What is control?

The act of producing a desired result

Occurs in all artificial process
215461: Automatic Flight Control I
Lecture for Week3
TFs, Block Diagrams & necessary
Why do we need it?
algebra and Models of engineering Not every system naturally gives us
systems for control
(many pictures are from Kuo, 2003 and reference therein) desired responses (results)

Sept 28, 2014 Aerospace Engineering, KU 1 Sept 28, 2014 Aerospace Engineering, KU 2

Open-loop vs. Closed-loop Example: Automobile Speed

Type of system Advantages Disadvantages Control
1. Open-loop 1. Simple construction and ease of 1. The system response very sensitive
Control = Sensing + Computation + Actuation
control systems maintenance. to external disturbance and
2. Less expensive than a corresponding internal variations in system
closed-loop control system
3. There is no stability problem
parameters.
2. Recalibration is necessary from time
Actuate Sense
4. Convenient when output is hard to to time in order to maintain the Gas pedal Vehicle Speed
measure or measuring the output required quality in the output
precisely is economically not feasible.
2. Closed-loop 1. Makes the system response 1. Risk instability
control systems relatively insensitive to external 2. Complexity in analysis and Compute
disturbance and internal implementation and expensive Control law
variations in system parameters.
2. Possible to use relatively
inaccurate and inexpensive
components to obtain the Goals
accurate control of a given plant. Stability: system maintains desired operating point (hold steady speed)
3. Better control of transient & Performance: system responds rapidly to changes (accelerate to 6 m/sec)
steady-state response
4. Increased accuracy
Robustness: system tolerates perturbations in dynamics (mass, drag, etc)
-Increased ability to reproduce output with
varied input.
Sept 28, 2014 Aerospace Engineering, KU 3 Sept 28, 2014 Aerospace Engineering, KU 4
 Transfer function (TF)
Remarks for the TF
of SISO systems
Transfer function : The TF is defined only for a linear time-invariant
Y ( s)
G ( s) =
(LTI) system (not for non-linear systems)
U ( s)
The TF is independent of the input
Laplace transform of the input/Laplace transform of the output
with all initial conditions set to zero The TF is proper when nm otherwise it is
improper
bm s m + bm 1s m 1 + K + b1s + b0 Numerator of the TF Characteristic Equation (CE) is defined as:
G ( s) = n =
s + an 1s n 1 + K + a1s + a0 Denometer of the TF Denometer of the TF =0
The stability of the system can be determined
by the root of the CE

Sept 28, 2014 Aerospace Engineering, KU 5 Sept 28, 2014 Aerospace Engineering, KU 6

Transfer function (TF) Example: a turboprop engine

of Multivariable systems (Kuo., pp 29-30)

Expanding the idea of TF of the SISO to MIMO :

Y( s) = G ( s)R ( s )
Y 1( s ) R1 ( s ) G11 ( s ) G12 ( s ) L G1 p ( s )
Y ( s ) R ( s)
where Y(s) =
2 , R (s) = 2 , G (s) = G21 ( s ) G22 ( s ) L G2 p ( s )
M M M M O M

Y q( s ) R p ( s ) Gq1 ( s ) Gq 2 ( s ) L Gqp ( s )
Y (s)
and Gij ( s ) = i , i.e. , Yi ( s ) = Gi1 ( s ) R1 ( s ) + Gi 2 ( s ) R 2 ( s ) + K + Gip ( s ) R p ( s )
R j (s)
The rotational speed of the shaft and the turbine inlet
for a p - input, q - output system temperature are controlled by both of these 2 variables:
fuel rate and propeller blade angle.

Sept 28, 2014 Aerospace Engineering, KU 7 Sept 28, 2014 Aerospace Engineering, KU 8
Example: a turboprop engine Remarks on LTI system
(Kuo., pp 29-30) representations
Suppose the governing equations can be found A linear time-invariant system can be represented
(experimentally and/or analytically): in many ways:
- differential equation
Y 1( s ) = G11 ( s ) R1 ( s ) + G12 ( s ) R 2 ( s ) - impulse response
Y 2( s ) = G21 ( s ) R1 ( s ) + G22 ( s ) R 2 ( s ) - transfer function
- state variable or state space
where Y 1( s ) = rotational speed of the engine - block diagram and flow graph
Y 2( s ) = turbine inlet temperature Each of the description can be converted to the
R1 ( s ) = fuel rate others. We will see how to present the system in
R 2 ( s ) = propeller blade angle terms of block diagrams and how to determine the
transfer function of systems of block diagrams.

Sept 28, 2014 Aerospace Engineering, KU 9 Sept 28, 2014 Aerospace Engineering, KU 10

Block Diagrams
Example: analysis of a simple system
(& Block Diagram Algebra)
n2
Let G ( s ) = , with less than 1, find the response
Block diagrams can be used to describe cause s 2 + 2 n s + n2
and effect relationship of any systems to the unit impulse of this system.
R(s) =1 Y(s) = ?
cause effect G(s)
System
(Plant)
n2
y (t ) = L1{G ( s ) R ( s )} = L1 2 (1)
2
For LTI
system

s + 2 n s + n
Using Partial Fraction expansion :
R(s) Y(s) n2
(1) = L1
C1 C2
L1 2 2
+
G(s) s + 2 n s + n s + j s + + j
where = n , = n 1 2
Sept 28, 2014 Aerospace Engineering, KU 11 Sept 28, 2014 Aerospace Engineering, KU 12
 Example: Find the time response of a system
Example: analysis of a simple system described by

n2 s2 + s 5
C1 = (s + j )G ( s ) s ( j ) = G (s) = due to the input r (t ) = u (t ) where u (t ) = unit step.
2 j s 2 + 3s + 2
n2
C2 = (s + + j )G ( s ) s ( + j ) = s2 s + 5 s2 s + 5
2 j Q Y ( s) = G ( s) R( s) = =

( )
s s + 3s + 2 s (s + 1)(s + 2 )
2

n2 t jt
y (t ) = L1
+
C1

+
+
C2
+
=

(
e e e j t ) =
5

5
+
3/ 2
s j s j 2 j
2s s + 1 2 + 2
=
n
1 2
(
e nt sin n 1 2 t) 5 3
y (t ) = 5e t + e 2t
2 2
which can be recognized as the impulse response function of
a spring - mass - damper system.

Sept 28, 2014 Aerospace Engineering, KU 13 Sept 28, 2014 Aerospace Engineering, KU 14

Quiz: Why do we need feedback?

Question: If we can model the plant perfectly,
Find the governing differential equation of why do we have to design a
the system in the previous example complicate algorithm on top of this?
Should we use open-loop control or
closed-loop control ?

Answer: Depend on the level of knowledge you

have of the system you want to control.

Feedback is the essential feature of an effective control

systems (it helps us deal with uncertainty in our knowledge
of the plant and changes in the plant behavior)
Sept 28, 2014 Aerospace Engineering, KU 15 Sept 28, 2014 Aerospace Engineering, KU 16
 Block Diagrams Block Diagrams of Control Systems
& Block Diagram Algebra : System interconnections and components
Block diagrams can be used to describe cause Summation points
and effect relationship of any systems

cause System effect

(Plant)

For LTI
system
Perform summation and/or
subtraction of the signals
R(s) Y(s)
coming through the junction
G(s)
(according to the signs)

Sept 28, 2014 Aerospace Engineering, KU 17 Sept 28, 2014 Aerospace Engineering, KU 18

Block Diagrams of Control Systems Block Diagrams of Control Systems

: System interconnections and components : System interconnections and components
Take-off points Systems can be interconnected in a variety of ways,
y(t)
e.g., the single block diagram (see page 2) and
Y(s)
y(t) Series interconnections
Y(s)
Y ( s ) = G1 ( s )G2 ( s ) R ( s )
Example: From a feedback speed control, write ec (t )
in terms of v(t),er (t), K, Kt . Parallel interconnections
Y ( s ) = (G1 ( s ) + G2 ( s )) R ( s )

Feedback interconnections
G (s)
Y (s) = R( s)
1 + G (s) H (s)
Sept 28, 2014 Aerospace Engineering, KU 19 Sept 28, 2014 Aerospace Engineering, KU 20
 Feedback Control Systems Quiz:
Basic terminology: reference input (command) Show (using some algebraic manipulation)
output (controlled variable)
that a feedback control system;
feedback signal
actuating signal R(s) + Y(s)
G1(s)
error signal _
feedback transfer function
Solution :
forward-path transfer function H1(s) Q Y ( s ) = G1 ( s )(R( s ) H1 ( s )Y ( s ) )
loop transfer function Y ( s ) + G1 ( s ) H1 ( s)Y ( s ) = G1 ( s ) R ( s )

can be simplified as:

G1 ( s )
L(s)=G(s)H(s) Y ( s) = R( s) = G ( s) R( s)
1 + G1 ( s ) H1 ( s )
closed-loop transfer function: G1 ( s )
where G ( s ) =
negative feedback system R(s) Y(s) 1 + G1 ( s ) H1 ( s )
M(s)=G(s)/(1+G(s)H(s)) G(s)
system transfer function

Sept 28, 2014 Aerospace Engineering, KU 21 Sept 28, 2014 Aerospace Engineering, KU 22

Block Diagram Reduction Block Diagram Reduction

using simple algebraic manipulations using simple algebraic manipulations
Moving a summation point ahead of a block Moving a take-off point ahead of a block

Moving a summation point beyond a block Moving a take-off point beyond a block

Sept 28, 2014 Aerospace Engineering, KU 23 Sept 28, 2014 Aerospace Engineering, KU 24
Block Diagram Reduction Example: Obtain the transfer function of the
using simple algebraic manipulations following system
H2(s)
Moving a take-off point ahead of a summation point _
+ Y(s)
R(s) + G1(s) + G2(s) G3(s) G4(s)
_ +
H1(s)

H3(s)

Moving a take-off point beyond a summation point

Solution:

R(s) G1 (s)G 2 (s)G 3 (s)G 4 (s) Y(s)

1 + G 2 (s)G 3 (s)H 2 (s) G 3 (s)G 4 (s)H1 (s) + G1 (s)G 2 (s)G 3 (s)H 3 (s)

Sept 28, 2014 Aerospace Engineering, KU 25 Sept 28, 2014 Aerospace Engineering, KU 26

Example: Find the transfer function of the

Systems with Multiple Inputs following system.
The procedure to obtain the relationship between the
inputs and the output for such systems is:
1. Set all inputs except one equal to zero
2. Determine the output signal due to this one non-zero
input Solution: First we set D(s)=0. The system is reduced to
3. Repeat the above steps for each of the remaining
inputs in turn
4. The total output of the system is the algebraic sum
(superposition) of the outputs due to each of the inputs
The output YR (s) due to the input R(s) is
G1 (s)G 2 (s)
YR (s) = R(s)
1 + G1 (s)G 2 (s)H(s)
Sept 28, 2014 Aerospace Engineering, KU 27 Sept 28, 2014 Aerospace Engineering, KU 28
 Put R(s)=0. The system is reduced to
Homework 2
1. Determine the transfer function of the
following block diagram.
The output YD (s) due to the input R(s) is
G 2 (s)
YD (s) = D(s)
1 + G1 (s)G 2 (s)H(s)
Y(s) = YD (s) + YD (s)
G1 (s)G 2 (s) G 2 (s)
= R(s) + D(s)
1 + G1 (s)G 2 (s)H(s) 1 + G1 (s)G 2 (s)H(s)

Sept 28, 2014 Aerospace Engineering, KU 29 Sept 28, 2014 Aerospace Engineering, KU 30

2. (Problem 3-1 of Kuo, 2003, page67)

The block diagram of the electric train control is shown: Example: Find C(s)/R(s).

er (t ) = voltage representing the desire speed, V

v(t ) = speed of the train, ft/s


The parameters and variables are M = mass of the train = 30,000 lb/sec
2

K = amplifier gain

K t = gain of speed indicator = 0.15V/ft/s

To determine the function of the a) Find the transfer function Gc(s) of the
controller, we apply a step function controller.
of 1 volt to the input of the controller, b) Derive the forward path TF: V(s)/E(s).
that is, ec (t ) = u s (t ). The output of the The feedback path is open in this case.
controller is measured and described c) Derive the closed-loop TF: V(s)/Er(s).
by the following function: d) Find the steady-state speed of the
f (t ) = 100(1 0.3e 0.6t 0.7e 10t )u s (t ) train when the input is er (t ) = u s (t ) Volt.

Sept 28, 2014 Aerospace Engineering, KU 31 Sept 28, 2014 Aerospace Engineering, KU 32
 Example: Solution:

Sept 28, 2014 Aerospace Engineering, KU 33 Sept 28, 2014 Aerospace Engineering, KU 34

Block Diagram Reduction by moving blocks

Example:

H1

Sept 28, 2014 Aerospace Engineering, KU 35 Sept 28, 2014 Aerospace Engineering, KU 36
 Quiz: Solution:

Sept 28, 2014 Aerospace Engineering, KU 37 Sept 28, 2014 Aerospace Engineering, KU 38

Engineering Systems Components

Electrical systems
Mechanical systems
Electro-mechanical systems
Hydraulic systems
Pneumatic systems
Thermal systems
etc

Sept 28, 2014 Aerospace Engineering, KU 39 Sept 28, 2014 Aerospace Engineering, KU 40
 In Continuous Time Domain Laplace Transform (S-domain)
In general: In general:

Resistor VR = Ri (t ) Resistor VR ( s ) = RI ( s )
di (t )
Inductor VL = L Inductor VL ( s ) = LsI ( s )
dt
t
1
VC = i ( )d 1 I (s)
VC (s) =
Capacitor C0 Capacitor C s
dV c (t )
KVL KCL iC = C I C ( s ) = CsV c ( s )
dt
Sept 28, 2014 Aerospace Engineering, KU 41 Sept 28, 2014 Aerospace Engineering, KU 42

Find the differential equation for the given Find the Transfer Function
system below: 1st method
Prevoiusly, we get:-

1
C
vi (t ) = i (t ) R + i (t )dt

dVc (t )
By applying KVL Thus, i (t ) = C Vi ( s ) = Vc ( s )[ RCs + 1]
1 dt
Vi (t ) = VR (t ) + VC (t ) Vi (t ) = i(t ) R + i(t )dt dvc (t )
C vi (t ) = RC + vc (t ) Finally,
Since dt
or
V R (t ) = i (t ) R and
di(t ) 1 1 dVi (t ) Change to Laplace Transform Vc ( s )
=
1
+ i(t ) = Vi ( s ) RCs + 1
1 dt RC R dt Vi ( s ) = RCsVc ( s ) + Vc ( s )
VC (t ) =
C i ( t ) dt
Sept 28, 2014 Aerospace Engineering, KU 43 Sept 28, 2014 Aerospace Engineering, KU 44
 Find the Transfer Function Find the differential equation for the given
system below
2nd method
R

1
Vi(s)
Cs
Construct KVL :- 1 By applying KVL
1 V (s) = ( R + )Vc ( s )Cs
V ( s) = ( R + ) I ( s)
Cs
Cs Vi (t ) = VR (t ) + VL + VC (t ) Thus,
V ( s ) = RVc ( s )Cs + Vc ( s ) Since di(t ) 1
dt C
Vc ( s ) =
1 Vi (t ) = i(t ) R + L + i(t )dt
Cs
I (s)
V ( s ) = Vc ( s )[ RCs + 1] V R (t ) = i (t ) R and

Change to:- Finally:- di ( t ) or

VL = L and
I ( s ) = Vc ( s )Cs Vc ( s ) 1 dt d 2i(t ) R di(t ) 1 1 dVi (t )
= + + i (t ) =
V ( s ) RCs + 1 1 dt 2 L dt LC L dt
Sept 28, 2014 Aerospace Engineering, KU 45
VC (t ) =
Sept 28, 2014 C i ( t ) dt
Aerospace Engineering, KU 46

Find the differential equation for the

Quiz: Find Transfer Function given system below

i2 (t )
i1 (t )

Loop 1 Loop 2
Apply KVL at Loop 1 Apply KVL at Loop 2

R2 (i2 (t ) i1 (t ) ) +
1
C
Using 1st and 2nd method Vi (t ) = R1i1 (t ) + R2 (i1 (t ) i2 (t )) i2 (t ) dt = 0 eqn b
Simplify it:

i1 (t ) =
1
[Vi (t ) + R2i2 (t )] eqn a
R1 + R2

Sept 28, 2014 Aerospace Engineering, KU 47 Sept 28, 2014 Aerospace Engineering, KU 48
 From previous: EXERCISE Find the
IN CLASS EXERCISE:
i2 (t ) dt + R2 (i2 (t ) i1 (t ) ) = 0
1
C
differential equation for the given
Substitute eqn a into eqn b, simplify it, will yield to system below:
1
1
( )
C
i ( t ) dt + R i ( t ) R V ( t ) + R i ( t ) =0
+
2 2 2 2 i 2 2
1
R R 2
Further simplify it:

1 RR R2
C i2 (t ) dt + 1 2 i2 (t ) =
R1 + R2 R1 + R2
Vi (t )

And after eliminating the integral, well get:

di 2 (t ) R1 + R2 1 dVi (t )
+ i2 ( t ) =
dt R1 R2 C R1 dt
If i2(0)=0, and Vi(t)=tu(t), with R1=R2=C=1, can you find i2(t) ???
Sept 28, 2014 Aerospace Engineering, KU 49 Sept 28, 2014 Aerospace Engineering, KU 50

From an R-C system From an R-L-C system

Its differential equation is given as:
Its differential equation is given as:
di(t ) 1 1 dVi (t )
+ i(t ) = 2
d i(t ) R di(t ) 1 1 dVi (t )
dt RC R dt 2
+ + i(t ) =
dt L dt LC L dt
Apply the Laplace transform yields to:
Apply the Laplace transform yields to:
1 1 R 1 1
sI (s) + I (s) = sVi (s) s 2 I ( s) + sI (s) + I (s) = sVi (s)
RC R L LC L
By heart, you should notice that: By heart, you should notice that:
I (s) I (s)
I ( s ) a output and TF = = G (s) I ( s ) a output and TF = = G (s)
Vi ( s ) Vi ( s )
1 s V i ( s ) a input 1 s
V i ( s ) a input R G(s) = L
Thus: G ( s) = Thus:
s+ 1 s +
2 R s+ 1
RC L LC
Sept 28, 2014 Aerospace Engineering, KU 51 Sept 28, 2014 Aerospace Engineering, KU 52
 From the R-L-C system Modeling of mechanical
systems
From the same circuit:

Mechanical Systems
Vc ( s )
How if I want the TF G ( s ) =
Vi ( s )
ZL(s)

By voltage divider:
Vc (s) Zc (s)
G(s) = = Translational Rotational
Zc(s) Vi (s) Zc (s) + ZL (s) + R
1 Newton
Newtons Third Law of Motion:
Where: Zc (s) = 1 For every reaction there is an equal and
cs So, the TF LC
ZL (s) = Ls G(s) = opposite reaction

s2 + R s + 1
L LC You MUST know how to draw Free Body Diagram (FBD)
Sept 28, 2014 Aerospace Engineering, KU 53 Sept 28, 2014 Aerospace Engineering, KU 54

Translational Mechanical Systems In Frequency domain

Force-
Force-Velocity Force-
Force-Displacement
(s-domain)
Component
Component Force-
Force-Velocity Force-
Force-Displacement
t
f (t ) = K v( )d f (t ) = Kx(t )
V ( s)
0 F ( s) = K F (s) = KX (s)
- spring constant
s

dx(t)
f (t) = fvv(t) f (t) = fv
dt F(s) = fvV (s) F(s) = fvsX(s)
- coefficient of
viscous friction

dv (t ) d 2 x(t )
f (t ) = M f (t ) = M
dt dt 2 F ( s ) = MsV ( s ) F ( s) = Ms2 X ( s)
M-mass
Sept 28, 2014 Aerospace Engineering, KU 55 Sept 28, 2014 Aerospace Engineering, KU 56
 Lets consider a simple system Free Body Diagram (FBD)
given below:
Step 1: Identify all possible forces on the left & right of the systems

Fdue to spring

Fdue to damper M
Fapplied force

Fdue to mass
a. Mass, spring, and damper system; b. block diagram

X (s) 1 Step 2: From the Newtons Third Law of Motion : For every reaction there is
G(s) = = 2 an equal and opposite reaction
F(s) Ms + fv s + K
Fdue to spring + Fdue to damper + Fdue to mass = Fapplied force
But, how to actually find G(s) ?????
Sept 28, 2014 Aerospace Engineering, KU 57 Sept 28, 2014 Aerospace Engineering, KU 58

Step 3: Replace all the forces with its differential equations

dx ( t ) d 2 x (t )
Now, lets consider another system
Kx ( t ) + f v +M = f (t )
dt dt 2
Step 4: Apply the Laplace transform
KX ( s ) + f v sX ( s ) + Ms 2 X ( s ) = F ( s )
Step 5: Identify input and output

X ( s ) a output
F ( s ) a input

Step 6: TF can be given as:

a. Two-degrees-of-freedom translational mechanical system
X (s) 1
G (s) = = b. block diagram
F ( s ) Ms + f v s + K
2
X 2 (s)
So, how to find G (s) =
F (s)
Sept 28, 2014 Aerospace Engineering, KU 59 Sept 28, 2014 Aerospace Engineering, KU 60
 For this case, you have to consider Step 2: From the Newtons Third Law of Motion : For every reaction there is
an equal and opposite reaction
each mass one-by-one (K1 + K2 ) X1(s) + ( fv1 + fv3 )sX1(s) + M1s2 X1(s) = K2 X2 (s) + fv3sX2 (s) + F(s)

Step 1: FBD on Mass 1

or [(K + K ) + ( f
1 2 v1 ]
+ fv3 )s + M1s2 X1(s) = [K2 + fv3s]X2 (s) + F(s)
Step 3: FBD on Mass 2

a. Forces on M1 due only to motion of M1 b. forces on M1 due only to motion of M2

a. Forces on M2 due only to motion of M2 b. forces on M2 due only to motion of M1

c. all forces on M1 (combination from a. & b.) c. all forces on M2 (combination of a. & b.)
Sept 28, 2014 Aerospace Engineering, KU 61 Sept 28, 2014 Aerospace Engineering, KU 62

Step 4: From the Newtons Third Law of Motion : For every reaction there is Step 7: Substitute eqn 3 into eqn 1, then simplify it, yields to:
an equal and opposite reaction
X 2 (s)
( K 2 + K 3 ) X 2 ( s ) + ( f v 2 + f v 3 ) sX 2 ( s ) + M 2 s 2 X 2 ( s ) = f v 3 sX 1 ( s ) + K 2 X 1 ( s ) G (s) =
F (s)
or [( K 2 ]
+ K 3 ) + ( f v 2 + f v 3 ) s + M 2 s 2 X 2 ( s ) = [ f v 3 s + K 2 ]X 1 ( s )
fv3s + K 2
G (s) =
Step 5: Now, we have 2 sets of equation, where
[M 1 s 2
+ ( f v1 + f v3 ) s + ( K 1 + K 2][
) M 2
]
2 s + ( fv2 + fv3 ) + ( K 2 + K 3 ) ( fv3s + K 2 )
2

[(K + K ) + ( f
1 2 v1 ]
+ fv3 )s + M1s2 X1(s) = [K2 + fv3s]X2 (s) + F(s) Eqn 1

[( K 2 ]
+ K 3 ) + ( f v 2 + f v 3 ) s + M 2 s 2 X 2 ( s ) = [ f v 3 s + K 2 ]X 1 ( s ) Eqn 2

Step 6: From eqn 2, well have

X 1 (s) =
[( K 2 + K3 ) + ( f v 2 + f v3 )s + M 2 s 2 ]
X 2 ( s) Eqn 3
[ f v3s + K 2 ]
Sept 28, 2014 Aerospace Engineering, KU 63 Sept 28, 2014 Aerospace Engineering, KU 64
 Rotational Mechanical Systems In Frequency domain
(s-domain)
Component Torque-
Torque-angular velocity Torque-
Torque-angular displacement
Component Torque-
Torque-angular velocity Torque-
Torque-angular displacement
t
T (t ) = K ( )d T (t ) = K (t ) ( s)
T ( s) = K T (s) = K (s)
0
s
-spring constant

d (t)
T(t) = D(t) T(t) = D T(s) = D(s) T(s) = Ds (s)
-coefficient of dt
viscous friction

d (t ) d 2 (t )
T (t ) = J T (t ) = J T ( s ) = Js ( s ) T ( s) = Js 2 ( s)
dt dt 2
-moment of inertia

Sept 28, 2014 Aerospace Engineering, KU 65 Sept 28, 2014 Aerospace Engineering, KU 66

Consider system below: By Newtons Third Law:

Js 2 ( s ) + Ds ( s ) = T ( s )
Thus, the transfer function:

(s) 1
By looking this direction
G (s) = =
Torque due to friction, Ds(s) (s) direction
T (s) Js 2 + D

FBD
J Applied Torque, T(s)

Torque due to moment of inertia, Js2(s)

Sept 28, 2014 Aerospace Engineering, KU 67 Sept 28, 2014 Aerospace Engineering, KU 68
 Find the transfer function, 2(s)/T(s), for the rotational
system shown below. The rod is supported by
bearings at either end and is undergoing torsion. A
torque is applied at the left, and the displacement is By looking this direction
measured at the right.
Consider only the inertia J1

a. Physical system
b. schematic

2 (s)
G ( s) = a. Torques on J1 due only to the motion of J1
b. torques on J1 due only to the motion of J2
T (s) c. final free-body diagram for J1(combine a. & b.)
Sept 28, 2014 Aerospace Engineering, KU 69 Sept 28, 2014 Aerospace Engineering, KU 70

From Figure (c), we apply the Newton Thirds Law

J1s 21 ( s ) + D1s1 ( s ) + K1 ( s ) = K 2 ( s ) + T ( s )

By looking this direction

OR Now, consider only the inertia J2

[J s +Ds +K] (s) = K (s) +T(s)

1
2
1 1 2
Eqn a

a. Torques on J2 due only to the motion of J2

b. torques on J2 due only to the motion of J1
c. final free-body diagram for J2
Sept 28, 2014 Aerospace Engineering, KU 71 Sept 28, 2014 Aerospace Engineering, KU 72
From Figure (c), we apply the Newton Thirds Law

J 2 s 2 2 ( s ) + D2 s 2 ( s ) + K 2 ( s ) = K1 ( s )

So, we have two sets of equations

OR

[J s ] Due to J1 Due to J2
2
+ D2 s + K 2 ( s )
[J s +Ds+K] (s) =K (s)+T(s)
Eqn b
(s) =
1
2
2
K Eqn a
1 1 1 2

(s) =
[J s 2
2
]
+ D2 s + K 2 ( s ) Eqn b
1
K
Substitute eqn b into eqn a, then simplify it, yields to:

K
G (s) =
( )(
J 1s + D1s + K J 2 s 2 + D2 s + K K 2
2
)
Sept 28, 2014 Aerospace Engineering, KU 73 Sept 28, 2014 Aerospace Engineering, KU 74

In class exercise: Find 3(s)/T(s) for Modeling a DC motor

Replaced by
the system shown below a permanent
magnet
if

Electrical Part Mechanical Part

Electromechanical part
Sept 28, 2014 Aerospace Engineering, KU 75 Sept 28, 2014 Aerospace Engineering, KU 76
 3 assumptions in modeling DC 3 assumptions in modeling DC
motor motor (cont.)
Assumption 1: The air gap flux is proportional to the field current Assumption 3: The back-emf is proportional to the velocity of the motor shaft

if = k 1i f
vb m
Assumption 2: The torque developed on the motor shaft is proportional to the
product of armature current and the air gap flux
Our objective is to find the transfer function of
ia = k 2 ia = k1k 2ia i f
m (s)
2 configurations can be made in controlling DC motor
Our focus
(1) Armature-controlled DC motor (if = constant)
(2) Field-controlled DC motor (ia = constant) Ea ( s )
Sept 28, 2014 Aerospace Engineering, KU 77 Sept 28, 2014 Aerospace Engineering, KU 78

Armature-Controlled DC motor Armature-Controlled DC motor

Electrical part:
Electro-
Electro-Mechanical part:
(cont.)
By applying KVL, we obtain the differential equation:
Relationship between Torque & armature current
di
ea (t ) = La a + Ra ia + vb m = k1k 2i f ia m = k m I a (s ) Equation 3
dt
Take the Laplace transform yields to:
k
Ea ( s ) = (La s + Ra )I a ( s ) + Vb ( s )
m
Equation 1 Motor torque constant

Relationship between back-emf & angular speed

Mechanical part:
vb m vb = kb m
d 2 m d m
m (t ) = J m 2
+ Dm
dt dt Back-emf constant
In Laplace

m ( s ) = J m s 2 m + D m s m = [J m s 2 + D m s ] m Equation 2 Vb = kb s m Equation 4

Sept 28, 2014 Aerospace Engineering, KU 79 Sept 28, 2014 Aerospace Engineering, KU 80
 G3 ( s )
Ia (s) =
1
[E (s) Vb (s)] G1 ( s ) G2 ( s )
(Las + Ra ) a Equation 1

m = km Ia Equation 3
m (s)
1 Ea (s) + 1
I a (s ) m (s) 1
m = m (s)
[
J m s 2 + Dm s ] Equation 2
sL a + R a
km
s2Jm + Dms

Vb = k b s m Equation 4
Vb (s)

We can draw in term of block diagram sk b

m (s)
Ea (s) + 1
I a (s ) m (s) 1
km H (s )
sL a + R a s2Jm + Dms
Thus, the transfer function for DC motor
Vb (s)

m (s) G1 ( s )G2 ( s )G3 ( s )

sk b =
Ea ( s ) 1 + G1 ( s )G2 ( s )G3 ( s ) H ( s )
Sept 28, 2014 Aerospace Engineering, KU 81 Sept 28, 2014 Aerospace Engineering, KU 82

In a nut shell
What weve learned so far ???
Electrical components & mechanical
components of the DC motor
Relationship between the electrical &
mechanical components of the DC motor
Block diagram for the DC motor
Transfer function of the DC motor

Sept 28, 2014 Aerospace Engineering, KU 83