Thermal System Closed-Loop Temperature Control

aluminum plate thin-film resistive heater ceramic insulation conduction and convection heat transfer AD590 temperature sensor microcontroller on-off closed-loop control with relay support analog electronics

Mechatronics Thermal System Case Study

Kevin Craig 1

Objective of the Case Study

Control the temperature of a thin aluminum plate, as measured by a temperature sensor positioned in the middle of the top of the plate, by regulating the voltage supplied to a resistive heater positioned under the plate. The temperature of the plate is to be regulated to a point 20 C above the temperature of the ambient air.

Mechatronics Thermal System Case Study

Kevin Craig 2

How Will We Accomplish The Objective ?

Apply the general procedure for a dynamic system investigation Understand the physical system, develop a physical model on which to base analysis and design, and experimentally determine and/or validate model parameter values Develop a mathematical model of the system, analyze the system, and compare the results of the analysis to experimental measurements Design a feedback control system to meet performance specifications Implement the control system and experimentally validate its predicted performance
Mechatronics Thermal System Case Study Kevin Craig 3

Physical System
Heated Thin Aluminum Plate 1/32 inch thick 2 in x 2 in k = 177 W/m-K

Temperature Sensor

Ceramic Tape Insulating Layer 1/8 inch thick 2 in x 2 in k = 0.055 W/m-K

Thin-Film Resistive Heater between Thin Aluminum Plate and Ceramic Insulation

Aluminum Support Plate

Mechatronics Thermal System Case Study

Kevin Craig 4

Properties of the Resistive Heater

Specification Manufacturer Model Number Heater Resistance Heater Area Heater Thickness Value Minco Products HK-5169-R185-L12-B 185 ohms 10% 4 sq. in. 0.010 in.

Mechatronics Thermal System Case Study

Kevin Craig 5

Properties of the AD590 Temperature Sensor

Specification Rated Temperature Range Power Supply (min) Power Supply (max) Nominal Output Current @ 298.2 K Temperature Coefficient Calibration Error Maximum Forward Voltage Maximum Reverse Voltage Case Breakdown Voltage
Mechatronics Thermal System Case Study

Value -55 C to 150 C 4 volts 30 volts 298.2 A 1 A/K 2.5 C 44 volts -20 volts 200 volts
Kevin Craig 6

Material Properties of 6061 Aluminum

Property Melting Point Density, Specific Heat, cp Thermal Conductivity, k Value 775 K 2770 kg/m3 875 J/kg-K 177 W/m-K

Mechatronics Thermal System Case Study

Kevin Craig 7

Temperature Sensor Circuit

15 V 1 A/K Sensor Voltage 1K

LF 411 Ksensor(volts/K) = (1A)(Rsensor)

Mechatronics Thermal System Case Study

Kevin Craig 8

Physical Model
Ambient air

Sensor Thin Aluminum Plate

Thin-Film Resistive Heater Ceramic Insulation

Heat input

heater

Mechatronics Thermal System Case Study

q convection

Kevin Craig 9

Simplifying Assumptions
Temperature of the plate is uniform No heat loss through the sides of the plate Thermal conductivity of the plate is constant Heat loss due to radiation is negligible compared to convective heat loss from the plate Convection coefficient is constant and is evaluated at the operating temperature of the plate Heat loss through the insulative layer is negligible Sensor dynamics are negligible Ambient air temperature is unaffected by the heat flux from the plate
Kevin Craig 10

Mechatronics Thermal System Case Study

Mathematical Modeling
Define system, system boundary, system inputs and outputs Define through and across variables Write physical relations for each element Write system relations of equilibrium and/or compatibility Combine system relations and physical relations to generate the mathematical model for the system

Mechatronics Thermal System Case Study

Kevin Craig 11

Define System, System Boundary, System Inputs and Outputs

Input: Voltage supplied to resistive heater Output: Plate temperature as measured by sensor on top of plate
q convection
Sensor Thin Aluminum Plate Thin-Film Resistive Heater Ceramic Insulation

Ambient air

Heat input q heater

Mechatronics Thermal System Case Study Kevin Craig 12

Define Through and Across Variables

Through Variable: heat flow rate q (J/s or W) Across Variable: temperature (K)
Assumption: All points in the body have the same average temperature and temperature deviations from the average at various points do not affect the validity of the single-temperature model.

Mechatronics Thermal System Case Study

Kevin Craig 13

Write Physical Relations for Each Element

Thermal Capacitance Thermal Resistance Conduction ( t ) q(t ) = Convection R Radiation
1 & = q in t q out t C
kA q(t ) = [ 1 ( t ) 2 ( t )] L q ( t ) = hA[ 1 ( t ) 2 ( t )]
4 q ( t ) = C[1 (t ) 4 2 ( t )] 2 = C[(1 + 2 2 )(1 + 2 )][1 ( t ) 2 ( t )]

af

af

Thermal Sources
Mechatronics Thermal System Case Study

Vh Vh2 P = Vh i h = Vh = Rh Rh
Kevin Craig 14

Write System Relations of Equilibrium and/or Compatibility

Select the temperature of each thermal capacitance as a state variable and use: & 1
= C q in t q out t

af

af

to obtain the corresponding state-variable equation. The net heat flow rate into a thermal capacitance depends on the heat sources and heat flow rates through thermal resistances. Use 1
qt =

af

1 t 2 t

af af

to express the heat flow rates through the resistances in terms of the systems state variables.
Mechatronics Thermal System Case Study Kevin Craig 15

Combine System Relations and Physical Relations to Obtain the Mathematical Model

af 1 q a tf = = hA R 1 & = q a tf q t) C O & = 1 L q a t f 1 a f PQ CM R N 1 1 1 & + = q at f + RC C RC

q in Vh2 t = heater input = Rh
ambient out in out in ambient in ambient
Mechatronics Thermal System Case Study

ambient

Kevin Craig 16

Predicted Dynamic Response

Solve the mathematical equation both analytically and numerically to predict the dynamic response of the physical system and gain physical insight. dq o + q o = Kq i dt 1 1 1 & + = q in t + ambient RC C RC & + = Rq in ( t ) + ambient = RC

af

Qo K s = Qi s + 1
Mechatronics Thermal System Case Study

af

F II F q = Kq G 1 eH K J H K Q K i f = tan a Q af + 1
t o is o i 2

Kevin Craig 17

Experimental Testing: Model Parameter Identification

Two physical parameters need to be determined, one by calculation and the other by experiment. Thermal Capacitance C = Mcp = Vcp = 4.96 J/K Thermal Resistance R = 1/hA Either measure = RC and solve for R or measure the steady-state temperature of the plate and use 2 ambient Vh q in = q out or = Rh R

Mechatronics Thermal System Case Study

Kevin Craig 18

Control System Design

Error To Workspace4 + u[1]^2/Rh Temperature Reference Sum Controller: Relay ElectricalThermal Conversion Qin To Workspace2

R(s) RC.s+1 System

Temp To Workspace

time

Matlab / Simulink Block Diagram Thermal Closed-Loop Control System

Mechatronics Thermal System Case Study

Clock To Workspace1

Kevin Craig 19

Control System Simulation

25

20 temperature (degrees C)

15

10

20

40

60 time (sec)

80

100

Simulation of Control about 20 C Setpoint with a 2 C Deadband

Mechatronics Thermal System Case Study Kevin Craig 20

Experimental Set-Up & Control Implementation

Mechatronics Thermal System Case Study

Kevin Craig 21

Microcontroller Software Design

Setpoint Read Sensor Signal Deadband
Initalize Variables

Read Sensor and Setpoint Signals

Sensor Signal Less Than Deadband Bottom

Sensor Signal Greater Than Deadband Top

Sensor Signal In Deadband

Compute Error Signal

Yes

Yes

Yes

Implement Logic

Heater On

Heater Off

Issue Previous Command

Command Actuator

Decision Logic
Mechatronics Thermal System Case Study

Flow Diagram
Kevin Craig 22

Microcontroller Real-Time Basic Program

100 REM set up Port A as an output, reset Port B 110 xby(0fd03h)=128:xby(0fd01h)=6 120 PRINT USING(####) 130 REM calibrate the A to D converters 140 XBY(0ff03h)=2 150 DO: SFR=XBY(0ff03h):WHILE SFR.AND.2 160 IF SFR.AND.40h THEN GOTO 140 165 MEMX=0 170 REM read current temperature 180 XBY(0ff00h)=1 190 CURTEMP=256*XBY(0ff01h) 200 CURTEMP=CURTEMP+XBY(0ff00h) 210 REM read setpoint temperature 220 XBY(0ff00h)=0 230 SETEMP=256*XBY(0ff01h) 240 SETEMP=SETEMP+XBY(0ff00h) 250 ERROR=SETEMP-CURTEMP 260 IF (ERROR<-82).AND.(MEMX<-82) THEN COMMAND=0 270 IF (ERROR>82).AND.(MEMX>82) THEN COMMAND=255 275 MEMX=ERROR 280 IF (ERROR>-82).AND.(ERROR<82) THEN COMMAND=COMMAND 290 XBY(0fd00h)=COMMAND 300 PRINT "CURRENT TEMP=",CURTEMP,"SET TEMP=",SETEMP 310 GOTO 170
Mechatronics Thermal System Case Study Kevin Craig 23

Comparison of Predicted Dynamic Behavior with Actual, Measured Dynamic Behavior

25

20 temperature (degrees C)

Predicted: red curve (-) Measured: blue curve (--)

15

10

50

100 time (sec)

150

200

Mechatronics Thermal System Case Study

Kevin Craig 24