsadditionalstringwhose
weight tends to lower the balloon, and when the balloon descends below its equilibrium
position it carries a lighter load of string which tends to drive the balloon upward. The
equivalentstinessintheequationofmotion(2)isk=g. Thisistheweightperunitlength
of the string. This example illustrates the fact that restoring-force elements do not always
look like springs.
(c) The second-order equation of motion (2) predicts oscillation if the damping is light enough
to produce a critical damping factor smaller than unity.
(d) A free vibration governed by (2) will eventually decay to zero if there is a positive damping
parameter b.
To obtain the behavioral parameters and
o
we use the equations

2
=
LDR
2
and
o
= _

d
, with
d
=
2

2
+LDR
2
1
2
T
d
271
_
(e) Fromthegivendataweseethattheratioofsuccessivepeakamplitudesis0.2. ThustheLDR
is ln(0.2) = -1.609, and the damping ratio is
(1.609)
2
_
1/2
= = 0.456

2
+ (1.609)
2
(f) From the given data we see that the damped natural period T
d
is 3.0 seconds, so that the
damped natural frquency is
d
= 2/T
d
= 2.094 rad/sec, and
2.094

o
= _ = 2.35 rad/sec
1(0.456)
2
To obtain the model parameters, we begin with the eective stiness k which is the weight per
unit length of the string. This is given as 0.5 ounces per foot which is 0.03125 pounds/foot in the
English system or
4.448
k= 0.03125 = 0.456 Newtons/meter
0.3048
(g) The eective mass M of the balloon plus the equilibrium length of string, is given by
k 0.456
M = = = 0.0826 kg

o
2
(2.35)
2
in the SI system. In the English system
W W 0.03125
M = = = = 0.00566 pound sec
2
/ft
g 32.2 (2.35)
2
from which we nd W = 0.1823 pounds or 2.92 ounces. The weight of 3 feet of string is 1.5
ounces so the weight of the balloon is 1.42 ounces. In the SI system, the mass of 3 feet of
stringis3(0.3048)(0.456)/9.81=0.0425kg,sothemassoftheballoonism=0.0826- 0.0425
=0.0401 kg.
(h) The eective damping coecient b= 2
o
M is
b= 2(0.456)(2.35)(0.0826)=0.1770 N/m/s
in SI units, or
b= 2(0.456)(2.35)(0.00566)=0.0121 pounds/ft/sec
(i) Theeectivestinesswaspreviouslyobtainedas0.456Newtons/meterinSIunitsor0.03125
pounds/foot in English units.
272
_

4.2 Bungee Jumper
Initially the bungee jumper free-falls under the inuence of gravity and air resistance, then when
theslackistakenupintheelasticcord,thecordexertsanincreasingretardingforcewhichreduces
her velocity to zero at level A, after which she bobs up and down with decreasing amplitude until
shecomestorestattheequilibriumpositionatlevelB.Forthepurposesofapreliminaryestimate
of the conditions up to instant at which level A is reached, neglect the eects of air resistance and
dampinginthecord. Underthisassumptionenergyisconserved. Intherstjumptheequilibrium
position B is 20 feet below the point where the cord begins to stretch. The stiness of the cord is
then estimated as
W 150
k= = = 7.5 pounds/foot
20
The undamped natural frequency of the mass-spring system consisting of the jumper and cord is
k k 7.5

o
= = = = 1.269 rad/sec
m W/g 150/32.2
Let the elevation of the upper attachment point of the cord be denoted by h
o
, the elevation of the
point where the slack is taken up and the cord begins to stretch be denoted by h
1
, the elevation
of the equilibrium point B be denoted byh
B
, and the elevation of the point A where the jumpers
velocity rst vanishes be denoted by h
A
.
(a) In the rst jump, h
o
h
1
=100 feet, and h
o
h
B
=120 feet. The level A can be located by
equating the potential energy lost in the fall to the elastic energy in the cord.
1 1
W(h
o
h
A
) = k(h
o
h
A
) = k(h1h
A
)
2
= k[(h
o
h
A
)100]
2
2 2
This reduces to a quadratic equation for (h
o
h
A
) whose solutions are
(h
o
h
A
)=12066.3 feet
The physically signicant root is (h
o
h
A
)=186.3 feet. The low point A is 66.3 feet below
the equilibrium level B, or 186.3 feet below the upper attachment point.
(b) The maximum downward acceleration is 32.2 feet/sec
2
during the initial free fall.
(c) The maximum upward acceleration occurs at point A and is
a
max
=(Max displacement from equilibrium)
o
2
= (66.3)(1.269)
2
=106.8ft/sec
2
which is 3.32 times the acceleration of gravity.
(d) Theprimaryassumptionmadeisthatdissipationofenergyhasbeenneglected. Thecordhas
been assumed to behave like a linear spring.
(e) In the second jump the slack length is doubled which cuts the stiness in half, and doubles
thedistanceto40feet. Thenaturalfrequencyisreducedto0.897rad/sec. Thelocationof
thenewlowpointAisobtainedbysolvingasimilarquadraticequationwiththenewvalues
of h
0
h
1
=200 feet, h
o
h
B
=240 feet, and =40 feet. The result is
(h
o
h
A
)=240132.7 feet
The physically signicant root is (h
o
h
A
)=373 feet. The low point A is 132.7 feet below
the equilibrium level B, or 373 feet below the upper attachment point.
273
0
50
100
150
200
250
300
350
400
E
l
e
v
a
t
i
o
n

[
f
e
e
t
]

LevelB
LevelA
LevelB1
LevelA1
(f) The level of the equilibrium point B is 240 feet below the attachment point.
(g) The maximum downward acceleration is still 32,2 feet/sec
(h) The maximum upward acceleration is
a
max
=(Max displacement from equilibrium)
o
2
=(132.7)(0.897)
2
=106.8ft/sec
2
Althoughthejumpinvolvesalongerfreefallandagreaterextensionofthecordthemaximum
acceleration does not change!
BungeeJumps
0 1 2 3 4 5
Time[seconds]
Figure 4.2.1: Time Histories of Bungee Jumps
(i) The time histories of the two jumps are shown in Fig. 4.2.1. For plotting purposes the
elevation of the upper attachment point was (arbitrarily) assigned the value 400 feet. The
trajectory of free fall is indicated by the dashed-line parabola. With 100 feet of slack the
jumpers trajectory departs from the free-fall parabola at h=300 ft and begins to oscillate
aboutthenalequilibriumlevelB.Withnodampingthemaximumexcursion(andmaximum
acceleration) is at Level A. With 200 feet of slack the jumpers trajectory departs from the
free-fall parabola at h = 200 ft and begins to oscillate about the nal equilibrium level B1.
With no damping the maximum excursion (and maximum acceleration) is at Level A1.
274
6
4.3 Elevator Model
Considerafree-bodydiagramoftheelevatorofmassmwhenitisacceleratingupwardwithaccel-
eration dv/dt. The equation of motion is
dv
T mg=m
dt
T
T
m
v
T
mg
O
3 ft/sec
1 sec Time
Velocity
Figure 4.3.1: Elevator Dynamics
When the elevator is at rest, or traveling at constant velocity (dv/dt = 0), the tension T in the
cable (and the load on the winch) would be simply the weight mg of the elevator and its cargo.
Designing the winch structure on the basis of this load would be static design. Including the
dynamic eect of accelerating the elevator results in T = mg+mdv/dt. To answer the question
of whether dynamics matters, one must compare the magnitudes of the static load mg and the
dynamic load mg+mdv/dt. The problem statement does not specify the explicit time behavior
of dv/dt. It merely states that a total change of velocity of 3 feet/second is accomplished in one
second. Thesketchaboveshowstwopossiblevelocityhistories. Theslopeofthevelocityhistoryis
theaccelerationdv/dt. Withthedashedhistorytheaccelerationisuniformandhasthemagnitude
dv
=
3ft/sec
=(3)(0.3048)=0.914m/sec
2
dt 1sec
This time history has the smallest maximum acceleration but the sharp corners at the beginning
and end of the one second acceleration are hard on the winch motor and are felt as jerks by
275
the passengers. The time history suggested by the solid curve in the gure produces a smoother
transitionattheexpenseofalargermaximumacceleration. Ifthesystemisdesignedtoaccelerate
accordingtothesolidcurvethemaximumaccelerationcanbe50%to100%largerthantheuniform
acceleration of the dashed history. In the latter case the ratio of the dynamic load to the static
load would be
g+ (dv/dt)
max
9.81+2(0.914)
= = 1.186
g 9.81
Here, dynamics does indeed matter! The dynamic load is 18.6% larger than the static load.
Engineers often make preliminary design calculations to determine ball-park estimates of size,
weight, and power. In these back-of-the-envelope calculations it is common to omit eects with
contributions of the order of ve to ten percent. However, these eects must be included in nal
design calculations, which have to be defended in safety investigations.
276
4.4 Hoisting for Engineers
The solution for this problem is not available.
277
_ _
4.5 Blocked Springs
Let the eective combined stiness of the four unblocked springs be k
u
, and let the eective com-
bined stiness of the four blocked springs be k
b
. In both cases the vehicle weight mg is supported
by the springs. Since only half of the spring is active in the blocked case the static deection

b
= mg/k
b
in the blocked case will be half the static deection
u
= mg/k
u
in the unblocked
case. Sincewearegiventhat
u

b
= 2.5inches,wededucethat
u
= 5.0 inches, and
b
= 2.5
inches.
(a) The eective stinesses are
2500 2500
k
u
= =500 pounds/inch and k
b
= =1000 pounds/inch
5 2.5
in British units. In SI units
4.448) 4.448)
k
u
=500 =87,560 Newton/meter and k
b
=1000 =175,100 Newton/meter
0.0254 0.0254
(b) Theundampednaturalfrequency
o
isgivenby
o
2
=k/mwheremisthemassofthevehicle.
In SI units m=2500(0.4536)=1134 kg. The natural frequencies are
87,560 175,100
(
o
)
u
= = 8.787rad/sec and (
o
)
b
= = 12.437rad/sec
1134 1134
(c) We are given that the unblocked suspension is critically damped (
u
= 1.0). The eective
damping coecient is
b= 2
u
(
o
)
u
m= 2(1.0)(8.787)(1134)=19,930 Newtons/meter/second
(d) Assuming that the shock absorbers are unchanged by the blocking operation, the damping
coecient remains b = 19,930 N/m/s and the new value of for the system with blocked
springs is
b 19,930

b
= = = 0.7071
2(
o
)
b
m 2(12.437)(11340
If y is measured vertically upward from the equilibrium position, and v = dy/dt, the constitutive
equations are
dv
f
m
=m , f
d
=bv, and f
k
=ky
dt
with appropriate subscripts u and b for the unblocked and blocked cases, respectively. and the
force-balance requirement is
f
m
=f(t)f
d
f
k
where f(t) is the upward driving force applied to the vehicle. These model requirements can be
organized into the following state equations for a state determined system with state variables y
and v.
_ _ _ _
_ _ _ _
0 1 0
d
_
y
_

_
y
_ _ _
dt_
v
_
=_

b
_
_
v
_
+

f(t)

_ _
m m m
278
The output equation for the displacement response y(t) is
_ _
_
y
_
y(t) = 1 0
_ _
v
The state equations can be integrated by adapting the scripts pos.m and eqpos.m for the Plate
on Springs problem of Assignment 5 to apply to the present problem. The adaptation of pos.m
involvedinputtingthespecicvaluesofm,b,f
a
,x
o
, and v
o
andeliminatingunecessaryplots. The
adapted version of pos.m is called blockedspringssol.m. It is located at the end of this solution.
TimeHistoryofDisplacement
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Time [Seconds]
Figure 4.5.1: Response of Unblocked Vehicle
0
0.01
0.02
0.03
0.04
0.05
0.06
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

279
1
TimeHistoryofDisplacement
0
0.005
0.01
0.015
0.02
0.025
0.03
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9


Time [Seconds]
Figure 4.5.2: Response of Blocked Vehicle
280
1
TimeHistoryofDisplacement
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
0
0.01
0.02
0.03
0.04
0.05
0.06
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

UnblockedVehicle
VehiclewithBlocks
Time [Seconds]
Figure 4.5.3: Comparison of Responses
The adaptation of eqpos.m involved only the name change to Eqblockedspringssol.m.
Using these scripts with the stiness inputs k
u
= 87560 N/m and k
b
= 175100 N/m, and an
integration time of 1.0 second produced the response curve shown in Fig. 4.5.1 for the unblocked
vehicle, and the response curve shown in Fig. 4.5.2 for the blocked vehicle. The two responses
are compared in Fig. 4.5.3. Note the quicker response to a smaller deection for the blocked
vehicle.
blockedspringssol.m
Download
% Produces plots of
% (i) position vs. time
% for the response of vehicle, with mass m, stiffness k,
% and damping parameter b, when the vehicle starts from initial conditions
% y = 0 and v = 0 under the action of a suddenly applied force
% fa = 1000 pounds ( 4448 Newtons) at t = 0.
clear variables
global m k b fa
281
% Input parameters
m = 1134;
k = input(Enter the stiffness "k" in Newtons/meter );
b = 19930;
fa= 4448;
% Input initial conditions.
y0= 0;
v0= 0;
tspan = input(Enter the duration "T" of the desired time history, in seconds );
X0 = [ y0 ; v0 ];
% Integrate equations of motion
[t,X] = ode45(Eqblockedspringssol, tspan, X0);
% Plot results
plot(t,X(:,1)), title(Time History of Displacement),
xlabel(Time [Seconds]), ylabel(Displacement [meters]), pause
Eqblockedspringssol.m
Download
% Eqblockedspringssol.m Adapted from eqpos.m. Provides equation of motion
% for vehicle with unblocked or blocked springs
% to be integrated by script blockedsprings_sol.m
function Xdot = Eqblockedspringssol(t,X)
global m k b fa
Xdot = [ 0 1 ; -k/m -b/m ]*X + [ 0 ; fa/m ];
282
4.6 Engine Vibration
The solution for this problem is not available.
283
4.7 Garage Door
k
m/2

g
x
y
T
T
T
mg/2
Figure 4.7.1: Garage Door support System
To model the pulley system, we assume that the inertia of the pulleys and cable is negligible in
comparison with the mass of the garage door. Furthermore we assume that the cable is exible
(it bends easily) but stretches very little in comparison with the extension of the spring. Let x
be the horizontal displacement of the pulley attached to the end of the spring, and let y be the
verticaldisplacementofthedoor,bothmeasuredfromtheequilibriumpositionofthedoorhanging
freelyundertheinuenceofgravity. Intheequilibriumposition,thetensionT
o
inthecablesmust
equaltheweightmg ofthedoor. Thetotalforceinthespringis2T
o
whichhasproducedaninitial
extension of the springs, where =2T
o
/k= 2mg/k.
In a small oscillation about the equilibrium position, the displacements x(t) and y(t) oscillate
about zero and the tension in the cable oscillates about the initial tension T
o
= mg. When the
cableremainstaut,thedisplacementsxandyarelinkedbythegeometriccompatibilityrequirement
y = 2x. This is the crucial insight for this system, and it is not immediately obvious to all. One
waytoseeitistoimaginethatthedoorandthecablesaretemporarilyfrozeninpositionwhilethe
pulleyattachedtothespringisdisplacedadistancextotheleft. Anemptyloopofcableappears,
totherightofthepulley,withastraightlengthxintheuppercableandastraightlengthxinthe
lowercable. Nowwiththepulleyxed,graduallylowerthedoortotakeuptheslackintheempty
loop. It will be necessary to lower the door a distance y= 2x to make the cable taut again.
(a) To formulate a model to analyze the door oscillations we need to assemble the geometric
compatibility requirements, the constitutive equations, and the force balance requirements.
284

The geometric compatibility requirement is


y= 2x
The constitutive equations for the door and the spring are
md
2
y
f
m
= and 2T =k(+x) = 2mg+kx
2 dt
2
and the force balance requirement is
f
m
=mgT
These relations constitute a mathematical model of the garage door system. They can be
organized into a single dierential equation for either x or y by inserting the constitutive
equationsforf
m
andT intheforcebalanceequationandusingthecompatibilityrequirement
to eliminate x in favor of y, or vice versa. The result is either
d
2
y k d
2
x k
m + y= 0 or m + x= 0
dt
2
2 dt
2
2
Notethatbecausetheoriginsforthedisplacementsxandywereestablishedattheequilibrium
position, the gravity force has canceled out of the equations of motion.
(b) The undamped natural frequency
o
for the preceding equations is given by

2
k kg
= =
o
2m 2W
where W is the weight of the door and g = 386 in/sec
2
is the acceleration of gravity. The
frequency of oscillation f
o
in Hz is

o
1 (5)(386)
f
o
= = = 0.350 Hz
2 2 2(200)
(c) It is assumed that the door only moves vertically in the small oscillation. The inertia of the
pulleys,cable,andspringisneglected. Theelasticityofthecableintensionisneglected,and
it is assumed that the tension T is uniform throughout the length of the cable.
285
4.8 Rotating Damped
A rotor with moment of inertia I and angular velocity is subjected to an accelerating torque T
and two retarding friction torques, each modeled linearly as B.
(a) The equation of motion is
d d 2B 1
I + 2B=T or = + T (1)
dt dt I I
(b) The steady-state angular velocity, when T = 10 Newton-meters, and I = 0.001 kg-m
2
, and
B = 0.005 N-m/r/s is
T 10

ss
= = =1000 r/s
2B 2(0.005)
(c) Consider the input T =T
a
sint to be the imaginary part of the complex input T
a
exp(it)
andlookforacomplexsolutionoftheform=Aexp(it). Aftersubstitutionin(2)wend
that this is indeed a solution , if
T
a
T
a
1
A= = _ exp(i)
2B+iI 2B
1 + (
I
)
2
2B
where the phase angle is xed by the relation
I
tan=
2B
Thesteady-statesolutiontotheinputT =T
a
sintisthentheimaginarypartofthecomplex
solution Aexp(it), which is
T
a
1
= _ sin(t+)
2B
1 + (
I
)
2
2B
The magnitude of the response amplitude is
T
a
1
M()= _
2B
1 + (
I
)
2
2B
and the phase angle is
=tan
1
I
2B
ForthecasewhereT
a
= 10 Newton-meters, the values of M()and()forthefrequencies
=50 rad/sec, =5 rad/sec, and =0.5 rad/sec are
(i) M(50) =
10 1
= 196.1 r/s,
0.01 (0.001)(50)
1+(
0.01
)
2
(50)=tan
1
(0.001)(50)/0.01=1.373r= 78.7 deg
10 1
= 894 r/s, (ii) M(5) =
0.01 (0.001)(5)
)
2
1+(
0.01
(5)=tan
1
(0.001)(5)/0.01=0.463r= 25.6 deg
286
(iii) M(0.5) =
10 1
= 999 r/s,
0.01
1+(
(0.001)(0.5)
)
2
0.01
(0.5)=tan
1
(0.001)(0.5)/0.01=0.0500r= 2.86deg
(d) The limiting value of the magnitude of the response amplitude M() when 0 is A
o
=
T
a
/2B =1000r/s. TheratioM()/A
o
is0.1961when=0.5rad/sec,0.894when=0.05
rad/sec, and 0.999 when =0.005 rad/sec.
MagnitudeofResponseAmplitude
10
-3
10
-2
10
-1
10
0
10
1
Frequency [radians/second]
Figure 4.8.1: Magnitude of Response of Rotating Inertia
10
0
10
1
10
2
10
3
M
a
g
n
i
t
u
d
e

o
f

R
e
s
p
o
n
s
e

A
m
p
l
i
t
u
d
e

[
r
a
d
i
a
n
s
/
s
e
c
o
n
d
]

287
PhaseAngleofResponse
-90
-80
-70
-60
-50
-40
-30
-20
-10
0
P
h
a
s
e

a
n
g
l
e

[
d
e
g
r
e
e
s
]

10
-3
10
-2
10
-1
10
0
10
1
Frequency [radians/second]
Figure 4.8.2: Phase Angle of Response of Rotating Inertia
(e) The low-frequency asymptote for M() is A
o
=T
a
/2B, while the high-frequency asymptote
is T
a
/I. These asymptotes intersect at the break frequency
break
= 2B/I=10 rad/sec.
(e) The Bode plots shown in Fig. 4.8.1 and Fig. 4.8.2 were obtaining by running a MATLAB
script similar to that used in Problem 4.21.
288
_
_
4.9 Car Suspension 1
The dierential equation for the system is
mx + cx + kx=cx
0
+kx
0
, (1)
where x
0
stands for the ground waveform input. For step input, the particular solution will be:
x
p
(t) = 1. (2)
The homogenous solution has the form:
x
h
(t) = Ae
nt
sin(
d
t) + Be
nt
cos(
d
t) (3)
To nd the initial velocity, integrate equation 1 from 0

to 0
+
,
m( x(0
+
)x (0

))+c(x(0
+
)x(0

))+k0 = cx
0
(0
+
)x
0
(0

) (4)
mx (0
+
) = cx
0
(0
+
) (5)
c c
x (0
+
) = x
0
(0
+
) = (6)
m m
Plugging the initial conditions into the full solution
x(t) = Ae
nt
sin(
d
t) + Be
nt
cos(
d
t) + 1 (7)
x(0
+
) = 0 = B+ 1 B =1 (8)
x (0
+
) =
c
=A
d

n
B A= (
c

n
) (9)
m m
d

d
For m=500kg, k= 5 10
4
N/m and c= 2 10
3
Ns/m,
k

n
= =10 (10)
m
c
= = 0.2 (11)
2 km

d
=
n
1
2
= 9.8 (12)
=
n
=2; (13)
A= 0.2 (14)
x(t) = 0.2e
2t
sin(9.8t)e
2t
cos(9.8t) + 1 (15)
The result is plotted in Figure 4.9.1. The passengers will not like this ride because there are too

many oscillations. when c= 2 km=10000, the system will be critically damped, in this case
the complete solution has the form:
x(t) = 1 + Ae
nt
+Bte
nt
(16)
x(0
+
) = 0 = A+ 1 A=1 (17)
c c
x (0
+
) = =
n
A+B B =
n
=10 (18)
m m
x(t) = 1 e
10t
+ 10te
10t
(19)
289
In order to nd the acceleration for the under-damped case, reformulate equation 16,
x(t) = 0.2e
2t
sin(9.8t)e
2t
cos(9.8t) + 1 (20)
=Re11.02e
2t+j(9.8t+0.197)
(21)
x(t) = Re1.02(2 + 9.8j)
2
e
2t+j(9.8t+0.197)
(22)
x(t)=102e
2t
cos(9.8t+ 0.6) (23)
The maximum acceleration could happen at t=0+ and t=(0.6)/9.8 = 0.26sec,
x(t= 0+) = 102cos(0.6) = 84 (24)
x(t= 0.26)=102e
20.26
=60 (25)
So the maximum acceleration is 84m/sec
2
.
Similarly, for the critically damped case, dierentiating equation 25 twice,
x(t) = 300e
10t
+1000te
10t
(26)
x(0+)=300 (27)
Sothepassengerwillhavebiggeraccelerationinthecriticallydampedcase. Fromthestepresponse,
the 5% settling time is 1.35sec in the under-damped case and 0.4sec in thecritically-damped case.
0 0.5 1 1.5 2 2.5 3
Time[sec]
Figure 4.9.1: Step Response
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
D
i
s
p
l
a
c
e
m
e
n
t

[
m
]

UnderdampedStepResponse
CrticallyDampedStepResponse
290
4.10 Disk Drive
J

b
=b
Figure 4.10.1: Free body diagram of disk drive spindle
(a) Figure 4.10.1 shows the free body diagram for the spindle. Note: that all constitutive equa-
tions are included with the diagram.
(b)
J =
m

b
J =
m
b
b
m
+ =
J J
(c) When the disk spins down
m
=0 therefore our equation of motion becomes
b
+ = 0
J
Weknowthatthesolutionforthismusttaketheform(t) = Ae
st
=Ae
t/
. If we substitute
into the equation of motion we get
b
Ase
st
+A e
st
= 0
J
b
s =
J
J J
= b=
b
0.001 kg m
2
N m s
b = = 0.002
0.5 s rad
291
0
50
100
150
200
250
S
p
e
e
d

(
r
/
s
)
(, 157.5)
Please note that angular velocity must always be in rad/s (the unit label rad=1) unless
otherwise noted.
vs time
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
time (s)
Figure 4.10.2: Plot vs time
(d) To determine (t), we need to determine the nal velocity of the disk. When = 0 the
characteristic equation become

b=
m

f
=
m
=250r/s
b
Since the initial velocity of the system is 0, we know the solution must be of the form
(t) =
f
(1e
t/
)=250(1e
2t
)
Figure 4.10.2 shows the (t)
292
vs time
t
1
-1 t
1
t
1
+1 t
1
+2 t
1
+3 t
1
+4
t
1
+5
time (s)
Figure 4.10.3: Plot vs time
(e) (t) for t > t
1
is
(t) =
f
e
(tt
1
)/
=250e
2(tt
1
)
Figure 4.10.2 shows the time response of the disk.
250
200
S
p
e
e
d

(
r
/
s
)
150
100
(t
1
+, 92.5)
50
0
293
4.11 Crashworthiness
The sketch in Fig. 4.11.1 shows the linear model used to study crashworthiness. The mass m of
thevehicleis(2000pounds)(0.4536kg/pound)=907.2kgandthevehiclespeedatimpactv
o
is(10
mi/hr)(5280ft/mi)(0.3048meters/ft)/(3600secs/hr)=4.470meters/sec. Whenthereisnoenergy
dissipation (b=0) the displacement of the mass goes through a half-cycle of undamped vibration
of the form
x(t) = Asin
o
t
The vehicle then retreats from the barrier with velocity v
o
. The amplitude A of the oscillation
is given as 6 inches which is (6 in)(0.0254 meters/in) = 0.1524 meters, but the undamped natural
frquency
o
is unknown.
m
k
b
v
o
Figure 4.11.1: Model of Vehicle Impacting a Barrier
One way to calculate
o
is to set the velocity obtained by dierentiating x(t) equal to v
o
at t= 0.
This leads to
v
o
4.470

o
= = = 29.33 rad/sec
A 0.1524
Another way to obtain
o
is to use conservation of energy,
KE=
1
mv
2
=
1
kA
2
2
o
2
to get k, and then use
o
2
=k/mto evaluate
o
.
(a) If the value of
o
is already known, then the eective stiness k is obtained from
k=m
o
2
=(907.2)(29.33)
2
=780,400 N/m
If
o
isnotavailable,thenthestinesscanbeobtaineddirectlyfromconservationofenergy.
_
v
o
_
2
_
4.470
_
2
k=m =907.2 =780,500 N/m
A 0.1524
(b) During the half-cycle of undamped sinusoidal motion x(t) = Asin
o
t the acceleration is
d
2
x/dt
2
=A
o
2
sin
o
t. Thepeakdecelerationoccursattheinstantofpeakdeectionofthe
spring and has the magnitude
Decel
peak
=A
2
= (0,1524)(29.33)
2
=131.1 meters/sec
2
o
294
The geometric compatibility requirements are satised by taking x to represent the displacement
ofthemassandthedeformationofthespring,andv=dx/dttorepresentthevelocityofthemass
and the relative velocity across the damper. The constitutive equations are
dv
f
m
=m f
d
=bv and f
k
=kx
dt
and the force-balance requirement is
f
m
=f
d
f
k
These model requirements can be used to construct the state equations for the state variables x
and v
_ _
_ _ _ _
0 1
d
_
x
_

_
x
_
=_
k b
_
dt_ _ _ _
v

v
m m
Thedesiredoutputsare(i)thedeectiondisp=xofthefender,(ii)thedecelerationofthevehicle
decel = (k/m)x+ (b/m)v), and (iii) the total force on the barrier f
barr
= kx+bv. In matrix
notatio
_ _ _ _
disp 1 0 _ _


_ _ _
x
_
decel =

k/m b/m

_ _

_ _
v

_ _
f
barr
k b
Thestateequationscanbeintegratedbyadaptingthescriptspos.mandeqpos.m. Theadapta-
tion of pos.m involved inputting the specic values of m, k, x
o
, and v
o
and letting the program
ask for the desired value of instead of the desired value of b. The adapted version of pos.m is
called crashworthinesssol.m and is found at the end of this solution.
295
TimeHistoryofDisplacement
-0.02
0
0.02
0.04
0.06
0.08
0.1
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
Time [Seconds]
Figure 4.11.2: Displacement Response for = 0.5
The script for the equation called for by crashworthinesssol.m is called Eqcrashworthinesssol.m.
It is similar toEqblockedspringssol.m except that here there is no applied force f
a
.
296
-40
-20
0
20
40
60
80
100
120
140
D
e
c
e
l
e
r
a
t
i
o
n

[
m
e
t
e
r
s
/
s
e
c

s
q
d
]

TimeHistoryofDeceleration
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
Time [Seconds]
Figure 4.11.3: Deceleration Response for = 0.5
Typical plots produced by crashworthiness.m are shown in Figures 4.11.2-4.11.4, which were ob-
tained for = 0.5.
297
-2
0
2
4
6
8
10
12
F
o
r
c
e

[
N
e
w
t
o
n
s
]

x10
4
TimeHistoryofForceonBarrier
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
Time [Seconds]
Figure 4.11.4: Barrier Force Response for = 0.5
(g) The plots obtained are only valid during the interval in which the fender remains in contact
withthebarrier. Intheundampedcase,thisperiodwasahalfcycleofsinusoidaloscillation.
Inthedampedcases,themodelresultsarevalidonlyaslongasthebarrierforceiscompressive.
If tension is required the fender separates from the barrier. For the case = 0.5 we see that
the model predicts tensile barrier force after t = 0.082 seconds. In all four cases there is
separation of the vehicle from the barrier after the impact.
(h) Peakdecelerationsareproportionaltopeakbarrierforces(thisisaconsequenceofNewtons
law F = ma). In the undamped case, the barrier force is delivered by the spring. Its
magnitude is f
barr,u
=mA
o
2
=mv
o

o
. In the damped case, the barrier force is delivered by
both the spring and the damper. At the initial instant of impact, when the fender begins to
movewithvelocityv
o
,buthasntyetdeectedthespring,theentirebarrierforceisdelivered
by the damper. Its magnitude is f
barr,d
= bv
o
which can be made as large as desired by
increasingtheparameterb. Thisinitialdampingforcewillbethesameasthepeakundamped
force if bv
o
=mv
o

o
. This occurs when = 0.5.
(i) The peak deceleration is greatest in Case (f), = 1.0.
298
(j) Bytrialanderrorwithcrashworthinesssol.monendsthatdecelhasafairlyatminimum
value of slightly more than 106 meters/second
2
over the range 0.25 < < 0.30.
crashworthinesssol.m
Download
% crashworthinesssol.m MATLAB Script adapted from pos.m which was
% MATLAB script for Problem 3 in 2.003 Assignment 5. Produces plots of
% (i) displacement vs. time
% (ii) deceleration vs. time
% (iii) force on barrier vs. time
% for the response of vehicle, with mass m= 907.2 kg,
% stiffness k = 780500 N/m, and damping ratio zeta,
% when the vehicle impacts barrier with initial conditions
% x = 0 and v = 4.470 m/s at t = 0.
clear variables
global m k b
% Input parameters
m = 907.2;
k = 780500;
zeta = input(Enter the damping ratio, zeta );
b = 2*zeta*29.33*907.2;
% Input initial conditions.
x0= 0;
v0= 4.470;
tspan = input(Enter the duration "T" of the desired time history, in seconds );
X0 = [ x0 ; v0 ];
% Integrate equations of motion
[t,X] = ode45(Eqcrashworthinesssol.m, tspan, X0);
% Obtain outputs
disp = X(:,1);
decel = (k/m)*X(:,1) + (b/m)*X(:,2);
fbarr = m*decel;
% Plot results
plot(t,disp), title(Time History of Displacement), grid,
xlabel(Time [Seconds]), ylabel(Displacement [meters]), pause
plot(t,decel), title(Time History of Deceleration), grid,
xlabel(Time [Seconds]), ylabel(Deceleration [meters/sec sqd]),pause
plot(t, fbarr), title(Time History of Force on Barrier), grid
xlabel(Time [Seconds]), ylabel(Force [Newtons]),
Eqcrashworthinesssol.m
Download
% Eqcrashworthinesssol.m Adapted from eqpos.m. Provides equation of motion
% for vehicle impacting barrier
% to be integrated by script crashworthinesssol.m
function Xdot = Eqcrashworthinesssol(t,X)
global m k b
Xdot = [ 0 1 ; -k/m -b/m ]*X ;
299
4.12 Nonlinear Rotational System
The solution for this problem is not available.
300
4.13 Developing Dierential Equations
We are interested in modeling the rotation system in Figure 4.13.1 and the translational system
in Figure 4.13.2. The free-body diagram for the rotational system is given in Figure 4.13.3. The
free-body diagram for the translation system is given in Figure 4.13.4.
C C
J
, output , input
Figure 4.13.1: SDOF Rotational System
C C
M
X, output
F, input
K K
Figure 4.13.2: SDOF Translational System
301
Figure 4.13.3: Free-body diagram for the Rotational System
Figure 4.13.4: Free-body diagram for the Translational System
302
Using the free-body diagram, the equation of motion for the rotational system is
J

+ (c
1
+c
2
)

=
The rotational system is a rst-order system because we can rewrite the governing equation using
the velocity (=

)such that there is only a single derivative.


J + (c
1
+c
2
)=
The governing equation for the translational system is
Mx + (c
1
+c
2
) x+ (k
1
+k
2
)x=F
This system is second-order.
303



4.14 Mass Spring System Frequency Response
(a) The dierential equation for this system is
mx + kx=F
Take the Laplace transform and rearrange to obtain the transfer function
X(s)(ms
2
+k) = F(s)
X(s)
=
1
F(s) ms
2
+k
M is simply the magnitude of the transfer function evaluated at s=j.

X s

( ) 1
M = =
F(s) [100
2
[
The phase is
0 0
= tan
1
tan
1
1 100
2
= 0

, <10 and 180

, >10
(d) The plot is shown in Figure 4.14.1.
Bode Diagram
0 1 2
10 10 10
Frequency (rad/sec)
Figure 4.14.1: Bode plot
304
M
a
g
n
i
t
u
d
e

(
d
B
)

-100
-50
0
50
100
150
P
h
a
s
e

(
d
e
g
)

-180
-135
-90
-45
0
4.15 Modeling Practice
The solution for this problem is not available.
305
4.16 Small Motion Transfer Function
The solution for this problem is not available.
306



4.17 Mass Spring Damper System Frequency Response 1
(a) The dierential equation for this system is
mx + bx + kx=F
Take the Laplace transform and rearrange to obtain the transfer function
X(s)(ms
2
+bs+k) = F(s)
X(s)
=
1
F(s) ms
2
+bs+k
M is simply the magnitude of the transfer function evaluated at s=j.

X(s)

1
M = = _
F(s)
(100
2
)
2
+
2
The phase is
0
= tan
1
tan
1
1 100
2
=tan
1

100
2
(d) The plot is shown in Figure 4.17.1.
307
Bode Diagram
M
a
g
n
i
t
u
d
e

(
d
B
)

-80
-70
-60
-50
-40
-30
-20
-10
P
h
a
s
e

(
d
e
g
)

-180
-135
-90
-45
0
0 1 2
10 10 10
Frequency (rad/sec)
Figure 4.17.1: Bode plot
308
_ _
4.18 Mass Spring Damper System Frequency Response 2
(a) Begin by writing out two force balance equations, one for each mass. This will yield two
dierential equations involving x
1
, x
2
and F. These simplied equations are
m
1
x
1
+ (b
1
+b
2
) x
1
+ (k
1
+k
2
)x
1
=b
2
x
2
+k
2
x
2
+F
m
2
x
2
+b
2
x
2
+k
2
x
2
=b
2
x
1
+k
2
x
1
Next take the Laplace transform of each equation
X
1
(s) m
1
s
2
+ (b
1
+b
2
)s+ (k
1
+k
2
) = X
2
(s) (b
2
s+k
2
) + F(s)
X
2
(s)
_
m
2
s
2
+b
2
s+k
2
_
=X
1
(s) (b
2
s+k
2
)
Toobtainthedesiredtransferfunctions,simplysubstituteoneequationintotheother. After
a lot of messy algebra, this yeilds
X
1
(s)
=
m
2
s
2
+b
2
s+k
2
F(s) m
1
m
2
s
4
+ (m
2
(b
1
+b
2
) + m
1
b
2
)s
3
+ (m
1
k
2
+m
2
(k
1
+k
2
) + b
1
b
2
)s
2
+ (b
1
k
2
+b
2
k
1
)s+k
1
k
2
X
2
(s)
=
b
2
s+k
2
F(s) m
1
m
2
s
4
+ (m
2
(b
1
+b
2
) + m
1
b
2
)s
3
+ (m
1
k
2
+m
2
(k
1
+k
2
) + b
1
b
2
)s
2
+ (b
1
k
2
+b
2
k
1
)s+k
1
k
2
(b) The bode plots and pole/zero plots are shown in Figures 4.18.1-4.18.4.
Bode Diagram
-1 0 1 2 3
10 10 10 10 10
Frequency (rad/sec)
Figure 4.18.1: Bode plot for H
1
(s)
M
a
g
n
i
t
u
d
e

(
d
B
)

-150
-100
-50
0
P
h
a
s
e

(
d
e
g
)

-180
-135
-90
-45
0
309
-100
-50
0
50
100
I
m
a
g
i
n
a
r
y

A
x
i
s

-0.5 -0.4 -0.3 -0.2 -0.1 0
Real Axis
Figure 4.18.2: Pole/zero plot for H
1
(s)
Bode Diagram
-1 0 1 2 3 4 5
10 10 10 10 10 10 10
Frequency (rad/sec)
Figure 4.18.3: Bode plot for H
2
(s)
M
a
g
n
i
t
u
d
e

(
d
B
)

-350
-300
-250
-200
-150
-100
-50
0
P
h
a
s
e

(
d
e
g
)

-360
-315
-270
-225
-180
-135
-90
-45
0
310
-100
-50
0
50
100
I
m
a
g
i
n
a
r
y

A
x
i
s

-10000 -9000 -8000 -7000 -6000 -5000 -4000 -3000 -2000 -1000 0
Real Axis
Figure 4.18.4: Pole/zero plot for H
1
(s)
For both bode plots, peaks appear at the frequencies of the pole locations, that is, the
imaginaryvalueofeachpole. Theimaginaryvalueofapoleissimpletheundampednatural
frequency, and it makes sense that peaks would occur at damped natural frequencies. m
1
is
at a relative maximum at about 1.96 [rad/sec]. Both masses oscillate with a magnitude of
-5.85[dB].m
1
isatarelativeminimumatabout100[rad/sec],oscillatingat-120[dB].m
2
is
oscillating at -80[dB], which is 100 times the amplitude of m
1
.
311
4.19 Propeller Shaft Vibration
Torsional vibration of coupler shaft between engine and propellor. Moment of inertia of propellor
isestimatedasI =mL
2
/12,wheremisthemassofanaluminumrod,2inches(0.0508meters)in
diameter and six feet (1.829 meters) in length. The density of aluminum is 2720 kg/m
3
, so
m=r
2
L=2720(0.0254)
2
(1.829)=10.08 kg
and
I =
mL
2
=
(10.08)(1.829)
2
= 2.811 kg-m
2
12 12
K

e
,
e

p
,
p
Figure 4.19.1: Torsional Vibration of Propellor Shaft
(a) The input is the given angular displacement of the engine end of the shaft

e
= t+sin t
2
Take the output to be the angular displacement
p
and angular velocity
p
= d
p
/dt of
the propellor end of the shaft. Assume that the that the torque T
shaft
transmitted to the
propellor from the shaft and the torque T
fric
applied to the propellor by the air resistance
can be modelled by the linear relations
T
shaft
=K(
p

e
) and T
fric
=B
p
The torque balance requirement is
d
p
T
shaft
T
fric
=T
I
=I
dt
312
_
These relations can be combined in the standard form for a state-determined system
_ _ _ _ _ _ _ _
d
_

p
_
0 1
_

p
_ _
0
_
K
_
dt_

p
_
=
_
K/I B/I
_

p
_
+
_
1
_ I
(t+sin
2
t)
or in the form of a single second-order equation
d
2

p
d
p

I +B +K
p
=K(t+sin t)
dt
2
dt 2
Theinputmotionconsistsofauniformrotationatangularrateplusasinusoidaloscillation
atfrequency/2. Becausetheequationislinearthesteady-stateresponsestothesetwoinputs
canbeobtainedseparatelyandsuperposedtoobtainthecompletesolution. Fortheuniform
rotation, try a response of the form

p,unif
= t+Constant
and nd that it satises
d
2

p,unif
d
p,unif
I +B +K
p,unif
=Kt
dt
2
dt
if the Constant = B/K. this means that, after an initial transient, uniform engine speed
results in a uniform propellor speed with a constant twist in the shaft just large enough for
the shaft torque to balance the friction torque. For the sinusoidal oscillation, the response

p,osc
=(t) must satisfy the equation
d
2
d
I +B +K=Ksin t
dt
2
dt 2
A solution is sought in the form

= ImA exp(i t)
2
The complex amplitude A must satisfy the relation
A
K

2
=
I
=
o

2
+i
B

2
+i
o

I 4 2I
o
4
where

2
=
K
and 2
o
=
B
o
I I
This result is very similar to the result obtained in Problem 16.9, with the exception that
heretheringfrequency/2playstheroleofthedrivingfrequencyinProblem16.9. The
ratio of the complex amplitude at frequency /2 to the response when 0 is
A(/2) 1
Z = =
A(0)
[1
(

/
2
2)
2
] + i2
(

/
o
2)
o
The peak value of [Z[ is
1
[Z[
peak
= _
2 1
2
and that the peak amplitude is realized when the forcing frequency is

peak
=
o
12
2
313
(b) To use the results of Problem 16.9, we need the ratio, [Z[
peak
, of the peak response to the
zero-frequencyresponseandthemagnitudeoftheforcingfrequencywhichproducesthepeak
response. Here we know that peak response occurs when the engine speed is 2200 rpm. The
corresponding forcing frequency is /2 = 1100 rpm or 115.2 rad/sec. We do not know the
zero-frequencyresponse,butwedoknowthattheratioofthepeakresponse,at2200rpm,to
the response at 500 rpm is 4. We also know from Fig.2 that the response magnitude for low
frequencies, well below the natural frequency, is not much greater than the zero-frequency
response. Therefor we can take the given ratio of 4 as an estimate of the ratio [Z[
peak
. With
theseestimates,wecanusetheresultsofProblem16.9toobtainestimatesofthebehavioral
parameters
o
and .

o
= _

peak
= _
115.2
and [Z[
peak
= 4 = _
1
12
2
12
2
2 1
2
The solution of these equations yields the estimates
= 0.1260 and
o
=117.1 rad/sec
from which follow
K =I
o
2
= 2.811(117.1)
2
= 38, 500 N-m/radian
and
B = 2
o
I = 2(0.126)(117.1)(2.811)=83.0 N-m/rad/sec
as estimates of the model parameters.
314
_
_
4.20 Safe Packaging
Thersttwoguresforthisproblemarecompletelyirrelevanttotheactualsolution,theyareonly
there to show the thought process you would go through if you were solving this problem on your
own.
f
g
m1
x
f
b
f
k
Figure 4.20.1: Free Body Diagram of instrument.
(a) From the FDB,
F = m
1
x = f
g
f
k
f
b
Where f
g
= m
1
g
f
k
= kx
f
b
= bx
m
1
x = m
1
gbx kx

b k
(x) + x + x = g
m
1
m
1
with x
0
= 0, x
0
=v
0
(b) For the system to oscillate must be less than 1.

2
k
=
n
m
b
2
n
=
m
1 b m b
= =
2m k
2 km
< 1

b < 2 km
(c) Since this is a pretty straight forward 2nd order system we can just add the forced and free
315
responses
x(t) =
x
forced
(t) =
x
free
(t) + x
forced
(t)
_ _
g
k
1
_
1
2
e
nt
sin(t
n
_
1
2
+) + 1
x
free
(t) =
v
0

n
_
1
2
e
nt
sin(t
n
_
1
2
)
316
4.21 Sliding Damped
Amassmissubjectedtoanacceleratingforcef andaretardingfrictionforce,modeledlinearlyas
bv.
(a) The equation of motion is
dv dv b 1
m +bv=f or = v+ f (1)
dt dt m m
(b) The steady-state velocity, when f = 10 Newtons, and m=1000 kg, and b=100 N/m/s is
f 10
v
ss
= = = 0.10 m/s
b 100
(c) Consider the input f = f
a
sint to be the imaginary part of the complex input f
a
exp(it)
andlookforacomplexsolutionoftheformv=Aexp(it). Aftersubstitutionin(1)wend
that this is indeed a solution , if
f
a
f
a
1
A= = _ exp(i)
b+im b
1 + (
m
b

)
2
where the phase angle is xed by the relation
m
tan=
b
Thesteady-statesolutiontotheinputf =f
a
sintisthentheimaginarypartofthecomplex
solution Aexp(it), which is
f
a
1
v= _ sin(t+)
b
1 + (
m
b

)
2
The magnitude of the response amplitude is
f
a
1
M()= _
b
1 + (
m
b

)
2
and the phase angle is
=tan
1
m
b
Forthecasewheref
a
= 10 Newtons, the values of M()and()forthefrequencies=0.5
rad/sec, =0.05 rad/sec, and =0.005 rad/sec are
(i) M(0.5) =
10 1
= 0.01961 m/s,
100
1+(
(1000)(0.5)
)
2
100
(0.5)=tan
1
(1000)(0.5)/100=1.373r= 78.7 deg
(ii) M(0.05) =
10 1
= 0.0894 m/s,
100 (1000)(0.05)
)
2
1+(
100
(0.05)=tan
1
(1000)(0.05)/100=0.463r= 25.6 deg
(iii) M(0.005) =
10 1
= 0.0999 m/s,
100 (1000)(0.005)
1+(
100
)
2
(0.005)=tan
1
(1000)(0.005)/100=0.050r= 2.86deg
317
(d) The limiting value of the magnitude of the response amplitude M() when 0 is A
o
=
f
a
/b= 0.10 m/s. The ratio M()/A
o
is 0.1961 when =0.5 rad/sec, 0.894 when =0.05
rad/sec, and 0.999 when =0.005 rad/sec.
(e) The low-frequency asymptote for M() is A
o
=f
a
/b, while the high-frequency asymptote is
f
a
/m. These asymptotes intersect at the break frequency
break
=b/m= 0.10 rad/sec.
10
-3
10
-2
10
-1
10
0
10
1
Frequency [radians/second]
Figure 4.21.1: Magnitude of Response of Sliding Mass
10
-3
10
-2
10
-1
MagnitudeofResponseAmplitude
M
a
g
n
i
t
u
d
e

o
f

R
e
s
p
o
n
s
e

A
m
p
l
i
t
u
d
e

[
m
e
t
e
r
s
/
s
e
c
o
n
d
]

318
PhaseAngleofResponse
-90
-80
-70
-60
-50
-40
-30
-20
-10
0
P
h
a
s
e

a
n
g
l
e

[
d
e
g
r
e
e
s
]

10
-3
10
-2
10
-1
10
0
10
1
Frequency [radians/second]
Figure 4.21.2: Phase Angle of Response of Sliding Mass
(f) TheBodeplotsshowninFig. 4.21.1andFig. 4.21.2wereobtainedbyrunningtheMATLAB
script slidingsol.m
slidingsol.m
Download
%Script for plotting Bode plots for Problem Sliding Damped
Om = logspace(-3, 1, 200);
unity = ones(1,length(Om));
M = 0.1 ./ sqrt( unity + ( 10 .* Om).^2);
loglog(Om, M), title(Magnitude of Response Amplitude),
xlabel(Frequency [radians/second]),
ylabel(Magnitude of Response Amplitude [meters/second] ),
pause
phi = -57.3*atan(10 * Om);
semilogx(Om, phi), title(Phase Angle of Response),
xlabel(Frequency [radians/second]),
ylabel(Phase angle [degrees])
319
4.22 Toy Flywheel
r
J
F
k
T
c
T
k
Figure 4.22.1: Cross-Sectional and Top view
a) Figure 4.22.1 shows the free body diagram for this system. Summing the torques
T = J

=T
k
Tc
T
k
= rF
k
F
k
= k(x
s
x)
T
c
= c
t

so J

+c
t

= rk(x
s
x)
J

+c
f

+rkx = rkx
s
x
Recall that r = x =
r
so
J
r
x +
c
r
t
x + rkx=rkx
s
(t)
This may be rewritten as
x +
c
t
x +
r
2
k
x=
r
2
k
x
s
(t)
J J J
(b) For this part we note the following relationships
2
n

c
t
J
2
k

r
n
J
For the string not to go slack the system must not overshoot x
0
, this means that 1. For
1 to be true

c
t
2 r
2
kJ
320
_
_ _
(c) For the string to go slack the system must be allowed to overshoot or < 1. In the case of
< 1
1
x(t) = 1 _ e
nt
cos(
d
t )
1
2

d
=
n
1
2
= tan
1
_

1
2
This equation can be expressed in terms of the system variables by substituting for
n
and
from 3. Figure 4.22.2 shows the time response for this system where t

indicates the time


after which the response is not valid.
0
x
o
Valid for t<t*
0
t*
Figure 4.22.2: Time Response
(d) When c
t
=0, the expression for x(t) becomes
x(t) = x
0
(1cos
n
t)
As we saw in part 3, t

is the time at which the response passes x


0
. In this case, it takes a
quarter cycle to get to x
0
, thus
1 2

t =
4
n
Figure 4.22.3 shows the response for this system.
321
t*=0.25*2* n
Figure 4.22.3: Step Response
322
4.23 Truck Shocks
The basic assumption here is that the back of the truck can be modeled as a plate of mass M
supported by an equivalent spring of stiness k and damping parameter b, onto which hops Uncle
Massive, of mass m, which applies a step-function of force to the system, while changing the total
massofthesystemtoM+m. Undertheseassumptionsthetheorydevelopedinclasscanbeused
toestimatetheeectivemassM ofthebackofthetruckfromtheobservedfrequencyofoscillation
and static deection.
(a) The statement that the oscillation contained more than ve clearly dened cycles is a tip-o
thatthesystemislightlydamped,whichsuggeststhatdampingmightbeneglectedinmaking
a preliminary estimate. The static deection under a load of w = mg tells us that the
equivalent stiness k can be estimated from w=k, or
w 250 pounds
k= = =250 pounds/inch
1.0 inch
Theobservedoscillationfrequencyof2Hzcanbetakenastheundampednaturalfrequency,

o
/2,ofthesystemwithmassM+mandstinessk. ThetruckmassM canbeestimated
by using the relation

2
k
=
o
M +m
in combination with the preceding expression for k to obtain
M =
mg
m (1)

2
o
IntheBritishsystemofunits,massissubordinatetoforce,andM isreplacedbyW/gwhere
W istheequivalentweightofthebackofthetruck. TouseBritishunits,Eq.(1)ismultiplied
through by g=386 in/sec
2
to get
W =
wg
2
w=
(250)(386)
250=361 pounds

o
(1)(4)
2
Alternatively the given data can be converted to SI units ( = 0.0254 m, and m = 113.4
kg),and Eq.(1) used to obtain M = 163.8 kg whose weight, Mg = (163.8)(9.81) = 1607
Newtons, is equivalent to the British weight of 361 pounds.
(b) It is assumed that the truck oscillation is a primarily vertical motion of the eective mass
of the rear end of the truck plus Uncle Massive on an eective vertical spring (the rear tires
plus suspension). It is also assumed that the observed damped natural frequency is a good
approximation for the undamped natural frequency required in the formulas above.
323
_
_
4.24 Car Suspension 2
(a),(b) The dierential equation for the system is
mx + bx + kx=F, (1)
The particular solution will be:
x
p
(t) = F/k. (2)
The homogenous solution has the form:
x
h
(t) = Ae
nt
cos(
d
t) + Be
nt
sin(
d
t) (3)
Plugging the initial conditions into the full solution
x(t) = Ae
nt
cos(
d
t) + Be
nt
sin(
d
t) + F/k (4)
0 = A+F/k (5)
A=F/k (6)
x (t) = e
nt
_
(A
d
sin
d
tA
n
cos
d
t) + (B
d
cos
d
tB
n
sin
d
t)
_
(7)
0 = A
n
+B
d
(8)
B =
A
n
(9)

d
For m=500kg, k= 5 10
4
N/m and b= 2 10
3
Ns/m,
k

n
= = 10 (10)
m
c
= = 0.2 (11)
2 km

d
=
n
1
2
= 9.8 (12)
=
n
=2; (13)
A=0.01 (14)
B = 2.0410
3
(15)
x(t) = 0.01e
2t
cos(9.8t).00204e
2t
sin(9.8t) + .01 (16)
The result is plotted in Figure 4.24.1.
(c) The passengers will not like this ride because there are too many oscillations. when b =

2 km=10000,thesystemwillbecriticallydamped,inthiscasethecompletesolutionhas
the form:
x(t) = Ae
nt
+Bte
nt
+F/k (17)
x(0)=A+F/k (18)
A=F/k (19)
A=0.01 (20)
x (0)=
n
A+
n
Bte
nt
+Be
nt
(21)
0 =
n
A+B (22)
B =
n
A (23)
B =0.1 (24)
x(t) = 0.01e
nt
0.1te
nt
+ 0.01 (25)
324
In order to nd the acceleration for the under-damped case, dierentiate equation 16 twice
with respect to time,
x(t) = 0.01e
2t
cos(9.8t).00204e
2t
sin(9.8t) + .01 (26)
x (t) = .10208e
2t
sin9.8t (27)
x(t) = e
2t
cos9.8t.20416
2t
sin9.8t (28)
The maximum acceleration will happen at t= 0
+
x(t= 0+) = 1 (29)
(30)
So the maximum acceleration is 1m/sec
2
.
Similarly, for the critically damped case, dierentiating equation 25 twice,
x(t) = 0.01e
nt
0.1te
nt
+ 0.01 (31)
x(t) = (1 10t)e
10t
(32)
x(0+)=1 (33)
So the accelerations will be the same.
(d) From the step response, the 5% settling time is 1.35 sec in the under-damped case and 0.47
sec in the critically-damped case.
0 0.5 1 1.5 2 2.5 3
t
Figure 4.24.1: Step Response
325
0
0.002
0.004
0.006
0.008
0.01
0.012
0.014
0.016
x
(
t
)

Underdamped Response
Critically Damped Response
carsuspension2sol.m
Download
clear all
close all
k=5*10^4
b=2*10^3
m=500
F=500
wn=sqrt(k/m)
zeta=b/(2*sqrt(k*m))
wd=wn*(1-zeta^2)^.5
sigma = zeta*wn
A=-F/k
B=A*sigma/wd
t=[0:.01:3];
hold on
x= A*exp(-sigma*t).*cos(wd*t)+ B*exp(-sigma*t).*sin(wd*t)+F/k;
plot(t,x)
xlabel(t)
ylabel(x(t))
cd= -.01.*exp(-10.*t)-.1.*t.*exp(-10*t)+.01
plot(t,cd,.)
legend(Underdamped Response,Critically Damped Response)
326
4.25 Kid-Skid
Figure 4.25.1:
327
Figure 4.25.2:
328
Figure 4.25.3:
329
4.26 Rolling Machine
Figure 4.26.1:
330
Figure 4.26.2:
331
4.27 Sunday Bike Ride
Figure 4.27.1:
332
Figure 4.27.2:
333
Figure 4.27.3:
334
g
4.28 Compound Mass Spring Damper System 1
(a) ThefreebodydiagramforthemassmisshowninFigure4.28.1andthefreebodydiagrams
for nodes n
1
and n
2
are shown in Figure 4.28.2.
. .
k
2
(x-n
1
) F
1
b
2
(x-n
2
)
m
x
k
3
x F
2 b
3
x
.
Figure 4.28.1: Free body diagram for mass m
b
2
n
.
2
k
2
n
1
n
2
n
1
. .
b
2
(x-n
2
) k
2
(x-n
1
)
Figure 4.28.2: Free body diagrams for nodes n
1
and n
2
(b) Webeginbywritingtheforcebalanceequationfornoden
1
,keepinginmindthattheeective
335
_ _ _ _
_ _ _ _
_ _ _ _
mass for this node is 0.
k
2
(xn
1
)k
1
n
1
=m
n
1
n
1
k
2
x(k
2
+k
1
)n
1
= 0
k
2
x= (k
2
+k
1
)n
1
k
2
x=n
1
(1)
k
2
+k
1
We do the same for node n
2
b
2
( xn
2
)b
1
n
2
=m
n
2
n
2
b
2
x (b
2
+b
1
) n
2
= 0
b
2
x = (b
2
+b
1
) n
2
b
2
x = n
2
(2)
b
2
+b
1
Now we write out the force balance equation for the mass m.
F
1
F
2
k
2
(xn
1
)b
2
( xn
2
)k
3
xb
3
x = mx
F
1
F
2
(k
2
+k
3
)x(b
2
+b
3
) x+k
2
n
1
+b
2
n
2
=mx (3)
We have equations for the positions n
1
(t) and n
2
(t), so we substitute equations (1) and (2)
into (3)
F
1
F
2
(k
2
+k
3
)x(b
2
+b
3
) x+k
2
k
2
x+b
2
b
2
x = mx
k
2
+k
1
b
2
+b
1
b
2
k
2
mx + b
2
+b
3

2
x + k
2
+k
3

2
x=F
1
F
2
b
1
+b
2
k
1
+k
2
mx +
b
2
(b
1
+b
2
)
+b
3

b
2
2
x +
k
2
(k
1
+k
2
)
+k
3

k
2
2
x=F
1
F
2
b
1
+b
2
b
1
+b
2
k
1
+k
2
k
1
+k
2
mx +
b
1
b
2
+b
3
x +
k
1
k
2
+k
3
x=F
1
F
2
b
1
+b
2
k
1
+k
2
mx + b
eq
x + k
eq
x=F
1
F
2
(4)
where
k
eq
=
k
1
k
2
+k
3
and b
eq
=
b
1
b
2
+b
3
k
1
+k
2
b
1
+b
2
Equation (4) is our nal 2
nd
order dierential equation, but it was a lot of algebra to get to.
There is a simpler way to derive the equation if you rst simplify the Free Body Diagrams.
Lookingbacktotheoriginalproblemdiagram,wecansimplifythetwospringsabovethemass
intooneequivalentspring. Rememberthattheeectivespringconstantforspringsinparallel
is their summation, and the eective spring constant for springs in series is the reciprocal of
of the summation of their inverses, simplied to be their product over their summation:
k
parallel
=k
1
+k
2
and k
series
=
k
1
k
2
(5)
k
1
+k
2
This equivalent model is also valid for damping coecients. We use equation (5) to redraw
the free body diagram as shown below in Figure 4.28.3.
336
g
_ _ _ _
_
_
k
1
k
2
b
1
b
2
.
k
1
+k
2
x
F
1 b
1
+b
2
x
m
x
k
3
x F
2 b
3
x
.
Figure 4.28.3: Force balance for mass m with equivalent constants
We now use this diagram to write out the force balance equation:
k
1
k
2
b
1
b
2
F
1
F
2
x x + k
3
x+b
3
x = mx
k
1
+k
2
b
1
+b
2
mx +
b
1
b
2
+b
3
x +
k
1
k
2
+k
3
x=F
1
F
2
b
1
+b
2
k
1
+k
2
mx + b
eq
x + k
eq
x=F
1
F
2
(6)
Equations (4) and (6) are identical, but equation (6) is much easier to derive.
(c) Notice that the dierential equation in (6) is almost in a standard form. We can devide
equation (6) through by the eective spring constant of the system toarrive at the standard
form in equation (7), where f(t) is some set of forces.
m b
eq
F
1
F
2
x + x + x=
k
eq
k
eq
k
eq
1 2
x + x + x=f(t) (7)

n
2

n
By comparing these equations, we can come up with expressions for
n
and , and plug in
the appropiate system values for the numberical answers.

n
=
k
eq
=100 [rad/sec]
m
= _
b
eq
=.05
2 k
eq
m

d
=
n
1
2
= 99.875 [rad/sec]
Thepolesforunderdampedsecondordersystemsarelocatedats=
n

d
. Thes-plane
plot of this system is shown in Figure 4.28.4.
337
-100
-80
-60
-40
-20
0
20
40
60
80
100
I
m
a
g
i
n
a
r
y

P
a
r
t

-7 -6 -5 -4 -3 -2 -1 0 1
Real Part
Figure 4.28.4: S-plane plot of system poles
(d) The general form for the solution x(t) of equation (6) is
x(t) = Ae
t
cos
d
t+Be
t
sin
d
t
where =
n
. We nd A by plugging in the initial condition x(0)=10
2
:
Ae
0
cos
d
0 +Be
0
sin
d
0 = 10
2
A= 10
2
B is found by plugging in the intial condition x(0)=0:
Ae
t
cos
d
t+Be
t
sin
d
t=x(t)
t
_ _
e (A
d
sin
d
tAcos
d
t) + (B
d
cos
d
tBsin
d
t) = x (t)
A+B
d
= 0
B =
A
= 10
2

d
Use these results to plot x(t) in matlab. The plot is shown in Figure 4.28.5.
338
0 0.5 1 1.5
-0.01
-0.008
-0.006
-0.004
-0.002
0
0.002
0.004
0.006
0.008
0.01
x
(
t
)

t
Figure 4.28.5: Plot of x(t) when x(0)=10
2
, x(0)=0
msdsol.m
Download
%Script for part D
keq=10000
beq=10
m=1
wn=100
zeta=.05
wd=wn*(1-zeta^2)^.5
sigma = zeta*wn
A=10^(-2)
B=A*sigma/wd
t=[0:.001:1.5];
x= A*exp(-sigma*t).*cos(wd*t)+ B*exp(-sigma*t).*sin(wd*t);
plot(t,x)
xlabel(t)
ylabel(x(t))
339
4.29 Compound Mass Spring Damper System 2
Beginbytakingequation(4)fromthesolutionofProblem4.28astheforcebalanceinthisproblem:
mx + b
eq
x + k
eq
x=F
1
F
2
where
k
eq
=
k
1
k
2
+k
3
and b
eq
=
b
1
b
2
+b
3
k
1
+k
2
b
1
+b
2
The problem states that F
1
= 10 [N], F
2
= 0 [N], and x(0)=x (0)=0. Rewrite the force balance
to be
mx + b
eq
x + k
eq
x= 10
Thesolutiontoadrivensecondorderdierentialequationwillbethesumoftheparticularsolution
andthehomogeneoussolution. Startbyndingtheparticularsolution. Guessthatx
p
(t) will have
the form x
p
(t) = C, x
p
(t) = 0.
mx
p
+b
eq
x
p
+k
eq
x
p
= 10
0 + 0 + k
eq
x
p
= 10
10
x
p
=
k
eq
We know the homogeneous solution of this equation from the solution of Problem 4.28:
x
h
=Ae
t
cos
d
t+Be
t
sin
d
t
The total solution is the sum of the particular and homogeneous solutions.
10
x(t) = Ae
t
cos
d
t+Be
t
sin
d
t+
k
eq
we nd the constants A and B by plugging in the initial conditions x(0) = 0 and x(0)=0:
10
Ae
0
cos
d
0 +Be
0
sin
d
0 + = 0
k
eq
10
A+ = 0
k
eq
10
A=
k
eq
and
10
Ae
t
cos
d
t+Be
t
sin
d
t+ =x(t)
k
eq
t
_ _
e (A
d
sin
d
tAcos
d
t) + (B
d
cos
d
tBsin
d
t) = x (t)
A+B
d
= 0
A 10
B = =

d
k
eq

d
340
A sketch of this solution is found in Figure 4.29.1
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
x
(
t
)

[
m
]

-3
x 10
0 0.2 0.4 0.6 0.8 1 1.2 1.4
t [s]
Figure 4.29.1: Plot of x(t) when x(0) = 0, x(0)=0, F
1
= 10 [N]
341
4.30 Wind Induced Building Vibrations
Figure 4.30.1:
342
Figure 4.30.2:
343
Figure 4.30.3:
344
5 Electrical Systems
5.1 Camera Flash Circuit
The solution for this problem is not available.
345
5.2 Batteries and Capacitors
is dependent of the battery internal resistance as follows:
+
(a) The short circuit current I
sc
5V
R
b
=
I
sc
1.
5V
R
b
=
5.7A
1.
R
b
(b) The power P dissipated will be as follows:
= 0.26
P = V I
sc
5V )(5.
P = 8.55W
= (1. 7A) P
The power is being dissipated at the battery internal resistance, heating up the battery as a
consequence.
(c) In one time constant =R
l
C the capacitor voltage should drop to 37 percent. So
= 10 min
= 600sec
600sec
12010
6
F
= 5 10
6

R
l
=
R
l
(d) Circuit diagram as follows:
+
+
-
-
Figure 5.2.1: Circuit diagram of battery charging capacitor
(e) Making the analysis using two loop currents we have the following:
346
1.5V v
c
I
1
=
R
b
v
c
I
2
=
R
l
dvc
I
1
I
2
= C
dt
Replacing I
1
and I
2
we have the following:
(R
b
R
l
)C
dv
c
+ (R
b
+R
l
)v
c
(1.5V)R
l
= 0
dt
156
dv
c
+ 5 10
6
v
c
7.510
6
= 0
dt
(f) Homogenous solution v
c
=Ae
st
and
dvc
=Ase
st
, replacing
dt
156Ase
st
+ 5 10
6
Ae
st
= 0
s = 32x10
3
dvc
Particular solution v
c
= Ae
st
+B and
dt
= Ase
st
, with boundary condition v
c
(t = 0) = 0,
replacing
A = B
B = 1.5V
A = 1.5V
So v
c
= 1.5(1e
32.110
3
t)
347
Capcitorvoltagevs.timewhenchargedbyAABattery
0 0.2 0.4 0.6 0.8 1 1.2 1.4
0
0.5
1
1.5
v
c

(
V
)

time(sec)
x10
4
Figure 5.2.2: Charging capacitor
The system pole is located at s=32.110
3
, as shown on gure 5.2.3
348
1
0.8
0.6
0.4
0.2
0
0.2
0.4
0.6
0.8
1
3.5 3
PoleZeroMap
2.5 2 1.5 1 0.5 0
RealAxis
x10
4
Figure 5.2.3: Pole location
I
m
a
g
i
n
a
r
y

A
x
i
s

(g) For the discharging capacitor we have the following


I
I
v
c
dv
c
dt
Ae
s(tt
1
)
s
s
dv
c
= C
dt
v
c
=
R
l
= Ae
s(tt
1
)
= Ase
s(tt
1
)
= R
l
CAse
s(tt
1
)
1
=
R
l
C
1
=
600
349
Capacitorvoltagevs.timewhendisconnected,withinitialchargeof1.5V
0 500 1000 1500 2000 2500
0
0.5
1
1.5
(tt1)(sec)
v
c

(
V
)

Figure 5.2.4: Discharging capacitor


The system pole is located at s=1.6710
3
, as shown on gure 5.2.5
1
0.8
0.6
0.4
0.2
0
0.2
0.4
0.6
0.8
1
PoleZeroMap
I
m
a
g
i
n
a
r
y

A
x
i
s

1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0


RealAxis
3
x10
Figure 5.2.5: Pole location
350
5.3 Loaded Motor
The solution for this problem is not available.
351
6 Fluid Systems
6.1 Rotational Fluid System
a)
J
T
1
T
2
(t)
2
c
1
J
c
2
(t)
1
(t)
2
Figure 6.1.1: Freebody diagram of Inertia and Schematic of system.
The free body diagram for the inertia is shown in Figure 6.1.1 along with a schematic of the
dynamic elements in this problem. As can be seen, the inertia has two torques acting on it:
T
1
generated by the damper c
1
between the input and the inertia and T
2
generated by the
damperc
2
betweentheinertiaandground. Theschematicdrawingishelpfulbutnotrequired
in the solution.
(b) Using the FBD from part a, we can write the following relationship:
T =J

2
=T
1
+T
2
This is where the schematic becomes helpful. By inspection, we can work out the following
relationships
T
1
= c
1
(
1

2
)
T
2
= c
2

2
Substituting for T
1
and T
2
yields the following
J

2
=c
1
(
1

2
)c
2

2
J

2
+ (c
1
+c
2
)
2
=c
1

1
352
_ _
_ _
(c) The input to the system has been dened as
1
(t) = u
s
(t) thus the dierential equation for
this system become
J

2
+ (c
1
+c
2
)
2
=c
1
u
s
(t)
This dierential equation (DE) matches that for a rst order forced response. We know the
solution to this DE is of the form

2
(t) =
h
(t) +
p
(t) where

p
(t) =
ss
=
c
1
(Particular solution)
c
1
+c
2

h
(t) = Ae
st
(Homogeneous solution)
c
1
+c
2
s =
J
Using initial conditions

2
(0) = 0 =A+
c
1
A=
c
1
c
1
+c
2
c
1
+c
2
c
1

c
1
+c
2
t

2
(t) = 1 e
J
c
1
+c
2
(d) We determined earlier that

ss
=
c
1
c
1
+c
2
Looking carefully at out schematic of the system, we know that the torque from the shaft
acting on the inertia is T
1
. Since, the torques on either side of the damper need to be equal,
we know the torque acting on the shaft is equal and opposite to T
1
. Since the system is in
steadystate,weknowthatthevelocityoftheshaftisconstantwhichmeansthatthesumof
torques acting on the shaft equal 0.
T
shaft
= 0 = T
in
T
1
T
in
=T
1
We know that
T
1
= c
1
(
1

2
)

1
= u
s
(t) = 1

2
=
ss
=
c
1
c
1
+c
2
Thus T
1
= c
1
1
c
1
=
c
1
c
2
c
1
+c
2
c
1
+c
2
More generally T
1
=
c
1
c
2

1
c
1
+c
2
353
6.2 Fluid Leak
The solution for this problem is not available.
354
6.3 Whered the water go?
(a) UsetheparametersR= 1.5 10
9
[Pasec/m
3
],andL= 7 10
6
[Pasec
2
/m
3
]. Beginbysketching
the equivalent circuit diagram as shown below.
P
s
(t)
L R
C
+
-
q
n
Figure 6.3.1: Shower equivalent circuit
q
n
is the water ow through the shower nozzle. We need a dierential equation in terms of
the nozzle ow. Note that the capacitance in this circuit can be neglected as it is eectively
short-circuited by the shower nozzle. Write out the loop equation for this circuit:
P
s
(t) = P
R
+P
L
Use the constitutive relationships for the uidic resistor and uid inertance:
P
R
=q
R
R
dq
L
P
L
=L
dt
note that q
R
=q
L
=q
n
. Substitute these equations back into the loop equation:
P
s
(t) = q
R
R+L
dq
L
dt
1 L dq
L
P
s
(t) = q
R
+
R R dt
the solution to this standard rst order dierential equation takes the form
q
n
(t) = Ae
t/
+B where =
L
R
Our initial condition is that P
s
(0)=310
5
[Pa]. Plugging this into our dierential equation
gives
q
n
(0) =
310
5
= 2 10
4
[m
3
/sec]
1.5 10
9

also, the nal condition must be that q


n
= 0. Plugging this into the solution to the
dierential equation gives A= 2 10
4
and B =0. The nal solution is:
q
n
(t) = 2 10
4
e
t/.004667
Theowwillfalltobelow1%ofitsinitalvalueafter5. Thisisapproximately.023[s]. The
sketch of the shower ow transient is shown in Figure 6.3.2.
355
-4
x 10
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035
0
0.5
1
1.5
2
2.5
q
n

[
m
3
/
s
e
c
]

time [s]
Figure 6.3.2: Shower ow transient
Becausetheshowernozzleismodeledasaswitch,thereisnopressuredropacrossthenozzle
and therefore the pressure at the nozzle is atmospheric pressure, and this remains constant.
The sketch of this plot is shown Figure 6.3.3.
-1
0
1
2
3
4
5
P
n

[
P
a
]

0 0.005 0.01 0.015 0.02 0.025 0.03
time [s]
Figure 6.3.3: Shower pressure transient
(b)-(d) Dont worry, nothing to answer here.
356
showersol.m
Download
%Shower Solution showersol.m
R= 1.5e9
L=7e6
P_0=3e5
tau=L/R
q_0=P_0/R
t=[0:.0001:.035];
q=q_0 * exp(-t/tau)
figure(1)
plot(t,q)
xlabel(time [s])
ylabel(q_n [m^3/sec])
figure(2)
xlabel(time [s])
ylabel(P_n [Pa])
line([0 0],[-1 5])
line([-.000 .030], [0 0])
357
7 Thermal Systems
7.1 Thermal Block Question
See Professor Trumpers Notes: Chapter 1 for some information on modeling thermal systems. A
diagram depicting the idealized block outside of the oven is shown in Figure 7.1.1.
T
amb
T
block
c=385 J/(kg
o
C)
m=0.1 kg
q
Figure 7.1.1: Diagram of copper block out of oven.
The ambient temperature surrounding the block is given as T
amb
, the internal temperature of the
block is given as T
block
, and the heat ow from the block to the environment is given as q. The
specic heat of copper (c) is given as 385
kg
J
o
C
, and the mass of the block is 0.1 kg.
The energy associated with the change of the blocks temperature is given by
CT
block
(1)
where C is the thermal capacitance of the block given by mc. The energy associated with heat
transfer for a period of time t is given by
qt. (2)
Inthisidealizedmodel,theheatlossisbalancedexactlybythechangeintemperatureoftheblock.
Thus, the instantaneous energy/heat balance is given by
mcdT
block
=qdt. (3)
The negative sign in front of the q is due to the fact that the heat is owing out of the block. If
you were to form a control volume around the block, heat owing in would have a positive sign.
According to the way the picture is drawn, the heat ow, q, is owing outwards. In addition,
the heat ow, q, will be modeled as linearly proportional to the dierence between the block
temperature and ambient temperature as
q=
T
block
T
amb
(4)
R
whereRisthethermalresistancebetweentheblockandtheenvironment. Bysubstitutionwehave
mcdT
block
= qdt (5)
=
T
block
T
amb
dt (6)
R
358
Therefore, the governing equation is
RC
dT
block
+T
block
=T
amb
(7)
dt
where the time constant, , for the system is given by RC.
200
300
400
500
600
700
800
900
1000
T
b
l
o
c
k
(
K
)

T
amb
0 500 1000 1500 2000 2500 3000
time(s)
Figure7.1.2: Theblocktemperature(T
block
)isplottedagainsttime(t). Theambienttemperature
is given by T
amb
.
Going back to Figure 7.1.2 we can calculate the time constant by estimating the amount of time
requiredfortheblocktodecayby63%ofthedierencebetweentheinitialblocktemperatureand
theambienttemperature. Thisoccursatapproximately559Kafteratimeintervalof600seconds.
Therefore, the time constant for the system is 600 seconds. With = RC and C=38.5
o
J
C
, the
o
C
thermal resistance, R, equals 15.6
J/s
.
359
7.2 The Hot Copper Block 1
(a) First, draw a diagram which shows heat ow through the system. It is useful to thinkabout
heat owing from heat capacitances to other heat capacitances through thermal resistances.
This model is shown in Figure 7.2.1.
T
w
T
b
R
bw
q
wa
T
a
R
wa
C
B
C
W
q
bw
Figure 7.2.1: Thermal model
Weneedtousetheconstitutiverelationshipsforthermalelementstowriteoutthegoverning
equations. Remember that
dT
C =q
dt
where C is thermal capacitance,
dT
refers to the time derivative of the temperature of the
dt
capacitance and q is the heat ow into the capacitance. Also
T =qR
where T is the temperature drop across the resistance and R is the thermal resistance. An
importantnote: Weneedtoincludethemassofthematerialtocalculatethethermalcapac-
itances in these equations, as the units for C are [J/

C] and you are given the specic heat


of the copper and water in [J/kg

C]. Thermal capacitance is derivied by multiplying the


specicheattimesthemass. UsingtheseequationsandFigure7.2.1,writeoutthegoverning
equations:
C
B
dT
b
=q
bw
(1)
dt
C
W
dT
w
=q
bw
q
wa
(2)
dt
(T
b
T
w
) = q
bw
R
bw
(3)
(T
w
T
a
) = q
wa
R
wa
(4)
combine equations (1) through (4) to the derivied two equations in state space form as:
dT
w
dt
=
1
R
bw
C
W
(T
b
T
w
)
1
R
wa
C
W
(T
w
T
a
) (5)
dT
b
dt
1
= (T
b
T
w
)
R
bw
C
B
(6)
360
(b) To transform equations (5) and (6) into a single 2
nd
-order dierential equation in T
w
(t), we
need to eliminate T
b
from equations (5) and (6). Start by re-arranging (5):
R
bw
C
W
dT
w
+
R
bw
(T
w
T
a
) + T
w
=T
b
(7)
dt R
wa
and take the time derivative
d
2
T
w
R
bw
_
dT
w
dT
a
_
dT
w
dT
b
R
bw
C
W
+ + = (8)
dt
2
R
wa
dt dt dt dt
combine equations (7) and (8) into (6), rembering that
dTa
= 0:
dt
_ _
d
2
T
w
_
R
bw
_
dT
w
R
bw
dT
a
1

dT
w
R
bw
R
bw

R
bw
C
W
+ + 1 =
_
R
bw
C
W
+ T
w
T
a
_
dt
2
R
wa
dt R
wa
dt R
bw
C
B
dt R
wa
R
wa
d
2
T
w
_
R
bw
C
W
_
dT
w
1 R
bw
dT
a
1
R
bw
C
W
+ + 1 + + T
w
= + T
a
dt
2
R
wa
C
B
dt R
wa
C
B
R
wa
dt R
wa
C
B
d
2
T
w
dT
w
R
bw
C
W
R
wa
C
B
+ (R
bw
C
B
+R
wa
C
B
+R
wa
C
W
) +T
w
=T
a
dt
2
dt
(9)
(c) Equation (9) is of the form
mx + bx + kx=f(t)
we need to nd the roots to gure out what the solution will look like.

b b
2
4mk
s
1
, s
2
=
2m
s
1
=0.2621[sec
1
]
s
2
=1.387810
5
[sec
1
]
This means the solution takes the form
T
w
(t) = Ae
s
1
t
+Be
s
2
t
+C
TondtheconstantsA,B andC, we need three known states. We know that T
w
(0)=T
a
=
25

C. The other initial condition is found by evaluating (5) at t= 0


dT
w
(0) 1 1
= (T
b
(0)T
w
(0)) (T
w
(0)T
a
(0))
dt R
bw
C
W
R
wa
C
W
T

w
(0) =
9525
=
1
[

C/sec]
R
bw
C
W
6
we also know that T

=T
a
, plugging in the initial conditions yields two equations:
A+B+C = 25
1
s
1
A+s
2
B =
6
C = 25
361
-0.25
-0.2
-0.15
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
I
m
a
g
i
n
a
r
y

P
a
r
t

thereforeourconstantsareA=0.6359,B = 0.6359,andC =25. Thetotalsolutionisthus
T
w
(t) = .6359(e
1.282910
5
t
e
0.0284t
) + 25
The system poles are shown in Figure 7.2.2
S-plane plot of poles for part (c)
-0.25 -0.2 -0.15 -0.1 -0.05 0 0.05
Real Part
Figure 7.2.2: System Poles
(d) Figure7.2.3showstheplotsofT
w
(t)overbothalongandashorttimescale. Theplotslook
like this because the copper block temperature equalizes with the water temperature very
quickly. Then,overamuchlongertime(ontheorderoftwodays;thismustbeareallygood
vacuumbottle!),thewatertemperatureequilizeswiththeambienttemperaturethroughheat
losses through the walls of the vacuum bottle. You can see the eects of the time constants
in the gures. The small time constant has no important eect beyond about 50 seconds,
where the longer time constant dominates the response after that.
362
T

(
t
)

[

C
]

(
t
)

[

C
]

w
w

PlotofT (t),shorttimescale
w
25.7
25.6
25.5
25.4
25.3
25.2
25.1
25
0 5 10 15 20 25 30 35 40
Time,[s]
PlotofT (t),longtimescale
w
25.7
25.6
25.5
25.4
25.3
25.2
25.1
0
25
20 40 60
Time,[hrs]
80 100 120
Figure 7.2.3: Plot of T
w
(t)
(e) If the block starts out at only 60

C, we need to change the initial condition found with


equation (5). It now becomes
T

w
(0) =
6025
=
1
R
bw
C
W
12
Figure 7.2.4 show the dierence from the original response. The original response is dotted.
Notethatsincetheblocktermperaturedierencetoambienthasbeenloweredbyafactorof
2, this response is one half the earlier response.
363
25.2
T

(
t
)

[

C
]

(
t
)

[

C
]

w
w

PlotofT (t),shorttimescale
w
25.7
25.6
25.5
25.4
25.3
25.1
25
0 5 10 15 20 25 30 35
Time,[s]
PlotofT (t),longtimescale
w
25.7
25.6
25.5
25.4
25.3
25.2
25.1
25
0 20 40 60 80 100 120
Time,[hrs]
Figure 7.2.4: Plots of T
w
(t), T
b
(0)=60

C
thermalblocksol.m
Download
%The Hot Copper Block thermalblocksol.m
%Part C
clear all
close all
cB=.1*385
cW=1 * 4200
rbw=.1
rwa=17
twinit = 25
twdiff = 1/6
m=rbw*cB*rwa*cW
c = rbw*cB + rwa*cB + rwa * cW
k=1
tw=[m c k]
s = roots(tw)
B = twdiff/(s(2)-s(1))
A= -B
C= twinit
364
40
figure(1)
plot(s ,0 , x)
axis([-.3 .05 -.25 .25])
xlabel(Real Part)
ylabel(Imaginary Part)
title(S-plane plot of poles for part (c))
%Part D
figure(2)
subplot(2,1,1)
t1=[0:.001:40];
T_w1 = A*exp(s(1)*t1) + B*exp(s(2)*t1) + C;
plot(t1,T_w1)
axis([0 40 25 25.7])
xlabel(Time, [s])
ylabel(T_w(t) [^\circC])
title(Plot of T_w(t), short time scale)
subplot(2,1,2)
t2=[0:5:500000];
T_w2 = A*exp(s(1)*t2) + B*exp(s(2)*t2) + C;
plot(t2/3600,T_w2)
axis([0 120 25 25.7])
xlabel(Time, [hrs])
ylabel(T_w(t) [^\circC])
title(Plot of T_w(t), long time scale)
%Part E
twdiff2 = 1/12
B2 = twdiff2/(s(2)-s(1))
A2 = -B2
T_w3 = A2*exp(s(1)*t1) + B2*exp(s(2)*t1) + C;
T_w4 = A2*exp(s(1)*t2) + B2*exp(s(2)*t2) + C;
figure(3)
subplot(2,1,1)
plot(t1,T_w1,:)
hold
plot(t1,T_w3)
axis([0 40 25 25.7])
xlabel(Time, [s])
ylabel(T_w(t) [^\circC])
title(Plot of T_w(t), short time scale)
subplot(2,1,2)
plot(t2/3600,T_w2,:)
hold
plot(t2/3600,T_w4)
axis([0 120 25 25.7])
xlabel(Time, [hrs])
ylabel(T_w(t) [^\circC])
title(Plot of T_w(t), long time scale)
365
7.3 The Hot Copper Block 2
(a) First, draw a diagram which shows heat ow through the system. It is useful to thinkabout
heat owing from heat capacitances to other heat capacitances through thermal resistances.
However,becausetheheaterisembeddedinthecopperblock,wemodeltheresistancebetween
the block and the heater to be 0. This model is shown in Figure 7.3.1.
T
w
T
b
R
bw
q
wa
T
a
q
in
R
wa
C
B C
W
q
bw
Figure 7.3.1: Thermal model
Weneedtousetheconstitutiverelationshipsforthermalelementstowriteoutthegoverning
equations. Remember that
dT
C =q
dt
dT
where C is thermal capacitance,
dt
refers to the time derivative of the temperature of the
capacitance and q is the heat ow into the capacitance. Also
T =qR
where T is the temperature drop across the resistance and R is the thermal resistance. An
importantnote: Weneedtoincludethemassofthematerialtocalculatethethermalcapac-
itances in these equations, as the units for C are [J/

C] and you are given the specic heat


of the copper and water in [J/kg

C]. Thermal capacitance is derivied by multiplying the


specicheattimesthemass. UsingtheseequationsandFigure7.3.1,writeoutthegoverning
equations:
C
B
dT
b
=q
in
q
bw
(1)
dt
C
W
dT
w
=q
bw
q
wa
(2)
dt
(T
b
T
w
) = q
bw
R
bw
(3)
(T
w
T
a
) = q
wa
R
wa
(4)
combine equations (1) through (4) to the derivied two equations in state space form as:
dT
w
1 1
= (T
b
T
w
) (T
w
T
a
) (5)
dt R
bw
C
W
R
wa
C
W
dT
b
1 q
in
= (T
b
T
w
) + (6)
dt R
bw
C
B
C
B
366
To transform equations (5) and (6) into a single 2
nd
-order dierential equation in T
w
(t), we
need to eliminate T
b
from equations (5) and (6). Start by re-arranging (5):
R
bw
C
W
dT
w
+
R
bw
(T
w
T
a
) + T
w
=T
b
(7)
dt R
wa
and take the time derivative
d
2
T
w
R
bw
_
dT
w
dT
a
_
dT
w
dT
b
R
bw
C
W
+ + = (8)
dt
2
R
wa
dt dt dt dt
combine equations (7) and (8) into (6), rembering that
dTa
= 0:
dt
d
2
T
w
_
R
bw
_
dT
w
R
bw
dT
a
R
bw
C
W
+ + 1 =
dt
2
R
wa
dt R
wa
dt
_ _
1

dT
w
R
bw
R
bw
q
in

_ R
bw
C
W
+ T
w
T
a
_+
R
bw
C
B
dt R
wa
R
wa
C
B
R
bw
C
W
d
2
T
w
+
_
R
bw
+ 1 +
C
W
_
dT
w
+
1
T
w
=
R
bw
dT
a
+
1
T
a
+
q
in
dt
2
R
wa
C
B
dt R
wa
C
B
R
wa
dt R
wa
C
B
C
B
d
2
T
w
dT
w
R
bw
C
W
R
wa
C
B
+ (R
bw
C
B
+R
wa
C
B
+R
wa
C
W
) +T
w
=T
a
+q
in
R
wa
(9)
dt
2
dt
Equation (9) is the governing dierential equation.
(b) Equation (9) is of the form
mx + bx + kx=f(t)
we need to nd the roots to gure out what the solution will look like.

b b
2
4mk
s
1
, s
2
=
2m
s
1
=0.2621[sec
1
]
s
2
=1.387810
5
[sec
1
]
This means the solution takes the form
T
w
(t) = Ae
s
1
t
+Be
s
2
t
+C
To nd the constants A, B and C, we need three known states. We know that T
w
(0) =
T
b
(0)=T
a
= 25

C. The other initial condition is found by evaluating (5) at t= 0


dT
w
(0) 1 1
= (T
b
(0)T
w
(0)) (T
w
(0)T
a
(0))
dt R
bw
C
W
R
wa
C
W
T

w
(0) =
2525
= 0[

C/sec]
R
bw
C
W
we also know that T
w
() = T
a
+q
in
R
wa
, plugging in the initial conditions yields two equa-
tions:
A+B+C = 25
s
1
A+s
2
B = 0
C = 25 + 4 17 = 93
367
therefore our constants are A= 3.6 10
3
, B =68.0036, and C =93. The total solution is
thus
T
w
(t) = 3.6 10
3
e
0.2621t
68.0036e
1.387810
5
t
+ 93
(c) Figure7.3.2showstheplotsofT
w
(t)overbothalongandashorttimescale. Thesmalltime
constant has no important eect beyond about 50 seconds, where the longer time constant
dominates the response after that. The script thermalblock.m nds the system poles and
plots the response on both a small and large time scale.
Plot of T (t), short time scale
w
25
25.005
25.01
25.015
25.02
T
w
(
t
)

[
q
C
]

0 2 4 6 8 10 12 14 16 18 20
Time, [s]
Plot of T (t), long time scale
w
30
40
50
60
70
80
90
T
w
(
t
)

[
q
C
]

0 20 40 60 80 100 120
Time, [hrs]
Figure 7.3.2: Plot of T
w
(t)
thermalblock2sol.m
Download
%The Hot Copper Block 2 thermalblock2sol.m
clear all
close all
cB=.1*385
cW=1 * 4200
rbw=.1
rwa=17
368
twinit = 25
twdiff = 0
m=rbw*cB*rwa*cW
c = rbw*cB + rwa*cB + rwa * cW
k=1
tw=[m c k]
s = roots(tw)
B = (68*s(1))/(s(2)-s(1))
A = -B-68
C = 93
figure(1)
subplot(2,1,1)
t1=[0:.001:20];
T_w1 = A*exp(s(1)*t1) + B*exp(s(2)*t1) + C;
plot(t1,T_w1)
%axis([0 40 25 25.7])
xlabel(Time, [s])
ylabel(T_w(t) [^\circC])
title(Plot of T_w(t), short time scale)
subplot(2,1,2)
t2=[0:5:500000];
T_w2 = A*exp(s(1)*t2) + B*exp(s(2)*t2) + C;
plot(t2/3600,T_w2)
axis([0 120 25 95])
xlabel(Time, [hrs])
ylabel(T_w(t) [^\circC])
title(Plot of T_w(t), long time scale)
369
7.4 Transistor on heat sink
(a) As always, rst begin by drawing a gure. In this case, draw the circuit equivalent of this
system. The equivalent circuit is shown in Figure 7.4.1.
q
in
(t)
T
d
(t) T
c
(t) T
s
(t)
T
a
(t)
R
dc
R
cs
R
sa
Figure 7.4.1: Transistor thermal system equivalent circuit
Because this is steady state, we do not need to worry about thermal capacitances at the
moment. We need the constitutive relationships for the devices:
T
d
T
c
= 50 R
dc
T
c
T
s
= 50 R
cs
T
s
T
a
= 50 R
sa
rearranging these equations and plugging in the numerical values leads to
T
s
=T
a
+ 50 R
sa
= 50

C
T
c
=T
s
+ 50 R
cs
= 60

C
T
d
=T
c
+ 50 R
dc
= 65

C
(b) Nowdrawanotherpicturewiththethermalcapacitanceoftheheatsinkaddedin. Thisnew
circuit equivalent is shown in Figure 7.4.2.
q
in
(t)
T
d
(t) T
c
(t) T
s
(t)
T
a
(t)
R
dc
R
cs
R
sa
C
s
Figure 7.4.2: Transistor thermal Equivalent circuit with heatsink thermal capacitance
370
Why is the thermal capacitance in parallel with its associated resistance? Imagine if they
were in series. At steady state, there is no heat ow through the capacitance. If they
were connected in series, there would be an open circuit with no heat ow, which would be
impossibleforthisthermalsystem. Thethermalcapacitanceissimplythespecicheattimes
the mass: C
s
= 0.90 500 = 450 [J/

C] (remember that

C =

K as far as magnitude is
concerned). The constitutive relation for the thermal capacitance is
dT
s
dT
a
dT
s
q=C
s
=C
s
dt dt
Write out the node equation at T
s
dT
s
T
s
T
a
q
in
=C
s
+
dt R
sa
dT
s
R
sa
C
s
+T
s
=R
sa
q
in
+T
a
dt
This rst order dierential equation is the governing equation for the system. The general
solution to this equation is
T
s
(t) = Ae
t/
+B where =R
sa
C
s
First solve for the time 0 t 1. During this time, q
in
= 500 [W]. The initial and nal
conditions for this time frame are
T
s
(0)=50

C
T
s
() = T
a
+500 R
sa
=275

C
These conditions yield A=225 and B =275. The solution is
T
s
(t) = 225e
t/225
+275
Plug this solution into the constitutive relationships for the other devices to obtain other
equations:
T
c
(t) = T
s
(t)+500 R
cs
=225e
t/225
+375
T
d
(t) = T
c
(t)+500 R
dc
=225e
t/225
+425
Now, there is something to note here. These equations dont match the initial conditions of
theirrespectivesteadystatevalues. Thatistosay,thesetemperaturesseemtojumpfrom
their steady state values to new values, and this doesnt make sense. This happens because
we have approximated these thermal capacitances to be zero, and therefore their associated
time constants will be zero, and their responses will be innitely fast. In the real world, we
would have to take these capacitances into eect.
Nowattackthetimeframe1 t . Allthatwillchangeisthevalueofq
in
to50[W].The
initial and nal conditions will be
T
s
(1)=225e
1/225
+275=51

C
T
s
() = T
a
+ 50 R
sa
= 50

C
371
These conditions yield A=1 and B =50. The solution is
T
s
(t) = e
t/225
+ 50
Plug this solution into the constitutive relationships for the other devices to obtain other
equations:
T
c
(t) = T
s
(t) + 50 R
cs
=e
t/225
+ 60
T
d
(t) = T
c
(t) + 50 R
dc
=e
t/225
+ 65
The peak device temperature happens at t= 1:
T
d
(1)=225e
1/225
+425=201

C
This would more than likely fry the transistor.
A bigger heatsink would not help in this case. In the limiting case as C
s
, T
d
(t) =
200

C. This is because there must be a temperature dierence that follows the constituent
relationshipT
d
T
s
=.3 500=150

C.Thedevicewillfryjustaseasilyat200

C as it would
201

C,soeventhebiggestheatsinkwonthelpthischip. Smallerthermalresistanceswould
though. The plots of the temperature transients are shown below.
372
201 66
d

200.8 65.8
d

q

T
(
t
)

[
q

C
]

T

(
t
)

[
q

C
]
200.6
200.4
T

(
t
)

[
q

C
]

T
(
t
)

[
q

C
]

T

(
t
)

[
q

C
]
65.6
65.4
c

200.2 65.2
200 65
0 0.2 0.4 0.6 0.8 1 0 200 400 600 800 1000
time [s] time [s]
151 61
150.8 60.8
c

150.6 60.6
150.4 60.4
ss

150.2 60.2
150 60
0 0.2 0.4 0.6 0.8 1 0 200 400 600 800 1000
time [s] time [s]
51 51
50.8 50.8
C
]
50.6 50.6
T

(
t
)

[

50.4 50.4
50.2 50.2
50 50
0 0.2 0.4 0.6 0.8 1 0 200 400 600 800 1000
time [s] time [s]
Figure 7.4.3: Plots of Temperature transients
transistorsol.m
Download
373
%Transistor Solution transistorsol.m
rdc=.1
rcd=.2
rsa=.5
cs=450
tau=rsa*cs
t1=[0:.001:1];
t2=[1:.5:1000];
gen_t1 = -225*exp(-t1/tau);
ts1=gen_t1 + 275;
tc1=gen_t1+375;
td1=gen_t1+425;
gen_t2 = exp(-t2/tau);
ts2=gen_t2 + 50;
tc2=gen_t2+60;
td2=gen_t2+65;
figure(1)
subplot(3,2,1)
plot(t1,td1)
xlabel(time [s])
ylabel(T_d(t) [^\circ C])
subplot(3,2,3)
plot(t1,tc1)
xlabel(time [s])
ylabel(T_c(t) [^\circ C])
subplot(3,2,5)
plot(t1,ts1)
xlabel(time [s])
ylabel(T_s(t) [^\circ C])
subplot(3,2,2)
plot(t2,td2)
xlabel(time [s])
ylabel(T_d(t) [^\circ C])
subplot(3,2,4)
plot(t2,tc2)
xlabel(time [s])
ylabel(T_c(t) [^\circ C])
subplot(3,2,6)
plot(t2,ts2)
xlabel(time [s])
ylabel(T_s(t) [^\circ C])
374
_ _
8 Circuits
8.1 RC Transfer Function
1
(a) BeginbycombiningR
2
andC intoanequivalentimpedanceZ
eq
=R
2
+
Cs
. Sumthecurrents
at the node between R
1
and R
2
V
i
(s)V
o
(s)
=
V
o
(s)
R
1 R
2
+
1
Cs
V
i
(s)
=V
o
(s)
1
+
1
1
R
1
R
1 R
2
+
Cs
V
o
(s)
=
R
1
1
V
i
(s)
R
1
1
+
R
2
+
1
1
Cs
V
o
(s)
=
R
2
Cs+ 1
V
i
(s) (R
1
+R
2
)Cs+ 1
This transfer function is an important result. When we nd the equivalent impedance, the
circuit becomes a voltage divider. In a voltage divider, the voltage drop across any resistor
is simply its resistance divided by the total resistance seen at the input terminals multiplied
by the input voltage. This is also true for impedances. The voltage we wanted to solve for
was across the equivalent impedance Z
eq
=R
2
+
1
. The total impedance seen by the input
Cs
voltage source is Z
total
=R
1
+R
2
+
1
. This becomes
Cs
1
V
o
(s) =
Z
eq
V
i
(s) =
R
2
+
Cs
1
Z
total
R
1
+R
2
+
Cs
V
o
(s)
=
R
2
Cs+ 1
V
i
(s) (R
1
+R
2
)Cs+ 1
This produces the same transfer function (with much less work).
(b)
1
z
1
= =8.33
R
2
C
1
p
1
= =6.67
(R
1
+R
2
)C
375
-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
Figure 8.1.1: Pole Zero Plot
(c) This solution is not available.
376
8.2 RLC Transfer Function
(a) Using the voltage divider relationship:
Z
eq
1
V
o
(s) = V
i
(s) =
Cs
1
V
i
(s)
Z
total
R+Ls+
Cs
V
o
(s)
=
1
V
i
(s) LCs
2
+RCs+ 1
(b)
1
C = = 10
8
[F]

2
L
n
2
R= =100 []

n
C
The characteristic equation is 10
10
s
2
+ 10
6
s+1=0. The roots (and therefore the poles)
are
s
1,2
=500099875i
In polar form, these poles become
s
1,2
= 10
5
1.52077
5
x 10
Figure 8.2.1: System Poles
-6000 -5000 -4000 -3000 -2000 -1000 0
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
(c) The step response for this system with values is shown in Figure 8.2.2.
377
Step Response
A
m
p
l
i
t
u
d
e

0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
System: sys
Time (sec): 3.2e-005
Amplitude: 1.85
System: sys
Time (sec): 9.4e-005
Amplitude: 1.62
System: sys
Time (sec): 0.000158
Amplitude: 1.45
System: sys
Time (sec): 0.00022
Amplitude: 1.33
System: sys
Time (sec): 0.000598
Amplitude: 1.05
0 0.2 0.4 0.6 0.8 1 1.2
Time (sec)
x 10
-3
Figure 8.2.2: Step Response
(d) This solution is not available.
378
_

8.3 Bandpass Filter


(a) Using KCL, the current is i=i
C
+i
R
+i
L
. The output voltage is v
R
=L
di
dt
L
. The current
=C
dvr
in the capacitor is i
C
dt
. The current in the resistor is i
R
=v
r
/R. The current in the
inductorisi
L
=ii
C
i
R
. Substitutingfori
L
intotheconstitutiverelationfortheinductor,
we obtain the governing dierential equation
d
2
v
R
dv
R
di
RLC +L +Rv
R
=RL
dt
2
dt dt
(b) Taking the Laplace Transform, we nd the transfer function is the impedance
V
R
(s) RLs
Z(s) = =
I(s) RLCs
2
+Ls+R
(c) To nd the damping ratio and natural frequency, we compare the characteristic equation
(RLCs
2
+Ls+R=0) to the canonical form (s
2
+ 2
n
s+
n
2
=0). Thus, we nd
1 1 L

n
= and =
LC
2R C
The zero is located at s
z
=0. The poles are located at
1
_
1
_
2
4
s
p
=
2RC RC LC
(d) Using the given values (R = 50k, C = 10pF, and L = 3.1481510
7
H, we nd
n
=
563.610
6
rad/sec = 89.7MHz and = 0.18.
(e) The Bode plot is shown in Figure 8.3.1 The circuit is called a bandpass lter because it
passessignalswithafrequencyinabandcenteredat89.7MHz. Frequencycontentawayfrom
this band is attenuated. The frequency 89.7MHz is frequency of the Boston radio station
WGBH-FM.
379

5
10
7
10
8
10
9
10
7
10
8
10
9
Frequency[Hz]
Figure 8.3.1: Bode Plot
(f) If radio stations are separated by at least 400kHz, we must examine the attenuation at
89.3MHz and 90.1MHz. The magnitude of the transfer function is
[H(j)[=
(RL)
2
(R(1LC
2
) + 1)
2
+ (L)
2
The peak magnitude (f = 89.7MHz) is 510
4
. The magnitude at 89.3MHz is 1.8510
4
.
The magnitude at 90.1MHz is 1.8510
4
. The attenuation of the neighboring radio stations
is shown in Figure 8.3.2.
10
1
10
2
10
3
10
4
10
M
a
g
n
i
t
u
d
e

100
50
0
50
100
P
h
a
s
e

[
d
e
g
r
e
e
s
]

380
10
4.1
10
4.2
10
4.3
10
4.4
10
4.5
10
4.6
M
a
g
n
i
t
u
d
e

10
7.95
10
7.951
10
7.952
10
7.953
10
7.954
10
7.955
10
7.956
Frequency[Hz]
Figure 8.3.2: Attenuation of neighboring frequencies
381
_
8.4 LC Circuit Dierential Equations
(a) Using KCL at the node between the inductor and capacitor with the assumed currents both
positive into the node gives the following:
i
L
+i
C
= 0 (1)
1
i
L
= v
L
dt
L
dv
c
i
C
=C
dt
Equation(1)mustbedierentiatedbeforesubstitutingforthecurrentsandfromthedirection
of our assumed currents, v
L
=v
i
v
o
and v
C
= 0 v
o
. The governing dierential equation
is then
d
2
v
o
v
o
v
i
+ = (2)
dt
2
LC LC

(b) TheLaplacetransformofEquation(2),with
n
= 1/ LC,initialrestandsomerearranging,
is

2
v
o
(s) =
n
v
i
(s)
s
2
+
2
n
With the input voltage a unit step and after performing partial fraction decomposition
1 s
v
o
(s) =
s s
2
+
2
n
The inverse Laplace transform reveals the output voltage as a function of time after substi-

tuting = 1/ LC
t
v
o
(t) = 1 cos
LC
Note in the period of the output voltage in Figure 8.4.1
0
1
2
O
u
t
p
u
t

V
o
l
t
a
g
e

(
v
o
)

2(LC)
1/2
0
Time (s)
Figure 8.4.1: Labeled plot of the output voltage
382
8.5 Equivalent Impedance
The equivalent impedance of the circuit is
Z
eq
=Z
L
+ (Z
R1
+Z
R2
)[[(Z
C1
[[Z
C2
) (1)
Butrealizetheseriesresistiveimpedance,Z
R1
+Z
R2
,canbesimpliedintoanequivalentresistive
impedancebyR
12
=R
1
+R
2
andZ
R12
=Z
R1
+Z
R2
. Also,theparallelcapacitorscanbereplaced
by an equivalent capacitance C
12
= C
1
+C
2
. The impedance of the equivalent capacitor is then
Z
C12
=Z
C1
[[Z
C2
. Equation (1) is now simplied to
Z
eq
=Z
L
+ (Z
R12
[[Z
C12
)
Substitute Z
L
=Ls, Z
R12
=R
12
, and Z
C12
= 1/C
12
s gives
R
12
(1/C
12
s)
Z
eq
=Ls+
R
12
+ 1/C
12
s
Now expand the R
12
and C
12
substitutions, then simplify and
L(R
1
+R
2
) (C
1
+C
2
)s
2
+Ls+ (R
1
+R
2
)
Z
eq
=
(R
1
+R
2
) (C1 +C2)s+ 1
383
8.6 Circuit Response
FromProf. Trumpershandoutsandlecture,weknowthatthesolutiontoanydierentialequation
is the sum of the homogenous and particular solutions.
(a) The dierential equation for i
c
is
di
c
RC
dt
+i
c
=
dv
in
C
dt
v
in
= u(t)
dv
in
dt
= (t)
thus for t > 0RC
di
c
dt
+i
c
= 0
thus for t > 0i
c
(t) = Ae

t
RC
where A = i(0
+
)
v
in
(0)v
c
(0)
and i(0
+
) =
R
1 t
i) i(t) = e

RC
R
9

t
ii) i(t) = e
RC
R
iii) i(t) = 0
0 1 2 3 4 5
9
8
7
6
5
4
3
2
1
0
1
i
ii
iii
Figure 8.6.1: Step Response for Part (a)
384
(b)
dv
L
dv
in
L +Rv
L
= L
dt dt
v
in
= u(t)
dv
in
= (t)
dt
dv
L
thus for t > 0L +Rv
L
= 0
dt
Rt
thus for t > 0v
L
(t) = Ae

L
where A = v
L
(0
+
)
and v
L
(0
+
) = v
in
(0)Ri
L
(0)
Rt
i) v(t) = e

L
Rt
ii) v(t) = 9e

L
iii) v(t) = 0
StepResponse
A
m
p
l
i
t
u
d
e

9
8
7
6
5
4
3
2
1
0
1
i
ii
iii
0 1 2 3 4 5
Time(sec)
Figure 8.6.2: Step Response for Part (b)
385
6
_ _
(c)
L
R
dv
L
dt
+v
L
for t > 0
L
R
dv
L
dt
+v
L
=
=
for t > 0v
L
(t)
i
in
(0)
v
R
(0)
=
=
=
v
L
(t) =
di
in
L
dt
0
Ae

Rt
L
i
R
(0)+i
L
(0)=i
R
(0)+0
v
L
(0)=A=i
in
R
Re

Rt
L
StepResponse
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 1 2 3 4 5
Time(sec)
Figure 8.6.3: Step Response for Part (c)
(d)
dv
c
RC +v
c
= Ri
in
dt
t
v
c
(t) = Ri
in
+Ae

RC
v
c
(0) = 0 A=R

t
v
c
(t) = 1 e
RC
R
A
m
p
l
i
t
u
d
e

386
6
A
m
p
l
i
t
u
d
e

0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
StepResponse
0 1 2 3 4 5
Time(sec)
Figure 8.6.4: Step Response for Part (d)
(e) (a)
1
V
c
=
Cs
=
1
V
in R+
1
RCs+ 1
Cs
I
c
CsV
c
Cs
= =
V
in
V
in
RCs+ 1
(b)
V
L
Ls
=
V
in
R+Ls
(c)
V
L
V
L
I
in
= +
R Ls
V
L
LRs
=
I
in
R+Ls
(d)
V
c
V
c
I
in
= +
1
R
Cs
V
c
R
=
I
in
RCs+ 1
387
6
8.7 LRC Circuit 1
-
+

Figure 8.7.1: LRC Circuit


(a) v
1
= 0.5 V, v
2
= 0 V, i
l
= 0.5 A, i
c
= 0
(b)
v
1
(0
+
) = v
c
(0

) = 0.5V
dv
1
(0
+
) = dv
1
dt(0
+
) =
i
c
(0
+
)
=
(i
L
(0
+
) + i
R
(0
+
)
dt C C
v
c
(0
+
)
i
R
(0
+
) = = 0.5A
R
i
L
(0
+
) = i
L
(0

) = 0.5A
dv
1
(0
+
) = 1
dt
(c)
v
1
= v
c
=v
c
(0)
i
c
C
dv
1
=i
c
sC dt
v
c
v
c
i
c
= i
R
+i
L
= +
R R+Ls
C
dv
1
=
v
c

v
c
RLC
d
2
v
1
+ (R
2
C+L)
dv
1
+ 2Rv
1
= 0
dt R R+Ls dt
2
dt
d
2
v
1
dv
1
+ 2 + 2v
1
= 0
dt
2
dt
v
1
(t) = 0.707e
t
sin(t+ 3/4)
388
8.8 LRC Circuit 2
(a) The node equations are
0 = i
s

v
1
i
c

v
1
v
2
(1)
1 1
0 =
v
1
v
2
i
L
(2)
1
fornodesv
1
andv
2
respectively. Theconstituentrelationshipsforthecapacitorandinductor
are
i
c
=C
dv
c
=
dv
1
(3)
dt dt
v
L
=L
di
L
=v
2
(4)
dt
These are the 4 main equations that we need to solve this problem. We need to nd v
2
in
terms of v
1
. Begin by dierentiating (2) and plugging in equation (4).
dv
1
dv
2
=v
2
(5)
dt dt
Substitute equation (3) into equation (1) and dierentiate
0 = i
s
v
1

dv
1
v
1
+v
2
(6)
dt
di
s
d
2
v
1
dv
1
dv
2
= + 2 (7)
dt dt
2
dt dt
plug equation (5) into (7)
di
s
d
2
v
1
dv
1
= + +v
2
dt dt
2
dt
v
2
=
di
s

d
2
v
1

dv
1
(8)
dt dt
2
dt
Finally, substitute (8) into (6) to obtain the nal dierential equation.
di
s
d
2
v
1
dv
1
+i
s
= + 2 + 2v
1
dt dt
2
dt
(b) i
s
= 1 A, v
1
= 0.5 V, v
2
= 0 V, i
l
= 0.5 A, i
c
= 0 A.
(c)
di
s
d
2
v
1
dv
1
+i
s
= + 2 + 2v
1
dt dt
2
dt
I(s)(s+ 1) i
s
(0

) = V
1
(s)
_
s
2
+ 2s+ 2
_
sv
1
(0

)v
1
(0

)2v
1
(0

)
I(s)i
s
(0

) + v
1
(0

)(s+ 2) + v
1
(0

)
V
1
(s) =
s
2
+ 2s+ 2
389
Plug in the initial conditions i
s
(0

) = 1, v
1
(0

) = .5, and v
1
(0

)=0. Also, I(s) = 0.


1 + .5(s+ 2)
V
1
(s) =
s
2
+ 2s+ 2
.5s
= (9)
s
2
+ 2s+ 2
Equation (9) does not appear in our Laplace transform table. We must manipulate it until
it matches a standard form.
.5s
V
1
(s) =
s
2
+ 2s+ 2
.5s
=
(s+ 1)
2
+ 1
2
s+ 1 1
=.5 .5
(s+ 1)
2
+ 1
2
(s+ 1)
2
+ 1
2
Now we have the transform in a form available on the table. The nal equation is then:
t
e
v
1
(t) = (costsint)
2
(d) The plot is shown in Figure 8.8.1.
0 1 2 3 4 5 6 7 8
Time [s]
Figure 8.8.1: Plot of v
1
(t)
390
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
v
1
(
t
)

[
V
]
,

t

t

0

_ _
(e) Begin by nding the transfer function
V
Is
1
(
(
s
s
)
)
by using a current divider relationship after
combining the inductor and resistor into a single impedance.
1 1 1
I
s
(s) = V
1
+ +
R+Ls R
1
Cs
V
1
1 +s
=
I
s
(s) s
2
+ 2s+ 2
To nd
I
L
(s)
, recognize that
Is(s)
I
L
(s)
=
I
L
(s)V
1
(s)
I
s
(s) V
1
(s)I
s
(s)
Thus, all we need is a relationship between V
1
and I
L
to obtain the correct transfer func-
tion. Impedance provide that when we combine the resistor and the inductor into a single
impedance.
I
L
(s)(R+Ls) = V
1
(s)
I
L
(s)
=
1
V
1
(s) R+Ls
This works out because the resistor and inductor are in series, and thus the current through
them are equal. Therefore
I
L
(s)
=
I
L
(s)V
1
(s)
=
1
I
s
(s) V
1
(s)I
s
(s) s
2
+ 2s+ 2
391
_ _
8.9 LR Circuit Step
V
i
L R
1
R
2
V
o
+
-
+
+
-
-
V
L
V
R1
i
Figure 8.9.1: Circuit Diagram
(a) First, write out constituent relationships for each element:
V
L
=L
di
L
(1)
dt
V
R1
=i
R1
R
1
(2)
V
R2
=V
o
=i
R2
R
2
(3)
From the circuit diagram in Figure 8.9.1, notice that
i=i
L
=i
R1
=i
R2
(4)
However, we want the dierential equation to be in terms of V
i
and V
o
. Apply KVL to the
entire circuit loop to obtain
V
i
V
L
V
R1
V
o
= 0 (5)
To get an expression for V
L
, dierentiate equation (3) with respect to time and substitute
into equation (1), taking into account (4):
V
L
=
L dV
o
(6)
R
2
dt
To get an expression for V
R1
, substitute equation (3) into equation (2):
V
R1
=
R
1
V
o
(7)
R
2
Substitute equations (6) and (7) back into (5) to obtain the dierential equation
V
i

L dV
o

R
1
V
o
V
o
= 0
R
2
dt R
2
L dV
o
+ 1 +
R
1
V
o
=V
i
R
2
dt R
2
392
_ _ __
-5
-4
-3
-2
-1
0
1
2
3
4
5
I
m
a
g
i
n
a
r
y

P
a
r
t

To nd the transfer, substitue sV(s) for each derivative to obtain:
V
o
(s)
L
s+ 1 +
R
1
=V
i
(s)
R
2
R
2
V
o
(s) 1
H(s) = = _ _
V
i
(s) L
s+ 1 +
R
1
R
2
R
2
R
2
1
H(s) =
L
R
1
+R
2
R
1
+R
2
s+ 1
The nal form of H(s) is the same as the form above it. However, it is in standard form for
easier manipulation later. The pole zero plot is shown below.
-15000 -10000 -5000 0 5000
Real Part
Figure 8.9.2: Pole Zero plot of circuit
(b) To nd R
1
, apply the Final Value Theorem to the transfer function, using the nal value .1
asobtainedfromthegraph. RememberthattheLaplaceTransformoftheunitstepfunction
is
1
.
s
R
2
1 1
V
o
() = .1 = lim s
s0
L
s R
1
+R
2
R
1
+R
2
s+ 1
R
2
.1 =
R
1
+R
2
R
1
=9000[]
L
To nd L, recognize that =
R
1
+R
2
from the standard form of a transfer function for an
exponential decay
s
1
+1
. is found from the graph to be 110
4
seconds.
1 10
4
=
L

R
1
+R
2
L= 1 [H]
393


= _
_ _
_ _
(c) The value M is the magnitude of the transfer function evaluated at s = j, and the vaule
is the phase of the transfer function evaluated at s = j. The magnitude of the transfer
function is

V
o
(j)

1
[H(j)[ =

V
i
(j)

= _
_ _
2
_ _
2
R
1
+ 1 +
L

R
2
R
2
1
100+
10
1
6

2
The phase of the transfer function is
H(j)
H(j) = tan
1
1H(j)
L

=tan
1 R
2
R
1
+ 1
R
2
=tan
1
_

_
10000
(d) ThebodeplotisshowninFigure8.9.3. TheDCgainisthemagnitudeofthetransferfunction
as 0, which is 0.1. The high-frequency asypmtotic gain is the magnitude of the transfer
function as , which is 0.
394
Figure 8.9.3: Bode Plot, M and
395
9 Op Amps
9.1 Opamp Arithmetic
The solution for this problem is not available.
396
9.2 Inverting Op-Amp
The solution for this problem is not available.
397
9.3 Opamp Transfer Function
The solution for this problem is not available.
398
9.4 Op-Amp Dierentiator 1
In this problem, we examine a practical dierentiator. Notice that the second op-amp circuit is
simply a unity gain inverter. The transfer function of the circuit is
V
o
(s)
=
RCs
V
i
(s) R
1
Cs+ 1
(a) The zero of the transfer function occurs at s
z
= 0. The pole occurs at s
p
= 1/R
1
C =
0.5rad/s.
(b) Themagnitudeandphaseofthetransferfunctionarefoundbyevaluatingthetransferfunction
at s=j. The magnitude is
RC 0.2
M = _ = _
(R
1
C)
2
+ 1 (2)
2
+ 1
The phase is

= arctanR
1
C= arctan(2)
2 2
(c) The step response is given in Figure 9.4.1. An ideal dierentiator has a transfer function of
1
. ThistransferfunctionisnotidealbecauseoftheresistanceR
1
. Anidealdierentiatorhas
s
no pole, only a zero. The step response of an ideal dierentiator is an impulse.
StepResponse
From:U(1)
0 2 4 6 8 10
Time(sec.)
Figure 9.4.1: Step Response
A
m
p
l
i
t
u
d
e

0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
T
o
:

Y
(
1
)

(d) The Bode plot is given in Figure 9.4.2. The breakpoint frequency, also known as the corner
frequency, is identical to the pole location. The corner frequency is used for sketching the
399
12
asymptotesofthemagnitudeportionofthe Bode plot. Atthecornerfrequency, the transfer

function has a value of 10


2
2
. At the corner frequency, actual plot is 3 decibels below the

asymptote (20log
2
2
3).
BodeDiagrams
From:U(1)
10
2
10
1
10
0
10
1
Frequency(rad/sec)
Figure 9.4.2: Bode Plot
P
h
a
s
e

(
d
e
g
)
;

M
a
g
n
i
t
u
d
e

(
d
B
)
60
55
50
45
40
35
30
25
20
0
20
40
60
80
100
T
o
:

Y
(
1
)

400
9.5 Op-Amp Proportional plus Integral Controller
We are asked to analyze a Proportional plus Integral (PI) circuit. The transfer function for this
circuit is
V
o
(s) R
f
Cs+ 1
=
V
i
(s) R
i
Cs
(a) The pole occurs at s
p
=0. The zero is located at s
z
=1/R
f
C =1.
(b) Themagnitudeandphaseofthetransferfunctionarefoundbyevaluatingthetransferfunction
at s=j. The magnitude is
_
(R
f
C)
2
+ 1
2
+ 1
M = =
R
i
C 0.1
The phase is
3 3
= arctan R
f
C = arctan
2 2
(c) The step response is given in Figure 9.5.1. The transfer function for a PI circuit may be
written as
H
PI
(s) = K
p
+
K
I
s
The transfer function may be re-written as
V
o
(s) R
f
1
=
V
i
(s) R
i
R
i
Cs
The step response will have a non-zero initial value and then grow linearly with time.
StepResponse
From:U(1)
A
m
p
l
i
t
u
d
e

25
20
15
10
5
0
T
o
:

Y
(
1
)

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1


Time(sec.)
Figure 9.5.1: Step Response
401
(d) The Bode plot is given in Figure 9.5.2. The breakpoint frequency, also known as the corner
frequency, is identical to the zero location.
BodeDiagrams
From:U(1)
10
2
10
1
10
0
10
1
Frequency(rad/sec)
Figure 9.5.2: Bode Plot
P
h
a
s
e

(
d
e
g
)
;

M
a
g
n
i
t
u
d
e

(
d
B
)
20
30
40
50
60
70
280
260
240
220
200
180
T
o
:

Y
(
1
)

402
_ _
9.6 Op-Amp Circuit Design 1
We want to design an non-inverting summing circuit whose output is
V
o
= 5V
1
+ 3V
2
10V
3
V3
R5
R4
Vout
R1
R2
R3
V2
V1
Figure 9.6.1: Summing Amplier
We should use the principle of superposition, and we must realize that for adding voltages, we
shouldhavetheinputsgotothenon-invertingterminalandsubtractingavoltageinputsshouldgo
totheinvertingterminal. Thenon-invertingcongurationgivestheadditionofV
1
andV
2
,whilethe
inverting conguration gives the subtraction of V
3
. Together, they can realize the given equation.
The circuit in Figure 9.6.1 gives the desired output for appropriately chosen resistors.
Asalways,westartwiththeequationV
o
=A(s)[V
+
V

]andndV
+
andV

. We then plug them


back into the equation. Using superposition (i.e. setting V
o
= 0 and solving for V

, then setting
V
3
=0 and solving for V

, and summing the two) gives


R
4
R
5
V

= V
o
+ V
3
R
4
+R
5
R
4
+R
5
Using the voltage divider and superposition, we nd
R
2
//R
3
R
1
//R
3
V
+
= V
2
+ V
1
R
2
//R
3
+R
1
R
1
//R
3
+R
2
R
2
R
3
R
1
R
3
= V
2
+ V
1
R
2
R
3
+R
1
(R
2
+R
3
) R
1
R
3
+R
2
(R
1
+R
3
)
Plugging these values back into the equation and solving for V
o
, with A(s) having innite gain,
gives
R
4
+R
5
V
o
=
R
4
R
1
R
3
R
2
R
3
R
5
V
1
+ V
2
V
3
R
1
R
3
+R
2
(R
1
+R
3
) R
2
R
3
+R
1
(R
2
+R
3
) R
4
+R
5
403
The V
3
coecient gives us the constraint that R
5
/R
4
=10. We arbitrarily choose R
4
= 1k and
this then sets R
5
= 10k. We then choose R
3
= 1k, and this leaves 2 equations and 2 unknowns
for the coecients of V
1
and V
2
. Solving the system of equations gives R
1
= 1k and R
2
=600.
The nal solution has the following form:
V
o
=K[C
1
V
1
+C
2
V
2
C
3
V
3
]
where K =11, C
1
= 0.4545, C
2
= 0.2727, and C
3
= 0.9091.
404
9.7 Op-Amp Block Diagram
The solution for this problem is not available.
405
9.8 Op-Amp Circuit Design 2
Vin
R2
R1
C1
C2
Vout
Figure 9.8.1: Circuit implementation of rst stage
ThecircuitinFigure9.8.1givesoneimplementationforthedesiredoutputforappropriatelychosen
resistors and capacitors. There is more than one possible design.
Aspresentedintutorials,thecircuitwithouttheC
2
capacitorgivesapoleattheoriginandazero.
This leaves us to gure out how to get an additional pole. The impedance of a capacitor is 1/Cs
so that may be a good starting point to achieving a pole. We add a capacitor in parallel with the
capacitor and resistor in series and this does the trick.
The impedance across the negative and output terminals is
R
2
Cs+ 1
Z =
s(R
2
C
1
s+C
1
+C
2
)
We know that for an inverting op using the innite gain model, V
out
/V
in
= Z/R
1
. The transfer
function then reduces to
V
o
1 R
2
Cs+ 1
=
V
in
R
1
(C
1
+C
2
)
R
C
2
1
C
+
1
C
C
2
2
s+ 1
We choose C
1
= 0.01F and that sets R
2
= 100k. Solving the rest of the equations, we get
C
2
= 1.11nF and R
1
=900k.
Finally, we add a negative unity gain buer to invert the signal and make it positive overall. This
circuit is cascaded onto the rst and it is shown in Figure 9.8.2.
406
R
Vin
Vout
R
Figure 9.8.2: Negative unity gain buer
407
9.9 Op-Amp Dierentiator 2
The op-amp output is V
out
=A(s)(V
+
V

).
(a) Because the op-amps have a low output impedance, we may model the system as two blocks
cascaded (see Figure 9.9.1).
R
1
R
C
+
_
V
i
+
_
V
1
V
o
R
2
R
2
+
_
Figure 9.9.1: Cascaded Block Circuit
First, consider the dierentiator block. Using the voltage divider relation and the principle
of superposition, the voltage at the inverting terminal is
R
1
+
1
R
V

=
Cs
V
1
+ V
i
R
1
+R+
1
R
1
+R+
1
Cs Cs
The voltage at the non-inverting terminal is zero. The amplier outputs V
1
= A(s)(V

).
The block diagram for this part of the circuit is given in Figure 9.9.2.
_
-V
-
V
i
RCs
(R+R
1
)Cs+1
_
A(s)
V
1
R
1
Cs+1
(R+R
1
)Cs+1
Figure 9.9.2: Block Diagram
Next,considertheinvertingamplierpartofthecircuit. ThevoltageinputisV
1
. Thevoltage
at the non-inverting terminal is zero. The voltage at the inverting terminal is
V
o
+V
1
V

=
2
The output voltage of the op-amp is V
o
=A(s)V

. The block diagram is shown in Figure


9.9.3.
408
A(s)
_
V
o
V
1
-V
-
_
2
Figure 9.9.3: Block Diagram for Inverting Amplier
To obtain the total block diagram, we simply cascade the two blocks. In general, we cannot
simplycascadetwosystemsbecausetherewillinteractionbetweenthetwosystems. Withtwo
cascaded blocks, the second block has no eect on the rst. However, because the op-amps
haveaverysmalloutputimpedance,wemaycascadethetwoblocks. Thetotalblockdiagram
is shown in Figure 9.9.4.
_
V
i
RCs
(R+R
1
)Cs+1
_
A(s)
R
1
Cs+1
(R+R
1
)Cs+1
_ _
A(s) V
o
2
Figure 9.9.4: Block Diagram of Cascaded System
(b) Assuming A(s) = g/s, the transfer function for the rst block (practical dierentiator) is
A(s)RCs
H
1
(s) = _ _
(R+R
1
)Cs+ 1 + A(s)(R
1
Cs+ 1)
A(s)
H
2
(s) =
2
1 +
A(s)
2
Substituting A(s) = g/s, we obtain
gRCs
H
1
(s) = _ _
s (R+R
1
)Cs+ 1 + g(R
1
Cs+ 1)
and
1
H
2
(s) =
2
s+ 1
g
The total transfer function is
V
o
(s)
=
1.25710
7
s
V
i
(s) 7.00310
8
s
3
+ 6.2s
2
+ 1.25710
8
s+ 6.28310
7
409
(c) The pole-zero plot is shown in Figure 9.9.5.
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
I
m
a
g
i
n
a
r
y

A
x
i
s

-6 -5 -4 -3 -2 -1 0
Real Axis 7
x 10
Figure 9.9.5: Pole Zero Plot
Thezeroislocatedats
z
=0. Thepolesarelocatedats
p
=0.5,3.143710
7
,5.709610
7
.
The dominant pole is s=0.5. The transfer function for the dominant roots is of the form
s
H(s) =
s+ 0.5
In other words, the approximate transfer function is a non-ideal dierentiator. The step
response for the approximate system is shown in Figure 9.9.6.
410
Step Response
A
m
p
l
i
t
u
d
e

0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0 2 4 6 8 10 12
Time (sec)
Figure 9.9.6: Step response ignoring higher-order dynamics
(d) The step response of the exact transfer function is shown in Figure 9.9.7. We see that step
response is the same as for the system including only the dominant dynamics except for the
rst.02seconds. Thereasonthattheactualsystemtakesaniteamountoftimetoapproach
the ideal behavior is because of the high frequency dynamics. As seen from the Bode plot
(see Figure 9.9.8), the system has a very small response at high frequencies (above 1MHz).
The step response excites all frequencies, but the contribution to the response is very small
when the dynamics of the op-amps are considered.
411
0.05
A
m
p
l
i
t
u
d
e

Step Response Step Response
0.1 0.1
0.09 0.09
0.08 0.08
0.07 0.07
0.06 0.06
A
m
p
l
i
t
u
d
e
0.05
0.04 0.04
0.03 0.03
0.02 0.02
0.01 0.01
0 0
0 2 4 6 8 0 0.005 0.01 0.015 0.02 0.025 0.03
Time (sec) Time (sec)
Figure 9.9.7: Step response of exact transfer function
Bode Diagram
-1 0 1 2 3 4 5 6 7 8 9
10 10 10 10 10 10 10 10 10 10 10
Frequency (rad/sec)
Figure 9.9.8: Bode Plot of full Transfer Function
412
M
a
g
n
i
t
u
d
e

(
d
B
)

-80
-70
-60
-50
-40
-30
-20
P
h
a
s
e

(
d
e
g
)

-180
-135
-90
-45
0
45
90
9.10 RC Op-Amp Frequency Response
Z
1
Z
2
V
o
V
i
5
8

-
+
+
-
V
i
R
1
+
-
V
o
R
2
R
3
C
2
C
1
-1
-1
A
200 2000 5000
Figure 9.10.1: RC Op-Amp Circuit
(a) Figure 9.10.1 shows the circuit for this problem. I solved this using impedance methods.
BothZ
1
andZ
2
sharethesamecurrenti. With a(s) = ,wecanmaketheassumptionthat
V

= 0.
_ _
1
1 C
1
s R
1
R
3
C
1
s+R
1
Z
1
= + =
R
1
R
3
C
1
s+ 1 C
1
(R
1
+R
3
)s+ 1
1 R
2
C
2
s+ 1
Z
2
= R
2
+ =
C
2
s C
2
s
C
1
(R
1
+R
3
)s+ 1
I = V
i
R
1
R
3
C
1
s+R
1
C
2
s
I = V
o
R
2
C
2
s+ 1
V
o
(R
2
C
2
s+ 1)(C
1
(R
1
+R
3
)s+ 1)
=
V
i
C
2
s(R
1
R
3
C
1
s+R
1
)
(b) Our transfer function has two poles and two zeros. We need to associate these poles and
zeros with the breakpoints on the frequency response. We have a pole at the origin which
is associated with the initial negative slope. We then have a zero at 200 r/s (slope goes to
413
zero). We then have a 2nd zero at 2000 r/s and the 2nd pole at 5000 r/s. To start with we
notethatthelowfrequencyresponseisdominatedbythepoleattheorigin. Wecanexpress
the low frequency magnitude as
1
M(j) =
C
2
R
1

SincewearegivenR
1
if we know the magnitude at some low frequency we can nd C
2
. Well
we know the magnitude at 200 is 5 and thus using just the low frequency asymptote
1 1
5 = C
2
= = 0.1F
R
1
C
2
200 510000200
Now we can tie the (R
2
C
2
s+ 1) zero or (C
1
(R
1
+R
3
)s+1) with the breakpoint at 200. I
chose the following:
1 1
200= R
2
= = 50 k.
R
2
C
2
2000.1E6
Now all that is left is the zero at 2000 and the pole at 5000, thus:
1
2000 =
C
1
(R
1
+R
3
)
1
5000 =
R
3
C
1
2000C
1
(R
1
+R
3
) = 5000R
3
C
1
3000R
3
=2000R
1
2
R
3
= R
1
=6667
3
1
C
1
= = 0.03F
5000R
3
Now we need to gure out A. Well we know that the slope of magnitude between 2000 and
5000 is +1. Thus
A = 10
log(5)+x
x = log(5000)log(2000)=3.6993.301=0.395
A = 10
(0.7+0.395)
= 12.5
Alternately, you could use the fact that at high frequencies
R
2
(R
1
+R
3
) 50000(16667)
M() = = = 12.5
R
1
R
3
100006667
Figure 9.10.2 shows the actual frequency response of this circuit with the asymptotes.
If you choose to set the rst zero using (C
1
(R
1
+R
2
)s+1) you would get the following
component values:
R
2
= 5000
R
3
= 417
C
1
= 0.48F
414
10
2
10
3
10
4
10
1
Figure 9.10.2: Frequency Response
415
10 Dierential Equations
10.1 Dierential Equations 1
(a) 500 y+1000y= 0
y
h
(t) = c
1
e
s
1
t
and
dy
h
=c
1
s
1
e
s
1
t
dt
500c
1
s
1
e
s
1
t
1000c
1
e
s
1
t
= 0
(500s
1
+1000)c
1
e
s
1
t
= 0
therefore s
1
=2
Im
-2
Re
x
Figure 10.1.1: Plot of poles for part (a)
(b) y(3)=10
y
h
(t) = c
1
e
s
1
t
c
1
e
6
= 10
10
c
1
=
e
6
10
y
h
(t) = e
2t
6
e
(c) y
h
(t)<10
6
10
e
2t
<10
6
e
6
e
2t
<10
7
e
6
2t < ln(10
7
e
6
)
t > 5.06
(d) Usingunitsofmetersfory,wehavethefollowingplotshowninFigure10.1.2,madewiththe
script di2sol.m found at the end of the solution.
416
500dy/dt+1000y=0 with y(-3 s)=10 m
-4 -3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1
time (s)
Figure 10.1.2: Plot for Part (d)
(e) One physical system would be a spring-dashpot system like the one modeled in Lab 1. The
variableywouldrepresentthedisplacementinunitsofmeters(m). Thedashpotwouldhave
aviscousdampingconstant(b)of500Ns/m,andthespringwouldhaveaspringconstant(k)
of 1000 N/m. First order systems are not just limited to springs and dashpots but describe
electrical systems, uid ow, and many other natural behaviors. Figure 10.1.3 is taken from
the ActivLab website.
0
10
20
30
40
50
60
70
80
p
o
s
i
t
i
o
n

(
m
)

y(t)
b
Figure 10.1.3: Physical System
diplot1sol.m
Download
t=[-4:.01:1];
y=10*exp(-6)*exp(-2*t);
417
figure(2)
plot(t,y)
title(500dy/dt+1000y=0 with y(-3 s)=10 m)
xlabel(time (s))
ylabel(position (m))
418
_
_
10.2 Dierential Equations 2
(a) y +10 y+10000y= 0
k b

n
= and = and
d
=
n
1
2
m
2 km
so

n
=100rad/s
= 0.05

d
= 99.9rad/s
(b) y(0.1)=3 and y(0.1)=0
From lecture for an underdamped system:
y
h
(t) = c
1
e
t
cos
d
t+c
2
e
t
sin
d
t
so
y
h
(t) = c
1
e
t
cos
d
tc
1

d
e
t
sin
d
tc
2
e
t
sin
d
t+c
2

d
e
t
cos
d
t
Frompart(a),weknow
n
,, and
d
. We also know =
n
. Thereforewecansolveforc
1
and c
2
based on the initial conditions.
c
1
= 4.051
c
2
= 2.848
2 2
d y/dt +10dy/dt+1000y=0 with dy/dt(0.1 s)=0 m/s and y(0.1 s)= -3 m
-0.5 0 0.5 1
time (s)
(c)
Figure 10.2.1: Plot of Position vs. time
We can zoom in on the plot shown in Figure 10.2.1 near the given initial conditions to look
at y and y at t=.01s. This zoomed plot is shown in Figure 10.2.2.
419
-60
-40
-20
0
20
40
60
p
o
s
i
t
i
o
n

(
m
)

2 2
d y/dt +10dy/dt+1000y=0 with dy/dt(0.1 s)=0 m/s and y(0.1 s)= -3 m
-4
-3
-2
-1
0
1
2
3
4
p
o
s
i
t
i
o
n

(
m
)

0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18
time (s)
Figure 10.2.2: Zoomed Plot of Position vs. time
If we look at Figure 10.2.2 at t = 0.1s, we can see that the position is approximately -3 m.
Because the position curve is at a peak, we can also conclude that the velocity y is 0.
(d) One physical system would be a mass-spring-dashpot system like the one shown below. The
variable y wouldrepresentthedisplacementinunitsofmeters(m). Thedashpotwouldhave
a viscous damping constant (b) of 10 Ns/m, the spring would have a spring constant (k) of
10000 N/m, and the mass (m) would be 1 kg. Second order systems are not just limited
to masses, springs, and dashpots but describe electrical systems, uid ow, and many other
natural behaviors. Figure 10.2.3 below is taken from the ActivLab website.
420
b
y(t)
Figure 10.2.3: Physical System
di2sol.m
Download
%diff2sol.m, script for Archive Problem Differential Equations 2
figure(1)
k=10000;
m=1;
b=10;
wn=sqrt(k/m);
zeta=b/(2*sqrt(k*m));
wd=wn*sqrt(1-zeta^2);
sigma=zeta*wn
s1=-sigma+j*wd
s2=-sigma-j*wd
t=[-.5:.001:1];
%To solve for c1 and c2, the following may be used:
%c1 and c2 can also be solved manually
%Bx=C
t1=.1;
B=[exp(-t1*sigma)*cos(wd*t1) exp(-t1*sigma)*sin(wd*t1);
-sigma*exp(-t1*sigma)*cos(wd*t1)-wd*sin(wd*t1)*exp(-sigma*t1) -sigma*exp(-t1*sigma)*sin(wd*t1)+wd*cos(wd*t1)*exp(-sigma*t1)]
C=[-3
0];
x=inv(B)*C
c1=x(1)
c2=x(2)
y=c1*exp(-sigma*t).*cos(wd*t)+c2*exp(-sigma*t).*sin(wd*t);
plot(t,y)
title(d^2y/dt^2+10dy/dt+1000y=0 with dy/dt(0.1 s)=0 m/s and y(0.1 s)= -3 m)
xlabel(time (s))
ylabel(position (m))
figure(2)
t=[.1:.001:1];
y=c1*exp(-sigma*t).*cos(wd*t)+c2*exp(-sigma*t).*sin(wd*t);
plot(t,y)
title(d^2y/dt^2+10dy/dt+1000y=0 with dy/dt(0.1 s)=0 m/s and y(0.1 s)= -3 m)
xlabel(time (s))
ylabel(position (m))
421
_
_
-30
-20
-10
0
10
20
30
I
m
a
g
i
n
a
r
y

A
x
i
s

10.3 Dierential Equations 3
(a) y +200 y+10000y= 0
k b

n
= and = and
d
=
n
1
2
m
2 km
so

n
=100 rad/s
= 1

d
= 0 rad/s
s plane plot of poles
-150 -100 -50 0 50
Real Axis
Figure 10.3.1: Poles for Part (a)
(b) y(0) = 1 and y(0)=1
For the critically damped case:
y
h
(t) = c
1
e
t
+c
2
te
t
y
h
(t) = c
1
e
t
+c
2
e
t
c
2
te
t
=
n
so
422
c
1
=1 given y(0)=1
c
2
=101 given y(0)=1
therefore
y
h
(t) = e
100t
+101te
100t
Looking at the last plot of Figure 10.3.2, we can see that both y(0) and y(0) are 1 at t= 0.
2 2
d y/dt +200dy/dt+1000y=0 with dy/dt(0 s)=1 m/s and y(0 s)= 1 m
-1000
-500
0
500
p
o
s
i
t
i
o
n

(
m
)

-0.05 -0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05
time (s)
0
0.5
1
p
o
s
i
t
i
o
n

(
m
)

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09
time (s)
0.999
0.9995
1
1.0005
1.001
p
o
s
i
t
i
o
n

(
m
)

-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
time (s)
x 10
-3
Figure 10.3.2: Plots for Part (b)
di3sol.m
Download
%diff3sol, a script for Problem Archive problem Differential Equations 3
figure(1)
plot(-100,0,x)
axis([-150 50 -30 30])
figure(6)
t=[-.05:.001:.05];
y=exp(-100*t)+101*t.*exp(-100*t);
subplot(3,1,1)
plot(t,y)
title(d^2y/dt^2+200dy/dt+1000y=0 with dy/dt(0 s)=1 m/s and y(0 s)= 1 m)
423
0.1
xlabel(time (s))
ylabel(position (m))
t=[0:.001:.1];
y=exp(-100*t)+101*t.*exp(-100*t);
subplot(3,1,2)
plot(t,y)
xlabel(time (s))
ylabel(position (m))
t=[-.01:.0001:.01];
y=exp(-100*t)+101*t.*exp(-100*t);
subplot(3,1,3)
plot(t,y)
xlabel(time (s))
ylabel(position (m))
axis([-.002 .002 .999 1.001])
424
10.4 Eulers Theorem 1
Three Forms of Eulers Theorem:
e
j
= cos +jsin
_ _
cos=
_
e
jt
+e
jt
_
2
_ _
sin=
e
jt
e
jt
_ _
2j
(a)
ce
jt
+c

e
jt
= Acost+Bsintwhere c=+j and c

=j
ce
jt
+c

e
jt
= c(cost+jsint)
= (+j)(cost+jsint) + (j)(costjsint)
= cost+jsint+jcostsint+costjsintjcostsint
= 2cost2sint
= Acost+Bsint
so
A B
= and =
2 2
A B

A B
c= j and c = +j
2 2 2 2
Or, if you dont like that method:
_ _ _ _

e
jt
+e
jt

e
jt
e
jt


Acost+Bsint = A_ _+B_ _
2 2j
A
_ _
B
_ _
= e
jt
+e
jt
+ e
jt
e
j
2 2j
_ _ _ _

A B
jt
+

A B
jt
=
_
+
_
e
_
+
_
e
2 2j 2 2j
therefore
A B

A B
c= + c =
2 2j 2 2j
A Bj A Bj
= + =
2 2jj 2 2jj
A B A B
= j = +j
2 2 2 2
where c

is the complex conjugate of c.


(b) c=
7
2
j and c

=
7
2
+j
(c) Using units of meters for A and B and rad/s for , we get the plots seen in Figure 10.4.1.
The following MATLAB script creates this plot.
425
-10
-5
0
5
10
p
o
s
i
t
i
o
n

(
m
)

-10
-5
0
5
10
p
o
s
i
t
i
o
n

(
m
)

Acos(wt)+Bsin(wt) where A=7 m, B=2 m, w=100 rad/sec
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
time (s)
ce
jwt
+c
*
e
jwt
where c=A/2-B/2j , c
*
=A/2+B/2j, w=100 rad/s
0.18 0.2
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
time (s)
Figure 10.4.1: Plots for Part (c)
euler1sol.m
Download
A=7;
B=2;
w=100;
t=[0:.0001:2*pi*3/100];
y=A*cos(w*t)+B*sin(w*t);
figure(1)
subplot(2,1,1)
plot(t,y)
title(Acos(wt)+Bsin(wt) where A=7 m, B=2 m, w=100 rad/sec)
xlabel(time (s))
ylabel(position (m))
subplot(2,1,2)
%or plotting ce^(jwt)+c*e^(-jwt) we have
c=A/2-B/2*j;
cstar=A/2+B/2*j;
y=c*exp(j*w*t)+cstar*exp(-j*w*t);
plot(t,y)
title(ce^{jwt}+c^*e^{jwt} where c=A/2-B/2j , c^*=A/2+B/2j, w=100 rad/s)
xlabel(time (s))
ylabel(position (m))
0.16 0.18 0.2
426
_ _
10.5 Eulers Theorem 2
(a)
_ _
j(t+)
+e
j(t+)
e

Mcost+=M

_
2
_
M
= e
jt
e
j
+e
jt
e
j
2
=
M
e
jt
e
j
+
M
e
jt
e
j
2 2
From Problem 10.4:
_ _ _ _

e
jt
+e
jt

e
jt
e
jt


Acost+Bsint=A_ _+B_ _
2 2j
so
_ _ _ _
A B M A B M
jt j jt jt j jt
_
+
_
e = e e and
_

_
e = e e
2 2j 2 2 2j 2
_ _ _ _
A B M A B M
j j
_
+
_
= e
_

_
= e
2 2j 2 2 2j 2
therefore
M A B
(cos+jsin) = j
2 2 2
Mcos=A
Msin=B
M
2
_ _
=A
2
+B
2
cos
2
+ sin
2
B
tan=
A
(b)
c
M = 4(1c)
2
+ 4(c)
2
andtan=
1c
M is the amplitude of the waveform in Problem 10.4.
427
_
_
_
10.6 Force Steps
Inthisproblem,weexaminethebehaviorofthethreebasicmechanicalelementswhenastepforce
is applied. Assuming zero initial position and velocity of the elements, for the mass
mx
m
=u
s
(t) (1)
t
1 1
x
m
=v
m
= u
s
(t)dt= t (2)

m m
t
1 1
2
x
m
= tdt= t (3)
m 2m

For the spring,


u
s
(t) = f
k
(t) = kx
k
(t) (4)
1
x
k
(t) = u
s
(t) (5)
k
d 1
v
k
(t) = x
k
(t) = (t) (6)
dt k
For the dashpot,
f
c
(t) = cx
c
=u
s
(t) (7)
1
x
c
=v
c
= u
s
(t) (8)
c
t
1 t
x
c
= u
s
(t)dt= (9)

c c
velocityofmass
positionofmass
positionofspring
velocityofdashpot
positionofdashpot
Figure 10.6.1: Step Response of Mass, Spring and Dashpot
0
5
10
0
50
0
1
2
velocityofspring
0
1
2
0
1
2
1 0 1 2 3 4 5 6 7 8 9 10
0
5
10
428
10.7 Input for Zero Output
The solution for this problem is not available.
429
_
_
_ _
_
10.8 Zero Input Response
In this problem we would like to solve for the Zero Input Response (ZIR) of the systems given.
First, we start with the rotational system. The governing equation is
d c
1
+c
2
+ = 0
dt J
Solving this equation, we obtain
(t) = Ae
t/
where = J/(c
1
+c
2
) and A is determined from initial conditions. Using the initial condition

(0)=(0)=
0
we nd A=
0
. To solve for the position (t), we integrate (t) to nd
(t) =

0
e
t/
+C

Using the initial condition (0)=


0
, we nd C =

0
+
0
. Thus, the total response is
(t) =
0
+

0
1e
t/

For the translational system, we begin with the governing equation


Mx + (c
1
+c
2
) x+ (k
1
+k
2
)x= 0
with the initial conditions x(0) = x
0
and x(0) = v
0
. Dividing the equation by M, we nd the
natural frequency to be
k
1
+k
2

n
=
M
The damping ratio is
c
1
+c
2
= _
2 M(k
1
+k
2
)
Assuming the system is underdamped ( < 1), we nd the damped natural frequency to be
_
k
1
+k
2
(c
1
+c
2
)
2

d
=
n
1
2
=
M 4M
2
To solve for x(t),weassumeaharmonicsolutionx(t) = Ae
st
. Thegoverningequationsimpliesto
an algebraic equation
ms
2
+ (c
1
+c
2
)s+ (k
1
+k
2
) = 0
or equivalently,
s
2
+ 2
n
s+
n
2
= 0
Solving this quadratic equation, we obtain
_
c
1
+c
2
(c
1
+c
2
)
2
k
1
+k
2
=
n
j
n
1
2
=
2M 4M
2
M
Assuming the underdamped solution, the response x(t) is
x(t) = A
1
e
s
1
t
+A
2
e
s
2
t
430
_ _
_ _
_
At t= 0, we have
x(0)=x
0
=A
1
+A
2
and x (0)=v
0
=A
1
s
1
+A
2
s
2
After some algebra, we arrive at the solution
_
v
0
+
n
x
0
_
x(t) = e
nt
x
0
cos( 1
2

n
t) + _ sin( 1
2

n
t)

n
1
2
Other convenient forms exist. For example, we could combine the cos and sin terms into a single
cos term with a phase.
x(t) = X
0
e
nt
cos 1
2

n
t
where

_ _
_ _
2

v
0
+
n
x
0
x
X
0
=
_
x
2
0
+

n
_
1
2
and = tan
1
_
v
0
+

0
nx
0
_
n 1
2
431
_ _
_ _
_ _
_ _ __
_ _
_ _
10.9 Zero Step Response 1
The forced-response with zero initial conditions is composed of the particular solution and the
homogeneous solution.
First, we solve the rotational system. The governing equation is
d c
1
+c
2
u
s
(t)
+ =
dt J J
The particular solution is given by
p
=
c
1
+
1
c
2
. The homogeneous solution is
h
= Ae
t/
. The
total solution is
(t) = Ae
t/
+
1
c
1
+c
2
Using the initial condition that (0)=0, we obtain
(t) =
1
1e
t/
c
1
+c
2
Integrating this equation, and using the initial condition that (0)=0, we obtain
t
(t) = + e
t/
1
c
1
+c
2
c
1
+c
2
where =J/(c
1
+c
2
).
For the translational system, the particular solution is x
p
=F/(k
1
+k
2
). The total solution is
F
x(t) = + A
1
e
s
1
t
+A
2
e
s
2
t
k
1
+k
2
Solving using the initial conditions, we obtain
A
1
=
F
_
j
1
2(k
1
+k
2
)
1
2
A
2
=
F
_
j
1
2(k
1
+k
2
)
1
2
After some algebra, we obtain
F
nt
sin
d
t
x(t) = 1e cos
d
t+ _
k
1
+k
2 1
2
c
1
+c
2
where =
2Mn
,
n
= (k
1
+k
2
)/M, and
d
=
n
1
2
. Again,variousequivalentformsmay
be found using trigonometric identities.
432
10.10 Zero Step Response 2
(a) The unit step response for the rotational system is given in Figure 10.10.1. To aid in the
sketching the response, nd the dierence between
p
t and x(t) as t . This quantity
representstheverticaldistancebetween
p
tandx(t)intheresponse. Theinitialslopeiszero
because the initial velocity is zero.
The unit step response for the translational system is given in Figure 10.10.2. To aid in
sketching, draw the envelope dened by the real part of s and use the damped natural
frequencytodeterminehowfasttheresponseoscillatesinsideoftheenvelope. Wealsoknow
from the initial conditions that the initial slope is zero because the initial velocity is zero.
4 StepResponse
x10
0 50 100 150 200 250
Time(sec)
Figure 10.10.1: Sketch of the step response for the rotational system
A
m
p
l
i
t
u
d
e

0
1
2
3
4
5
6
J/(c1+c2)
433
P
o
s
i
t
i
o
n

[
m
m
]

0 0.1 0.2
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
3
x10
Time[s]
Figure 10.10.2: Sketch of the step response for the translational system
(b) TheunitstepresponsefortherotationalsystemisgiveninFigure10.10.3. Thestepresponse
for the translational system is given in Figure 10.10.4. Both responses were made with the
Matlab script zsrsol.m found at the end of this solution.
x10
4
A
n
g
l
e

[
r
a
d
i
a
n
s
]

0
1
2
3
4
5
6
0 50 100 150 200 250
Time[s]
Figure 10.10.3: Unit step response (t) for the rotational system
434
P
o
s
i
t
i
o
n

[
m
m
]

0 0.1 0.2
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
3
x10
Time[s]
Figure 10.10.4: Unit step response x(t) for the translational system
zsrsol.m
Download
%Justin Verdirame
%%--Rotary System
J=0.1;
C1=1E-3;
C2=3E-3;
num_rot=[1];
den_rot=[J C1+C2 0];
sys_rot=tf(num_rot, den_rot);
step(sys_rot);
%%--Translational System
M=10;
K1=10^5;
K2=8E5;
C1=300;
C2=C1;
num_trans=[1e3];
den_trans=[M C1+C2 K1+K2];
t=0:0.0001:0.2;
figure(2)
sys_trans=tf(num_trans,den_trans);
step(sys_trans,t);
435
10.11 Nonlinear String System
The solution for this problem is not available.
436

10.12 Homogenous and Particular Solutions


(a) Figure 10.12.1 shows the model for inertia and damper system. The arrows in the gure
show the reference directions for the physical variables. Separating the model into parts for
each element, we can get the free body diagram as in Figure 10.12.2. Note the two T
d
have
equalvaluesbutoppositedirections,becausetheyareapairofactingforceandreactingforce.
According to Newtons second law, we can derive the dierential equation for the inertia
element:
I = T =T
in
T
d
. (1)
For the damping, the constitute relation is:
T
d
=C (2)
Combining equation 1 and 2, we can get
I + C=T
in
(3)

in
Figure 10.12.1: System Model
437

in
d
d

Figure 10.12.2: Free Body Diagram
(b) The characteristic equation is
Is+C = 0 (4)
The solution is s=
C
I
, so the homogenous solution is
h
=Ae
st
=Ae

C
I
t
. The parameter
A will be determined by the initial condition. To get the particular solution for input T
in
=
Ku
s
(t),weguessthatthesolutionhastheform
p
=B. Pluggingitintoequation 3,weget
BC =K. So B =K/C,andtheparticularsolutionis
p
=K/C. Thecompletesolutionis
K
+Ae

C
I
t
(5) =
C
With the initial condition, we have

0
=
K
C
+A. (6)
A=
0

K
C
. (7)
The complete solution is
K K
= (
0
)e

C
I
t
+
C C
. (8)
When
0
=
K
, there will be no transient.
C
438
0
5
10
15
20
25
30
35
40
a
n
g
u
l
a
r

v
e
l
o
c
i
t
y

[
r
a
d
/
s
e
c
]

Initialvelocity=0[rad/sec]
Initialvelocity=40[rad/sec]
(c) For K = 10
2
Nm, I = 0.000317kgm
2
and c= 0.0011N msec/rad,
(i) when
o
= 0rad/sec, =9.0909e
3.4700t
+ 9.0909,
(ii) when
o
= 40rad/sec, = 30.9091e
3.4700t
+ 9.0909.
The result is plotted in Figure 10.12.3.
0 0.5 1 1.5 2 2.5 3
time[sec]
Figure 10.12.3: Step Response
(d) (i) Forcosineinput,weguessthattheparticularsolutiontakestheformof
p
=B
1
cos(t)+
B
2
sin(t). Plugging it back into the dierential equation 3,
I(B
1
sin(t) + B
2
cos(t))+C(B
1
cos(t) + B
2
sin(t))=Kcos(t) (9)
IB
1
+CB
2
= 0 IB
2
+CB
1
=K (10)
KC
B
1
=
C
2
+I
2

2
(11)
KI
B
2
=
C
2
+I
2

2
(12)
The complete solution is
KC KI

p
=
C
2
+I
2

2
cos(t) +
C
2
+I
2

2
sin(t) (13)
(ii) Incomplexfomat,thecosineinputcanbeexpressedasT
in
=Kcost=ReKe
jt
, we
guess that the particular solution takes the form of
p
=ReBe
jt
. Plugging it back
into the dierential equation 3,
Re(BI(j)+CB)e
jt
=ReKe
jt
(14)
K
B = (15)
C+jI
439
The complete solution is

p
=Re
K
e
jt
(16)
C+jI
KC KI
=
C
2
+I
2

2
cos(t) +
C
2
+I
2

2
sin(t) (17)
(e) The complete solution is
=
h
+
p
(18)
KC KI
=Ae

C
I
t
+ cos(t) + sin(t) (19)
C
2
+I
2

2
C
2
+I
2

2
KC
For zero initial condition, we can get A=
C
2
+I
2

2
,
KI KC

C
I
t
) + (cos(t)e sin(t) (20) =
C
2
+I
2

2
C
2
+I
2

2
(21)
The total solution is plotted in Figure 10.12.4. From the gure we can see that the out-
put amplitude at = 0.5rad/sec is much larger than the output at = 20rad/sec. For
= 0.5rad/sec, the homegeous solution have signicant eect in the rst cycle. For =
20rad/sec,theeectofthehomegeoussolutionisnegligible. Fromthetotalsolutionexpres-
sion, at high frequency the system signicantly attenuate the output sinusoidal magnitude ,
sin component will dominate the output and thus the phase shift will be 90 degree.
10
8
6
4
2
0
2
4
6
8
10
a
n
g
u
l
a
r

v
e
l
o
c
i
t
y

[
r
a
d
/
s
e
c
]

frequency=0.5[rad/sec]
frequency=20[rad/sec]
0 2 4 6 8 10 12 14 16 18 20
time[sec]
Figure 10.12.4: Sinusoidal Response
440
11 Frequency Response
11.1 Circuit Bode Plots
The solution for this problem is not available.
441


_ _


_ _
11.2 Sketch Bode Plots
1
-2 -1 0 1 2 3
10 10 10 10 10 10
10
-1
10
0
10
M
a
g
n
i
t
u
d
e

-2 -1 0 1 2 3
10 10 10 10 10 10
Frequency (rad/s)
Figure 2: Frequency response for G(s) =
5s+1
s+10
x
2
(t) = A

X
2
(j)

sin t+
X
2
(j)

X
2

k
2

(j)

= _
U (k
1
k
2
b
1
b
2

2
)
2
+ ([k
1
b
2
+b
2
(k
1
+k
2
)])
2

X
2
(j) = tan
1
[k
1
b
2
+b
1
(k
1
+k
2
)]
U k
1
k
2
b
1
b
2

2
Problem 2
Figures 2-6 show the frequency response plots for this problem. The solid
black line on the plots is the asymptotic frequency response, while the
dashed red line is the actual frequency response. In order to sketch, the
frequency response you need to determine the frequencies of the poles and
zeros of the transfer function. You also need to nd the magnitude of at
least one frequency often either at = 0 or .
A) G(s) =
5s+1
= 5
s+0.2
, pole @ -10, zero @ -0.2. [G(0)[= 0.1,
s+10 s+10
[G()[= 5.
5s+1 s+0.2
B) G(s) =
s
2
+3s+1
= 5
(s+0.38)(s+2.62)
, poles @ -.38 and -2.62, zeros @ -0.2,
[G(0)[= 1, G() = 0
C) G(s) =
s+10
, poles @ -2 and -58.66j,
n1
=10 , zeros @
(s+2)(s
2
+10s+100)
-10, [G(0)[= 1/20, G() = 0
Figure 11.2.1:
442
0
20
40
60
80
P
h
a
s
e

(
d
e
g
)

1
-2 -1 0 1 2
10 10 10 10 10
10
-1
10
0
10
M
a
g
n
i
t
u
d
e

-50
0
50
P
h
a
s
e

(
d
e
g
)

-2 -1 0 1 2
10 10 10 10 10
Frequency (rad/s)
5s+1
Figure 3: Frequency response for G(s) =
s
2
+3s+1
-1
-1 0 1 2
10 10 10 10
10
-4
10
-3
10
-2
10
M
a
g
n
i
t
u
d
e

-150
-100
-50
0
P
h
a
s
e

(
d
e
g
)

-1 0 1 2
10 10 10 10
Frequency (rad/s)
s+10
Figure 4: Frequency response for G(s) =
(s+2)(s
2
+10s+100)
Figure 11.2.2:
443
_
_ _
_
11.3 LC Circuit Frequency Response
(a) Itiseasiesttouseimpedancesforthisproblem. Combinethecapacitorandinductorintoan
equivalent impedance Z.
Ls
Z =
Cs
=
Ls
Ls+
1
LCs
2
+ 1
Cs
TheconstitutiverelationshipforthisequivalentimpedanceisV
z
(s) = ZI
Z
(s). Bydenition,
V
Z
(s) = V
c
(s) and I
Z
(s) = I
s
(s). Combine these equations and take the inverse Laplace
transform to derive the dierential equation.
Ls
V
c
(s) = I
s
(s)
LCs
2
+ 1
(LCs
2
+ 1)V
c
(s) = LsI
s
(s)
LCv
c
+v
c
=LI

s
ThecharacteristicequationforthisdierentialequationisLCs
2
+ 1. This is of the standard
form

1
2
s
2
+

2
n

+ 1. This yields
n
1

n
= [rad/sec]
LC
= 0
1 1

d
= 1
2
= [rad/sec]
LC LC
=
n
=0 [1/sec]
The pole zero plot is shown in Figure 11.3.1
LC
1
LC
1
_
Re
Im
Figure 11.3.1: Pole Zero Plot
(b) Start with the dierential equation and take the Laplace transform, including initial condi-
444
_
_ _
tions.
LCv
c
+v
c
=LI

s
V
c
(s)(LCs
2
+ 1) LCv
c
(0

)LCv
c
(0

) = LsI(s)I(0

)
LsI(s) + LCv
c
(0

)
V
c
(s) =
(LCs
2
+ 1)
We need an expression for v
c
(0

). We know that i
L
(0

) = 1 [A]. Summing currents at the


top node with I
s
= 0, we nd that i
c
= i
L
= 1 [A]. By using the constitutive relationship
for a capacitor we nd that v
c
(0

) = 1/C. Therefore
LsI
s
(s) + L
V
c
(s) =
(LCs
2
+ 1)
SubstituteI
s
(s) = 1/sandtaketheinverseLaplacetransformtondanexpressionforv
c
(t).
2L
V
c
(s) =
(LCs
2
+ 1)
2
=
C
s
2
+
1
LC
_
1
L LC
= 2
C s
2
+
1
LC
L 1
v
c
(t) = 2 sin t
C LC
This function is plotted in Figure 11.3.2
2
C
LC
2
C
LC
_
V
c
[V]
t[s]
T=2 LC
Figure 11.3.2: Step Response with v
c
(0) = 0 [V] and i
L
(0)=1 [A]
445



(c) M is simply the magnitude of the transfer function evaluated at s=j.

V
c
(s)

L
M = =
I
s
(s) [1LC
2
[
The phase is
= tan
1
L
tan
1
0
0 1LC
2
1 1
= 90

,
2
< and 90

,
2
>
LC LC
(d) The plot is shown in Figure 11.3.3
Bode Diagram
M
a
g
n
i
t
u
d
e

(
d
B
)

10 L
99 C
P
h
a
s
e

(
d
e
g
)

90
45
0
45
90
.1
1
10
LC
LC
LC
Frequency (rad/sec)
Figure 11.3.3: Bode plot
446
_
_

_
11.4 LRC Circuit Frequency Response
(a) Itiseasiesttouseimpedancesforthisproblem. Combinetheresistor,capacitorandinductor
into an equivalent impedance Z.
1
Z =R//Ls//
Cs
1 1 1
= + +Cs
Z R Ls
1
Z =
1 1
+ +Cs
R Ls
Ls
=
LCs
2
+
L
s+ 1
R
The constitutive relationship for this equivalent impedance is V
z
(s) = ZI
z
(s). By denition,
V
z
(s) = V
c
(s) and I
z
(s) = I
s
(s). Combine these equations and take the inverse Laplace
transform to derive the dierential equation.
Ls
V
c
(s) = I
s
(s)
LCs
2
+
L
s+ 1
R
(LCs
2
+
L
s+ 1)V
c
(s) = LsI
s
(s)
R
LCv
c
+
L
v
c
+v
c
=LI

s
R
The characteristic equation for this dierential equation is LCs
2
+
L
s+ 1. This is of the
R
standard form
1
2
s
2
+

2
+ 1. This yields

n
n
1

n
= [rad/sec]
LC
L
n
1 L
= =
2R 2R C
_ _
_
_
2
_
1 1 L 1 1

d
=
LC
1
2R C
=
LC

4R
2
C
2
[rad/sec]
1
=
n
= [1/sec]
2RC
The pole zero plot is shown in Figure 11.4.1
447
_
_ _
_
Im
1 _ 1
LC 4R
2
C
2
_ 1
Re
2RC
_ 1 _ 1
LC 4R
2
C
2
Figure 11.4.1: Pole Zero Plot
(b) Withzeroinitialconditions,thetransferfunctionofthissystemissimplythesystemsequiv-
alent impedance. Substitute I
s
(s) = 1/s and take the inverse Laplace transform to nd an
expression for v
c
(t).
V
c
(s)
=
Ls
I
s
(s)
LCs
2
+
L
s+ 1
R
L
V
c
(s) =
LCs
2
+
L
s+ 1
R
1
=
C
s
2
+
1
s+
1
RC LC
1 1
1 LC

4R
2
C
2
= _
1 1
_
1
_
2
_
1 1
_
C
LC

4R
2
C
2
s+
2RC
+
LC

4R
2
C
2
1 t 1 1
v
c
(t) = _ e

2RC
sin t
1 1 LC 4R
2
C
2
C
LC

4R
2
C
2
This function is plotted in Figure 11.4.2
448


_
_
-500
-400
-300
-200
-100
0
100
200
300
400
500
v
c
(
t
)

[
V
]

0 1 2 3 4 5 6
time [s]
-4
x 10
Figure 11.4.2: Step Response with v
c
(0) = 0 [V] and i
L
(0)=1 [A]
(c) M is simply the magnitude of the transfer function evaluated at s=j.

V
c
(s)

L
M = = _
I
s
(s)
(1LC
2
)
2
+
L

_
2
R
The phase is
R
= tan
1
L
tan
1
L

0 1LC
2
L

= 90 tan
1 R
1LC
2
(d) TheplotisshowninFigure11.4.3. Thepeakfrequencyoccurswhenthesystemisexcitedat
its natural frequency w
n
=
1
= 10
5
[rad/sec]. The peak magnitude is found by plugging
LC
in = 10
5
into the magnitude expression. This evaluates to R.
449
Bode Diagram
M
a
g
n
i
t
u
d
e

(
d
B
)

30
35
40
45
50
55
60
65
70
P
h
a
s
e

(
d
e
g
)

-90
-45
0
45
90
4 5 6
10 10 10
Frequency (rad/sec)
Figure 11.4.3: Bode plot
450
11.5 Pole-Zero Plots
The solution for this problem is not available.
451
12 Signals, Transforms, and Transfer Functions
12.1 First Order Zeros
In this problem, we are asked to consider a system with a dierential equation
y + y=a u+u (1)
The system has pole at s=1/ and a zero at s=1/a.
(a) Integrate the equation from 0

to 0
+
, we get
_
t=0
+
t=0

(y + y)dt=
_
t=0
+
t=0

(a u+u)dt
(y(0
+
)y(0

))=a(u(0
+
)u(0

))
y(0
+
) =
a

u(0
+
) =
a

(2)
(3)
(4)
The particular solution and the homogenous solution are:
y
p
(t) = 1 (5)
y
h
(t) = Ae

(6)
The complete solution is
y(t) = y
p
(t) + y
h
(t) = 1 + Ae

(7)
Combining the initial condition
y(0
+
) =
a

= 1 + A (8)
A=
a

1 (9)
y(t) = 1 + (
a
1)e

(10)

(b) Figures 12.1.1 and 12.1.2 give the response and pole-zero plots.
452
X
(
t
)

1
0.8
0.6
0.4
1
0.9
1
(b) 0.5
(a)
0
0.5
1
0 0.1 0.2 0.3 0.4 0.5 20 18 16 14 12 10
1
0.5 (d)
(c)
I
m
(
s
)
0
0.5
1
0 0.1 0.2 0.3 0.4 0.5 11.5 11 10.5 10
1
(f)
(e)
0.5
0
0.5
1
0 0.1 0.2 0.3 0.4 0.5 10 9.8 9.6 9.4 9.2 9
Time[s] Re(s)
Figure 12.1.1: Step response and pole-zero plots for a =0.05, 0.09, 0.11. (a) Step response for
a = 0.05. (b) Pole-zero plot for a = 0.05. (c) Step response for a = 0.09. (d) Pole-zero plot for
a= 0.09. (e) Step response for a= 0.11. (f) Pole-zero plot for a= 0.11.
1.2
1.1
1
453
10
5
0
1
(b)
(a) 0.5
0
0.5
1
0 0.1 0.2 0.3 0.4 0.5 10 8 6 4 2 0
5 1
0.5
(d)
(c) 5
0
I
m
(
s
)
X
(
t
)
0
0.5
10 1
0 0.1 0.2 0.3 0.4 0.5 10 5 0 5
1
0.5
0
0.5
1
1
(f)
0.5
0
(e)
0.5
1
0 0.1 0.2 0.3 0.4 0.5 10 5 0 5 10
Time[s] Re(s)
Figure 12.1.2: Step response and pole-zero plots for a =1, -1, -0.1. (a) Step response for a = 1.
(b) Pole-zero plot for a=1. (c) Step response for a=1. (d) Pole-zero plot for a=1. (e) Step
response for a=0.1. (f) Pole-zero plot for a=0.1.
(c) Figure 12.1.3 shows that as a, H(s)1.
454
0 0.1 0.2 0.3 0.4 0.5
0.5
0.6
0.7
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
20 15 10 5
1
0.8
0.6
0.4
0.2
0
0.2
0.4
0.6
0.8
1
Increasingafora<
Decreasingafora>
a=,zero
andpolecoalesce
Figure 12.1.3: Step response and pole-zero plots for a = 0.1.
(d) We want to show that a zero causes certain exponential inputs to have zero response. We
consider an exponential input of the form u(t) = e
s
1
t
. We begin the equation in the time
domain
y + y=au + u (11)
If we desire y= 0, then au + u must equal zero for all time. Therefore, the input must be
u(t) = Ae
t/a
(12)
Comparing this input with u(t) = e
s
1
t
, we nd that s
1
=1/a. In the complex plane, the
zero and s
1
are in the same location.
(e) The response is the same as part (a).
455
_ _

_ _
12.2 Laplace Practice
The Laplace Transform of x(t)is dened as
_

L[x(t)]=X(s) = x(t)e
st
dt
0
therefore
_

_
T
_

W(s) = e
st
dt 0dt+ e
st
dt
0 0 T
1
st

1
st

= e 0 + e
s

0 T
1 1
sT
= e
s s
456
12.3 Step-Zero
In this problem, we consider the second-order system with a real zero
s/+ 1
H(s) = (1)
s
2
+ 2s+ 1
A useful way method of analyzing the behavior of this transfer function is to write it as a sum of
two transfer functions
s/ 1
H(s) = + (2)
s
2
+ 2s+ 1 s
2
+ 2s+ 1
If we dene H
1
(s) as
1
H
1
(s) = (3)
s
2
+ 2s+ 1
and H
2
(s) as
s
H
2
(s) = = sH
1
(s) (4)
s
2
+ 2s+ 1
then we notice that H
2
(s) is the derivative of H
1
(s) in the Laplace domain. Thus, we may write
the transfer function as
_ _
s
H(s) = + 1 H
1
(s) (5)

For our purposes, we consider only negative real zeros. If the zero is close to the imaginary axis,
then the transfer function closely resembles the derivative term only. If the zero is far away from
the origin, then the H(s) behaves primarily like H
1
(s). Using partial fractions, we nd that step
response is given by
__ _ _
y(t) = 1 + e
t
1
a
_
1
2

_
1
2
sin
_
1
2
tcos
_
1
2
t (6)
457
1
0.5 (a)
a=1
0
a=10
a=0.1
0.5
1
0.1 0.05 0
1
0.5 (b)
a=1
0
a=10 a=0.1
0.5
1
1 0.8 0.6 0.4 0.2 0
1
0.5 (c)
a=1
0
a=10 a=0.1
0.5
1
7 6 5 4 3 2 1 0
RealPartofs
Figure 12.3.1: Pole Zero Plots. (a) /zeta= 0.01. (b) /zeta= 0.1. (c) /zeta= 0.7.
I
m
a
g
i
n
a
r
y

P
a
r
t

o
f

458
_ _
_ _
_ _
_
12.4 Laplace to Time
(a) Thistransformissimpleenoughtobeinmosttables. However,wecansimplifyitwithpartial
fractions:
1 C D
Y(s) = = +
(s+a)(s+b) s+a s+b
nd the constants C and D by setting s=a and s=b
1 C D
= +
(s+a)(s+b) s+a s+b
1 = C(s+b) + D(s+a)
1
C =
ba
1
D=
ab
therefore
1 1 1 1
Y(s) =
ba s +a ba s +b
By looking up the inverse Laplace Transform of
s+
1
b
, we nd the total solution y(t)
1
y(t) = e
at
e
bt
ba
(b) First, note that the transform is
s
Y(s) =

s
2
2
+

2
n
s+ 1
n

2
=s
n
s
2
+ 2
n
s+
n
2
This Laplace Transform is in a standard form, but not on all tables. We will solve this
problem using the property
df
=sF(s)f(0)
dt
therefore
y(t) =
d
_

n
e
nt
sin
n
1
2
t
dt
1
2
_
_
_

2 _
_
_
=
n
2
e
nt
cos
n
1
2
t _
n
e
nt
sin
n
1
2
t
1
2
remember that for this form to be correct, must be less than 1.
459
12.5 Time to Laplace
(a) Thisfunctionisoneofthemostwidelyusedindynamicsystems,somemorizeitstransform!
1
Y(s) =
s+a
(b)
Y(s) =

d
(s+)
2
+
d
2
(c)
s+
Y(s) =
(s+)
2
+
d
2
460
12.6 Laplace Transform
Each term of X(s) can be evaluated directly using the Table of Laplace Transforms.
L
1
b=b(t)
and
_ _
L
1
a
= 1 e
at
s(s+a)
The nal result is then
L
1
X(s)=b(t) + 1 e
at
461
_ _
_ _
12.7 Signal in Time and Frequency Domain
(a) Sketch of x(t).
0.7
0.6
0.5
0.4
x
(
t
)

0.3
0.2
0.1
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
time(sec)
Figure 12.7.1: Signal x(t) plotted as a function of time
(b) Simplify the expression in to a sum of terms,
x(t) = e
t
u
s
(t)e
1
u
s
(t)e
t
u
s
(t1)+e
1
u
s
(t1)
Now take the Laplace transform of the rst, second and fourth terms,
1 e
1
t
e
1
e
s
X(s) = Le u
s
(t1) + (1)
s+ 1 s s
The third term requires some massaging to get it in a form available on the table. The term
can be modied into the form of a time delay, by factoring out e
1
.
L e
t
u
s
(t1) =e
1
L e
(t1)
u
s
(t1)
Now applying the Laplace Transform for a time delay from the table
_ _ 1 s
e
1
L e
(t1)
u
s
(t1) =
e e
s+ 1
Substituting this piece back into Equation (1) gives the solution
1 e
1
e
1
e
s
e
1
e
s
X(s) = +
s+ 1 s s+ 1 s
462
12.8 Transfer Function and Time Constants
(a) The transfer function for the circuit is easily found using the impedance method.
R
1
Z
eq1
=R
1
[[C
1
=
R
1
C
1
s+ 1
and Z
eq2
is similar. The circuit is now a voltage divider with complex impedances.
v
o
(s) =
Z
eq1
v
i
(s)
Z
eq1
+Z
eq2
Simplifying results in the transfer function
v
o
(s)
=
R
1
(R
2
C
2
s+ 1)
v
i
(s) R
1
R
2
(C
1
+C
2
)s+R
1
+R
2
To get the transfer function in the desired form we factor R1 + R2 out of the denominator,
while the numerator is already in the desired form. We now have
v
o
(s)
=
R
1
R
2
C
2
s+ 1
v
i
(s) R
1
+R
2
R
1
R
2
(C
1
+C
2
)
+ 1
R
1
+R
2
so
R
1
K =
R
1
+R
2

1
=R
2
C
2
R
1
R
2
(C
1
+C
2
)

2
=
R
1
+R
2
Using the Laplace initial value theorem with v
i
= 1/s
1 R
1
R
2
C
2
s+ 1
v
o
(0
+
) = lim sv
o
(s) = lim s
s s s R
1
+R
2
R
1
R
2
(C
1
+C
2
)
s+ 1
R
1
+R
2
After simplifying, we can now factor s out of the numerator and denominator, which cancel,
resulting in
R
1
R
2
C
2
+ 1/s
v
o
(0
+
) = lim
s R
1
+R
2
R
1
R
2
(C
1
+C
2
)
+ 1/s
R
1
+R
2
Now when taking the limit, the second terms in the numerator and denominator go to zero,
and after simplifying
v
o
(0
+
) =
C
2
C
1
+C
2
Using the Laplace nal value theorem with v
i
= 1/s
1 R
1
R
2
C
2
s+ 1
v()= limsv
o
(s)= lims
R
1
R
2
(C
1
+C
2
)
s0 s0 s R
1
+R
2
s+ 1
R
1
+R
2
and simplifying
v() =
R
1
R
1
+R
2
463
_ _
ThexinFigure12.8.1showsthepolelocationandtheoshowsthezerolocation. Plottingis
made possible by assuming
1
>
2
, but assuming
1
<
2
would have been just as valid for
plotting purposes.
Re
Im
-1/
2
-1/
1
Figure 12.8.1: Plot of pole and zero of system in terms of
1
and
2
When R
1
C
1
=R
2
C
2
the circuit becomes a voltage divider independent of frequency.
(b) Letting v
i
(s) = 1/s, a unit step, and assuming initial rest conditions, and continuing with
the K,
1
,
2
form,
K(
1
s+ 1)
v
o
(s) =
s(
2
s+ 1)
In order to calculate v
o
(t), we rst perform partial fraction decomposition, which results in
the following:
K K(
1

2
)
v
o
(s) = +
s
2
(s+ 1/
2
)
WecannowcalculatetheinverseLaplacetransformforeachtermdirectlyfromtheTableof
Laplace Transforms.
L
1
v
o
(t)=Ku
s
(t) +
K(
1

2
)
e
t/
2

2
After substituting K,
1
, and
2
in we have v
o
(t) as a function of R
1
, R
2
, C
1
, and C
2
for a
step input.
R
1
R
2
(C
1
+C
2
)
v
o
(t) =
R
1
u
s
(t) +
C
2

R
1
e
t(R
1
+R
2
)
R
1
+R
2
C
1
+C
2
R
1
+R
2
The step response and pole locations are now plotted for the following sets of parameter
values:
(i) R
1
= 1k, R
2
= 9k, C
1
= 1F, C
2
= 9F
(ii) R
1
= 1k, R
2
= 9k, C
1
= 9F, C
2
= 1F
(iii) R
1
= 9k, R
2
= 1k, C
1
= 9F, C
2
= 1F
464
Step Response
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05
Time (sec)
Pole locations for case (iii)
0
0.2
0.4
0.6
0.8
1
A
m
p
l
i
t
u
d
e

pole location
zero location
-1
-0.5
0
0.5
1
R
e
a
l

A
x
i
s

-120 -100 -80 -60 -40 -20 0
Imaginary Axis
Figure 12.8.2: Step response and pole locations for case (i)
Step Response
0
0.2
0.4
0.6
0.8
1
A
m
p
l
i
t
u
d
e

0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05
Time (sec)
Pole locations for case (ii)
pole location
zero location
1
0. 5
0
0.5
1
R
e
a
l

A
x
i
s

120 119 118 117 116 115 114 113 112 111 110
Imaginary Axis
Figure 12.8.3: Step response and pole locations for case (ii). Pole and zero are collocated.
465
Step Response
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05
0
0.2
0.4
0.6
0.8
1
A
m
p
l
i
t
u
d
e

Time (sec)
Pole locations for case (iii)
pole location
zero location
-1
-0.5
0
0.5
1
R
e
a
l

A
x
i
s

-1100 -1000 -900 -800 -700 -600 -500 -400 -300 -200 -100 0
Imaginary Axis
Figure 12.8.4: Step response and pole locations for case (iii)
466
12.9 Fourier Series Input
Figure 12.9.1:
467
Figure 12.9.2:
468
%2.003 Problem Set #9
%Problem 3
T=0.628;
w=2*pi/T*[1 3 5 7 9];
%Defining laplace variable
s=tf('s');
%G1(s)
K=10;
tau=20;
G1s=K/(tau*s+1);
figure(1);
bode(G1s,{10^-3,100});
title('Bode plot of G1(s)');
%G2(s)
K=1;
tau1=25;
tau2=10;
G2s=K/(tau1*s+1)/(tau2*s+1);
figure(2);
bode(G2s,{10^-3,100});
title('Bode plot of G2(s)');
%G3s
zeta=0.5;
wn=50;
G3s=s*wn^2/(s^2+2*zeta*wn*s+wn^2);
figure(3);
bode(G3s);
title('Bode plot of G3(s)');
%G4s
K=1;
tau=0.01;
zeta=1;
wn=100;
zetaz=0.05;
wzn=10;
G4s=K/(tau*s+1)*(s^2+2*zetaz*wzn*s+wzn^2)/(s^2+2*zeta*wn*s+wn^2);
figure(4);
bode(G4s);
title('Bode plot of G4(s)');
Figure 12.9.3:
469
G1(s)
BodeDiagram
3 2 1 0 1 2
10 10 10 10 10 10
Frequency(rad/sec)
G2(s) BodeDiagram
3 2 1 0 1 2
10 10 10 10 10 10
Frequency(rad/sec)
M
a
g
n
i
t
u
d
e

(
d
B
)

60
50
40
30
20
10
0
10
20
P
h
a
s
e

(
d
e
g
)

90
45
0
M
a
g
n
i
t
u
d
e

(
d
B
)

150
100
50
0
P
h
a
s
e

(
d
e
g
)

180
135
90
45
0
Figure 12.9.4:
470
G3(s)
BodeDiagram
1 2 3
10 10 10
Frequency(rad/sec)
G4(s)
BodeDiagram
0 1 2 3
10 10 10 10
Frequency(rad/sec)
M
a
g
n
i
t
u
d
e

(
d
B
)

5
10
15
20
25
30
35
P
h
a
s
e

(
d
e
g
)

90
45
0
45
90
M
a
g
n
i
t
u
d
e

(
d
B
)

80
70
60
50
40
30
20
10
0
P
h
a
s
e

(
d
e
g
)

90
45
0
45
90
135
180
Figure 12.9.5:
471
u(t)
0 0.2 0.4 0.6 0.8 1 1.2 1.4
1
0.8
0.6
0.4
0.2
0
0.2
0.4
0.6
0.8
1
u(t)with3terms
0 0.2 0.4 0.6 0.8 1 1.2 1.4
1
0.8
0.6
0.4
0.2
0
0.2
0.4
0.6
0.8
1
Figure 12.9.6:
472
Outputy(t)ofsystemG1(s)
0 0.2 0.4 0.6 0.8 1 1.2 1.4
0.04
0.03
0.02
0.01
0
0.01
0.02
0.03
0.04
x10
5
Outputy(t)ofsystemG2(s)
0 0.2 0.4 0.6 0.8 1 1.2 1.4
4
3
2
1
0
1
2
3
4
Figure 12.9.7:
473
G(s)
BodeDiagram
1 2
10 10
Frequency(rad/sec)
Outputy(t)ofsystemG(s)approximatingapuresinewavewithfrequency3*wo
0 0.2 0.4 0.6 0.8 1 1.2 1.4
M
a
g
n
i
t
u
d
e

(
d
B
)

30
20
10
0
10
20
30
P
h
a
s
e

(
d
e
g
)

90
45
0
45
90
1.5
1
0.5
0
0.5
1
1.5
Figure 12.9.8:
474
_ _
_
_
_
_
12.10 Spring Mass System
(a)
f(t) = u(t2)u(t10)
F(s) =
1
(e
2s
e
10s
)
s
(b)
mx + bx + kx = f(t) = u(t2)u(t10)
The Laplace transform of this system is
m[s
2
X(s)sx(0)x (0)] +b[sX(s)x(0)]+kX(s) =
1
(e
2s
e
10s
)
s
Assuming x(0)=x (0) = 0
(ms
2
+bs+k)X(s) =
1
(e
2s
e
10s
)
s
1
X(s) =
s(s
2
+s+ 1)
(e
2s
e
10s
)
Theinverselaplacetransformof1/(s(s
2
+s+1))caneasilybeseentobeadecayingsinusoid.
e
2s
and e
10s
are just time delays in the real time. Thus
1 1
L
1
= 1 e
0.5t
sin( 10.5
2
t+)
s(s
2
+s+ 1)
10.5
2
1
x(t) = u(t2)[1 e
0.5(t2)
sin( 10.5
2
(t2)+)]
10.5
2
1
u(t10)[1 e
0.5(t10)
sin( 10.5
2
(t10)+)]
10.5
2
= tan
1
( 10.5
2
/0.5)
Figure 12.10.1: x(t) for Part c
475
0 2 4 6 8 10 12 14 16 18 20
0.2
0
0.2
0.4
0.6
0.8
1
1.2
(c) WellthereareacoupleofwaystogetthisplotusingMatlab. Oneistousethelsimcommand
inwhichyouspecifythesystemtransferfunction(wehavenottalkedaboutthisinclassyet),
the input vector (0 t<2, 1 2<t<10, 0 t>10), and the calculation time vector (t=[0:0.01:20]).
Notethismethoddeterminesthesolutionbyintegratingfromonetimesteptothenext,thus
for Good results you need to have a relatively ne time vector. The advantage of this
method is a relatively compact .M le.
I calculated the response by nesting an if loop inside a for loop and calculated the exact
response for each time index. I used the script springmasssol.m at the end of this solution.
Figure 12.10.1 shows the resulting plot.
(d) Theplotisbasicallythesumoftwoseparatedecayingsinusoids. BythisImean,thedynamic
responsetotherststepinputhaslargelydiedawaybeforetheresponsetothenegativestep
begins.
(e) Ifthepulsewere2sec. insteadof8sec., thetwosinusoidswouldactdestructivelywitheach
other. The exact response would depend on the phase of the two sinusoid relative to one
anotherbutthesystemwouldnotreach1butstilloscillateaboutzero. Figure12.10.2shows
this response.
Figure 12.10.2: x(t) for Part e
springmasssol.m
Download
clear all; close all; %clears the workspace and closes all figures
t=[0:0.01:20]; %creates time vector
%Next line creates key constants
wd=1*sqrt(1-0.5^2); C=1/sqrt(1-0.5^2);
phi=atan2(sqrt(1-0.5^2),0.5);
X=[]; %Creates an empty matrix where I can store each value x(t)
for i=1:length(t) %starts a for loop index from 1 to the length of t
%If loop should be clear
if t(i)<2
476
0 2 4 6 8 10 12 14 16 18 20
0.2
0
0.2
0.4
0.6
0.8
1
1.2
x=0;
elseif t(i)<10
x=1-C*exp(-0.5*(t(i)-2))*sin(wd*(t(i)-2)+phi);
else
x=1-C*exp(-0.5*(t(i)-2))*sin(wd*(t(i)-2)+phi)...
-(1-C*exp(-0.5*(t(i)-10))*sin(wd*(t(i)-10)+phi));
end
X=[X,x]; %Adds new x(t_i) to matrix x(t)
end plot(t,X)
477
I
m

A
x
i
s

12.11 Transfer Function


(a) The zeros are located at
s
1
= 1
s
2
= 10
s
3
= 99
The poles are located at
s
1,2
= 0.50.866j
s
3,4
= 0.57.053j
-100 -90 -80 -70 -60 -50 -40 -30 -20 -10 0
8
6
4
2
0
-2
-4
-6
-8
Real Axis
Figure 12.11.1: Pole zero plot
(b) Figure 12.11.1 shows the pole/zero plot for this system.
(c) Fromthepolelocations,weknowthattheresponsewillbeasumofsinusoidswithadamped
naturalfrequencyof0.87r/sand7r/s. Thehigherfrequencysinusoidismorelightlydamped
but since is the same for both sinusoids, they will decay out in the same time period.
In this case, we would might assume that the energy contained in each of the sinusoids is
approximatelyequal. Sincekineticenergyis
1
mv
2
andv=
dx
=
d
cos
d
t,wewouldexpect
2 dt
themagnitudeofthefastersinusoidtobesmallerthanthemagnitudeoftheslowersinusoid.
Thustheresponsewilllooklikealargeslowsinusoidwithasmallerfastersinusoidaddedon
top. Theimpactofthezerosisalittlehardertopredict. Thezeroat-100issofastthatany
eect will be over so quickly that it will not eect the shape of the response. The zeros as
-1and-10willimpactthesystemresponse. Generally,zerosnearthefrequencyofacomplex
pole pair cause the system to overshoot more than predicted by the pole pair . Thus the
478
0
5
10
15
20
25
30
A
m
p
l
i
t
u
d
e

zeroat-10willcausethe7r/ssinusoidtoovershootmorethanpredictedwhilethezeroat-1
will cause the 0.86 r/s sinusoid to have more overshoot. Note: if the zeros were in the right
half plane they would cause the system to initially move in the negative direction.
(d) Figure 12.11.2 shows the step response for this system. My Matlab code for this problem is
shown at the end of the solution in an m-le called tfsol.m.
Step Response
0 2 4 6 8 10 12
Time (sec)
Figure 12.11.2: Step Response
tfsol.m
Download
clear all; close all;
num=[1 110 1100 1000]; den=[1 2 52 51 50];
A=roots(num); Ar=real(A);Ai=imag(A);
B=roots(den);Br=real(B);Bi=imag(B);
plot(Ar,Ai,o,Br,Bi,x)
grid on;
figure(2)
step(num,den)
479
13 Controls
13.1 Rooftop Antenna
(a) Tostartthisproblemweneedtobreaktheanalysisintosteps. Instep1,weanalysisantenna
system. Figure 13.1.1-A shows shows the free body diagram for the antenna. The control
torque T
m
and a disturbance T
d
act on the antenna inertia:
J
o
= T
m
+T
d

o
=
T
m
+T
d
Js

o
=
s
TheequivalentblockdiagramisshownbesidetheFBD.Next,weneedtoevaluatethemotor.
Figure 13.1.1-B shows the circuit model of the motor:
V
m
= R
a
I
a
+ 10
o
V
m
10
o
I
a
=
R
a
T
m
= K
m
I
a
Theequivalentblockdiagramforthemotorisshownnexttothecircuit. Wecannowcombine
theblockdiagramsinFigure13.1.1-A&BtoformtheblockdiagraminFigure13.1.1-C.The
amplier has the following characteristics:
V
m
= G(
i

o
)
with the equivalent block diagram in Figure 13.1.1-D. Combining Figure 13.1.1 -C&D yields
the block diagram for the complete system Figure 13.1.1-E.
In order to nd the transfer function V
o
/V
i
(s), we set the disturbance torque T
d
= 0. We
start by reducing the inner loop of the block diagram to:
Km

o
(s) =
RaJs
=
K
m
V
m 1 +
10Km
R
a
Js+ 10K
m
RaJs
We can now address the outer loop
GKm
s(RaJs+10Km)
GK
m

o
i
(s) =
1 +
GKm
=
R
a
Js
2
+ 10K
m
s+GK
m
s(RaJs+10Km)
To determine the transfer function
o
/T
d
(s), we set
i
= 0. It is helpful to reorganize the
block diagram into the form shown in Figure 13.1.1-F.
Js
2 R
a

o
(s) =
1
=
T
d 1 +
Km
(G+ 10s)
R
a
Js
2
+ 10K
m
s+GK
m
RaJs
2
480
T
d
T
m
+
1

J s
I
a
+
R
a
+
Vm
Vm
+ -
10

-
-
B)
T
d

,

J
A)
T
d
T
m
+ 1
s



1
R
a


I
a
10
K
m
T
m

1
R
a
Vm
+
-
I
a
10
K
m
T
m
+
+

1
J s
1
s


C)
D)
-
+
G


Vm
Vm
+
-
G



1
R
a
Vm
+
-
I
a
10
K
m
T
d
T
m
+
+

1
J s
1
s



+
-
G

E)

+
-
10s
T
d
1
J s
1
s


F)
G

+
+
K
m
R
a
Figure 13.1.1: Block diagrams for antenna analysis
481
(b)

o
100
(s) =

i
10s
2
+100s+100
s
1
= 8.873
s
2
= 1.127

o
(t) = 1 + 0.145e
8.873t
1.145e
1.127t
Figure 13.1.2 shows the pole map and step response for this system
5
I
m0
-5
-10 -5 0
Re
x(t)
0
0.2
0.4
0.6
0.8
1
A
m
p
l
i
t
u
d
e

r
a
d
i
a
n
s

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Time (s)
Figure 13.1.2: Pole map
o
/
i
(s) and step response
o
(t) for u(t)
i
(c) The solid line in Figure 13.1.3 shows the frequency response.
482
10
1
10
0
10
1
10
2
10
3
10
1
10
0
10
1
10
2
10
3
Frequency r/s
Figure 13.1.3: Frequency response
o
/
i
(s) (solid line) and
o
/T
d
(s) (dashed line)
(d)

o
5
(s) =
T
d
10s
2
+100s+100

o
(t) =
1
(1 + 0.145e
8.873t
1.145e
1.127t
)
20
The step response is identical to that shown in Figure 13.1.2 except the magnitude is scale
down to 0.05. The frequency response is the dashed line in Figure 13.1.3.
(e) The negative of the loop transmission for this system is
GK
m
100
L(s) = =
s(R
a
Js+ 10K
m
) 10s
2
+100s

c
= 1 r/s

c
= 95.8
o

m
= 84.2
o
120
100
80
60
40
20
0
M
a
g
n
i
t
u
d
e

d
B

150
100
50
0
P
h
a
s
e

o

483
P
h
a
s
e

(
d
e
g
)

M
a
g
n
i
t
u
d
e

(
d
B
)

50
0
-50
-100
-90
-135
-180
-1 0 1
2
3
10 10 10 10 10
Frequency (rad/sec)
Figure 13.1.4: Bode plot for the negative of the loop transmission L(s) for G=10.
Figure13.1.4showsthebodeplotofL(s). Thecrossoverfrequency
c
isthefrequencywhere
the magnitude of L(s)=1 (0 dB). The phase margin
m
=180
o

c
where
c
is the phase
of L(s) at
c
.
(f)

o
1000
(s) =

i
10s
2
+100s+1000
s
1,2
= 58.66j

o
(t) = 1 + 1.15e
5t
sin(8.66t2.094) for u(t)
i

o
5
(s) =
T
d
10s
2
+100s+1000
1

o
(t) = (1 + 1.15e
5t
sin(8.66t2.094)) for u(t)T
d
200
1000
L(s) =
10s
2
+100s

c
= 7.8r/s

c
= 128
o

m
= 52
o
Figure 13.1.5 shows the pole map and step response for
o
/
i
with a G=100. The
o
/T
d
(s)
stepresponseis1/200the
o
/
i
stepresponse. Figure13.1.6showsthefrequencyresponsefor

o
/
i
(solidline)and
o
/T
d
(s)(dashedline). Figure13.1.7showsthebodeplotofthenegative
of the loop transmission L(s). As we can see from the step response, increasing G speeds up
theclosed-loopdynamicsandreducesthedampingratiooftheresponse. IncreasingGimpacts
the frequency response by increasing the range of frequencies for which the magnitude is at
ornearone. Thismeansthattherangeoffrequencieswherewegetgoodcommandfollowing
is increased. We also notice in the frequency response that the phase transition is sharper
484
for the higher gain system reecting the reduced damping ration. Lastly we note, that the
higher gain system rejects disturbance much better (10X better to be exact). Increasing the
gain has transformed the two real poles into a complex pole pair. Further increasing the G
will move the poles further up along the imaginary axis (i.e. increase
d
) without changing
the real component of the pole (i.e. further reduce the damping ratio).
10
5
I
m0
-5
-10
-10 -5 0
Re
x(t)
0
0.5
1
1.5
A
m
p
l
i
t
u
d
e

r
a
d
i
a
n
s

0 0.2 0.4 0.6 0.8 1 1.2
Time (s) (sec)
Figure 13.1.5: Pole map
o
/
i
(s) and step response
o
(t) for u(t)
i
for G=100.
485
-150
-100
-50
0
50
M
a
g
n
i
t
u
d
e

(
d
B
)

1
10
0
10
1
2
10
3
-180
-135
-90
-45
0
P
h
a
s
e

(
d
e
g
)




d
10 10
Frequency (rad/sec)
Figure 13.1.6: Frequency response
o
/
i
(s) (solid line)and
o
/T
d
(s) (dashed line) for G=100
-80
-40
0
40
80
M
a
g
n
i
t
u
d
e

(
d
B
)

-1 0
10
1
10
2
10
3
-180
-135
-90
P
h
a
s
e

(
d
e
g
)

10 10
Frequency (rad/sec)
Figure 13.1.7: Bode plot for the negative of the loop transmission L(s) for G=100.
486
(g)

o
10000
(s) =

i
10s
2
+100s+10000
s
1,2
= 531.2j

o
(t) = 1 + 1.01e
5t
sin(31.2t1.73) for u(t)
i

o
5
(s) =
T
d
10s
2
+100s+10000
1

o
(t) = (1 + 1.01e
5t
sin(31.2t1.73)) for u(t)T
d
2000
10000
L(s) =
10s
2
+100s

c
= 30.7r/s

c
= 162
o

m
= 18
o
Figure13.1.8showsthepolemapandstepresponsefor
o
/
i
withaG=1000. The
o
/T
d
(s)
step response is 1/200 the
o
/
i
step response. Figure 13.1.9 shows the frequency response
for
o
/
i
(solid line) and
o
/T
d
(s) (dashed line). Figure 13.1.10 shows the bode plot of the
negative of the loop transmission L(s).
I
m

-10 -5 0
-30
-20
-10
0
10
20
30
Re
x(t)
0
0.5
1
1.5
2
A
m
p
l
i
t
u
d
e

r
a
d
i
a
n
s

0 0.2 0.4 0.6 0.8 1 1.2
Time (s) (sec)
Figure 13.1.8: Pole map
o
/
i
(s) and step response
o
(t) for u(t)
i
for G=1000.
487
-100
-80
-60
-40
-20
0
20
M
a
g
n
i
t
u
d
e

d
B

10
0
10
1 2 3
-150
-100
-50
0
P
h
a
s
e

o

10
10
Frequency r/s
Figure 13.1.9: Frequency response
o
/
i
(s) (solid line) and
o
/T
d
(s) (dashed line) for G=1000.
-80
-40
0
40
80
M
a
g
n
i
t
u
d
e

(
d
B
)

10
-1
10
0
10
1
10
2
10
3
-180
-135
-90
P
h
a
s
e

(
d
e
g
)

Frequency (rad/sec)
Figure 13.1.10: Bode plot for the negative of the loop transmission L(s) for G=1000.
488
13.2 Lead Controller
The solution for this problem is not available.
489
13.3 Third Order
Symmetrical mechanical drive. See Fig. 13.3.1. Two equal inertias I with equal linear friction
coecients B are coupled by a shaft with torsional stiness K
c
and torsional damping coecient
B
c
.
Coupler
I

l
I
Motor Load
T
m

m
Figure 13.3.1: Symmetrical Mechanical Drive
(a) The fundamental requirements are:
Geometrical Compatibility: =
m

l
, d/dt=
m

l
Constitutive Equations: T
k
=K
c
, T
d
=B
c
d/dt, T
f,m
=B
m
,
T
f,l
=B
l
, T
I,m
=Id
m
/dt, T
I,l
=Id
l
/dt.
Torque Balance: Motor: T
I,m
=T
m
K
c
B
m
B
c
d/dt,
Load: T
I,l
=K
c
B
l
+B
c
d/dt
In the following state equations , the rst equation expresses the geometric compatibility
requirement, the second equation is the result of inserting the constitutive equations into
the torque balance for the motor, and the third is the result of substituting the constitutive
equations into the torque balance for the load.
_ _ _ _ _ _ _ _
0 1 1 0


_ _ _ _ _ _
d

m
=


Kc

B+Bc Bc


m
+
1
T
m
dt _
I I I
_
I

l
_
Kc Bc

B+Bc
_

l
_ _
0
_
I I i
490
_ _
(b) The new variables and
avg
satisfy the following equations:
2
avg
=
m
+
l
=
m

l
By adding and subtracting these equations we nd

m
=
avg
+ /2

l
=
avg
/2
(c) We can construct the state equations for the new variables , and
avg
one at a time.
For , we have
d
=
m

l
=
dt
For , we write
d d
m
d
l
_
K
c
B+B
c
B
c
_
T
m
= = (
avg
+ /2) + (
avg
/2) +
dt dt dt I I I I

K
c

B+B
c
(
avg
/2)+
B
c
(
a
vg+ /2)
I I I
2K
c
B+ 2B
c
T
m
= +
I I I
Similarly, for
avg
, we write
d
avg
=
1
(
d
m
+
d
l
) =
1
_

K
c

B+B
c
(
avg
+ /2)+
B
c
(
avg
/2)
_
+
T
m
dt 2 dt dt 2 I I I I
1
_
K
c
B+B
c
B
c
_
+ (
avg
/2) + (
a
vg+ /2)
2 I I I
B T
m
=
avg
+
I I
These state equations for the new variables can be arranged in matrices of a third-order
sytem:
_ _ _ _ _ _ _ _
0 1 0 0


_ _ _ _ _ _
d
=

2
Kc

B+2Bc
0

+
1
T
m
dt _
I I
_
I



_ _
B
_ _ _
1
_

avg
0 0
I

avg
I
or, as a combination of a second-order sub-system
_ _ _ _ _ _ _ _
_

_ 0 1 _

_ _
0
_
d
_ _
= + T
m
dt_

_
2
Kc

B+2Bc
_

_ _ 1 _
I I I
describing the relative motion of the motor and the load, and a rst-order sub-system
d
avg
=
B

avg
+
1
T
m
dt I I
describing the average speed of motor and load. Both sub-systems have the motor torque
as input, but the second-order sub-system is independent of the average speed
avg
, and the
rst-ordersub-systemisindependentoftherelative motionbetweenthemotorandtheload.
491
_
(d) The rst-order sub-system equation can be rewritten as
I d
avg
T
m
+
avg
=
B dt B
from which we identify the decay time-constant
I 5e-5
= = = 0.5 sec
B 1.0e-4
and the steady-state speed
T
m
6e-3

ss
= = =60 rad/sec
B 1.0e-4
(e) From the second-order sub-system equations we identify
2K
c
B+ 2B
c

o
2
= and 2
o
=
I I
from which we obtain the behavioral parameters of the second-order sub-system
2(1.24e-2) 1.0e-4+2(2e-5)

o
= = 22.3 rad/sec and = = 0.0629
5e-5 2(22.3)(5e-5)
492
13.4 Error Constants
A)
E
(s) = 1
Y
B)
E 1
(s) =
R 1 +KG(s)
1
Step E(s) =
s(1+KG(s))
1
Ramp E(s) =
s
2
(1+KG(s))
1
Parabola E(s) =
s
3
(1+KG(s))
C) For this section G(0)=K
g
.
Type 0 Type 1 Type 2
Step Input
1
1+KK
g
0 0
Ramp
1
KK
g
0
Parabola
1
KK
g
Figure 13.4.1:
493
13.5 System Type
A) Figure 1 shows the negative loop transmission for this system.
B) Note: To get a phase margin of 50
o
, we would need the crossover
frequency to be 14.9 r/s. It is not possible to get a crossover at 14.9 r/s
because of the peak from the complex pair of poles. If we set K=8.6 (the
gain needed to raise the magnitude of the loop transmission to 1 at 14.9
r/s), we nd that the real crossover occurs at 16.9 r/s with a resulting
phase margin of 44
o
.
K
g
= 1
1 1
Error = = = 0.11
KK
g
8.6
C) Figure 2 shows the loop transmission for the type 2 system. As we can
see, this is not a good thing to do since we now have system where the
phase is always below 180
o
. Now it is possible to get stable closed loop
systems where the loop transmission phase is below 180
o
but these
systems must have magnitudes less than 1 when the phase passes through
180
o
. Since this system has 2 poles at =0 we essentially have an open
loop gain of when we pass through 180
o
resulting in an unstable
closed loop system.
Figure 13.5.1:
494
BodeDiagram
0 1 2 3
10 10 10 10
Frequency (rad/sec)
Figure 1: Bode diagram for Problem 2-A
BodeDiagram
0 1 2 3
10 10 10 10
Frequency (rad/sec)
Figure 2: Bode diagram for Problem 2-C
Figure 13.5.2:
495
150
100
50
0
50
M
a
g
n
i
t
u
d
e

(
d
B
)

270
225
180
135
90
P
h
a
s
e

(
d
e
g
)

200
150
100
50
0
50
M
a
g
n
i
t
u
d
e

(
d
B
)

360
315
270
225
180
P
h
a
s
e

(
d
e
g
)

270
225
180
135
90
P
h
a
s
e

(
d
e
g
)
System:untitled1
Frequency(rad/sec):14.9
Phase(deg):133
150
100
50
0
50
100
M
a
g
n
i
t
u
d
e

(
d
B
)
System:untitled1
Frequency(rad/sec):14.8
Magnitude(dB):38.7
BodeDiagram
3 2 1 0 1 2 3
10 10 10 10 10 10 10
Frequency (rad/sec)
Figure 3: Intermidiate Bode plot for problem 2-D
D To get an error=2%, 0.02
K
1
K 50. From part B, we know that
without the lag compensator is 8.6, so if we reduce the magnitude of the
transfer function by a factor of ten we will raise the required gain to 86
(this is greater than 50). The transfer function for a lag compensator is
s+ 1
G
lag
(s) = K
s+ 1
For a lag compensator the pole occurs at a lower frequency than the zero,
thus for the region between the pole and the zero the lag compensator has
a -1 slope. We want to drop the magnitude by a factor of 10 (1 dec) so that
tells us that =10 so that the lag compensator acts over 1 decade. Now
we need to gure out . We do not want the phase of the lag compensator
to eect the phase at our desired crossover, so can be any value which
places the zero of the lag a decade or more below the crossover frequency.
1
= 0.67
1.49
To make life easy, I have elected to make =1. Figure 3 shows the bode
plot for tau=1. As we can see, the magnitude is the where we wanted it
(required gain 86) but we have actually lost some phase. This is because
thelagcompensatoriscloseenoughtohavesomeimpactonthephase. IfI
were designing this for myself, I would just leave it but to x this for the
problem set I have set tau=2 and recovered all but 1
o
of phase. Figure 4
shows the negative loop transmission of this system with tau= 2, K =86,
and =10.
Figure 13.5.3:
496
P
h
a
s
e

(
d
e
g
)

M
a
g
n
i
t
u
d
e

(
d
B
)

BodeDiagram
100
10 10 10 10 10 10 10
3 2 1 0 1 2 3
50
0
50
100
150
90
135
180
225
270
Frequency (rad/sec)
Figure 4: Bode plot for problem 2-D
On a design note here, I could get 0 steady state error by using the
following lag compensator
s+ 1
G
c
(s) = K .
s
You may ask yourself, how is this dierent than changing the system to a
type 2 system. The answer is that because we have added a zero in
addition to the pole the phase of the system never actually equals180
o
at low frequencies, just gets really close. The major issue with this
approach is that if there are any un-modelled dynamics such as time delay
which cause the phase to decrease, the system will be unstable.
E Figure 5 shows the step response of the error function for the controller
designed in part d. Well the damping is much lower than expected and the
natural frequency is higher as well. The crossover frequency is higher than
expected due to the resonant peak as discussed in part b. In the case of
the lag controller, we nd that not only is the actual crossover frequency
about 16.5 r/s but that lag controller has further reduced our phase
margin to 36
o
.
F Figure 6 shows the error of my controller to a ramp input. Since, I
designed my controller to have an error less than 2% and my actual error
is about 1.6% I am happy. If customer (Prof. Hardt) really wanted 1% he
should have told me that before I built the thing.
Figure 13.5.4:
497
StepResponse
0.4
0.2
0
0.2
0.4
0.6
0.8
1
A
m
p
l
i
t
u
d
e

0 1 2 3 4 5 6 7 8 9
Time(sec)
Figure 5: Step response of the error for lag controller design in D
StepResponse
0
0.02
0.04
0.06
0.08
0.1
0.12
A
m
p
l
i
t
u
d
e

0 1 2 3 4 5 6 7 8 9
Time(sec)
Figure 6: Error to a ramp input
Figure 13.5.5:
498
10
10
13.6 Unity Feedback
Figure 13.6.1:
499
Figure 13.6.2:
500
13.7 P and PI Controllers
The solution for this problem is not available.
501
14 Motors and Transformers
14.1 Locked Load
With the load clamped so that it cannot rotate, the application of motor torque causes twisting
of the coupler and oscillation of the rotor. The system consists of the motor rotor with moment
of inertia I
r
and the coupler with linear torsional stiness K
c
and linear damping with damping
coecientB
c
,actedonbythemotortorqueT
m
andafrictionalmotortorquewithlineardamping
coecient B
m
, as shown in Fig. 14.1.1.
T
m
I
r

m
Locked
Load
Coupler

m
,
m
I
r
T
m
K
c

m
B
c

m
B
m

m
Figure 14.1.1: Motor Rotor connected to Locked Load by Elastic Coupler
(a) The fundamental requirements are:
Geometric compatibility:
m
=d
m
/dt
Constitutive equations: T
I
=I
r
d
m
/dt, T
f
= (B
m
+B
c
)
m
, T
k
=K
m

m
Torque balance: T
I
=T
m
T
f
T
k
Take
m
and
m
asstatevariables. Thenoneequationisprovidedbythegeometriccompati-
bilityrequirement. Thesecondequationisobtainedbysubstitutingtheconstitutiveequations
into the torque-balance requirement to get
I
r
d
m
=T
m
(B
m
+B
c
)
m
K
c

m
dt
WithT
m
astheinputand
m
asthedesiredoutput,thestandardformforastate-determined
system
d
x = Ax+Bu
dt
y = Cx+Du
becomes _ _ _ __ _ _ _
_

m
_ 0 1 _

m
_ _ 0 _
d
=
_ _
+ T
m
dt_ _ Kc Bm+Bc _ _ _ 1 _

Ir

Ir

m
Ir
502
_ _
_ _
with output
m
given by
_ _
_

m
_
y=Cx+Du= 1 0 + 0 =
m
_

m
_
(b) The behavioral parameters
o
and appear in the matrix A of second-order systems in the
following pattern
_ _
0 1
_ _
A= (1)

o
2
2
o
BycomparingthematrixAintheequationforthelocked-loadresponsewith(1)weidentify
the the following relationships for the behavioral parameters

2
=
K
c
and 2
o
=
B
m
+B
c
o
I
r
I
r
which lead to
K
c
1.24e-2 B
m
+B
c
1e-4+2e-5

o
= = = 15.75 rad/sec and = = _ = 0.0762
I
r
5e-5 2 K
c
I
r 2 (1.24e-2)(5e-5)
(i) The damped natural frequency is

d
=
o
1
2
= 15.75 1(0.0762)
2
= 15.70 rad/sec
(ii) The decrement ratio for a damped sinusoidal oscillation, as dened in the Notes for
Lecture 6, is the ratio of the amplitudes of two successive peaks. The time increment
betweentwosuchpeaksishalfoftheperiodofthedampedoscillation. Theenvelopeof
thedampedoscillationdecaysinproportiontoexp(
o
t),sowitht=
1
2
T
d
=/
d
the
decrement ratio is
0.0762
dec ratio=exp(
o
/
d
) = exp(_ ) = exp(_ ) = 0.787
1
2
1(0.0762)
2
(c) The dimensions of G
1
are [torque/angular velocity]. Its units are N-m/r/s or N-m-s/r in the
SI system of units. When the control algorithm (with
sensed
=
m
),
T
M
=G
1
(r
1

m
),
is substituted in the state equations of the uncontrolled locked-load system, the resulting
equations for the velocity-controlled system are
_ _ _ _ _ _ _ _
d
_

m
_ 0 1 _

m
_ _ 0 _
=
_ _
+ r
1
dt_

m
_

K
Ir
c

Bm+
I
Bc+G
1 _

m
_ _ G
Ir
1 _
r
TheinputischangedfromthemotortorqueT
m
tothereferencespeedr
1
,andtheMatricesA
andBnowdependonthegainG
1
. Theundampednaturalfrequency
o
remainsunchanged
butthedampingrationowdependsonthegainG
1
. BycomparingthematrixAwith(1)we
identify
2
o
=
B
m
+B
c
+G
1
I
r
503
_
ToobtainthevalueofG
1
requiredtoyieldacritical;lydampedresponse,set =1andsolve
for G
1
G
1
= 2I
r

o
B
m
B
c
=2(5e-5)(15.75)1e-42e-5=14.55e-4 N-m-s/r
(d) The dimensions of G
2
are [torque/angle]. Its units are N-m in the SI system of units. When
the control algorithm (with
sensed
=
m
),
T
M
=G
2
(r
2

m
),
is substituted in the state equations of the uncontrolled locked-load system, the resulting
equations for the position-controlled system are
_ _ _ _ _ _
_

m
_ 0 1 _

m
_ _ 0 _
_ _
d
_ _
= + r
2
dt_ _

Kc+G
2

Bm+Bc _ _ _ G
2 _

m
Ir Ir

m
Ir
BycomparingthematrixAinthisequationwith(1),weidentifythefollowingrelationships
for the behavioral parameters

o
2
=
K
c
+G
2
and 2
o
=
B
m
+B
c
I
r
I
r
Both the undamped natural frequency
o
and the damping ratio now depend on the gain
G
2
.
(i) The undamped natural frequency is
K
c
+G
2

o
=
I
r
(ii) The damping ratio is
B
m
+B
c
B
m
+B
c
= = _
2I
r

o 2 (B
m
+G
2
)I
r
(e) The control algorithms can be implemented by ordinary mechanical elements connected be-
tween a desired-motion driver and the motor rotor.
(i) The velocity feedback control algorithm
T
m
=G
1
(r
1

m
)
states that a torque is applied to the rotor which is proportional to the dierence in
speeds between the motion driver and the rotor. A linear friction element connected
between the motion driver and the rotor would apply this same torque if its damping
coecient were G
1
.
(ii) The position feedback control algorithm
T
m
=G
2
(r
2

m
)
states that a torque is applied to the rotor which is proportional to the dierence in
angular position between the motion driver and the rotor. A linear torsional spring
connected between the motion driver and the rotor would apply this same torque if its
stiness were G
2
.
Adrawbackofthesemechanicalimplementationsisthenecessityofconstructingamechanical
motion driver to provide the information about the desired motion.
504
_
_
_ _
14.2 Motoring with a Capacitor
Figure14.2.1showsthefreebodydiagramforthisproblem. Note: Ihaveappliedallofthetorques
in the positive direction.

m

in

J
v
m
=K
v
c
=q/C
Figure 14.2.1: Free Body Diagram and Circuit Diagram
(a) Based on the free body diagram
J =
in
+
m

m
=Ki
m
J =
in
+Ki
m
(b) Evaluating the circuit results in
0 = V
m
+V
c
V
m
=V
c
q= i
m
dt
q 1
_
V
c
= V
c
= i
m
dt
C C
V
m
=K
1 i
m
i
m
dt=K =K
C C
i
m
=KC
(c)
J =
in
K
2
C
J+K
2
C =
in
(d) The equivalent mechanical model would be J
eq
=
in
where J
eq
= J +R
2
C. This is true
because the capacitor acts to increase the apparent inertia of the motor rotor.
505
_ _
(e)
__ _ _ _ _ _ _
L J +K
2
C = (s) J +K
2
C + J +K
2
C (0)
(0)=0
1
Lu(t)=
s
(s)
=
1
(s) (J +K
2
C)s
2
L
1
(s) t
=
(s) (J +K
2
C)
506
= _ _
_ _ _ _
14.3 Leadscrew with Translating Stage
Whenthesystemismoving,thelead-screwwillexertaforceF
1
retardingthemotionofthemass.
The mass will in turn exert a torque T on the lead-screw retarding its motion. The equations of
motion for the carriage and lead-screw will be
Mx = F F
1
c
2
x k
2
x (1)
J

=T c
1

k
1
(2)
Weneedarelationshipbetweenxand. Fromthedenitionofthepitchp, eachrevolutionofthe
lead-screw will advance the carriage 10
2
[m]. One revolution is 2 [rad], therefore
x=
p
=Nwhere N =
p
(3)
2 2
Because the lead-screw is said to be ideal (no friction present), there is 100% eciency in power
conversion from the linear to the rotational domain. Therefore
F
1
x=T
T
F
1
= (4)
N
Take the Laplace transforms of equations (1) and (2) to get
F(s)F
1
(s) =
_
Ms
2
+c
2
s +k
2
_
X(s) (5)
T(s) =
_
Js
2
+c
1
s +k
1
_
(s) (6)
Combine equations (3) through (6) and solve for the transfer function
_ _
1
_ _
F(s) Js
2
+c
1
s +k
1
N
2
X(s) = Ms
2
+c
2
s +k
2
X(s)
X(s) 1
F(s)
m+
N
J
2
s
2
+ c
2
+
N
c
1
2
s+ k
2
+
N
k
1
2
507
14.4 Mass Pulley System
(a) As the pulley rotates clockwise, we assume a no-slip interface between the string and the
pulley, and therefore a length of string equal to R must be unwound on the right hand
side. The is clearly the distance the pulley moves upward, so x
2
=R.
(b) Againweassumeano-slipinterfaceandlookatarotationofradians. Wealreadyestablished
both that the center of the pulley moves upward a distance R and that the string length
increases on the right hand side by a distance R. That means the string length on the left
hand side has decreased by R. The position of the mass with respect to the center of the
pulleyisshorter(furtherupward)byR,andthecenterofthepulleyhasmovedupwarditself
by R, so x
1
= 2R= 2x
2
.
(c) We have already derived geometric equations which show (together) that all three variables
(x
1
,x
2
, and )mustbelinearlycoupledatalltime. Thismeanswecandescribethedynamics
of the system with a single equation of motion. Applying apply Newtons law at each inertia
and mass (there are no other nodes):
F

x
2
T
2
x
1
T
1
m
1
m
2
g
m
2
,J
R
m g
1
Figure 14.4.1: Pulley Free Body Diagrams
m
1
x
1
=T
1
m
1
g
m
2
x
2
=F T
1
T
2
m
2
g
J

= (T
2
T
1
)R
508
_ _
From the force balance at x
1
and the geometric relation between x
1
and x
2
, we can get an
expression for T
1
in term of x
2
: T
1
= 2m
1
x
2
+m
1
g. Using this new expression, the torque
balance, and the geometric relation between x
2
and , we can write T
2
as:
J J
T
2
=T
1
+
R
2
x
2
=m
1
g+ 2m
1
+
R
2
x
2
Combining, we get:
x
2
=
F (2m
1
+m
2
J
)g
4m
1
+m
2
+
R
2
The result is (of course) simply a constant acceleration (i.e. no damping or spring eect
anywhere).
509
14.5 DC Motors
(a)
V
m
= K
m

K
V
m
= i
m
R i
m
=
m

R
T = K
m
i
m
Note: in this case the current leaving the motor thus negative
K
2
T =
m

R
(b)
T
1
T
1
= K
m1
I I =
K
m1
K
m2
T
2
= K
m2
I T
2
= T
1
K
m1
V = K
m1

1
=IR+K
m2

2
1

2
= (K
m1

1
IR)
K
m2
1 T
1
1 T
2
= (K
m1

1
R ) = (K
m1

1
R )
K
m2
K
m1
K
m2
K
m2
(c) Thissystemisjustatransmissionwithlosses. Inthecaseofthemotors,thesystemoutputs
equivalenttorquesandshowslossesinthemotorvelocities. Inmostmechanicaltransmissions,
the velocities will be equivalent with the losses showing up in the output torque.
510
14.6 Gear Train

2 J
T
b
T
2 T
k
Figure 14.6.1: Freebody Diagram
Figure 14.6.1 shows the freebody diagram for the inertia of the system. For this problem we
can ignore all of the dynamics of the transformer except for the actual transformer. T
2
is the
transformed input torque T
1
. In my case I have drawn T
2
such that T
2
=
r
r
2
1
T
1
. If you drew T
2
2
acting in the positive direction you would get T
2
=
r
T
1
.
r
1
(a) The equation of motion for this system is
T
b
= b

2
T
k
= k
2
T
2
=
r
2
T
1
r
1

J
2
= T
2
T
b
T
k
J

2
+b

2
+k
2
=
r
2
T
1
r
1

2
+ 1.25

2
+ 10
2
=
r
2
T
1
r
1
To determine and
n
, we ignore the input thus

n
= 10 = 3.16
2
n
= 1.25 = 0.1980.2
(b) To determine the system equation in terms of
1
, we only need to note the characteristics
of the transformer. With the directions as dened on the picture, we know that
2
=
r
r
2
1

1
(Note: this is the inverse of the torque relationship). Since r
1
and r
2
are constants all the
derivativesof
1
and
2
sharethesamerelationship. Thusthecharacteristicequationbecomes

r
1
(J

1
+b

1
+k
1
) =
r
2
T
1
r
2
r
1
_ _
2
(

1
+ 1.25

1
+ 10
1
) =
r
2
T
1
r
1
You end up with the same coecients in the characteristic equation, thus and
n
are the
same as part a.
511
(c) Figure14.6.2showsasketchofthetimeresponsefor
1
(t) and
2
(t) (for my sketch r
2
/r
1
= 2).
Note the two responses are in opposite directions with the same phase and frequency. The
magnitudes are dierent by r
2
/r
1
.
A
m
p
l
i
t
u
d
e
0
(r
2
/r
1
)
(r
2
/r
1
)
2

2
0 1 2 3 4 5 6 7 8 9
Time (sec)
Figure 14.6.2: Sketch of
1
(t) and
2
(t)
(d) Since the system only contains dynamic elements on one side of the transformer, the trans-
former only eects the magnitude and sign of the system dynamics. If like in problem 1, the
system contained dynamic elements on both sides of the transformer, the transformer would
eect the dynamics of the system by scaling both the magnitude of the response and the
equivalent values of the inertia, damping, and stiness (see problem 1 where the equivalent
1
inertia of the transformer is J
1
+
r
r
2
J
2
).
512
14.7 Non-Ideal Transformer

T
b1

T
b2
2 J
T
T
1
T
J2
T
J2
1
J
1
3
J
2
T
k2
T
b3
T
k2
Figure 14.7.1: Freebody Diagrams
Tosolvethisproblemweneedtomakeafewassumptionsaboutthemodelelements. First,weneed
to consider the springs connecting the inertia elements in the model. We are told to assume that
k
1
isverylarge,thustheinputtorqueT
1
isassumedtoactdirectlyuponJ
1
. Thenextquestionwe
need consider is what to do with the damping caused by the bearings. In the case of the bearing
intheupperleftofthepicture,thelossesassociatedwiththisbearingclearlyactdirectlybetween
ground (

= 0) and the inertia J


1
. Less clear is how to handle the bearing in the center of the
picture. Our rst instinct would be to set this bearing to act between J
2
and J
L
. We can see that
this would be incorrect since the bearing drag would go to 0 when

2
=

3
(

3
is the rotational
velocity of J
2
). This is clearly incorrect since the bearing still generates a drag between the shaft
and ground. There are three dierent assumptions we can make instead. One, we can have the
bearing drag act between J
2
andground. Two,wecouldinsteadhavethebearingdragactbetween
J
L
andground. Lastly,wecouldsplitthedragandhaveitactbetweengroundandJ
L
andJ
2
. All
are equally valid but I will solve the problem by splitting the drag between the two inertia.
(a) Figure 14.7.1 shows the free body diagrams for the three inertias in the model. Note the
inclusion of the torque T
J2
on inertias J
1
and J
2
. This torque represents the torque through
thegeartransformerandwillbesolvedalgebraicallylater. Alltheothertorquesareexpressed
as follows (T
1
is an input thus not accounted for here):
T
b1
= b
1

1
T
b2
= b
2

3
T
k2
= k
2
(
2

3
)
T
b3
= b
3

2
513

_ _ _ _
_
The equations of motion then become

J = T

J
1

1
= T
1
T
b1
T
J2
J
1

1
+b
1

1
= T
1
T
J2

J
2

3
= T
J2
T
b2
T
k2
J
2

3
+b
2

3
+k
2

3
= k
2

2
T
J2

J
L

2
= T
k2
T
b3
J
L

2
+b
3

2
+k
2

2
= k
2

3
(b) Nowthatwehavewrittentheequationsofmotion,weneedtoeliminatetheunknowntorque
T
J2
. Todothisweapplyourknowledgeaboutthetransformerwhichequate
1
and
3
. Specif-
ically:
r
1

3
=
1
r
2
where r
1
is the radius of J
1
and r
2
is the radius of J
2
.
J
1

1
+b
1

1
= T
1
T
J2
J
2

3
+b
2

3
+k
2

3
= k
2

2
T
J2
r
1
(J
2

1
+b
2

1
+k
2

1
) = k
2

2
+T
J2
r
2

J
1
+
r
1
J
2

1
+ b
1
+
r
1
b
2

1
+
r
1

1
= T
1
k
2

2
r
2
r
2
r
2
J
L

2
+b
3

2
+k
2

2
= k
2

3
=k
2
r
1

1
r
2
We can determine the transfer function by taking the Laplace transforms of remaining two
equations of motion.
__ _ _ _ _

1
(s) J
1
+
r
r
1
2
J
2
s
2
+ b
1
+
r
r
1
2
b
2
s+
r
r
1
2
=T
1
(s)k
2

2
(s)

1
(s) =
T
1
(s)ks
2
(s)
J
1
+
r
1
J
2
s
2
+ b
1
+
r
1
b
2
s+
r
1
k
2
r
2
r
2
r
2

2
(s) J
L
s
2
+b
3
s+k
2
=k
2
r
r
1
2

1
(s)
k
2
r
1

1
(s)
r
2

2
(s) =
J
L
s
2
+b
3
s+k
2
For simplicity, I am going to dene the following transfer functions:
1
G
1
(s) = _ _ _ _
J
1
+
r
1
J
2
s
2
+ b
1
+
r
1
b
2
s+
r
1
k
2
r
2
r
2
r
2
k
2
r
r
2
1
G
3
(s) =
J
L
s
2
+b
3
s+k
2
(c) Figure 14.7.2 shows the block diagram for this system with the transfer functions as dened
above.
514

2
G
1
(s)

1
G
3
(s)
+
T
1
-
k
2
Figure 14.7.2: Block Diagram
(d) Using the block diagram, we see that
G
1
(s)G
3
(s)
G(s) =
1 +k
2
G
1
(s)G
3
(s)
k
2
r
r
1
G(s) = _ _ _ _
2
[ J
1
+
r
r
2
1
J
2
s
2
+ b
1
+
r
r
2
1
b
2
s+
r
r
2
1
k
2
][J
L
s
2
+b
3
s+k
2
] + k
2
2
r
r
2
1
515

14.8 Equivalent Mass and Inertia


(a) This problem is a little tricky. To solve it correctly you need to replace the mass with an
equivalent force (f
m
) as shown in the gure.
f
v
m
a b
f
v
v'
fm
a
b
Figure 14.8.1: Free Body Diagram for 1-a
f
m
= mv

v v
=
a b

a
v = v
b
a
f
m
= m v
b
af = bf
m
2
a
f = m v
b
2
2
a
m
eq
= m
b
2
(b) This problem should be solved using a free-body diagram. You should note that all of the
masses travel at the same velocity, thus you only need a free body diagrams for one wheel
and the combined masses.
Cart
Wheel
f
m,J
M
f
j
f
j

r
f
j
Figure 14.8.2: Free Body Diagram for 1-b
Note the force f
j
, this force is not a friction force, although it is created by the friction
between the tire and road surface, it is the force required to accelerate the rotational inertia
516
_ _
of the wheel. From the gure,
On the cart F = (M + 4m) x=f 4f
j
On one wheel T = J =f
j
r
x
No slip means v = x=r =
r
J
f
j
=
2
x
r
J
M + 4m+ 4 x = f
2
r
J
M
eq
= M + 4m+ 4
2
r
(c) Onceagainyoushouldbesolvingthisproblemusingfreebodydiagrams. Notethattherst
pulley does not have any inertia. From the FBD,
Pulley #1 T = 0 = r1(f
1
f
2
) + T
Pulley #2 T = J

Pulley 1 Pulley 2
f
1
f
2
r
1

f
2
f
1
'
r
2
J
T
Figure 14.8.3: Free Body Diagram for 1-c
517
14.9 Inertia in Geartrain
The solution for this problem is not available.
518
14.10 Motor Driving Inertia Through Gear Train
The solution for this problem is not available.
519
15 MATLAB and Simulink
15.1 Matlab Spirograph
(a) and (b) Agearofradiusr
2
turnsinsideagear-facedholeofradiusr
0
. Theinnergearhasapenxed
at a radius r
3
from its center. The resulting path of the pen is most easily described as that
ofthecenterofthegear(whichjustmovesinacircularpath)superimposedwiththeposition
of the pen with respect to the center of the gear. The angle between the xed center of the
outer hole and the center of the gear at any given time is
1
, while the corresponding angle
fromthecenterofthegeartothepenis
3
.
3
and
1
areproportionaltooneanother. Figure
15.1.1 shows the geometry of the gears (above) and the resulting path (below):
r
0
r
2
r
3
pen
C
C
0
P
0
1
P
1
Figure 15.1.1: Geometry of gears and resulting paths
Above,
1
isshownasadashedlinefortwolocations,0and90degrees. Here,r
0
= 7, r
2
= 4,
and r
3
= 3, so that
2
= (4/3)
1
, shown as the lines connecting gear center, c, with pen
location, p, for each of two locations of the inner gear. The gure below overlays the gear
geometry with the resulting spirograph:
520
r
0
r
2
r
3
pen
Figure 15.1.2: Overlay of geometry of gears and resulting spirograph
(c) The script spirographsol.m is avaliable at the end of the problem.
(d) Your plots will vary.
(c) You can convert the MATLAB script given into a MATLAB function by completing the
following 2 steps:
1) changetheTOPLINEofthem-leintoafunctiondeclaration(beginningwiththeword
function)
2) take out the lines in the m-le that dened the various radii (r0, r2 and r3)
The m-le spisol.m below denes a spirographing function called spisol this le is found at
the end of the problem. Note that the function has no output. (It will still plot the picture,
however.) If we wished to dene the vector of complex numbers describing the path as the
output of the function, we could rewrite the top line of the function above as:
function [position]=spisol(r0,r2,r3)
Therestof the m-lewouldremainthe same. Toplottheexampleshownearlier, you would
call the function as shown in Figure 15.1.1:
521
spi(7,4,3)
subplot(131); spi(4,3,2/3); axis off
subplot(132); spi(4,3,1); axis off
subplot(133); spi(4,3,4/3); axis off
subplot(131); spi(9,4,3); axis off
subplot(132); spi(9,5,4); axis off
subplot(133); spi(9,5,3); axis off
Figure 15.1.3: Spisol.m call and resulting plot
spirographsol.m
Download
% This MATLAB script plots a spirograph.
r0 = 2; % r0 is the outside (hole) radius
r2 = .84; % r2 is the inside gear radius
r3 = .7; % r3 is the radius at which the pen is
% held (wrt the gear center)
theta1 = [0:pi/1000:100*pi]; % theta1 defines an arbitrarily
% long list of angle positions which describes the path of
% the CENTER of the gear as it rotations in a ccw direction.
% Theta1 is long enough to ensure the path will have been
% traced completely at least once.
theta3 = -theta1*((r0/r2) - 1); % theta3 defines the angle
% of a line from the gear center to the gear pen. The geometry
% forces theta3 to be proportional to theta1 over time, just
% as with any standard gear system. Note, however, that the
% ratio of angles is the ratio of the two radii
% (hole_radius/gear_radius) MINUS 1, because for each time the
% gear makes a complete circuit around the hole, the surface
% on which it has travelled has "looped in on itself". Also,
% theta3 will be in the opposite sense as theta1: hence the
% minus sign.(i.e. If the gear center is making a ccw circuit
% in the hole, it is rotating in a cw direction.)
522
position = (r0-r2)*exp(j*theta1) + r3*exp(j*theta3);
% position defines a vector of complex numbers describing
% the complete path of the pen as the sum of the paths of
% the center of the gear and of pen wrt the gear.
figure(1) % make figure 1 the current figure
plot(position) % plots the real vs. imag parts of position
axis equal % forces the same scaling in x and y dirs.
spisol.m
Download
function []=spisol(r0,r2,r3)
theta1 = [0:pi/1000:100*pi];
theta3 = -theta1*((r0/r2) - 1);
position = (r0-r2)*exp(j*theta1) + r3*exp(j*theta3);
figure(1)
plot(position)
axis equal
523
15.2 MATLAB Plotting
10
5
0
5
10
(a)
0 0.05 0.1 0.15
0 0.5 1 1.5 2 2.5 3 3.5 4
1
0.5
0
0.5
1
x
(
t
)

(b)
0 0.5 1 1.5 2 2.5 3 3.5
1
0.5
0
0.5
1
t
(c)
Figure 2: Problem 5: (a) 10sin100t, (b) e
t
cos100t, (c) Re{e
st
}; s =
1 + 10i.
Problem 5: The objective of this problem is to recognize the time scales
involvedsothatpropertimerangeandsamplingratesmaybeselected. For
sinusoidal functions or exponential functions with a complex exponent of
the form ae
it
, the period is given by
2
T = (36)
||
For decaying exponential functions of the form ae
bt
the time constant is
given by
1
= (37)
b
The function will be within 1% of its nal value in 5. Therefore, one
must recognize the slowest timescale when plotting a function so that all
signicant behavior will be shown. The graphs are shown in Figures 2 and
3.
Figure 15.2.1:
524
0 0.1 0.2 0.3 0.4 0.5
0 0.1 0.2 0.3 0.4 0.5
0 0.02 0.04 0.06
t
Figure 3: Problem 5: (d) 5e
10t
, (e) 5e
10t
, (f) 8e
100t
5e
1000t
.
Figure 15.2.2:
0
2
4
6
(d)
0
200
400
600
800
x
(
t
)

(e)
0
2
4
6
(f)
525
15.3 Simulink Introduction
The solution for this problem is not available.
526
15.4 Wackygraph
(a) Them-lewackygraph.mcreatesastrangelookinggraphwithmanyloops(seeFigure15.4.1).
We are interested in understanding the eects of each parameter on the shape of the graph.
Begin by examining the function being plotted:
position= (r
0
+r
1
)e
j
1
+r
2
e
j
3
Recallthatacomplexexponentialisacirclewhenplottedinthethecomplexplane. r
0
andr
2
areconstants, butr
1
isafunctionof
1
. Substitutingrelationsgiveninthem-le, weobtain
position= 2r
0
e
j
1
+g
2
sin(N
1
)e
j
1
+r
2
e
jg
1

1
(1)
Therefore, we recognize that the diameter of the main gure is 2r
0
. This circle of main
diameterismodulatedbytwoadditionalterms,g
2
sin(N
1
)e
j
1
andr
2
e
jg
1

1
. g
2
sin(N
1
)e
j
1
hasthesameangle
1
asthemaincirclebutthemodulationchangesastheanglechanges. For
N > 1, the sizeoftheperturbationchangesfaster than
1
. Theotherperturbationr
2
e
jg
1

1
hasaprimaryradiusr
2
. Theperturbationrotatesintheoppositedirectionfromthedirection
oftracingoutthelargecircleofradiusr
0
andwithadierentspeedbasedonthemagnitude
of g
1
(see Figure 15.4.2 for a diagram explaining the variables). r
1
is a dependent variable
because it is calculated in the m-le.
(b) Using Equation (1) we can separate position into its real and imaginary parts using z =
re
j
=rcos+jrsin. We nd the real part is
Re(position) = 2r
0
cos
1
+g
2
sin(N
1
) sin
1
+r
2
cos(g
1

1
)
and the imaginary part is
Im(position) = 2r
0
sin
1
+g
2
sin(N
1
)sin
1
r
2
sin(g
1

1
)
(c) TocommentaMATLABle, usethe%symbolbeforethecomments. Acommentedversion
of wackygraph.m is given below.
20
15
10
5
0
5
10
15
20
25 20 15 10 5 0 5 10 15 20 25
Figure 15.4.1: Wackygraph with default parameters.
527
2r
0

1
r
2
g
1

1
g
2
sinN
1
Figure 15.4.2: Diagram of the Wackygraph system.
% wackygraph %
DL Trumper 9/9/02
r0 = 10; %define r0
r2 = 1; %define r2
N = 10; %define N
g1 = 40; %define g1
g2 = 0.2*r0; %define g2
theta1 = [0:pi/1000:100*pi]; %define the vector theta1
theta3 = -g1*theta1; %define the vector theta3 as proportional to theta1
r1 = r0 + g2*sin(N*theta1); %calculate the vector r1
position = (r0+r1).*exp(j*theta1) + r2*exp(j*theta3); %calculate position
figure(1) %select the window Figure 1
plot(position) %plot the variable position
axis equal %force the x and y axes to be scaled equally
Wemustuse.*insteadof. forelement-by-elementmultiplication. Otherwise,Matlabreturns
the following error:
??? Error using ==> *
Inner matrix dimensions must agree.
528
(d) Various plots are given in Figure 15.4.3.
(e) A possible script is given below
function wackyscript(r0, r2, g1, g2, N)
%wackyscript(r0, r2, g1, g2, N)
%Justin Verdirame
%2.003 Problem Set 2
%Fall 2002
%
%This function takes as input r0, r2, g1, g2, N and outputs
%a plot based on wackygraph.m written by Prof Trumper.
theta1 = [0:pi/1000:100*pi]; %define vector
theta3 = -g1*theta1; %calculate vector theta3
r1 = r0 + g2*sin(N*theta1); %calculate vector r1
position = (r0+r1).*exp(j*theta1) + r2*exp(j*theta3);
figure(1) %creates window for the plot if it does not exist
plot(position) %plot vector
axis equal %forces the x and y axes to have the same scaling
529
Figure 15.4.3: Wackygraph plotted with various parameters.
530
_ _
16 Case Studies
16.1 CD Player 1
(a) LettheouterradiusoftheCDbeR,andlettheradiusoftheholeber. Letthemassdensity
of the disc be per unit area, and let the total mass of the CD be m. Then
m=(R
2
)(r
2
) = (R
2
r
2
)
The moment of inertia is
1 1 1
I
CD
=(R
2
)( R
2
)(r
2
)( r
2
) = (R
4
r
4
)
2 2 2
and
1 1
_
m
_
1
(R
4
r
4
) = (R
4
r
4
) = m(R
2
+r
2
)
2 2 (R
2
r
2
) 2
Substitute R= 0.060 m, r= 0.0075 m, and m= 0.028 kg to get I
CD
= 5.12e-5 kg-m
2
.
(b) At steady state, the motor torque T
m
is balanced by the frictional torque T
fric
=B
m

ss
. If

ss
isthesameforthediscmountedandnotmounted,thedampingcoecientB
m
mustalso
bethesameinbothcases. Inthetransient,itisnecessarytoconsiderthecompletedynamic
analysis. The constitutive equations are
T
I
=I
d
m
and T
fric
=B
m

m
dt
where I is the total moment of inertia of the rotating parts, and the torque balance is
T
I
=T
m
T
fric
which lead to the dierential equation
d
m
I d
m
T
m
I +B
m

m
=T
m
or +
m
=
dt B
m
dt B
m
fromwhichwelearnthatthedecaytime-constantis =I/B
m
. Now we are given the time it
takes the spindle to reach 95% of
ss
which is known to be 3 time-constants, for the spindle
alone, and for spindle plus the mounted CD, so
I
r
0.3 I
r
+I
CD
2.0

noCD
= = and
withCD
= =
B
m
3 B
m
3
and, by subtraction to eliminate I
r
, we nd
I
CD
2.00.3 3I
CD
= or B
m
= = 9.04e-5 N-m-s/r
B
m
3 1.7
This is the answer for both parts (i) and (ii).
(c) To estimate I
r
, return to
noCD
and solve for I
r
0.3
I
r
=B
m

noCD
= 9.04e-5 =9.04e-6 kg-m
2
3
531
(d) The standard form for state-determined representation of a dynamic system is
d
x = Ax+Bu
dt
y = Cx+Du
wherexisacolumnmatrixofthestatevariables, uisacolumnmatrixoftheinputvariables,
and y is a column matrix of the desired output variables. The parameters which govern the
natural response of the system are contained in the square matrix A. The parameters which
describehowtheinputisdeliveredarecontainedinB,theparameterswhichdescribehowthe
outputdependsonthestatevariablesarecontainedinC,andtheparameterswhichdescribe
how the output depends directly on the input are contained in D.
In the present case there is only one state variable
m
, one input T
m
,andoneoutput
m
, so
all the matrices reduce to scalars. The state equation for
m
is
d
m
B
m
1
=
m
+ T
m
dt I
total
I
total
where I
total
= I
r
+I
CD
= (0.904+5.12)e-5 = 6.02e-5 kg-m
2
. The matrices of the standard
form reduce to the following scalars:
B
m
1
x=
m
, A= , u=T
m
, B= , y=
m
, C= 1, D= 0
I
total
I
total
532
16.2 CD Player 2
(a) The dimensions of the parameter G are [torque]/[angular velocity], or [FLT], or [ML
2
/T]. In
the SI system, the common unit is the Newton-meter per radian per second, or N-m-s.
(b) Assume that
sensed
=
m
, and insert T
m
= G(r
m
) in the state equation in Problem
16.1(d) to get
d
m
B
m
+G G
=
m
+ r
dt I
total
I
total
as the state equation for the controlled system. Comparing this equation for the controlled
system with the equation in 1(d) above, note that the input u has changed from a torque
T
m
to a speed r, and that, A has changed from B
m
/I
total
to (B
m
+G)/I
total
, and B has
changed from 1/I
total
to G/I
total
. The system is still a rst-order dynamic system but its
behavior can easily be modied by adjusting the gain G.
(c) The steady-state speed corresponding to 4 revolutions/second is
ss
= 8 = 25.1 radi-
ans/second.
(i) For the uncontrolled system, the motor torque required to maintain this speed is
T
m
=B
m

ss
= (9.04e-5)(8) = 2.27e-3 N-m
(ii) For the controlled system, the reference speed required to maintain the same speed is
B
m
+G 9.04e-5+G
r=
ss
= 25.1
G G
(d) Thedecaytime-constantfortheuncontrolledsystemis
u
=I
total
/B
m
,whilethedecaytime-
constant for the controlled systgem is
c
=I
total
/(B
m
+G). If the controlled system is to be
ve times faster than the uncontrolled system, it is necessary for
c
=
u
/5, or
I
total
1I
total
=
B
m
+G 5 B
m
from which we obtain
G= 4B
m
= 4(9.04e-5)=36.2e-5 N-m-s
(e) The scripts for this problem are found below. The symbol v in the script stands for the
angular velocity
m
. After the equation is integrated to obtain
m
(t), the motor torque T
m
is constructed from the control algorithm
T
m
=G(r
m
)
and plotted. See Fig. 16.2.1.
533
2
3
4
5
6
7
8
9
10
11
12
M
o
t
o
r

T
o
r
q
u
e

[
N
e
w
t
o
n

m
e
t
e
r
s
]

x10
-3
TimeHistoryofMotorTorque
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time[seconds]
Figure 16.2.1: Transient Motor Torque under Velocity Feedback Control
Note that the maximum motor torque occurs at t= 0 when
m
=0. The magnitude of the
maximum torque is
T
m
(0)=Gr= (B
m
+G)
ss
= (B
m
+ 4B
m
)
ss
= 5B
m

ss
= 11.35e-3 N-m
Note that the ve-fold increase in response speed is accompanied by a ve-fold increase in
maximum torque.
EqVelFdbksol.m
Download
% EqVelFdbksol.m
% This function, which is called by the program VelFdbk.m, contains the
% equation of motion of a CD player with velocity feedback.
%
% Introduce V_dot, the time derivative of v.
%
function V_dot = EqVelFdbksol(t,v)
%
% Declare global parameters (a MATLAB detail).
%
global Itot Bm r G
%
% The mathematical model of velocity feedback applied to the CD player
% provides an equation for V_dot.
534
%
V_dot = -((Bm + G)/Itot)*v + (G / Itot) * r;
%
% Thats all there is to it!
velfdbksol.m
Download
% VelFdbksol.m
% This program calculates the torque in CD player with velocity feedback.
% The differential equation used to model the system is:
%
% Itot dv/dt = - (Bm + G) v + G r
%
% This equation is stored in a separate m-file EqVelFdbk_sol.m called by this
% program (a tedious MATLAB detail).
%
clear variables
%
% Declare global parameters (a MATLAB detail).
%
global Itot Bm r G
%
% Input parameter values.
%
Itot = input(Enter the total moment of inertia "I(CD) + I(rotor)" in SI units: );
Bm = input(Enter the linear friction coefficient Bm in SI units: );
ss = input(Enter the steady-state speed in rad/sec );
G = input(Enter the value of the gain G in SI units: );
%
% Input initial condition.
%
v0 = input(Enter the initial angular velocity in rad/sec: );
%
% Input integration time.
%
tspan = input(Enter time interval of integration in seconds: );
%
% Calculate r
%
r= (( Bm + G ) / G) * ss;
%
% Call a numerical integration algorithm.
%
[t,v] = ode45(EqVelFdbksol, tspan, v0);
%
% Calculate motor torque
%
Tm = G * ( r - v );
% Plot the results.
%
plot(t,Tm)
xlabel(Time [seconds])
ylabel(Motor Torque [Newton meters])
title(Time History of Motor Torque)
grid\\
535
16.3 CD Player 3
The equation of motion for the CD-player is
d
m
d
m
B
m
1
I +B
m

m
=T
m
or =
m
+ T
dt dt I I
where: (i)I =I
r
whennodiscismountedonthespindle,or(ii)I =I
r
+I
d
when a disc is mounted
on the spindle.
(a) UsingtheresultofProblem16.1withB
m
= 2B,themagnitudeM()oftheangularvelocity
response amplitude and the phase angle () for an input torque of T
m
=T
a
sint are
M()=
T
a
_
1
and ()=tan
1
I
B
m
1 + (
I
)
2
B
m
Bm
The zero-frequency limit of the magnitude M() is
M(0)=
T
a
B
m
Since no specic value of the input torque amplitude T
a
was given, we will express the the
magnitude of the response in terms of the ratio M()/M(0). For the three frequencies 1
rad/sec, 5 rad/sec, 15 rad/sec the magnitude ratios and phase angles are:
(i) for no disc on the spindle (I =I
r
= 9e-6e-6 kg-m
2
, B
m
=9e-5 N-m/r/s),
M(1) 1 (9e6)(1)
M(0)
= _
1 + (
(9e6)(1)
= 0.995 and = tan
1
9e5
= 5.71deg
)
2
9e5
M(5) 1 (9e6)(5)
M(0)
= _
1 + (
(9e6)(5)
)
2
= 0.894 and = tan
1
9e5
= 26.6 deg
9e5
M(15) 1 (9e6)(15)
M(0)
= _
1 + (
(9e6)(15)
= 0.555 and = tan
1
9e5
= 56.3 deg
)
2
9e5
(ii) for a disc on the spindle (I =I
r
+I
d
= 5.9e-5 kg-m
2
, B
m
=9e-5 N-m/r/s),
M(1) 1 (5.9e5)(1)
M(0)
= _
1 + (
(5.9e5)(1)
= 0.647 and = tan
1
9e5
= 33.2 deg
)
2
9e5
M(5)
= _
1
= 0.1671 and = tan
1
(5.9e5)(5)
= 73.0 deg
M(0)
1 + (
(5.9e5)(5)
)
2
9e5
9e5
M(15) 1 (5.9e5)(15)
M(0)
= _
1 + (
(5.9e5)(15)
= 0.0564 and = tan
1
9e5
= 84.2 deg
)
2
9e5
536
(b) The low-frequency asymptote for M()/M(0) is unity, and the high frequency asymptote
is M()/M(0) = B
m
/I. These two asymptotes intersect at the break frequency
break
=
B
m
/I. The numerical values are:
(i)
break
=9e-5/9e-6 = 10 rad/sec when no disc is mounted on the spindle.
(ii)
break
=9e-5/5.9e-5 = 1.525 rad/sec when a disc is mounted on the spindle.
10
-1
10
0
10
1
10
2
10
3
10
-3
10
-2
10
-1
10
0
MagnitudeofResponseAmplitude
M
a
g
n
i
t
u
d
e

R
a
t
o

[
d
i
m
e
n
s
i
o
n
l
e
s
s
]

(i)NoDiscMounted
(ii)WithDiscMounted
Frequency [radians/second]
Figure 16.3.1: Magnitude Ratio for CD-Player
537
PhaseAngleofResponse
-90
-80
-70
-60
-50
-40
-30
-20
-10
0
P
h
a
s
e

a
n
g
l
e

[
d
e
g
r
e
e
s
]

(i)NoDiscMounted (ii)WithDiscMounted
10
-1
10
0
10
1
10
2
10
3
Frequency [radians/second]
Figure 16.3.2: Phase Angle for CD-Player
(c) Bode plots for Case (i) and Case (ii) are shown in Fig. 16.3.1 for magnitude and in Fig.
16.3.2forphaseangle. TheseplotswereobtainedbyrunningMATLABscriptssimilartothe
one shown in the solution to Problem 4.21.
(d) The decay time-constant for the CD-player is =I/B
m
. The time for 98% of the transient
to decay is 4 = 4I/B
m
. This time is a natural property of the system and is independent
of the frequency of the applied oscillating torque T
m
. This time does, however, depend on
the value of I which depends on whether a disc is mounted or not.
(i) With no disc mounted the time to reach steady state is 4(9e-6)/(9e-5) = 0.4 seconds.
(ii) With a disc mounted the time to reach steady state is 4(5.9e-5)/(9e-5) = 2.62 seconds.
538
_
_ _ _
16.4 Pinewood Derby 1
(a) ThepotentialenergyPEofamassmelevatedadistancehinthegravityeldisPE=mgh.
In SI units the 5 ounce mass is 5/16 pounds = (5/16)(0.4536 kg/pound) = 0.1418 kg. The
3 foot elevation h= 3(0.3048 meters/foot) = 0.914 meters, and the acceleration of gravity is
9.81 m/s
2
. The energy available to move the racecar is
mgh= (0.1418)(9.81)(0.914)=1.271kg (m/s)
2
= 1.271newton meters=1.271Joules
1
6
f
e
e
t
6
in
c
h
e
s

3

f
e
e
t

alp
h
a
Figure 16.4.1: Pinewood Derby
(b) If the center of mass is at the front of the car, its elevation is h= 0.914m and the available
potential energy is as calculated in (a). When the car is rolling on the level portion of the
track this energy has been transformed into kinetic energy KE=
1
2
mv
2
, so the maximum
velocity is
v= 2PE/m= 2gh= 2(9.81)(0.914)=4.24 m/s
(c) If the center of mass is at the rear of the car its elevation is
6 6 3
h= 0.914 + (0.3048)sin(alpha)=0.914 + (0.3048)( ) = 0.943 m
12 12 16
and the maximum velocity on the level portion of the track is
v= 2(9.81)(0,943)=4.30 m/s
a 1.4 % increase in speed over the case where the mass center is at the front of the car.
(d) The energy relations are silent regarding the time taken to eect the energy transformation.
To introduce the time variable it is necessary to consider the equation of motion of the car.
On the sloping section of track the force which accelerates the mass along the track is the
component of the weight parallel to the track: mgsin(alpha). With no friction, this is the
only force, and the equation of motion is
dv
mgsin(alpha)=m
dt
539

Integrating this equation from the initial condition v= 0 at t=0 yields


v=gtsin(alpha)
Let s represent distance down the track from the starting point, so that ds/dt = v. Then
integrating the equation
ds
=gtsin(alpha)
dt
from s= 0 at t= 0 to s=L at time t=T yields
1
L= gT
2
sin(alpha)
2
from which it follows that

2L
T =
gsin(alpha)
The time T to reach the bottom of the inclined track in case (c) is obtained by substituting
L= 16.5 feet = (16.5)(0.3048)m=5.09m and sin(alpha)=3/16 to get
2(5.09)
T = = 2.35 sec
9.81(3/16)
540
16.5 Pinewood Derby 2
Atthelimitingvelocityv
ss
,theretardingforcebv justequalsthecomponentoftheweightparallel
to the track, mgsin(alpha), so
mgsin(alpha)
b=
v
ss
If v
ss
= 2(4.30)=8.60 m/s, then
(0.1418)(9.81)(3/16)
b= = 0.0303 kg/sec
8.60
541
16.6 Pinewood Derby 3
Considertheequationofmotionofaracecarwithmassmonan inclineofangle(alpha), actedon
by the weight component parallel to the track, and by a viscous retarding force bv, as shown in
Figure 16.6.1.
v

m

b
v

m
g

s
i
n
(
a
l
p
h
a
)

Figure 16.6.1: Racecar Dynamics Diagram
dv
m =mgsin(alpha)bv
dt
or, after dividing through by m, and rearranging
dv b
+ =gsin(alpha)
dt m
This is the equation treated in the MATLAB scripts carsol.m and carviscsol.m.
(a) For part (a) use the data m= 0.1418 kg, (alpha) = inv sin 3/16=10.81 degrees, b= 0.0303
kg/sec, and T = 2.35 sec to get the plot for Part (a).
542
Speedofcaronincline
0 0.5 1 1.5 2 2.5
0
0.5
1
1.5
2
2.5
3
3.5
S
p
e
e
d
,

[
m
e
t
e
r
s
/
s
e
c
o
n
d
]

Time[seconds]
Figure 16.6.2: Plot for Part (a)
Attheendof2.35secondsthespeedofthecaris3.40m/s. Toindicatetheeectofincreas-
ing the weight, you can run the program again with all the same data, except with the mass
increasedto0.1701kg(6ounces)togetanalvelocityof3.53m/sattheendof2.35seconds.
(b) The limiting speed or terminal velocity established in Problem 16.5 is v
ss
= 8.60 m/s. The
speedwhichis99.9%ofthisis8.591m/s. Todeterminethetimeittakestoreachthisspeed,
xthevaluesofm= 0.1418kg,(alpha)=10.81degrees,andb= 0.0303kg/sec,andrunthe
program for a sequence of time intervals, iterating toward a nal velocity of 8.591 m/s. For
example,
Try T = 10 secs, get V = 7.59 m/s
Try T = 20 secs, get V = 8.49 m/s
Try T = 30 secs, get V = 8.596 m/s
Try T = 29 secs, get V = 8.593 m/s
Try T = 28 secs, get V = 8.589 m/s
Try T = 28.5 secs, get V = 8.591 m/s
The time history of velocity in the 28.5 second interval is displayed in the plot for Part (b).
The previous result is not very accurate because of round-o error. If we test the MATLAB
progams with the given data, we nd that for the very large times, T = 70 to T = 100, the
nal velocity is essentially constant at v = 8.61045 m/s; i.e., the program believes that the
terminal velocity is not 8.60 m/s, but 8.61045 m/s. If we then iterate to a nal velocity of
8.602 m/s, which is 99.9% of 8.61045 m/s, we nd that the time required is T = 32.5 sec,
which is 14% larger than our previous result of 28.5 sec.
543
0
1
2
3
4
5
6
7
8
9
S
p
e
e
d
,

[
m
e
t
e
r
s
/
s
e
c
o
n
d
]

Speedofcaronincline
0 5 10 15 20 25
Time[seconds]
Figure 16.6.3: Plot for Part (b)
carsol.m
Download
% carsol.
% This program calculates the velocity of a car rolling down an incline
% under the influence of viscous friction
% The differential equation used to model the car i
% m dv/dt = m g sin(alpha) - b
%
% This equation is stored in a separate m-file car_visc_sol called by this
% program (a tedious MATLAB detail
% Declare global parameters (a MATLAB detail
global g alpha b
% Input parameter value
g= 9.81
m = input(Enter the mass of the car in SI units: )
alpha = input(Enter the angle of the incline in degrees: )
b = input(Enter the value of the friction coefficient "b" in SI units:
%Input initial conditions.
v0 = input(Enter the initial velocity in SI units:
% Input integration time.
tspan = input(Enter time interval of integration in seconds: );
% Call a numerical integration algorithm.
[t,v] = ode45(carviscsol, tspan, v0);
% Plot the results.
plot(t,v)
xlabel(Time [seconds])
ylabel(Speed, [meters/second])
title(Speed of car on incline)
grid
544
30
% Print out speed at final instant.
%
N = max(size(v));
Finalspeed = v(N)
carviscsol.m
Download
% carviscsol.m
% This function, which is called by the program car.m, contains the
% equation of motion of a car on an incline subjected to viscous friction.
%
% Introduce V_dot, the time derivative of v.
%
function V_dot = carviscsol(t,v)
%
% Declare global parameters (a MATLAB detail).
%
global g alpha b m
%
% The mathematical model of the car provides an equation for v_dot.
%
V_dot = g * sin(alpha * pi/180) - (b/m) * v;
%
% Thats all there is to it!
545
_ _ _ _
_ _
A
C
16.7 Pinewood Derby 4
(a) Replace f
viscous
by f
drag
to get the free body diagram of the car mass acted on by the
component of the weight, and the retarding form-drag force, shown below.
d

v

2
s
g
n
(
v
)

v

m

m
g

s
i
n

Figure 16.7.1: Model Diagram


The equation of motion for the model is
m
dv
=mgsinAC
d
v
2
sgnv
dt
(b) The dimensions of the force f
drag
are [F] = [M][L]/[T
2
]; the dimensions of the area A are
[L
2
]; the dimensions of the velocity squared v
2
= [L
2
]/[T
2
]; and the signum function is
dimensionless. The dimensions of the parameter C
d
are
f
drag
ML/T
2
[C
d
] =
Av
2
sgnv
=
(L
2
)(L
2
/T
2
)
= [M/L
3
]
In SI units, the dimensions of C
d
are kg/m
3
.
(c) Onahorizontaltrack=0,andtheonlyhorizontalforceactingonthemassistheretarding
form-drag force. Starting with a positive velocity v
o
at t = 0, the mass will slow down, but
will not reverse direction; i.e., v(t) will be non-negative, so the equation of motion can be
written
dv
2
m =AC
d
v
dt
The initial condition is v = v
o
at t = 0. The dierential equation can be integrated by
separating the variables v and t
v
dv
t
AC
d
= dt
v
2
0
m
vo
1 1 AC
d
+ = t
v v
o
m
546
The time history of the speed of the car is obtained by solving the above equation for v(t).
v(t) =
v
o
1 +
AC
d
vo
t
m
Note that v(t) decreases smoothly with time, and asymptotically approaces the limit v = 0
as t approaces innity.
547
16.8 Pinewood Derby 5
The dierential equation for the model in which a car of mass m descends a very long inclined
track with angle under the inuence of gravity and form drag is
m
dv
=mgsinAC
d
v
2
sgnv
dt
as given at the beginning of Problem 16.7.
(a) At the terminal velocity v
ss
there is no longer any acceleration (dv/dt = 0), and the drag
force is equal and opposite to the accelerating force component, so that
mgsin
2
v =
ss
AC
d
(b) If m= 0.1418 kg, = 10.81 degrees = sin
1
(3/16), and v
ss
=8.60 m/s, then the product of
the frontal area A and the drag coecient C
d
is
mgsin (0.1418)(9.81)(3/16)
AC
d
= = = 0.00353 kg/m
v
2
(8.60)
2
(c) To use the scripts car2sol.m and cardragsol.m to determine the time T for V to reach
(0.999)(8.60)=8.5914m/sxtheinputsofm= 0.1418kg,= 10.81deg,andAC
d
= 0.00353
kg/m
3
nandthencallcar2-solandinputvariousguessesforT,iteratingtowardFinalspeed
= 8.5914 m/s. For example:
Try T = 10 secs, get V = 8.362 m/s
Try T = 20 secs, get V = 8.594 m/s
Try T = 19 secs, get V = 8.5919 m/s
Try T = 18.9 secs, get V = 8.5917 m/s
Try T = 18.8 secs, get V = 8.5915 m/s
Try T = 18.7 secs, get V = 8.5913 m/s
Try T = 18.75 secs, get V = 8.5914 m/s (close enough)
Thisresultisnotveryaccurate,becauseofround-oerror. Thevalue0.00353kg/mobtained
in(b)abovefortheparameteraC
d
wasroundedotothreesignicantgures. Ifweassume
thatAC
d
isexactlyequalto0.00353kg/m,andworkbackwardswendthatv
ss
isnotexactly
equal to 8.60 m/s but is actually 8.5958 m/s correct to ve signicant gures. If we repeat
the exercise we nd that at T = 17.45 sec, the nal speed is 8.5872 m/s which is 99.9 % of
8.5958 m/s. Both of the previous results assume that there are no errors in the integration
carried out by MATLAB. MATLAB is pretty good, but it does make small errors which can
accumulateoveralongintegration. IfweletitrunforlongtimesT,wendthatthroughout
the range from T = 35 secs to T = 70 secs, the nal speed is essentially constant (to ve
signicant gures) at 8.5970 m/s; i.e., MATLAB disagrees, in the fourth signicant gure,
with both v
ss
= 8.6000 m/s and v
ss
= 8.5958 m/s.
(d) In Problem 4(b) of Problem 16.7, it was found that with viscous friction the time to reach
99.9% of the terminal velocity was in the neighborhood of 28.5 to 32.5 seconds. Here, with
form drag, the same speed is reached in about 17.45 to 18.75 seconds. With form drag
(quadratic speed dependence) the terminal velocity is approached more quickly than is the
caseforviscousfriction(linearspeeddependence). Thetimehistoriesofvelocityforthetwo
models are shown in the Fig. 2(d)-1.
548
Speedofcaronincline
0
1
2
3
4
5
6
7
8
9
S
p
e
e
d
,

[
m
e
t
e
r
s
/
s
e
c
o
n
d
]
ViscousFriction
FormDrag
0 5 10 15 20 25 30
Time[seconds]
Figure 16.8.1: Fig. 2(d)-1
To explain why a car with quadratic drag approaches the terminal velocity more rapidly than a
car with linear drag, compare the constitutive equations for the retarding forces as shown in Fig.
16.8.2. At all speeds below the common terminal velocity there is less retardation with form drag
than there is for viscous friction. This means that, when both cars have the same velocity, the
car with form drag accelerates faster than the car with viscous friction. Now the the slopes of the
velocity curves in Fig. 16.8.1represent the accelerations ofthe cars. Thus at each speed, the form
drag curve will have a steeper slope than the viscous friction curve. As a consequence the form
drag curve must approach the terminal velocity more quickly than the viscous friction curve.
549
0 1 2 3 4 5 6 7 8 9 10
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
R
e
t
a
r
d
i
n
g

f
o
r
c
e

[
N
e
w
t
o
n
s
]

FormDrag
ViscousFriction
ConstitutiveEquationsforRetardingForces
Velocity[meters/second]
Figure 16.8.2: Fig. 2(d)-2
NOTE re MATLAB technique: To plot two graphs in the same Figure, as in Fig. 2(d)-1. Run
car2soltoobtainaplotofviscousfrictioncase. Then,intheMATLABCommandWindow,type
hold on. (this keeps the plot in the Figure screen). Next run car2sol (with the same time
interval T). The plot for the form-drag case then appears on top of the existing plot. To print
captions such as Viscous Friction on the interior of the plot, type, in the MATLAB Command
Window, gtext(Viscous Friction). When you hit return the cursor on the Figure turns into a
giantcrosswhichcanbemovedbythemouse. Thenwhenyouclickthemousethecaptionappears
on the Figure starting from the intersection point of the cross. You can make as many additional
plotsasyouwish,aslongasholdison. Toreturntonormalone-at-a-timeplots,typeholdo
in the MATLAB Command Window.
car2sol.m
Download
% car2sol.m
% This program calculates the velocity of a car rolling down an incline
% under the influence of form drag.
% The differential equation used to model the car is:
%
% m dv/dt = m g sin(alpha) - A Cd v^2
%
% This equation is stored in a separate m-file car-drag-sol called by this
% program (a tedious MATLAB detail).
%
% Declare global parameters (a MATLAB detail).
550
%
global g alpha ACd m
%
% Input parameter values.
%
g= 9.81;
m = input(Enter the mass of the car in SI units: );
alpha = input(Enter the angle of the incline in degrees: );
ACd = input(Enter the value of the parameter "A*Cd" in SI units: );
%
% Input initial conditions.
%
v0 = input(Enter the initial (non-negative) velocity in SI units: );
%
% Input integration time.
%
tspan = input(Enter time interval of integration in seconds: );
%
% Call a numerical integration algorithm.
%
[t,v] = ode45(cardragsol, tspan, v0);
%
% Plot the results.
%
plot(t,v)
xlabel(Time [seconds])
ylabel(Speed, [meters/second])
title(Speed of car on incline)
grid
%
% Print out speed at final instant.
%
N = max(size(v));
Finalspeed = v(N)
cardragsol.m
Download
% cardragsol.m
% This function, which is called by the program car2_sol.m, contains the
% equation of motion of a car on an incline subjected to form drag. It
% is assumed that the car velocity is nonnegative.
%
% Introduce V_dot, the time derivative of v.
%
function V_dot = cardragsol(t,v)
%
% Declare global parameters (a MATLAB detail).
%
global g alpha ACd m
%
% The mathematical model of the car provides an equation for v_dot.
%
V_dot = g * sin(alpha * pi/180) - (ACd/m) * v^2;
%
% Thats all there is to it!
551
16.9 Engine Block Vibration 1
Weight of engine block W = 200 pounds, stiness k = 18,000 pounds/inch, damping coecient
b=2 pounds /inch/second. Input force is f(t) = f
a
sint with f
a
=2 pounds.
f (t )
m
y
b k
Figure 16.9.1: Engine Block Subjected to Sinusoidally Varying Force
Let y(t) be displacement of engine block from equilibrium position. Geometric compatibility re-
quires that the block velocity v(t) satisfy v = dy/dt. Restoring force is f
k
= ky, damping force
is f
fric
= bv, input force is f(t) = f
a
sint, and Newtons second law is f
m
= mdv/dt with
m = W/g. When these constitutive equations are substituted in the force-balance requirement
f
m
=f(t)f
k
f
fric
and v is replaced by dy/dt the result is the equation of motion
m
d
2
y
+b
dy
+ky=f
a
sint
dt
2
dt
(a) Iftheinputforceisconsideredtobetheimaginarypartofthecomplexexcitationf
a
exp(it),
itisappropriatetoconsiderthesteady-stateresponsetobetheimaginarypartofthecomplex
responseAexp(it) where Aisthecomplexamplitudeoftheresponse,whichistobefound
by requiring the complex response to satisfy the equation of motion. Substitution of y =
Aexp(it) into
d
2
y dy
m +b +ky=f
a
exp(it)
dt
2
dt
yields
f
a
A=
k+ibm
2
(b) (i) At very low frequencies Af
a
/k.
(ii) At very high frequencies Af
a
/m
2
.
(iii) The magnitudes of (b) and (c) become equal when
2

break
2
=k/m=
o
2
.
552
_
_
_
(c) In terms of the behavioral parameters
o
and dened by

2
=
k
and 2
o
=
b
o
m m
the ratio Z of A() to A(0),
A() 1
Z = =

2
A(0)
(1

o
2
) + i2

o
is a convenient dimensionless form of the complex amplitude A. The engine displacement
amplitude has its peak magnitude when [Z[
2
is a maximum. Now
1

2
[Z[
2
=
(1 )
2
+ 4
2

2

2
o o
and its maximum occurs when its derivative with respect to
2
/
o
2
vanishes; i.e., when
d Z
2

2
[ [
= 2(1 ) + 4
2
= 0
d
2
/
o
2

o
2
The solution of this equation for yields

peak
=
o
12
2
(d) Toevaluate[Z(
peak
)[,insert
2
=
o
2
(12
2
)intheprecedingexpressionfor[ [ Z
2
toget
peak
1 1
[Z(
peak
)[
2
=
4
4
+ 4
2
(12
2
)
=
4
2
(1)
2
so that
1
[Z(
peak
)[ = _
2 1
2
For comparison, note that [Z(
o
)[ = 1/2. The actual peak occurs at a frequency that is
smallerthantheundampednaturalfrequency
o
by a factor of 12
2
andthemagnitude
ofthepeakisgreaterthanthemagnitudeoftheresponseat=
o
bythefactor1/ 1
2
.
553

MagnitudeofResponseAmplitudeRatio
10
-2
10
-1
10
0
10
1
10
2
-80
-60
-40
-20
0
20
40
FrequencyRatio,OMEGA/omega0 [dimensionless]
|
Z
(
O
M
E
G
A
)
/
Z
(
0
)
|
,

D
e
c
i
b
e
l
s

[
d
i
m
e
n
s
i
o
n
l
e
s
s
]

Figure 16.9.2: Bode Plot of Magnitude of A()/A(0) in dB vs. Frequency Ratio /


o
PhaseAngleofResponse
-200
-180
-160
-140
-120
-100
-80
-60
-40
-20
0
P
h
a
s
e

a
n
g
l
e

[
d
e
g
r
e
e
s
]

10
-2
10
-1
10
0
10
1
10
2
FrequencyRatio,OMEGA/omega0 [dimensionless]
Figure 16.9.3: Bode Plot of Phase Angle vs. Frequency Ratio /
o
554
(e) TheBodeplotsshowninFig.2andFig.3wereobtainedbyrunningthescriptenginebodesol.m.
enginebodesol.m
Download
% enginebodesol.m makes Bode Plots for Engine Block Vibration,
% in terms of the amplitude ratio Z = A(OMEGA)/ A(omega0) and the
% frequency ratio OMEGA/omega0 (OmRa = Omega Ratio).
clear variables
zeta = 0.01035;
OmRa = logspace(-2, 2, 200);
unity = ones(1,length(OmRa));
Re = unity - OmRa.^2;
Im = 2* zeta *OmRa;
Dsq = Re.^2 + Im.^2;
MagZsq = unity ./ Dsq;
dB = 10* log10(MagZsq);
phi = 57.3 * atan2( -Im , Re );
semilogx(OmRa,dB), title(Magnitude of Response Amplitude Ratio),
xlabel(Frequency Ratio, OMEGA/omega0 [dimensionless]),
ylabel(|Z(OMEGA)/Z(0)|, Decibels [dimensionless] ),
pause
semilogx(OmRa, phi), title(Phase Angle of Response),
xlabel(Frequency Ratio, OMEGA/omega0 [dimensionless]),
ylabel(Phase angle [degrees])
555
16.10 Engine Block Vibration 2
Repeat Problem 16.9, but with f(t) = C
2
sint with C = 0.005 pound-sec
2
.
(a) Substitution of y=Aexp(it) into
m
d
2
y
+b
dy
+ky=C
2
exp(it)
dt
2
dt
yields
C
2
A=
k+ibm
2
(b) (i) At very low frequencies AC
2
/k.
(ii) At very high frequencies AC/m.
(iii) The magnitudes of (b) and (c) become equal when
2

2
=k/m=
2
.
break o
(c) In terms of the behavioral parameters
o
and dened by
k b

o
2
= and 2
o
=
m m
the ratio Z of A() to A(),

2
A()

o
2
Z = =
(1

2
) + i2

A()

o
o
is a convenient dimensionless form of the complex amplitude A. The engine displacement
amplitude has its peak magnitude when [Z[
2
is a maximum. Now

2
[Z[
2
=
(1 )

2
o
4
+ 4
2

2

2
o o
and its maximum occurs when its derivative with respect to
2
/
o
2
vanishes; i.e., when

2
2

2
[(1

2
)
2
+ 4
2

2
]

4
[2(1

2
) + 4
2
] = 0
o o o o o
The solution of this equation for
2
/
o
2
yields

2
1
=

2
12
2
o
so that

peak
= _

o
12
2
(d) To evaluate [Z(
peak
)[, insert
2
/
2
= 1/(12
2
) in the preceding expression for[Z[
2
to
peak o
get
[Z(
peak
)[
2
=
(
1
1
2
2
)
2
2
=
1
(1
1
1
2
2
)
2
+
1
4
2
2
4
2
(1)
2
556

_
_
_
so that
1
[Z(
peak
)[=
2 1
2
ComparingthisresultwiththatofProblem16.9,wenotethattheactualpeakhereoccursata
frequencythatisgreaterthantheundampednaturalfrequency
o
byafactorof1/ 12
2
,
but the magnitude of the peak is the same. It is greater than the magnitude of the response
at =
o
by the same factor 1/ 1
2
.
MagnitudeofResponseAmplitudeRatio
10
-2
10
-1
10
0
10
1
10
2
FrequencyRatio,OMEGA/omega0 [dimensionless]
Figure 16.10.1: Bode Plot of Magnitude of A()/A(0) in dB vs. Frequency Ratio /
o
-80
-60
-40
-20
0
20
40
|
Z
(
O
M
E
G
A
)
/
Z
(
0
)
|
,

D
e
c
i
b
e
l
s

[
d
i
m
e
n
s
i
o
n
l
e
s
s
]

557

-200
-180
-160
-140
-120
-100
-80
-60
-40
-20
0
P
h
a
s
e

a
n
g
l
e

[
d
e
g
r
e
e
s
]

PhaseAngleofResponse
-2 -1 0 1 2
10 10 10 10 10
FrequencyRatio,OMEGA/omega0 [dimensionless]
Figure 16.10.2: Bode Plot of Phase Angle vs. Frequency Ratio /
o
(e) TheBodeplotsshowninFig. 16.10.1andFig. 16.10.2wereobtainedbyrunningthefollowing
script (adapted from the Script for Problem 16.9):
enginevibration2sol.m
Download
% enginevibration2sol.m makes Bode Plots
% in terms of the amplitude ratio Z = A(OMEGA)/ A(omega0) and the
% frequency ratio OMEGA/omega0
clear variables
zeta = 0.01035;
OmRa = logspace(-2, 2, 200);
unity = ones(1,length(OmRa));
Re = unity - OmRa.^2;
Im = 2* zeta *OmRa;
Dsq = Re.^2 + Im.^2;
Nsq = OmRa.^4
MagZsq = Nsq ./ Dsq;
dB = 10* log10(MagZsq);
phi = 57.3 * atan2( -Im , Re );
semilogx(OmRa,dB), title(Magnitude of Response Amplitude Ratio),
xlabel(Frequency Ratio, OMEGA/omega0 [dimensionless]),
ylabel(|Z(OMEGA)/Z(0)|, Decibels [dimensionless] ),
pause
semilogx(OmRa, phi), title(Phase Angle of Response),
xlabel(Frequency Ratio, OMEGA/omega0 [dimensionless]),
ylabel(Phase angle [degrees])
558
16.11 Engine Block Vibration 3
The equation of motion
d
2
y dy
m +b +ky=f(t) = f
a
sint
dt
2
dt
obtained in Problem 16.9 can be rewritten in the standard form for state-determined systems by
taking y(t) and v(t) = dy/dt as state varibles and writing the state equations in matrix form
_ _ _ _ _ _ _ _
_
y
_ 0 1 _
y
_ _
0
_
d
_ _
= +
f
a
sint
dt_
v
_

b _
v
_ _
1
_m
m m
x10
-6
HistoryofDisplacement
x10
-3
HistoryofPowerInput
5 3
-5
0 2 4 6
-3
Time [Seconds] Time [Seconds]
x10
-5
PowerDissipated
x10
-3
HistoryofPowerStored
0 2 4 6
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]
2
1
0
-1
-2
P
o
w
e
r

F
l
o
w

i
n
t
o

S
t
o
r
a
g
e

[
W
a
t
t
s
]

P
o
w
e
r

I
n
p
u
t

[
W
a
t
t
s
]
0
3.5 3
2
1
0
-1
-2
P
o
w
e
r

D
i
s
s
i
p
a
t
e
d

[
W
a
t
t
s
]
3
2.5
2
1.5
1
0.5
0 -3
0 2 4 6 0 2 4 6
Time[Seconds] Time[Seconds]
Figure 16.11.1: Engine Start-Up when =
break
/3
Oneofthedesiredoutputsisthestatevariabley(t). TheotheroutputsdesiredareP
in
=f(t)v(t);
559
P
diss
=bvv; and P
stored
=P
in
P
diss
. The values of the physical parameters are:
m=200 pounds = 90.7 kg
k= 18,000pounds/inch = 3.152e6 Newton/meter
b=2 pounds/inch/second = 350 Newtons/meter/second
f
a
=2 pounds = 8.90 Newtons

break
/3 = 62.1 rad/sec = 62.1 rad/sec

break
=186.4 rad/sec = 186.4 rad/sec
3
break
=559 rad/sec = 559 rad/sec
TheplotsshowninFigs. 6-8wereobtainedbyrunningthefollowingMATLABscripts. Theinput
script is called engstartsol.m and the script called by engstartsol.m is called eqengstartsol.m.
x10
-4
HistoryofDisplacement HistoryofPowerInput
1.5 0.3
-1
0
-1.5 -0.05
0 2 4 6 0 2 4 6
Time [Seconds] Time [Seconds]
PowerDissipated HistoryofPowerStored
0.25
1
P
o
w
e
r

D
i
s
s
i
p
a
t
e
d

[
W
a
t
t
s
]

D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

P
o
w
e
r

F
l
o
w

i
n
t
o

S
t
o
r
a
g
e

[
W
a
t
t
s
]

P
o
w
e
r

I
n
p
u
t

[
W
a
t
t
s
]
0.5
0.2
0.15
0
0.1
-0.5
0.05
0.25
0.2
0.15
0.1
0.05
0.06
0.05
0.04
0.03
0.02
0.01
0
0 -0.01
0 2 4 6 0 2 4 6
Time[Seconds] Time[Seconds]
Figure 16.11.2: Engine Start-Up when =
break
560
x10
-6
HistoryofDisplacement
x10
-3
HistoryofPowerInput
1.5 2
1
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

P
o
w
e
r

F
l
o
w

i
n
t
o

S
t
o
r
a
g
e

[
W
a
t
t
s
]

P
o
w
e
r

I
n
p
u
t

[
W
a
t
t
s
]
1
0
-1
0.5
0
-0.5
-1
-1.5 -2
0 2 4 6 0 2 4 6
Time [Seconds] Time [Seconds]
x10
-5
PowerDissipated
x10
-3
HistoryofPowerStored
3.5 2
1
0
-1
P
o
w
e
r

D
i
s
s
i
p
a
t
e
d

[
W
a
t
t
s
]
3
2.5
2
1.5
1
0.5
0 -2
0 2 4 6 0 2 4 6
Time[Seconds] Time[Seconds]
Figure 16.11.3: Engine Start-Up when =3
break
engstartsol.m
Download
% EngStartsol.m, a MATLAB script that produces plots of
% (i) position vs. time
% (ii) input power vs. time
% (iii) dissipated power vs. time
% (iv) power flow into storage vs. time
% for the response of an engine block, with mass m, stiffness k,
% and damping parameter b, when the system starts from initial conditions
% y = 0 and v = 0 under the action of a suddenly applied sinusoidal force
% fa sin OMEGA t, at t = 0.
clear variables
global m k b fa Om
% Input parameters
m = 90.7; %input(Enter the mass "m" in kilograms );
k = 3.152e6; %input(Enter the stiffness "k" in Newtons/meter );
b = 350; %input(Enter the damping constant "b" in kilograms/sec );
fa= 8.90; %input(Enter the magnitude "fa" of the suddenly applied force in Newtons );
Om= input(Enter the frequency OMEGA of the suddenly applied force in rad/sec );
% Input initial conditions.
561
y0= 0;
v0= 0;
tspan = input(Enter the duration "T" of the desired time history, in seconds );
X0 = [ y0 ; v0 ];
% Integrate equations of motion
[t,X] = ode45(EqEngStartsol, tspan, X0);
% Construct power variables
Pin = fa * sin(Om*t) .* X(:,2);
Pdiss = b*X(:,2).^2;
Pstored = Pin - Pdiss;
% Plot results
subplot(221),plot(t,X(:,1)), title(History of Displacement),
xlabel(Time [Seconds]), ylabel(Displacement [meters]), pause
subplot(222),plot(t,Pin), title(History of Power Input),
xlabel(Time [Seconds]), ylabel(Power Input [Watts]), pause
subplot(223),plot(t, Pdiss), title(Power Dissipated),
xlabel(Time [Seconds]), ylabel(Power Dissipated [Watts]), pause
subplot(224),plot(t, Pstored), title(History of Power Stored),
xlabel(Time [Seconds]), ylabel(Power Flow into Storage [Watts])
\end{verbatim}
EqEngStartsol.m
Download
% EqEngStartsol.m Provides equation of motion for initial transient
% to be integrated by script EngStart-sol.m
function Xdot = EqEngStartsol(t,X)
global m k b fa Om
Xdot = [ 0 1 ; -k/m -b/m ]*X + [ 0 ; fa/m*sin(Om*t)];
\end{verbatim}
562
16.12 Plate On Springs 1
m
y
k
k
c
k
c
Figure 16.12.1: Steel Plate on Springs
The displacement y
k
measures the vertical extension of the four springs from the equilibrium
conguration where the weight of the plate mg is balanced by initial compressive forces in the
springs. Theexcesstensileforceineachspringduetoadisplacementy
k
isf
ki
=k
c
y
k
(i= 1,,4).
The total vertical spring force is f
k
= 4f
ki
= 4k
c
y
k
=ky
k
. The constitutive equation for the mass
is Newtons law f
m
= mdv
m
/dt. When friction is absent and there is no external load, the force
balance equation is simply f
k
=f
m
, or
dv
m
dv
k
ky
k
=m , or ky
k
=m
dt dt
on using the geometric compatibility condition v
m
= v
k
. The standard form of the dierential
equation (See Lecture notes of 9/15/99) is
d
2
y
k
k
+ ( )y
k
= 0
dt
2
m
Guess a solution of the form y
k
=Aexp(t), which leads to the characteristic equation

2
+
2
= 0
o
where
2
= k/m. The roots of the characteristic equation are +i
o
and i
o
, and the general
solution of the dierential erquation is
y
k
=Aexp(i
o
t) + Bexp(it)
AnalternativeformofthegeneralsolutioncanbeobtainedbyintroducingEulersformulaexp(i
o
t) =
cos
o
t+isin
o
t.
y
k
=C
1
cos
o
t+C
2
sin
o
t
These two forms of solution are equivalent if the constants of integration are related as follows:
1
C
1
= A+B A= (C
1
+C
2
)
2
1
C
2
= AB B = (C
1
C
2
)
2
563
Thegiveninitialconditionsarey
k
(0) = 0anddy
k
/dt(0)=v
o
. Usingthetrigonometricformofthe
general solution, we have
y
k
=C
1
cos
o
t+C
2
sin
o
t and
dy
k
=C
1

o
sin
o
t+C
2

o
cos
o
t
dt
Setting t=0 yields
C
1
=0 and C
2
=
v
o

o
so the displacement history of the plate, starting from rest at the equilibrium position with the
initial velocity v
o
is
y
k
(t) =
v
o
sin
o
t

o
564
_
_
16.13 Plate On Springs 2
TheunloadedplatesystemhereisthesameasthatconsideredinProblem16.12. Themassofthe
plateismandtheeectivestinessofthefourspringsisk. Thenaturalfrequencyoftheunloaded
system is
o
where
2
=k/m. When the book of mass m/2 is attached to the plate, the resulting
system has mass 3m/2 and stiness k, and its natural frequency of free oscillation is
1
where

1
2
= 2k/3m. The oscillations in both Case I and Case II involve motions with the book attached
to the plate. The dierence between Cases I and II lies in the initial conditions, not in the basic
system.
(a) TheratiooffrequenciesofCaseIoscillationstounloadedplateoscillationsis
1
/
o
= 2/3 =
0.816.
(b) The ratio of frequencies of Case II oscillations to unloaded plate oscillations is
1
/
o
=
2/3 = 0.816.
(c) In the (book + plate) system there is an additional gravity load of mg/2, so the equilibrium
position of the (book + plate) system is beneath the equilibrium position of the unloaded
plate by a distance of =mg/2k.
(d) In Case I the system starts from rest at the equilibrium position of the unloaded plate.
Viewed from the equilibrium position of the (book + plate) system, the initial displacement
y
k
(0)=, and the initial velocity is dy
k
/dt(0)=0.
(e) In Case II there is an initial velocity dy
k
/dt(0) = v
1
imparted by the impact of the book
upon the plate. This initial velocity can be determined by applying conservation of linear
momentumtotheimpact. Immediatelybeforetheimpacttheplatehasnomomentumwhile
the book has a downward momentum of 1/2mv
o
. Immediately after the impact the (book
+ plate) has the downward momentum 3mv
1
/2. For these two momenta to be equal it
is necessary that v
1
= v
o
/3. The initial conditions for Case II are thus: y
k
(0) = and
dy
k
/dt(0)=v
o
/3.
(f) The general solution for both Case I and Case II is the same as the general solution for
Problem16.12,exceptthat
o
iseverywherereplacedby
1
. ForCaseI,theinitialconditions
require
y
k
(0)==C
1
and
dy
k
dt
(0)=0=C
2

1
so that
y
k
(t) = cos
1
t
where =mg/2k, and
1
2
= 2k/3m.
For Case II, the initial conditions require
y
k
(0)==C
1
and
dy
k
(0)=v
1
=C
2

1
dt
so that
y
k
(t) = cos
1
t
v
1
sin
1
t

1
where =mg/2k, v
1
=v
o
/3, and
1
2
= 2k/3m
565
16.14 Mousetrap Dynamics 1
The solution for this problem is not available.
566
16.15 Mousetrap Dynamics 2
The solution for this problem is not available.
567
16.16 Mousetrap Dynamics 3
The solution for this problem is not available.
568
16.17 Mousetrap Dynamics 4
The solution for this problem is not available.
569
16.18 Hydraulic Elevator Design
(a) Astheelevatorfalls,thereisboththeconstantforceduetotheaccelerationofgravityanda
damping force which is proportional to (and in a direction opposing) the velocity.
F =ma
mgcv=ma
cx + mx = mg
Note that the x coordinate is dened to be positive in the up direction. Therefore, mg is in
the negative x direction. The cv term must always act in opposition to the direction of the
velocity. (Ifthemassismovinginthenegativexdirection,velocityisnegative,andtheforce
acts in the positive direction to slow the fall, for instance.)
When the elevator suddenly fails, whatever force has been compensating for the force of
gravity (from tension in elevator cables or force applied by the hydraulic lift, etc) suddenly
disappears,resultinginanetstepinforce. Thus,thefalloftheelevatorcanbewell-modeled
as a 1st order system with a step input:
cx + mx = mgu
s
(t)
1
(cs+ms
2
)X(s) = mgU(s) = mg
s
X(s) =
mg
s
2
(c+ms)
Note that the velocity and acceleration are just the rst and second derivatives of x(t):
V(s) = sX(s) =
mg
s(c+ms)
A(s) = s
2
X(s) =
mg
c+ms
Theacceleration,A(s),shouldlookfamiliar;itisjusttheLaplacedomainrepresentationofa
1stordersystem. Weknowtheresponsewillstartatsomevalue(whichwecanndbyusing
theI.V.T.)andgotosomesteadystatevalue(whichwecanndbyusingtheF.V.T.)viaan
exponential decay with some time constant. Rewriting A(s),
A(s) = s
2
X(s) =
m
mg/c
s+ 1
c
it should be clear to you that the time constant is:
m 2000kg
= = = 1 [s]
c 2000kg/sec
570
Using the Initial Value Theorem:
A(0)= lim s
mg/c
=
gm/c
=g=9.8 [m/s
2
]
s
m
c
s+ 1 m/c
This is what wed expect from the basic physics. Using the Final Value Theorem:
A()= lims
mg/c
= 0 [m/s
2
]
s0
m
s+ 1
c
therefore
a(t) = 9.8e
t
[m/s
2
]
v(t) = 9.8e
t
+C
1
[m/s]
To match the initial condition that v(0)=0,
v(t) = 9.8e
t
9.8 [m/s]
Position can be obtain by taking another integral:
x(t) = 9.8e
t
9.8t+C
2
Then we need to match i.c.s again:
x(t) = 9.8e
t
9.8t+ 9.8
Youcan usethestep()functioninMATLABtoplotthe stepresponseofasystem(givenits
transfer function), rather than deriving it by hand (as above).
Figure 16.18.1 are plots of position, velocity and acceleration. We can rst solve to nd the
time at which the elevator has reached 10 meters. This happens at about t= 1.87 seconds
571
Position
0
-5
-10
-15
-20
-25
0 0.5 1 1.5 2 2.5 3
t
hit
=1.8656
Vel oci ty
0
-2
-4
-6
-8
-10
0 0.5 1 1.5 2 2.5 3
v
hit
=-8.2829
Acceleration
0
-2
-4
-6
-8
-10
Time(seconds)
0 0.5 1 1.5 2 2.5 3
a
hit
=-1.5171
Figure 16.18.1: Plots of Position, Velocity and Acceleration
The total force on a passenger must satisfy F = ma, where the acceleration is shown (over
time) in the plot above, and the mass is 100 kg.
(b) Wecancalculatetheresponseoftheelevatorhittingthespring-dampingsystematthebottom
of the shaft by calculating a FREE RESPONSE from the initial conditions and a FORCED
RESPONSE due to a step in force (of the mass of the elevator times gravity). For conve-
nience, well redene time such that t=0 when the elevator hits and x=0 where the elevator
hits.
Togetthevelocitytosettlewithin.1m/sasquicklyaspossible,wellaimforaresponsewhich
overshootsbyjustunder0.1m/s. Thiswouldbeclosetocriticallydamped. (Thevelocityat
impact is about 8.3 m/s, so a response which slows down and then reverse to move upward
at no more than 0.1 m/s would have an overshoot of about 1.2%.)
572
A
c
c
e
l
e
r
a
t
i
o
n

(
m
/
s

2
)

V
e
l
o
c
i
t
y

(
m
/
s
)

P
o
s
i
t
i
o
n

(
m
)

Its ne to have a response which overshoots by more than this amount, but note that this
will generally increase the time needed for the system to settle.
Figure 16.18.2 shows the response with the following parameters:
k=6990 [N/m]
c=2000+5300 [Ns/m]
0 2 4 6 8 10
-5
0
5
10
15
20
a

(
m
/
s

2
)

Accel erat i on
0 2 4 6 8 10
-10
-5
0
5
v

(
m
/
s
)

Vel oci t y
0 2 4 6 8 10
-3
-2
-1
0
x

(
m
)

Ti me(seconds)
Posi t i on
Figure16.18.2: PlotofPosition,VelocityandAccelerationwithk=6990[N/m]andc=2000+5300
[Ns/m]
573
Note that the acceleration requirement is barely met (under 20 m/s
2
) at t=0, and that the
position goes down to a minimum value of almost 3 meters!! (Thats a large of deection in
the system...)
The system settles to within 0.1 m/s in about 1.24 seconds.
Foraproblemlikethis,youmaynditconvenienttowriteaMATLABscriptorfunction,so
youcaniteratethroughpossiblesolutionsanddevelopatintuitionaboutthedynamicsofthe
system. (Here, since the system is allowed to overshoot, you can calculate the free and forced
responses for an underdamped second order system as in 2.27, given the initial conditions
(v=vhit) and step in force (m*g).
Below is a second alternative which puts more importance on minimizing the maximum
deection. The settling time is a bit longer (2.4 sec; about 2x that of the previous solution).
The maximum deection is cut by about 40% (from 3m to 1.8m). Figure 16.18.3 shows the
responses when k= 22,500 [N/m] and c=2000 [Ns/m].
574
20
15
10
5
0
-5
5
0
-5
-10
0
Acceleration
Vel oci ty
Position
0 1 2 3 4 5 6 7 8 9 10
v

(
m
/
s
)

(
m
/
s

2
)

0 1 2 3 4 5 6 7 8 9 10
-0.5
-1
-1.5
-2
0 1 2 3 4 5 6 7 8 9 10
Time(seconds)
x

(
m
)

Figure 16.18.3: Plot of Position, Velocity and Acceleration with k = 22,500 [N/m] and c = 2000
[Ns/m]
575
16.19 Servo Position Control
The solution for this problem is not available.
576
16.20 Servo Frequency Compensation
The solution for this problem is not available.
577
16.21 Servo Torque Disturbance
The solution for this problem is not available.
578
16.22 Plate On Springs Damped 1
Therigidplateofmassmissupportedbyfoursprings,eachofstinessk
c
anddampingparameter
b
c
. If y
k
is measured from the equilibrium position where
f(t)
k
c
k
c
b
c
bc
m
y
k
Figure 16.22.1: Plate on Springs with Damper
the weight of the plate is balanced by the initial compression of the springs, the constitutive
equations are
f
m
=m
dv
m
f
ki
=k
c
y
ki
f
bi
=b
c
v
bi
(i= 1, ,4)
dt
Geometric compatibility requires that, for (i= 1, ,4),
dy
ki
dy
k
y
ki
=y
k
v
m
=v
bi
= =
dt dt
Because there are four springs,
4 4
f
k
=

f
ki
= 4k
c
y
k
=ky
k
andf
b
=

f
bi
= 4b
c
dy
k
=b
dy
k
dt dt
1 1
Finally force balance requires that
f(t)f
k
f
b
=f
m
which leads to the dierential equation
d
2
y
k
dy
k
m +b +ky
k
=f(t)
dt
2
dt
When the behavioral parameters
o
and dened by
k b b

o
2
= 2
o
= or =
m m
4mk
579
_
_
are introduced, the equation takes the form
2
d
2
y
k
+ 2
o
dy
k
+
o
y
k
=
f(t)
dt
2
dt m
(a) The desired behavioral parameters are
o
= 5Hz(2 rad/cycle) = 31.4 rads/sec and (i)
= 0.1, (ii) = 0.5, (iii) = 1.0, and (iv) = 1.5. With the plate mass xed at 5 pounds,
or 5(0.4536) = 2.268 kilograms, the spring constant k must also be xed at
k=m
o
2
= 2.268(31.4)
2
=2236Newtons/meter
in order to keep the undamped natural frequency, given by
o
2
= k/m, equal to 5 Hz. To
provide the desired values of , the damping parameter b must be chosen to satisfy
b b
= = or b= 2m
o
= 2(2.268)(31.4) =142.4
2 km
2m
o
The b-values for the four cases are: (i) 14.24 kg/sec; (ii) 71,2 kg/sec; (iii) 142.4 kg/sec; (iv)
213.6 kg/sec. The decay time constants, = 1/
0
, for the rst three cases are: (i) 0.318
secs: (ii) 0.0637 secs; (iii) 0.0318 secs. In Case (iv) the roots of the characteristic equation
are =
o
(
2
1). The time constants are the negative reciprocals of the -values,
so the longest time constant is the negative reciprocal of the smallest
_
_
_
1

longest
= 31.4(1.5 + (1.5)
2
1) = 0.0834secs
The duration of the time histories in the four cases are 5 times the corresponding time
constants: (i) 1.59 secs; (ii) 0.319 secs; (iii) 0.159 secs; (iv) 0.417 secs.
The damped natural frequency
d
, for cases with 0 < < 1, is given by

d
=
o
1
2
For Case (i)
d
= 31.26 rad/sec for = 0.1, and for case (ii)
d
= 27.21 rad/sec. For Cases
(iii) and (iv), the eigenvalues are purely real, and
d
= 0.
The data for the four cases in Part (a) are assembled in the Table below.
Case k [N/m] b [N/m/sec] [sec] 5 [sec]
d
[r/s]
o
(i) 0.1 2239 14.25 0.3183 1.592 31.26 3.124
(ii) 0.5 2239 71.25 0.06365 0.3183 27.21 15.62
(iii) 1.0 2239 142.5 0.03183 0.1592 0.0 31.42
(iv) 1.5 2239 213.8 0.08332 0.4166 0.0 47.13
In addition to these data, every case has m = 2.268 kg, f
a
= (5)(4.448) = 22.24 Newtons,
and the initial conditions, y
k
(0) = 0 and v
k
(0)=0.
ThetimehistoriesareobtainedbyinputtingtheabovedataintotheMATLABscripts. When
0< < 1,Cases(i)and(ii),thescripttobeusedisMassSprgDmpr1.m. When =1,Case
(iii), the script to be used is MassSprgDmpr2.m, and when 1 < , Case (iv), the script to
be used is MassSprgDmpr3.m.
580
Case(i) Case(ii)
0.02 0.012
0.01
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]
0.015
0.01
0.005
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]
0.008
0.006
0.004
0.002
0 0
0 0.5 1 1.5 2 0 0.1 0.2 0.3 0.4
Time[seconds] Time[seconds]
Case(iii) Case(iv)
0.01 0.01
0 0.1 0.2 0.3 0.4 0.5
0.008
0.006
0.004
0.002
0.008
0.006
0.004
0.002
0 0
0 0.05 0.1 0.15 0.2
Time[seconds] Time[seconds]
Figure 16.22.2: Part (a)
581
0
0.002
0.004
0.006
0.008
0.01
0.012
0.014
0.016
0.018
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

Case(i)
Case(ii)
Case(iii)
Case(iv)
TimeHistoryofDisplacementResponse
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Time[seconds]
Figure 16.22.3: Part (a)
The results are displayed in Fig. 16.22.2. In Fig. 16.22.2 the duration of each response was
5timesthedecaytimeconstantofeachcase. AnalternativecomparisonofthefourCasesin
Part(a)isdisplayedinFig. 16.22.3whereallresponsesareplottedonthesametimeaxisfor
a duration of 0.5 seconds.
(b) The desired behavioral parameters are
d
= 31.42 rad/sec (5 Hz) with the following values
ofdampingratio: Case(i)has = 0.1;Case(ii)has = 0.3;Case(iii)has = 0.5;Case(iv)
has = 0.7. From the denitions of
d
and follow the formulas
m
d
2
2m
d
k= and b= _
1
2
1
2
The data for Part (b) are assembled in the following Table.
Case k [N/m] b [N/m/sec] [sec] 5 [sec]
d
[r/s]
o
(i) 0.1 2262 14.32 0.3167 1.584 31.42 0.3167
(ii) 0.3 2460 44.02 0.1012 0.5060 31.42 0.1012
(iii) 0.5 2985 82.28 0.05513 0.2757 31.42 0.0551
(iv) 0.7 4390 139.7 0.03247 0.1624 31.24 0.0325
582
In Part (b) all four Cases are handled by the MATLAB script MassSprgDmpr1.m. The
plots for durations of 5 times the decay time constant are displayed in Fig. 16.22.4.
Case(i) Case(ii)
0.02 0.014
0.012
0.01
0.008
0.002
0 0
0 0.5 1 1.5 2 0 0.2 0.4 0.6 0.8
Time[seconds] Time[seconds]
Case(iii)
x10
-3
Case(iv)
0.01 6
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]
0.015
0.01
0.006
0.004
0.005
5
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]
0.008
4
0 0
0 0.1 0.2 0.3 0.4 0 0.05 0.1 0.15 0.2
Time[seconds] Time[seconds]
Figure 16.22.4: Part (b)
0.006
3
0.004
2
0.002
1
583
0
0.002
0.004
0.006
0.008
0.01
0.012
0.014
0.016
0.018
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

Case(i)
Case(ii)
Case(iii)
Case(iv)
TimeHistoryofDisplacementResponse
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Time[seconds]
Figure 16.22.5: Part (b)
An alternative plot in which all four cases are plotted against the same time scale for a
duration of 0.5 seconds is displayed in Fig. 16.22.5.
(c) The desired behavioral parameters are = 0.10 seconds with (i) = 0.3, (ii) = 0.5, (iii)
= 0.7, and (iv) = 0.9. From the denitions of and
1 b
= 2
o
=

o
m
follow the formulas
m 2m
k= and b=

The data for Part (c) are assembled in the following Table
Case k [N/m] b [N/m/sec] [sec] 5 [sec]
d
[r/s]
o
(i) 0.3 2520 45.36 0.1 0.5 31.80 10
(ii) 0.5 907.2 45.36 0.1 0.5 17.32 10
(iii) 0.7 462.9 45.36 0.1 0.5 10.20 10
(iv) 0.9 280.0 45.36 0.1 0,5 4.843 10
584
In Part (c) all four cases are handled by the MATLAB script MassSprgDmpr1.m. The
responsesfordurationsof5timesthedecaytimeconstantareplottedagainstthesametime
axis in Fig. 16.22.6.
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
TimeHistoryofDisplacementResponse
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

Case(i)
Case(ii)
Case(iii)
Case(iv)
Time[seconds]
Figure 16.22.6: Part (c)
585
_ _
16.23 Plate On Springs Damped 2
TheeigenvaluesforthecaseshandledbyMassSprgDmpr1.m(0< < 1)arecomplexconjugates
of the form
=
o
+i
d
and =
o
i
d
The eigenvalues for the case handled by MassSprgDmpr2.m ( = 1) are a pair of repeated real
roots: =
o
and=
o
. TheeigenvaluesforthecasehandledbyMassSprgDmpr3.m(1< )
are
1
=
o
(+
2
1)and
2
=
o
(
2
1). Thevaluesof
o
and
d
forthevarious
cases are listed in the Tables in Problem 16.22. The eigenvalues in the upper half of the complex
planeforPart(a)areplottedinFig. 16.23.1. ThecomplexrootsforCases(i)and(ii)havemirror
image roots in the lower half-plane.
Figure 16.23.1: Eigenvalues for Part (a)
TheeigenvaluesintheupperhalfofthecomplexplaneforPart(b)areplottedinFig. 16.23.2,and
the eigenvalues in the upper half of the complex plane for Part (c) are plotted in Fig. 16.23.3.
-90 -80 -70 -60 -50 -40 -30 -20 -10 0 10
-5
0
5
10
15
20
25
30
35
40
Real
Imag
Case(i)
Case(ii)
Case(iii) Case(iv) Case(iv)
586
-90 -80 -70 -60 -50 -40 -30 -20 -10 0 10
-5
0
5
10
15
20
25
30
35
40
Real
Imag
(i)
(ii) (iii) (iv)
Figure 16.23.2: Eigenvalues for Part (b)
Figure 16.23.3: Eigenvalues for Part (c)
ItisinstructivetocomparetheresponsesfromProblem16.22withthelocationsoftheeigenvalues
for Parts (a)-(c).
-90 -80 -70 -60 -50 -40 -30 -20 -10 0 10
-5
0
5
10
15
20
25
30
35
40
Real
Imag
(i)
(ii)
(iii)
(iv)
587
16.24 Plate On Springs Damped 3
HerethesystemofProblem16.22isexcited,notbyasuddenlyappliedforce,butbytheimpactof
a rubber ball at t=0, which produces an initial velocity v(0)=v
o
.
(a) The initial velocity is obtained by applying conservation of linear momentum to the impact.
Downward momentum before impact = (2pounds)(10 ft/sec)
Downward momentum after impact = (5pounds)(v
o
)+(2pounds)(-3 ft/sec)
Equating these, yields v
o
= 5.2 ft/sec = 1.5850 m/s (for convenience, it has been assumed
that the positive direction for y
k
and v=dy
k
/dt is downwards).
(b) A damped natural frequency of 5 Hz with = 0.3 is obtained with the parameters found for
Case (ii) of Part (b) in Problem 16.22: k=2460 N/m, and b= 44.02 N/m/s.
(c) When these parameters, plus the conditions f
a
= 0, y
k
(0) = 0, and v(0) = 1.5850 m/s, are
inputtotheMATLABscriptMassSprgDmpr1.mtheresponsehistorylabeledResponseto
impact of ball in Fig. 16.24.1 is obtained.
TimeHistoryofDisplacementResponse
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Time[seconds]
Figure 16.24.1: Comparison of responses
588
-0.015
-0.01
-0.005
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

Responsetoimpactofball
Responsetosuddenlyappliedforce
(d) InFig. 16.24.1theresponsetotheimpactoftheballiscomparedtotheresponseofthesame
systemtoastep-functionforceof5pounds(Case(ii)ofPart(b)inProblem16.22). Inboth
casesthemotionstartsfromtheoriginalequilibriumpositionofthesystem. Inbothcasesthe
massoscillatesforafewcyclesandcomestorest. Thedampednaturalfrequencyanddamp-
ing ratio is the same in both oscillations. The nal rest position is the same as the original
equilibriumpositionintheball-impactcase,whilethenalrestpositioninthestep-function
force case is the new equilibrium position under a constant load of 5 pounds. The initial
velocityinthestep-functionforcecaseiszero(displacementresponsestartswithzeroslope),
while the initial velocity in the ball-impact case is large due to the impact (displacement
response starts with a steep slope). For the particular magnitudes of the excitations given
theoscillationintheball-impactcaseismoreintensethantheoscillationduetothesuddenly
applied force.
589
16.25 Plate On Springs Damped 4
Due to the application of a 1.0 Newton step-force, the system oscillates at its damped natural
frequency
d
about the new equilibrium position y
k
=. From the given Displacement Response
plot,theequilibriumosetcanbeestimatedtobe=4.4x10
4
meters. Thedampednatural
frequency can be estimated by counting the cycles, 10, in 2.0 seconds, which gives a damped
period T
d
= 0.2 seconds, and
d
= 2/T
d
= 31.4 rad/sec. The logarithmic decrement LDR can
be estimated by measuring the ratios of successive peak amplitudes. This is generally a dicult
measurement to make accurately. One way to increase the accuracy is to make a large number of
measurements based on dierent pairs of successive peaks, and average the results. Draw a line
across the plot at y = and measure the amplitudes of several peaks. Since only the ratio of
successive peaks will be used, the measurements can be in terms of any convenient length unit.
The measurements below are in millimeters.
Station Amplitude Ratio
0 6.22
1 4.86 0.781
2 4.00 0.823
3 3.08 0.770
4 2.56 0.831
5 1.94 0.758
6 1.68 0.866
The average of these rst six ratios is 0.805, so the estimate for the LDR is ln0.805=0.217.
(a) The estimated stiness is
f
a
1.0
k= = =2270 Newtons/meter
4.410
4
(b) Toestimatethemassm, when the stiness k isknown,itisnecessarytoknowtheundamped
natural frequency
o
. At this point we know only the damped natural frequency
d
. We
can postpone estimation of the mass m until the damping ratio is estimated in (c) below,
or we can assume, since more than ten cycles of oscillation are visible, that the damping is
suciently light to permit us to make the approximation that
o

d
, in which case
k 2270
m = = 2.30 kg

d
2
(31.4)
2
(c) To estimate the damping parameter b, we need to know the damping ratio . The damping
ratio follows from the log decrement ratio, LDR, according to the formula:
LDR
2

2
=

2
+LDR
2
Insertion of the estimation LDR = 0.217 yields 0.069. (Note: With this value of the
undamped natural frequency is estimated as follows

d
31.4

o
= _ _ = 31.5 rad/sec
1
2
1(0.069)
2
With this estimate for
o
the mass m, previously estimated in (b) above at 2.30 kg, would
now be estimated to be 2.29 kg. The damping parameter b is then estimated as
b= 2
o
m2(0.069)(31.5)(2.29)=9.95 kg/sec, or 9.95 N/m/sec
590
_ _
16.26 Plate On Springs Damped 5
With two plate-on-springs units face-to-face, the eective stiness of the combined system is k =
1000 N/m, and the eective damping parameter is b = 10 N/m/sec. For free vertical motion of
the mass m= 2.0 kg, the displacement y
k
from the equilibrium position of the model satises the
equation
m
d
2
y
k
+b
dy
k
+ky
k
= 0
dt
2
dt
(a) The undamped natural frequency
o
for the model is
k 1000
o

o
= = = 22.4 rad/sec, or f
o
= = 3.56 Hz
m 2 2
(b) The damping ratio for the model is
b 10
= = = 0.1116
2
o
m 2(22.4)(2)
(c) The decay time constant for the model is
1 1
= = = 0.400sec

o
(0.1116)(22.4)
591
_ _
16.27 Plate On Springs Damped 6
Toobtainastate-determinedsystem,introducetheverticalvelocityvoftheplate. Thegeometrical
compatibility requirement is
dy
=v
dt
The constitutive equations are
f
springs
=ky and f
damping
=bv
and the force balance requirement is
f
m
=f
springs
f
damping
The dynamic state equations are two rst-order dierential equations for y and v
dy
= v
dt
dv 1 k b
= f(t) y v
dt m m m
(a) The deection under the weight mg is =mg/k so the mass m is given by
k (3000)(0.007)
m= = = 2.14 kg
g 9.81
The undamped natural frequency is
o
= k/m = 3000/2.14 = 37.4 rad/sec, and the
damping parameter is
b= 2
o
m= 2(0.5)(37.4)(2.14)=80.0 N/m/s
(b) The dynamic state equations can be rewritten in matrix form
_ _ _ _
_ _ _ _
0 1 0
d
_
y
_

_
y
_ _ _
=_ _ +
dt_ _
k b
_ _
f
a
v

v
_ _
m m m
or
d
x=Ax+Bf
a
dt
where
_ _ _ _
_ _
0 1 0
_
y
_ _ _

x= A=_
k b
_ B=
1
_ _
_ _
v

m m m
where f
a
is the magnitude of the constant applied force. These equations can be integrated,
from prescribed initial values of y and v at t = 0 by the MATLAB command ode45. A
possible script for doing this is displayed at the end of the problem.
(c) When the scripts above are run with the following inputs
m = 2.14 kg
592
k = 3000 N/m
b = 80.0 N/m/s
f
a
= mg = (2.14)(9.81) = 21.0 Newtons
the following three plots are obtained for a time interval of T = 0.35 seconds. There is very
little activity in the interval between t = 0.35 and t = 2.0 seconds.
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
Time [Seconds]
Figure 16.27.1: Plot of Position vs. Time
0
1
2
3
4
5
6
7
8
9
x10
-3
TimeHistoryofDisplacement
D
i
s
p
l
a
c
e
m
e
n
t

[
m
e
t
e
r
s
]

593
0.16
0.14
0.12
0.1
0.08
0.06
0.04
0.02
0
-0.02
-0.04
TimeHistoryofVelocity
V
e
l
c
i
t
y

[
m
e
t
e
r
s
/
s
e
c
o
n
d
]

V
e
l
c
i
t
y

[
m
e
t
e
r
s
/
s
e
c
o
n
d
]

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35


Time [Seconds]
Figure 16.27.2: Plot of Velocity vs. Time
Velocity vs.Displacement
0.16
0.14
0.12
0.1
0.08
0.06
0.04
0.02
0
-0.02
-0.04
0 1 2 3 4 5 6 7 8 9
Displacement [meters]
x10
-3
Figure 16.27.3: Plot of Velocity vs. Position
eqpossol.m
Download
% eqpossol.m Provides equation of motion for plate on springs
% to be integrated by script POS_sol.m
594
function Xdot = eqpossol(t,X)
global m k b fa
Xdot = [ 0 1 ; -k/m -b/m ]*X + [ 0 ; fa/m ];
possol.m
Download
% possol.m A MATLAB script for Plate On Springs. Produces plots of
% (i) position vs. time
% (ii) velocity vs. time
% (iii) velocity vs. position
% for the response of a steel plate on springs, with mass m, stiffness k,
% and damping parameter b, when the plate starts from initial conditions
% y = y0 and v = v0 under the action of a suddenly applied force fa at t = 0.
clear variables
global m k b fa
% Input parameters
m = input(Enter the mass "m" in kilograms );
k = input(Enter the stiffness "k" in Newtons/meter );
b = input(Enter the damping constant "b" in kilograms/sec );
fa= input(Enter the magnitude "fa" of the suddenly applied force in Newtons );
% Input initial conditions.
y0= input(Enter the initial displacement from equilibrium, in meters );
v0= input(Enter the initial velocity, in meters/second );
tspan = input(Enter the duration "T" of the desired time history, in seconds );
X0 = [ y0 ; v0 ];
% Integrate equations of motion
[t,X] = ode45(eqpossol, tspan, X0);
% Plot results
plot(t,X(:,1)), title(Time History of Displacement),
xlabel(Time [Seconds]), ylabel(Displacement [meters]), pause
plot(t,X(:,2)), title(Time History of Velocity),
xlabel(Time [Seconds]), ylabel(Velcity [meters/second]), pause
plot(X(:,1), X(:,2)), title(Velocity vs. Displacement),
xlabel(Displacement [meters]), ylabel(Velcity [meters/second])
595
16.28 Shipping Crate 1
Letxbethedisplacementoftheenginewithrespecttothestationarycrate. Theeectivestiness
of the two end elements is 2k, and the eective damping parameter is 2b.
(a) The equation of motion for the engine is
d
2
x dx
m + 2b + 2kx= 0
dt
2
dt
(b) The engine weight is W = 500 pounds. Its mass is m = W/g. The values of k and b can
be deduced from the given values of damped natural frequency
d
= 2 and damping ratio
= 0.707 from the equations

2
2k

2
=
1
d

2
=
W/g
and 2b= 2
o
W/g
o
The undamped natural frequency is
2

o
= _ = 8.88 rad/sec
1(0,707)
2
and
W
o
2
(500)(8.88)
2
k= = = 51.2 pounds/inch
2g 2(386)
The damping parameter is
W 500
b=
o
= (0.707)(8.88) = 8.13 pounds/in/sec
g 386
596
16.29 Shipping Crate 2
Fig. 16.29.1 shows an idealized model of the engine of mass m = W/g with shock-absorbing
packaging, represented by the springs with stiness k and damping parameter b, in a crate which
is xed to a truck. The velocity of the truck, with respect to an inertial reference frame, is v
truck
,
and the velocity of the engine, with respect to an inertial reference frame, is v
engine
. The relative
displacement of the engine with respect to the crate is x
rel
. The origin for x
rel
is taken to be the
equilibrium position of the engine in the motionless crate. In problems like this, where relative
motion is involved, it is important to remember that Newtons law only applies to motions with
respect to an inertial reference frame
v
truck
v
engine
x
rel
k
b b
k
m
Figure 16.29.1: Model of Engine in Crate on Moving Truck
(a) The geometric compatibility conditions are
dx
rel
=v
rel
and v
engine
=v
truck
+v
rel
dt
The constitutive equations are
f
m
=m
dv
engine
, f
spring
=kx
rel
, and f
damping
=bv
rel
dt
and the force balance condition is
f
m
=2f
spring
2f
damping
The six preceding equations constitute a mathematical model for the longitudinal motion of
the engine with respect to the crate.
(b) A single dierential equation in terms of x
rel
can be derived by inserting the spring force
f
spring
= kx
rel
and the damping force f
damping
= bdx
rel
/dt in the force-balance equation,
597
along with f
m
expressed as f
m
=m(dv
truck
/dt+d
2
x
rel
/dt
2
). The result is
m
d
2
x
rel
+ 2b
dx
rel
+ 2kx
rel
=m
dv
truck
dt
2
dt dt
Theformofthisequationissimilartothatfortheequationforthedisplacementofthesteel
plateonsprings. Heretheunknownresponseisarelative displacementandthedrivingforce
is the negative of the force that would be required to give the engine the same acceleration
asthetruck. Asingledierentialequationfortherelativevelocityof the engine with respect
to the crate is obtained by dierentiating every term in the preceding equation.
m
d
2
v
rel
+ 2b
dv
rel
+ 2kv
rel
=m
d
2
v
truck
dt
2
dt dt
2
(c) Astate-determinedsystemforthestatevariablesx
rel
andv
eng
can be obtained by expressing
the forces in the force-balance equation of (a) in terms of x
rel
and v
rel
=v
eng
v
truck
. The
two rst-order dierential equations for the state-determined system are
dx
rel
= v
eng
v
truck
dt
dv
eng
k b b
= x
rel
v
eng
+ v
truck
dt m m m
and the desired output is obtained from
v
rel
=v
eng
v
truck
In matrix form
dX
=AX+Bu with y=CX+Du
dt
with
_ _ _ _ _ _
_
x
rel
_ 0 1 _ 1 _
_ _
X= A= B= C= 0 1 D= [1]
_ _ k b _ b _
v
eng

m m m
and y=v
rel
and u=v
truck
.
598
17 Quiz Problems
17.1 Fun with Block Diagrams
(a)
K
T(s) =
s
2
+ 20s+K

n
= K

2
n
= 2 K = 20
K =100
(b)
V
out
(s+3)(6s+ 1)
=
V
1
(s+3)(6s+ 1) + (8s+ 7)
V
out
(6s+ 1)
=
V
2
(s+3)(6s+ 1) + (8s+ 7)
(c) Solve using superposition
V
out
(6s
2
+ 27s+10)=V
1
(6s
2
+ 19s+ 3) + V
2
(6s+ 1)
6V

out
+ 27V

out
+ 10V
out
= 6V

1
+ 19V

1
+ 3V
1
+ 6(

V)
2
+V
2
599
17.2 Complex Translation
The solution for this problem is not available.
600

_
17.3 Cylinder Step Response
The transfer function for this system is
x(s)
=
1
f(s) m
eq
s
2
+cs+k
k

n
=
m
eq
c
2
n
=
m
eq
(a) From graph, we measure the following
2
T 1.0s
d
= = 6.28r/s
T

d
=
n
1
2
0.750.5
M
p
100 = 50
0.5
A
= = 0.215

2
+A
2
100
A=ln = 0.693
M
p

n
= 6.43r/s

2
m
eq
=
n
= 4.85kg
k
c= 2
n
m
eq
= 13.8Ns/m14Ns/m
Alternately, you could determine using the log decrement method.
(b)
I
m
eq
=m+
r
2
I = 0.5kgm
2
601
_
_
17.4 JKC Frequency Response 1
The transfer function for this system is
(s)
=
k
(s) Js
2
+cs+k
k
Thus
n
=
J
c
2
n
=
J
(a) There are a couple of ways to solve this part of the problem. First, you can read
r
= 9 r/s
and M
p
= 5 dB from the bode plot and use the following relationships
1
M
p
= _
2 1
2

r
=
n
12
2
to nd 0.3 and
n
10r/s. Or you can read
n
= 10 r/s directly from the phase plot
(=90

)
(b) k=1500Nm/r, c= 90 Nms/r
(c)
= 1.1r/s (t)sin(1.1t+ 0)
= 10 r/s (t)1.58sin(10t/2)
= 20 r/s (t)0.3 sin (20t2.75)
602
17.5 Mass Spring Damper Dynamics
The solution for this problem is not available.
603
17.6 Op-Amp Analysis
(a)
v
o
R
2
C
1
s
=
v
i
(R
2
C
2
s+ 1)(R
1
C
1
s+ 1)
(b)
M() = _
R
1
C
1

(1R
1
R
2
C
1
C
2

2
)
2
+ ((R
1
C
1
+R
2
C
2
))
2
() = 90

tan
1
_
(R
1
C
1
+R
2
C
2
)
_
(1R
1
R
2
C
1
C
2

2
)
(c) Figure 17.6.1 shows the Bode plot for this system.
Bode Diagram
1 0 1 2 3 4 5
10 10 10 10 10 10 10
Frequency (rad/sec)
Figure 17.6.1: Bode Diagram
(d) Figure 17.6.2 shows the pole-zero plot for this system.
M
a
g
n
i
t
u
d
e

(
d
B
)

45
40
35
30
25
20
P
h
a
s
e

(
d
e
g
)

90
45
0
45
90
604
PoleZero Map
I
m
a
g

A
x
i
s
1
0.8
0.6
0.4
0.2
0
0.2
0.4
0.6
0.8
1
10000 9000 8000 7000 6000 5000 4000 3000 2000 1000
Real Axis
Figure 17.6.2: Pole-Zero Plot
605
0
17.7 RLC Circuit Analysis
(a)
V
o
1
=
V
i
LCs
2
+RCs+ 1
(b)
1

n
= 2 5000=31,400r/s=
LC
1
L= = 0.001H = 1 mH

2
C
n
R
= 2
n
= 2 0.707 31,400
L
R= 44.4
(c) There are no zeros, poles at roots of
1000 1
s
2
+ s+ = 0
0.001 1e6 1e3
s
1
1e6
s
2
1e3 dominant pole
x
ss
= 1
Step Response
0 1 2 3 4 5 6
Time (sec)
x 10
3
A
m
p
l
i
t
u
d
e

0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
System: sys
Time (sec): 0.001
Amplitude: 0.633
Figure 17.7.1: Step Response for part (c)
606
(d) The bode plot is shown in Figure 17.7.2
P
h
a
s
e

(
d
e
g
)

M
a
g
n
i
t
u
d
e

(
d
b
)

0
20
40
60
80
100
0
50
100
150
10
0
10
1
10
2
10
3
10
4
10
5
10
6
10
7
10
0
10
1
10
2
10
3
10
4
10
5
10
6
10
7
Frequency (r/s)
Figure 17.7.2: Bode plot for part (d)
(e) s=1e3 = s
2
(f) (i)
v
o
R
2
=
v
i
R
2
LCs
2
+ (R
1
R
2
C+L)s+R
1
+R
2
(ii)
v
o
R
2
Cs+ 1
=
v
i
LCs
2
+ (R
1
+R
2
)Cs+ 1
607
17.8 Sailing for Engineers
The solution for this problem is not available.
608
17.9 Second Order Step Response
The solution for this problem is not available.
609
17.10 Spring Damper Dynamics
The solution for this problem is not available.
610
17.11 Derive Blocks for Op-amp Circuit
The solution for this problem is not available.
611
17.12 Automobile Fender Spring/Damper System
The solution for this problem is not available.
612
17.13 Cu Flywheel with Eddy Current Damper
The solution for this problem is not available.
613
17.14 Current Driven RC Circuit
The solution for this problem is not available.
614
17.15 Driven Mass Spring System
The solution for this problem is not available.
615
17.16 Flywheel Driven by Hanging Mass
The solution for this problem is not available.
616
17.17 Homogeneous Second Order DE
The solution for this problem is not available.
617
17.18 Match Pole/Zero Plots with Step Response
The solution for this problem is not available.
618
17.19 Opamp Block Transfer Functions
The solution for this problem is not available.
619
17.20 Piston with 2nd Order Translation
The solution for this problem is not available.
620
17.21 Pole Zero Bode Matching
The solution for this problem is not available.
621
17.22 Power Semiconductor Thermal Problem
The solution for this problem is not available.
622
17.23 Submersible Capsule Hoist System
The solution for this problem is not available.
623
17.24 Transfer Function from Pole Zero Plot
The solution for this problem is not available.
624
17.25 Vaccine Cooler
The solution for this problem is not available.
625
17.26 Voltage Driven RRC Circuit
The solution for this problem is not available.
626
17.27 Linear Mechanical System with Position Input
The solution for this problem is not available.
627
17.28 Tank with Pump Inlet Lower than Outlet
The solution for this problem is not available.
628
17.29 Thermal Power Chip Analysis
The solution for this problem is not available.
629
18 Math Techniques
18.1 Complex Expression Reduction
(a)
(2+3i)(45i) = 23 + 2i (1)
2 + 3i
=0.1707+0.5366i (2)
45i
(6 + 7i)(3+7i) = 67.000021.0000i (3)
6 + 7i
= 0.5345+1.0862i (4)
3 + 7i
(b) First write the complex numbers in

2 + 3i= 13e
iarctan1.5
13e
0.9828i
(5)

45i= 41e
iarctan1.25
41e
0.8961i
(6)

6 + 7i= 85e
i(arctan(7/6)+)
85e
2.2794i
(7)

3 + 7i= 58e
iarctan(7/3)
58e
1.1659i
(8)
Tocomputeproductsofcomplexnumbersinpolarform,wecanusethemnemonicMMAA,
ModuliMultiplyArgumentsAdd,wherethemodulusofz=re
i
isr andtheargumentis.
We nd
(2+3i)(45i) = 23.0868e
0.0867i
(9)
2 + 3i
= 0.5631e
i1.8788
(10)
45i
(6 + 7i)(3+7i) = 70.2140e
2.8379i
(11)
6 + 7i
= 1.2106e
1.1135i
(12)
3 + 7i
630
_
_
18.2 Complex Expressions
(1+2i)(34i) = 11 + 2i (22)
1 + 2i 1 + 2i3 + 4i 1 2
= = + i (23)
34i 34i3 + 4i 5 5
(6 + 4i)(5+7i) = 5822i (24)
6 + 4i 1 31
= + i (25)
5 + 7i 37 37
Now we wish to write the complex numbers a+bi in polar form re
i
using
r= a
2
+b
2
(26)
b
= arctan (27)
a
Whencomputing the angle, onemust makesure that proper angle is cal-
culatedbasedonwhichquadrantofthecomplexplanecontainsthecomplex
number (In Matlab, one may prefer to use the command atan2 instead of
atan.).

1 + 2i= 5e
iarctan2
5e
1.1071i
(28)
34i= 5e
iarctan(4/3)
5e
0.9273i
(29)

6 + 4i= 2 13e
iarctan(2/3)
2 13e
2.5536
(30)

57i= 74e
iarctan(7/5)
74e
0.9505i
(31)
To compute products of complex numbers in polar form, we can use the
mnemonic MMAA, Moduli Multiply Arguments Add, where the modulus
of z=re
i
is r and the argument is . We nd

(1+2i)(34i) = 5 5e
0.1798i
(32)

1 + 2i
=
5
e
2.0344i
(33)
34i 5

(6 + 4i)(5+7i) = 2 962e
1.4931i
(34)
6 + 4i 13
2.7791i
= 2 e (35)
5 + 7i 74
Figure 18.2.1:
631
18.3 Matrix Operation Practice
Part (a)
5

1
2
3

5
10
15

Part (b)

1 2 3

9
8
0

= 25
Part (c)

1
2
3

9 8 0

9
18
27
8
16
24
0
0
0

Part (d), the roots of s


2
+ 2s+ 3 are
s=1

2i
Figure 18.3.1:
632
19 Recitation Problems
19.1 Recitation 1 Problem
633
_ _
_
19.2 Recitation 2 Problem
(a) From class notes the characteristic equation is:
mx + bx + kx= 0
The roots/eigenvalues of this equation are at (using a shorter notation than class):

b b
2
4mk
s
1,2
=
2m 2m
A simple check shows that the roots are distinct and imaginary (b
2
4mk = 410
6
), so
we can write the solution in the form:
x(t) = c
1
e
s
1
t
+c
2
e
s
2
t
= 2e
t
(cos
d
tsin
d
t)
Evaluating the homogeneous response at t= 0 we have:
1
x(0)=x
0
= 2 or = x
0
2
Dierentiating the homogeneous response and evaluating at t=0 (see Chapter 1 for all the
details):
x
0
+x
0
x (0)=x
0
= 2(
d
) or =
2
d
Case 1: x
0
= 0 and x
0
=1 Plugging in values we get =0 and =
2
1
d
which leads to:
1
x(t) = e
t
sin
d
t

d
Case 1: x
0
= 1 and x
0
=0 Plugging in values we get =.5 and =

which leads
2 1
2
to:

x(t) = e
t
cos
d
t+ _ sin
d
t
2 1
2
(b)
b 1
= = =.707
2 km 2
k

n
= = 2 2 = 2.8242 [rad/sec]
m
_

d
=
n
1
2
= 2 2( 1.5)=2 [rad/sec]
1

=
n
= 2 2=2 [rad/sec]
2
1
= =.5 [sec]

634
x(0)=0, v(0)=1
-0.05
0
0.05
0.1
0.15
0.2
d
i
s
p
l
a
c
e
m
e
n
t

(
m
e
t
e
r
s
)

0 0.5 1 1.5 2 2.5 3
time (seconds)
Figure 19.2.1: Nominal case, x
0
= 0, x
0
= 1, = 0.707
Scaled View of x(0)=0, v(0)=1
d
i
s
p
l
a
c
e
m
e
n
t

(
m
e
t
e
r
s
)

1.2
1
0.8
0.6
0.4
0.2
0
-0.2
Initial slope = 1
0 0.5 1 1.5
time (seconds)
Figure 19.2.2: Close up Nominal case, x
0
= 0, x
0
= 1, = 0.707 show that the initial velocity is
equal to the slope
635
x(0)=1, v(0)=0
d
i
s
p
l
a
c
e
m
e
n
t

(
m
e
t
e
r
s
)

1.2
1
0.8
0.6
0.4
0.2
0
-0.2
Note zero slope at t=0
0 0.5 1 1.5 2 2.5 3
time (seconds)
Figure 19.2.3: Close up Nominal case, x
0
= 1, x
0
= 0, = 0.707 showing non-zero initial position,
zero initial velocity
(c) The damping ratio = .707 is between zero and one so by denition the system is under-
damped. Professor Dubowsky jokes that the value of .707 is just right because it is the
best compromise between speed of response and overshoot.
(d) The exponential envelope term is e
t
. The attenuation is always greater than zero, so
this decaying exponential is stable and will not increase without bounds, which matches our
physical intuition that a shock absorbers purpose is to decrease motion.
(e) See the following plots, which all have the same mass and spring constant, but dierent
dampers b. Initial conditions are x
0
= 0, x
0
= 1. All the following curves are drawn to the
same scale.
636
Nominal Case: x(0)=0, v(0)=1
0 2 4 6 8
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
d
i
s
p
l
a
c
e
m
e
n
t

(
m
e
t
e
r
s
)

time (seconds)
Figure19.2.4: Nominalcase,x
0
= 0, x
0
= 1, = 0.707,b= 2 10
3
[Ns/m]. Replottedtoshowthe
same scale as the other plots with varying damper b.
Underdamped, b=1/4 * nominal b, x(0)=0, v(0)=1
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
d
i
s
p
l
a
c
e
m
e
n
t

(
m
e
t
e
r
s
)

0 2 4 6 8
time (seconds)
Figure 19.2.5: Smaller damping than nominal case, x
0
= 0, x
0
= 1, = 0.18, b
new
= .25b = 500
[Ns/m].
637
d
i
s
p
l
a
c
e
m
e
n
t

(
m
e
t
e
r
s
)

Undamped, b=0, x(0)=0, v(0)=1


0.4
0.3
0.2
0.1
0
-0.1
-0.2
-0.3
-0.4
0 2 4 6 8
time (seconds)
Figure 19.2.6: Undamped case, x
0
= 0, x
0
= 1, = 0, b
new
= 0 [Ns/m].
Critically damped, b = 2 * sqrt(k * m), x(0)=0, v(0)=1
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
d
i
s
p
l
a
c
e
m
e
n
t

(
m
e
t
e
r
s
)

0 2 4 6 8
time (seconds)

Figure19.2.7: Criticallydampedcase,x
0
= 0, x
0
= 1, = 1, b
critical
= 2 km= 2.8310
3
[Ns/m].
638
Overdamped, b = 2 * b critical, x(0)=0, v(0)=1
d
i
s
p
l
a
c
e
m
e
n
t

(
m
e
t
e
r
s
)

0.4
0.3
0.2
0.1
0
-0.1
-0.2
-0.3
-0.4
0 2 4 6 8
time (seconds)
Figure 19.2.8: Overdamped case, x
0
= 0, x
0
= 1, = 1, b
new
= 2b
critical
= 5.6610
3
[Ns/m].
(f) A pothole acts as an impulse, so the forced response is the impulse response.
(g) Sports car: Want a fast response and tight handling.
Grandmas Caddy: Want a soft, cushy ride, could accept slower response.
Pickup Truck: The load (mass) may increase, so pick a damping ratio that still gives a good
response if the mass is larger. Could do some calculations...
639
19.3 Recitation 3 Problem
(a) The second-order dierential equation for this system is:
J

+b

+k=T
The characteristic equation is the unforced part:
J

+b

+k= 0
This can also be written in the standard form (from class):
1 2 T(t)

+ +=f(t) =

2
n

n
k
(b)
b
=
2 kJ
_
k

n
=
J
_

d
=
n
1
2
(c) Shown below is the MATLAB plot, using T
0
=10 [Nm]
200
150
100
50
0
time (seconds)
Figure 19.3.1: Angular displacement vs time for the swinging door.
a
n
g
u
l
a
r

d
i
s
p
l
a
c
e
m
e
n
t

(
r
a
d
i
a
n
s
)

0 5 10 15 20 25
640
_ _ _ _
(d) For this class, M
p
is overshoot (not percent overshoot as elsewhere). Could calculate by
nding derivative and looking for rst zero, but too complicated for everyday use.

M
p
=e
/
d
=e
1
2
=.53
Compare to M
p
approximation (valid because 0< < 0.6)

M
p
1 = 0.67
0.6

t
p
= = 3.2 [s]

d
(e) Rise time is the time it takes for the response to go from 10% to 90% of the nal value.
1.8
t
r
= = 1.8 [s]

n
(f) For this gure, the 1% settling time is
4.6

t
s
= = 23 [s]

n
(g)
(t) =
p
(t) +
h
(t)

p
(t) =
T
0
u
s
(t)
k

h
(t) =
T
0
e
t
cos
d
t+

sin
d
t =
T
0
e
t

n
sin
d
t+
k
d
k
d
641
19.4 Recitation 4 Problem
(a) Just one. We say this system has one degree-of-freedom (DOF). There are multiple choices
for the independent variable such as h
1
, h
2
or the pressure at the bottom of tank 1 P
1
.
(b) Answers will vary.
(c) The volume of uid V is constant, so h
1
A+h
2
A=V = 2Ah
eq
where h
eq
is the equilibrium
position of the uid. We can rewrite this as h
2
= 2h
eq
h
1
. Note also that h

1
+h

2
= 0.
(d) Thepressureatthebottomofthetank1isP
1
P
a
=gh
1
. Againfortank2P
2
P
a
=gh
2
.
NotethatatmosphericpressureP
a
hasaslightlydierentnotationthanPascals,theSIunits
for pressure are [N/m
2
].
(e) Dene q positiveinward,q
1
owing into tank 1, q
2
owing into tank 2. Flow into and out of
tube is conserved q
1
+q
2
=0. Resistance in the tube is modeled as
1 g
q
1
= (P
1
P
2
) = (h
1
h
2
)
R R
(f) Change in volume in tank 1 V
1
= h
1
A equals ow q
1
= h

1
A. The same is true for tank 2
q
2
=h

2
A. Is the sign correct? Yes, if q
1
goes up, so does h
1
. Combining:
RA
h

1
+h
1
=h
eq
2g
Could also dene h
1
=h
eq
+h

which leads to h

=h
1
h
eq
and h

=h

1
Or
RA
h

+h

= 0
2g
(g) First-order system so no oscillations.
(h) Time constant = RA/2g doesnt depend on height, so initial amount of uid doesnt
change how quickly the system responds.
To double ,doubleresistanceintubeordoubleareaoftanks,orchangetheuidtoonethat
has half the density (or move experiment to another planet and reduce g).
(i) ThissystemhasadissipativeelementRandaspring-likeelementduetothepotentialenergy,
buteventhoughsomemassmovesquicklythroughthetubethebulkofthemassmovesvery
slowly, kinetic energy is small, and we essentially have a spring-damper rst-order system.
All the liquid in the tube in class moved together, so that system was an equivalent spring-
mass-damper.
642
_
_
_
19.5 Recitation 5 Problem
(a) x=R
(b) The results are summarized below.
(c)
Spring PE=
1
2
k
t
x
2
F
k
=k
t
x
Mass KE=
1
mx
2
F =mx
2
1

Inertia KE= J

2
T =J
2
Translational Damper P =b
t
x
2
F
b
=b
t
x
Rotational Damper P =b

2
T
b
=b

KE=
2
1
mx
2
+
2
1
J

2
=
2
1
mR
2

2
+
2
1
J

2
=
2
1
_
mR
2
+J
_

2
=
2
1
J
eq

2
PE=
1
k
t
x
2
=
1
k
t
R
2

2
=
1
_
k
t
R
2
_

2
=
1
k
eq

2
2 2 2 2
P =b
t
x
2
+b

2
=b
t
R
2

2
+b

2
= (b

+b
t
R
2
)

2
=b
eq

2
(d)
J
eq

=b
eq

k
eq
+T
(J+mR
2
)

+ (b

+b
t
R
2
)

+ (k
t
R
2
)=T
(e)
J
eq
=J+mR
2
b
eq
=b

+b
t
R
2
k
eq
=k
t
R
2
(f)
k
t
R
2

n
= k
eq
/J
eq
=
J+mR
2
2
=
b
eq

n
k
eq
b
eq
b

+b
t
R
2
= _ = _
2 k
eq
J
eq
2 k
t
R
2
(J+mR
2
)
(g) For large R, translational terms dominate
k
t
b
t

n
and
m 2 k
t
m
(h)
J
eq
x +
b
eq
x +
k
eq
x=T
R R R

n
and remain the same.
643
(i) The response is shown in Figure 19.5.1.
T/k
a
n
g
l
e

time
Figure 19.5.1: Underdamped Step Response
644
_ _
_ _
19.6 Recitation 6 Problem
(a) Start by redrawing the circuit with current denitions and node names.
i
R
1
R1
Node A Q
$ +
Q
R1
-
+
+
Q
R2
-
i
C
i
Q
+
R2
C
Q
R C R
2
-
+
Node B
Q
%
- V
in
i
-
+
R3
Q
R
3
R3
-
Figure 19.6.1: RC Circuit with node names
Elemental Equations:
v
R1
dv
C
i
R1
= i
C
=C
R
1
dt
v
R2
v
R3
i
R2
= i
R3
=
R
2
R
3
Voltages:
v
o
=v
C
v
B
=v
R3
v
C
=v
A
v
B
v
R1
=v
R2
=V
in
v
A
Node Equations:
i
R1
+i
R2
i
C
= 0
v
R1
+
v
R2
C
dv
C
= 0
R
1
R
2
dt
i
C
i
R3
= 0 C
dv
C

v
R3
= 0
dt R
3
After the algebra:
R
1
R
2
dv
o
R
3
+ C +v
o
=V
in
R
1
+R
2
dt
(b)

t
v
o
(t) = V
in
+Ae
ReqC
v
o
(0)=V
in
+A= 0 A=V
in
t
v
o
(t) = V
in
1 +e

ReqC
645
(c)
R
1
R
2
R
eq
=R
3
+
R
1
+R
2
646
19.7 Recitation 7 Problem
647
19.8 Recitation 8 Problem
648
19.9 Recitation 9 Problem
649
19.10 Recitation 10 Problem
650
20 Recitation Quizzes
20.1 Recitation 1 Quiz
(a) Combining and rearranging the 4 given equations, we get
AR

h+h= 0
g
(b) The form of the equation is h

+h=0. Therefore
AR
=
g
(c) Theformofthesolutionish(t) = h
0
e
t/
. Rememberingthate
1
=.37ande
2.3
=.1,after
one time constant h(t=) = .37h
0
, after 2.3 time constants h(t= 2.3) = .1h
0
.
651
20.2 Recitation 2 Quiz
a
m
p
l
i
t
u
d
e

(
m
i
l
l
i
m
e
t
e
r
s
)

40
30
20
10
0
-10
-20
-30
-40
W
T
0 2 4 6 8 10 12 14 16 18
time (seconds)
Figure 20.2.1: Second Order Response
The period T 2 seconds, so
d
=
2
T

=. The natural frequency


n

d
, particularly for this
system which is underdamped, so
n
=. The damping ratio is found from the time constant of
the exponential envelope, =

1
=

1
n
, where 6.25 seconds taken from the graph, leading to
1 1
=
n
=
6.25
=.05. The initial position is read directly from the graph, x(0) = 20 mm.
The system is underdamped.
652
20.3 Recitation 3 Quiz
(a) The sketch is shown below.
v(t) F
0
/b(1e
t/
) m/b
0.63 F0/b
F0/b
V
e
l
o
c
i
t
y

0 tau 2 tau 3 tau 4 tau
time
Figure 20.3.1: Plot of velocity vs time
(b) v() = F
0
/b
653
20.4 Recitation 4 Quiz
(a) The free body diagram for the pulley is shown below (along with the free body diagrams for
the rest of the system). Note that as drawn, F
k
=kxand F
b
=bx .
Spring-Damper
Pulley
x
F
k
F
b
F
T
1 T
2
Support Hanging Mass
F
k
F
b T
1
x
y
mg
F
Figure 20.4.1: Free body diagrams
(b) y= 2x
654
20.5 Recitation 5 Quiz
(a) The free body diagrams are shown below.
T
T
Pinion
F
t
T
t
=RF
t
b
Z
T
x
F
t
Rack
b x b
.
x
.
k
t
x
tt
Figure 20.5.1: Free body diagrams
(b) x=R
(c) Start with the standard force balancing equations.
J

=T b

RF
t
and
mx + b
t
x + k
t
x=F
t
use part (b) to modify this equation
mR

+b
t
R

+k
t
R=F
t
Combine this and the rst equation to obtain
(J+mR
2
)

+ (b

+b
t
R
2
)

+ (k
t
R
2
)=T
655
_ _
20.6 Recitation 6 Quiz
Elemental Equations:
v
R1
dv
C
i
R1
= i
C
=C
R
1
dt
v
R2
i
R2
=
R
2
Voltages:
v
o
=v
C
v
R1
=v
R2
=V
in
v
o
Node Equations:
i
R1
+i
R2
i
C
= 0
v
R1
+
v
R2
C
dv
C
= 0
R
1
R
2
dt
After the algebra:
R
1
R
2
dv
o
C +v
o
=V
in
R
1
+R
2
dt
656
20.7 Recitation 7 Quiz
657
20.8 Recitation 8 Quiz
658
20.9 Recitation 9 Quiz
659
20.10 Recitation 10 Quiz Spring 2005
660

You might also like