0% found this document useful (0 votes)
18 views23 pages

Week 4

The document discusses transfer function (TF) models for dynamic systems. TF models characterize the dynamic relationship between an output and input variable of a linear system. The document provides an illustrative example of deriving TF models for a CSTR process. It also discusses properties of TF models such as order, gain, time constants, and how TF models can be combined for multiple input systems. Linearization is introduced as a way to approximate nonlinear dynamic models with linear TF models around an operating point.

Uploaded by

candan.erken
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
18 views23 pages

Week 4

The document discusses transfer function (TF) models for dynamic systems. TF models characterize the dynamic relationship between an output and input variable of a linear system. The document provides an illustrative example of deriving TF models for a CSTR process. It also discusses properties of TF models such as order, gain, time constants, and how TF models can be combined for multiple input systems. Linearization is introduced as a way to approximate nonlinear dynamic models with linear TF models around an operating point.

Uploaded by

candan.erken
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 23

Transfer Functions

• In KMM342E: Dynamic models considered in ODE form


• In KMM411E: Alternative model form
based on Laplace transforms
Transfer Function Model
• Note: Both types of models can be used to determine the dynamic
behavior of a process
• TF: Characterize dynamic relationship of 2 process variables
• A dependent (output) variable Y ( s)
G ( s) 
• An independent (input) variable U ( s)
• Example: CSTR Important:
• IV: Ffeed → Cause Applicable to linear systems.
• OV: Cexit → Effect/Result What if the system is nonlinear?
1
Illustrative Example

• Mass balance:

• Component mass balance for A:

• :

Assumptions:
1. ρ & V = const.
2. w1, w2, w = const.

2
Illustrative Example
Case I: x1 ≠ const. & x2 = const.
• Aim: Derivation of TF model between x & x1
Component mass balance (steady-state):

Dynamic – Steady-state:

• Deviation (perturbation) variables:


• Advantage: The initial condition term becomes zero
Simplifies the later analysis (Independent from IC)

Laplace transform:

3
IE: Case I
• Assumption: Initially system is at SS

• Rearrange in TF form between output and input

• Standart form:

• Process gain:

• Time constant:

• Physical interpretations of KP & τI

4
IE: Case II
Case II: x1 & x2 ≠ const.
• Two input variables: x1 & x2
• Derivation of 2 TF models between x & x1
x & x2
Component mass balance (steady-state):

Dynamic – Steady-state:

Laplace transform:

Assumption: Initially system is at SS

5
IE: Case II
• Rearrange in TF form between output and inputs

• Standart form:

• Process gains:

• Time constants:

• In order to derive TF between x & x1


• Assumption: x2 at SS value
• In order to derive TF between x & x2
• Assumption: x1 at SS value
6
IE: Case II
• G1(s) & G2(s): 1st order TF
• Denominators are 1st order in Laplace variable s
Three important aspects:
1. Principle of Superposition for linear models
• Effects of individual IVs on OV are additive
2. Simultaneous changes in both inputs can be analyzed
3. Using TF model:
• Output response for a specified input change can be calculated
• Example X I ( s )  G1 ( s ) X 1I ( s )
• If we know x1I (t )
• X 1I ( s ) can be determined
I
• x (t ) can be derived
7
Example V
• Consider Case I
• SS conditions: w1 = 600 kg/min, w2 = 2 kg/min, x1 = 0.05, x2 = 1
• ρ & V = const: ρ = 900 kg/m3, V = 2 m3
a) Calculate nominal exit concentration:

b) x1 suddenly change from 0.05 to 0.075 at t = 0


Derive an expression for response x(t):

8
Properties of TFs
1. Direct calculation of SS output change for a sustained input
change

9
Properties of TFs
2. Order of denominator polynomial (in s) = order of equivalent DE
• Form of the general linear nth order DE:
dny dy n1 dy d mu
an  an1  K a1  a0 y  bm m 
dt n
dt n1 dt dt
d m1u du
bm1 m1
 K  b1  b0u (4-39)
dt dt

• TF (all initial conditions = 0):


m

Y  s  i 0
bi si
G s   (4-40)
U s n
 ai si
i 0

• Order of TF is defined to be order of denominator polynomial.


10
Properties of TFs
• SS gain:

• Characteristic (denominator) polynomial:

• Can be factored into product:

• Time constant form:


• Provide information about speed and qualitative features of
system response.

11
Properties of TFs
3. Order of numerator and denominator polynomials are restricted
by physical reasons:

• Suppose a real process is modeled by

• “Physical realizability” condition:

• m > 0: TFs exhibit “numerator dynamics”


• m = 0: Many important cases
12
Properties of TFs
4. Additive property:
• A single output might be affected by more than one input

• Total output change:

13
Properties of TFs
5. Multiplicative property:
• Valid for sequential processes/process elements:

• Important in designing process control systems


• Since process units will be connected

14
Example VI
Stirred-tank heating process:
w = 200 lb/min, ρ = 62.4 lb/ft3,
C = 0.32 Btu/lbF, V = 1.60 ft3
SS values: Ti = 70 F, Q = 1920 Btu/min
Instantly: Ti = 90 F, Q = 1600 Btu/min
Question: Response of T?

dT
V C  wC Ti  T   Q (1)
dt
0  wC Ti  T   Q (3)

 wC Ti  Ti   T  T    Q  Q 
dT
V C (4)
dt

15
Example VI Cont.
Stirred-tank heating process:

16
Example VII
Two liquid surge tanks in series:
• Outlet flow from each tank is
proportional to height of liquid in tank
• Assumptions:
• Cross-sectional areas: A1 & A2
• Valve resistances: R1 & R2
• Question: TF relating q2 to qi?

17
Example VII Cont.
Two liquid surge tanks in series:

• Block diagram representation

18
Linearization
• So far, we have emphasized linear models
• Can be directly transformed into TF models
• Most physical processes and physical models are nonlinear.
- But over a small range of operating conditions, the behavior
may be approximately linear.
- Linear approximations can be useful
- Especially for purpose of analysis
• Approximate linear models can be obtained analytically by a
method called “linearization”.
• It is based on a Taylor Series Expansion of a nonlinear function
around a specified operating point.
19
Linearization
• Consider a nonlinear, dynamic model relating two process
variables, u and y:

• Perform Taylor Series Expansion around reference points: u  u


and truncate after the first order terms, yy

• Note that the partial derivative terms are actually constants because
they have been evaluated at the nominal operating point (at SS),
• In terms of deviation variables:

20
Example I
• Consider the stirred-tank blending system example:
• Assumptions:
• V = const & x2 = 1
• Derive TFs relating output variable (x) to input variables (w1, w2, x1)

• What is the reason of nonlinearity?

21
Example

22
Example II

• Consider a tank liquid-level system


• Outflow passes through a valve
• Valve discharge rate is related to
square root of liquid level

• Derive appropriate dynamic model for process by linearization

23

You might also like