Time Response of Second Order Systems

d 2 y(t ) dy(t )
M 2
= f (t ) − B − Ky(t )
dt dt
M (s 2Y (s) − sy(0) − y' (0)) = F (s) − B(sY (s) − y(0)) − KY (s)
Let y' (0) = 0

Ms2Y (s) + BsY (s) + KY (s) = Msy0 + By0 + F (s)

Or
(Ms + B) y0 F ( s)
Y ( s) = + 2
Ms + Bs + K Ms + Bs + K
2

Consider the first term only:

(s + B / M ) y0 (s + 2ςωn ) y0
Y ( s) = 2 = 2
s + ( B / M )s + K / M s + 2ςωn s + ωn 2
ς = dimensionless damping ratio
ωn = the natural frequency

Where
K B
ωn = ; ς=
M 2 KM
2
The characteristic equation s 2 + 2ςω n s + ω n has two roots :
s1 = −ςω n + jω n 1 − ς 2
s2 = −ςω n − jω n 1 − ς 2
If ς > 1 ⇒ roots are real
If ς < 1 ⇒ roots are complex (under damped)
If ς = 1 ⇒ same roots and real (critically damped)

For unit step response :

1
y (t ) = 1 − e −ςω nt sin( βω nt + θ )
β
β
β = 1 − ς 2 , θ = tan −1 ( ) = cos −1 ς
ς
 1 −ζω nt
Step response: y (t ) = 1 − e sin(ω n βt + θ )
β
where β = 1 − ζ 2 , θ = cos −1 ζ
Showing the step response with different damping
coefficients

Standard Performance measures

„ Performance measures are usually defined in terms of the step response of a
system as below:
„ Swiftness of the response is measured by rise time Tr , and peak time Tp
„ For underdamped system, the Rise time Tr (0-100% rise time) is useful,
„ For overdamped systems, the the Peak time is not defined, and the (10-90 % rise
time) T is normally used
r1

„ Peak time: T p
„ Steady-state error: ess
„ Settling time: T s

„ Percent of
Overshoot:
M pt − fv
P.O. = × 100 %
fv
is the peak value
M pt
fv is the final value
of the response
„ Percentage overshoot measures the closeness of the
response to the desired response.
„ The settling time T s is the time required for the
system to settle within a certain percentage δ of the
input amplitude.
„ For second order system, we seek T s for which the
response remains within 2% of the final value. This
occurs approximately when:
e −ζω nTs < 0.02
or : ζω nTs ≅ 4
Therefore :
4
Ts ≅ 4τ =
ζω n
„ Hence the settling time is defined as 4 time constants.

„ Explicit relations for M pt and T p :

„ To find Tp we can either differentiate y(t) directly or

indirectly through Laplace Transform of y’(t):

π π
Tp = =
ωn β ωn 1− ς 2
The peak response is :
1−ς 2
M pt = 1 + e −ζπ /
Percentage overshoot :
1−ς 2
P.O. = 100e −ζπ /
Impulse response of the second order system:

„ Laplace transform of the unit impulse is R(s)=1

ω n2
Y (s) =
„ Impulse response: (s 2
+ 2 ζω n s + ω
2
n )

„ Transient response for the impulse function, which is simply is the

derivative of the response to the unit step:
ω n − ζω
y (t ) = e nt
sin( ω n β t )
β

„ Responses and pole locations

„ Time Responses and Pole Locations:
2
The characteristic equation s 2 + 2ςω n s + ω n has two roots :
s1 = −ςω n + jω n 1 − ς 2
s2 = −ςω n − jω n 1 − ς 2

„ s1 , s2 are the poles.

4
Ts ≅ 4τ =
ζω n

„ Therefore the settling time is inversely proportional to the real part of the poles.