MECE 3350U

Control Systems

Lecture 4
Transfer Functions

1/41
MECE 3350 - C. Rossa 1 / 41 Lecture 4
Videos in this lecture

Lecture 4: https://youtu.be/KP6DATDCIdc

Exercise 16: https://youtu.be/2BBO3lcdm5U

Exercise 17: https://youtu.be/DoL2rarunKg

Exercise 18: https://youtu.be/Oo8Woo0oIhs

Exercise 19: https://youtu.be/HZMWeGR0RFo

Exercise 20: https://youtu.be/bz7P2ROXYMw

2/41
MECE 3350 - C. Rossa 2 / 41 Lecture 4
Outline of Lecture 4

By the end of today’s lecture you should be able to

• Understand the concept of transfer functions

• Find the transfer function of a given system

• Find the temporal response of a system for a given input

3/41
MECE 3350 - C. Rossa 3 / 41 Lecture 4
Applications

If the spring is stretched to a point x (t) = 5 mm, held, then released at time
t = 0, how does the position of m1 evolve in time?

4/41
MECE 3350 - C. Rossa 4 / 41 Lecture 4
Applications

What torque must be applied to each of the robot joints so that end-effector
moves along a given trajectory with a given speed?

5/41
MECE 3350 - C. Rossa 5 / 41 Lecture 4
Input/output relation
Transfer function: A relation between the input and output of a linear system
Example: Input: force f (t), output: displacement x (t)

d 2x dx
m +b + kx = f (t)
dt dt
Taking the Laplace transform:

The transfer function is

X (s)
H(s) = = (1)
F (s)

6/41
MECE 3350 - C. Rossa 6 / 41 Lecture 4
Input/output relation

Example: Input: force f (t), output: velocity v (t)

The dynamic model is

(2)
Laplace transform of (2) is

The transfer function is

V (s)
H(s) = = (3)
F (s)

7/41
MECE 3350 - C. Rossa 7 / 41 Lecture 4
Input/output relation
Input: Voltage v (t), output: Current i(t)

di(t)
v (t) = Ri(t) + L (4)
dt
Taking the Laplace transform of (4):

The transfer function is

I(s)
H(s) = =
V (s)

8/41
MECE 3350 - C. Rossa 8 / 41 Lecture 4
Input/output relation
Input: Voltage v (t), output: vL (t)

di(t) di(t)
v (t) = Ri(t) + L , vL = L (5)
dt dt
Taking the Laplace transform of (5):

The transfer function is

I(s) 1 VL (s)
= , =
V (s) Ls + R V (s)

9/41
MECE 3350 - C. Rossa 9 / 41 Lecture 4
Input/output relation

The equation of motion of the simple pendulum is

d 2 θ(t) g
m + m sin θ = T (t) (6)
dt 2 `
For small angles, sin θ ≈ θ, in the frequency domain:

θ(s)
= (7)
T (s)

10/41
MECE 3350 - C. Rossa 10 / 41 Lecture 4
Transfer function poles and zeros

Transfer function: A rational function in the complex variable s = σ + jω:

N(s) bm s m + bm−1 s m−1 + . . . + b1 s + b0

H(s) = = (8)
D(s) an s n + an−1 s n−1 + . . . + a1 s + a0

The zeros zi are the roots of

N(s) = 0
Thus:
lim N(s) = 0 (9)
s→zi

The poles pi are the roots of

D(s) = 0
Thus:
lim H(s) = ∞ (10)
s→pi

11/41
MECE 3350 - C. Rossa 11 / 41 Lecture 4
Poles and zeros
Consider the following function:

s(s + 3)
F (s) =
s 2 + 2s + 5
→ Poles: −1 + 2j, −1 − 2j

→ Zeros: 0, −3

1
Imaginary Axis

-1

-2
-3 -2.5 -2 -1.5 -1 -0.5 0
Real Axis

12/41
MECE 3350 - C. Rossa 12 / 41 Lecture 4
First order transfer functions

Order: The number of the highest derivative in the denominator (power of s)

Standard form of a first order system:

1
G(s) = k
τs + 1
Characteristic equation: The denominator of the transfer function

I(s) 1
H(s) = = =
V (s) Ls + R

Time constant: Characterizes the response to a step input of a first-order

system.

L
τ = (11)
R
→ The denominator must be in the form of τ s + 1

13/41
MECE 3350 - C. Rossa 13 / 41 Lecture 4
Second order transfer functions
Standard form
ωn2
H(s) = (12)
s2 + 2ζωn s + ωn2
Where: ωn is the natural frequency, ζ is the damping ratio.
We will come back to these definitions in the next lecture.

X (s) 1
H(s) = = =
F (s) ms 2 + bs + k

⇒ζ= √b : Dimensionless damping ratio

2 mk
pk
⇒ ωn = m
: Natural frequency (rad/s)

14/41
MECE 3350 - C. Rossa 14 / 41 Lecture 4
Temporal response

Step 1: Define the input signal in Laplace domain

input signal time domain frequency domain

impulse r (t) = δ(t) R(s) = 1

step r (t) = a R(s) = a 1s

ramp r (t) = at R(t) = a s12

a is a constant

15/41
MECE 3350 - C. Rossa 15 / 41 Lecture 4
Temporal response

Step 1: Define the input signal in Laplace domain

input signal time domain frequency domain

polynomial r (t) = at n n!
R(s) = a s n+1

a
sine r (t) = sin(at) R(s) = s 2 +a2

s
cosine r (t) = cos(at) R(t) = s 2 +a2

a is a constant

16/41
MECE 3350 - C. Rossa 16 / 41 Lecture 4
Temporal response
Step 2: Replace the input signal in the transfer function

X (s) 1
H(s) = = (13)
F (s) ms 2 + bs + k

For a impulse input f (t) = δ(t), F (s) = 1 and the temporal response is
1
X (s) = (14)
ms 2 + bs + k

For a step-type input f (t) = 1 N, F (s) = 1/s and the temporal response is
1 1
 
X (s) = (15)
ms 2 + bs + k s

For a sinusoidal input f (t) = 5 sin(t) N:

1 5
 
X (s) = (16)
ms 2 + bs + k s2 + 1
17/41
MECE 3350 - C. Rossa 17 / 41 Lecture 4
Temporal response
Step 3: Calculate the inverse transform of the resulting function

For a impulse input f (t) = δ(t), F (s) = 1 and the temporal response is
1
n o
x (t) = L −1 (17)
ms 2 + bs + k

For a step-type input f (t) = 1 N, F (s) = 1/s and the temporal response is
1 1
n  o
x (t) = L −1 (18)
ms 2 + bs + k s
and so on.
1.5
displacement [m]

0
0 time [sec] 15

18/41
MECE 3350 - C. Rossa 18 / 41 Lecture 4
Steady-state value

Final value theorem: Gives the steady-state value without computing the
inverse transform.

If the function converges, i.e., the poles of sX (s) have negative real parts, then:

lim x (t) = lim sX (s) (19)

t→∞ s→0

For a step type input, the mass spring damper system settles at
1 1
n  o
lim x (t) = lim sX (s) = lim s = (20)
t→∞ s→0 s→0 ms 2 + bs + k s

For a impulse input, the RL system converges at

1
n o
lim i(t) = lim sI(s) = lim s = (21)
t→∞ s→0 s→0 Ls + R

19/41
MECE 3350 - C. Rossa 19 / 41 Lecture 4
Exercise 16

Find the transfer function H(s) between the input voltage V and the output
voltage Vout .

Procedure:
→ Find the differential equation for the current
→ Find the equation for the output voltage
→ Calculate the transfer function

20/41
MECE 3350 - C. Rossa 20 / 41 Lecture 4
Exercise 16 - continued

21/41
MECE 3350 - C. Rossa 21 / 41 Lecture 4
Exercise 17

V (s)
Find the transfer function H(s) = R(s)
between the force r (t) and the velocity
of mass m1 .

Procedure:
→ Find the differential equation the velocity of each mass
→ Calculate the Laplace transform
→ Calculate the transfer function

22/41
MECE 3350 - C. Rossa 22 / 41 Lecture 4
Exercise 17 - continued

23/41
MECE 3350 - C. Rossa 23 / 41 Lecture 4
Exercise 17 - continued

[m1 s + (b1 + b2 )]V (s) − b1 V2 (s) = R(s)

k
 
− b1 V (s) + m2 s + b1 + V2 (s) = 0
s

24/41
MECE 3350 - C. Rossa 24 / 41 Lecture 4
Exercise 18
A high precision positioning slide is shown in the figure. The drive shaft friction
is bd = 0.65, the drive shaft spring constant is k = 1.8, mc = 1, and the slide
friction is bs = 0.9.

Determine:
→ Find the transfer function H(s) = X (s)/Y (s).
→ Calculate the natural frequency, damping ratio, the poles, and zeros of H(s)
→ Find the steady-state value for a step input
→ Plot the step response of using Matlab
25/41
MECE 3350 - C. Rossa 25 / 41 Lecture 4
Exercise 18 - continued

26/41
MECE 3350 - C. Rossa 26 / 41 Lecture 4
Exercise 18 - continued

27/41
MECE 3350 - C. Rossa 27 / 41 Lecture 4
Exercise 18 - continued - Using Matlab
H = tf([0.7 2],[1 2.8 2]) → Transfer function
damp(H) → Natural frequency and damping
step(H,10) → Step response
pzplot(H) → Location of zeros and poles

1 0.3

Imaginary Axis (seconds-1 )

displacement [m]

0 -0.3
0 10 -2 0 2
displacement [m] Real Axis (seconds -1 )

28/41
MECE 3350 - C. Rossa 28 / 41 Lecture 4
Exercise 19

Calculate the natural frequency and damping ratio of the following transfer
function

1.05 × 107
T (s) =
2s 2 + 1.6 × 103 s + 1.05 × 107

Determine:
→ Write the transfer function in standard form
→ Find the steady-state value for a step input
→ Calculate the natural frequency and damping ratio

29/41
MECE 3350 - C. Rossa 29 / 41 Lecture 4
Exercise 19 - continued

1.05 × 107
T (s) =
2s 2 + 1.6 × 103 s + 1.05 × 107

30/41
MECE 3350 - C. Rossa 30 / 41 Lecture 4
Exercise 20

Find the transfer function G(s) = VI2 (s)

(s)
of the circuit shown. Then, calculate
the step response of the circuit using Matlab. Take R = 10 Ω, C = 0.001 F,
L = 0.1 H, V = 5 V.

Determine:
→ Find the transfer function H(s) = I2 (s)/V (s).
→ Find the steady-state value for a step input
→ Plot the step response of using Matlab

31/41
MECE 3350 - C. Rossa 31 / 41 Lecture 4
Exercise 20 - continued

32/41
MECE 3350 - C. Rossa 32 / 41 Lecture 4
Exercise 20 - continued - Using Matlab

0.25 60

0.2
40

Imaginary Axis (seconds-1 )

0.15
20
Current

0.1
0
0.05
-20
0

-0.05 -40

-0.1 -60
0 0.1 -60 -40 -20 0
time [sec] Real Axis (seconds -1 )

33/41
MECE 3350 - C. Rossa 33 / 41 Lecture 4
Skills check 5 - From 2018 midterm

Find the transfer function between the acceleration A(s) and the force F (s) of
the mass-spring system shown. The coefficient of viscous friction between the
mass and the plane is b (4 marks)1 .

1
H(s) = s 2 /(ms 2 + bs + k) 34/41
MECE 3350 - C. Rossa 34 / 41 Lecture 4
Skills check 6 - From 2018 final examination

Find the transfer function between the input voltage v (t) and the electric
charge q(t) in the LRC circuit (3 marks out of 100 )2 .

2
H(s) = 1/[L(s 2 + (R/L)s + 1/(LC )] 35/41
MECE 3350 - C. Rossa 35 / 41 Lecture 4
Skills check 7 - From 2018 final examination

Find the transfer function between the input voltage V (s) and the voltage
across the inductor VL (s)3 .

3
Consult instructor or TA for answer 36/41
MECE 3350 - C. Rossa 36 / 41 Lecture 4
Skills check 8 - Strongly recommended

Consider the RC network shown, where v (t) is the input voltage and vc (t) is
the circuit output voltage. R is the same for all resistors. Using the transfer
function and its inverse Laplace transform, determine the temporal response of
vc (t) for a unit step-type input of v (t).4

4
v (t) = 0.5(1 − e −2t/(3RC ) ) 37/41
MECE 3350 - C. Rossa 37 / 41 Lecture 4
Skills check 9 - From 2018 assignment 1
The figure shows an automobile suspension system composed of two masses m1
and m2 , a spring of stiffness ks , and a damper with a coefficient of viscous
friction b. The lower spring kw represents the tire compressibility5 .

(a) Find the transfer function Y (s)/R(s).

(b) Calculate the final value of y (t) for a unit step-type displacement of the
road surface, i.e., r (t) = 1.
5
Answer see next slide 38/41
MECE 3350 - C. Rossa 38 / 41 Lecture 4
Skills check 9 - Answer

(a)

s + kbs
 
Y (s) kw b
= (22)
R(s) m1 m2 s 4 + as 3 + bs 2 + cs + d
where
b b
a= m1
+ m2
ks ks kw
b= m1
+ m2
+ m1
kw b
c= m1 m2
kw ks
d= m1 m2

(b) limt→∞ y (t) = 1

39/41
MECE 3350 - C. Rossa 39 / 41 Lecture 4
Skills check 10

Consider the three-mass system shown. The systems has three input forces i.e.,
u1 (t), u2 (t), u3 (t), and three outputs x1 (t), x2 (t), and x3 (t).

Obtain the equations of motion in the frequency domain. All initial conditions
are zero6 .

6
Final answer in the next slide 40/41
MECE 3350 - C. Rossa 40 / 41 Lecture 4
Skills check 10 - Answer

For mass 1

U1 (s) + (k2 + sb2 )X2 (s) = [m1 s 2 + (b1 + b2 )s + k1 + k2 ]X1 (s)

For mass 2

U2 (s) + (k3 + sb3 )X3 (s) + (k2 + b2 s)X1 (s) = [m2 s 2 + (b2 + b3 )s + k2 + k3 ]X2 (s)

For mass 3

U3 (s) + (k3 + sb3 )X2 (s) = (m3 s 2 + b3 s + k3 )X3 (s)

41/41
MECE 3350 - C. Rossa 41 / 41 Lecture 4
Next class...

• Effect of pole locations

42/41
MECE 3350 - C. Rossa 42 / 41 Lecture 4