2.

Mathematical Modeling of Physical Systems

Definition:
 A physical system is a collection of physical
objects connected in some designed pattern to
serve some prescribed objective.
Example:
Industrial Plant, electronic amplifier, governing
mechanism of steam turbines, satellites, transportation
machineries etc.
N.B. No physical system can be represented in its
full physical details. Hence ideal assumptions are
always made for the purpose of analysis and
synthesis of systems. 1
• A Physical Model is an idealized representation of
physical system.
• Complexity of the model depends on the type of
study to be carried out and on the level of
accuracy needed.
• N.B. Making a model more complex is not
proportional or commensurate with the gain of
accuracy.

• Making simple and manipulative mathematical

model which gives a greater insight into the
dynamic behavior of the physical system is
advisable
2
• A mathematical model is developed from the
physical model using appropriate physical laws.
• Mathematical model can be:
 Scalar differential equation
 State variable vector-matrix differential equation
• When the mathematical model is solved for
various input conditions, then the result
represents the dynamic response of the physical
system.
• Linear Mathematical model: obeys the principles
of super-position and homogeneity
1 x1( t )   2 x2 ( t )  1 y1( t )   2 y2 ( t )
3
Inputs Outputs
• For analysis and synthesis of the mathematical
models available mathematical tools have to be
utilized.
Example: Laplace, Fourier Transforms & Linearization
• Unfortunately no physical system in nature is
perfectly linear.
• Certain assumptions must always be considered
to get a linear model which is a compromise
between the simplicity of the mathematical
model and the accuracy of results obtained.

4
2.1. Modeling of Mechanical Systems
• The building blocks are:
 Spring: represents the stiffness of a system (energy
storage media)
F  kx or T  k
 Damper (dashpot) : represents the force which
opposes motion (frictional or damping effects)
F  Cx or T  C
 Mass: represents the inertia or resistance to
acceleration
m F  mx or T  I
• This building blocks represent three essential
phenomena which occur generally in
mechanical systems.
5
• Dampers (with equivalent damping constant)
will be modeled to represent either of the
following:
 Coulomb Friction Force (Dry Friction): force of
sliding friction between dry surfaces and it is
substantially constant.
 Viscous Friction Force: force of friction between
moving surfaces separated by viscous fluid or
between solid surface and a fluid medium.

 Structural or Hysteretic Damping Force: force

due to internal friction in the material as
relative slipping or sliding of internal planes
occur during deformation.
6
• Energy is needed:
 to stretch the spring, hence an elastic potential
energy will be stored in the spring
1 2
Ve  kx
2
 to accelerate the mass and hence a kinetic
energy will be accumulated in the moving mass
1 2
T  mx
2
 to move the piston against the reaction force
and hence energy will be dissipated
x2

U 12    Cxdx
x1 7
• Simple Model:
Input Output Input Output
Spring Dashpot
F or T x or  F or T x or 

• Such models are applied for:

Models of machine tools mounted on the ground
and is a basis for studying the effects of ground
disturbances on the displacement of the machine
bed (Base Excitation).

Models of the wheel and its suspension for a car or

track and can be used for the study of the behavior
that could be expected when the vehicle is driven
over a rough road and hence used as a basis for the
design of the vehicle suspension.
8
 Gear Train Modeling
• Used in control systems to attain the mechanical
matching of motor to load.
Example: Servo-motor operates at high speed
but low torque.
• Multiplication of Torque and reduction of speed is
achieved by the application of gear trains.
Primary Gear
N1
Input Torque
from Motor θ1
T1
TM
Input Shaft θ2
I1, C1 Load Torque

TL
Secondary Gear
N2 Load Shaft
9
I2, C2
• Gear 1
I11  C11  T1  TM .....................(1)
Where T1 is the load Torque on Gear 1 due to the
rest of Gear Train
• Gear 2 I 2 2  C2 2  TL  T2 ......................(2)
Where T2 is the load transmitted to Gear 2 via Gear 1
Using the relation
 2 N1
1r1  2 r2  1 N1  2 N2  
1 N 2
and assuming there is no loss in power transfer
T1 2 N1
T11  T22     N11  N22
T2 1 N2
10
• Differentiating the above twice:
N11  N 2 2 and N11  N 2 2
• Replacing in to equations 1 and 2 and rearranging:
 N1 
I11  C11  T2    TM
 N2 
 I11  C11 
N1
N2
 
I 22  C22  TL  TM ......(3)
• Eliminating θ2:
  N1 
2
   N1 
2
 N1
 I1    I 2  1   C1    C2  1  TL  TM
  N2     N2   N2
   
N1 I1eq,C1eq
 I1eq1  C1eq1  TL  TM N1
TL
N2 TM N 2 11
• Similarly eliminating θ1 from equation 3:
2 2
 N2   N2 
I 2eq  I2    I1 and C2eq  C2    C1
 N1   N1 
• Hence, the Torque equation to the load shaft will
then be expressed as:
 N2 
I 2eq2  C2eq2  TL    TM
 N1 
I2eq,C2eq
 N2 
  TM TL
 N1 
12
2.2. Equation of Electrical Networks

• The basic building blocks are:

 Inductor
 Capacitors
 Resistors i L

1. Inductor: V(t)

 Inductance is the property of a coil of wire that

opposes a change in current

di
V(t )  L where L is inductance
dt
1
 i   Vdt
L 13
 C
2. Capacitor:
• Capacitance is the measure of energy storing
ability of a capacitor in the form of electric field
• The potential difference across it depends on the
charge q on the capacitor plates at the instant
concerned
q
V  , where C is the capacitance
C
dq 1 dV
but i   q   i.dt  V   idt  i  C
dt C dt
3. Resistor: i

V
• A resistor dissipates energy
V  IR      Ohms Law 14
Kirchoff’s Law:
• Describes how the electrical building blocks can
be combined

1. KCL: the algebraic sum of the currents at the

junction is zero.

i  0
2. KVL: In a closed circuit or loop, the algebraic
sum of the potential difference across each part
of the circuit is equal to the applied emf.

V  0 15
• Mesh and node analysis are used to analyze circuits
Mesh Current Method:
a. Assign a current in CW direction
b. Indicate the voltage drop polarities
c. Apply KVL around each closed loop
d. Solve the resulting equations
Nodal Voltage Method:
a. Determine the number of nodes
b. Select one node as a reference. Assign voltage
designation to each node where voltage is unknown
c. Assign current at each node
d. Apply KCL to each node
e. Express the current equations in terms of voltages
and solve the equations.
16
Examples

17
2.3. Transfer Functions
• Defn: The transfer function of a linear time-
invariant (LTI) system is defined as the ratio of
the Laplace transform of the output variable to
the Laplace transform of the input variable, with
all initial conditions assumed to be zero.

θi(s) θ0(s)
G(s)
Input Output

0 ( s )
Transfer Function  G( s ) 
i ( s )
18
• If θi is related to θo by the following differential
equation:
d 20 d 0
a2 2  a1  a00  b1i
dt dt
where a2 ,a1 ,a0 and b1 are constants.
• The Laplace Transform becomes:
a2 s 0 ( s )  a1 s0 ( s )  a00 ( s )  b1i ( s )
2

And hence the Transfer function is:

0 ( s ) b1
TF  
i ( s ) a2 s  a1 s  a0
2

19
2.4. Block diagram and Signal Flow Graph
• Transfer function is an expression in S domain,
relating the output and input of the LTI system in
terms of the system parameters and is
independent of the input.

• TF doesn’t give any information concerning the

internal structure of the system.

• Thus, the system can be represented by a block, a

short-hand pictorial representation of the cause
and effect relationship between input and output
of the system.
20
• A complex system comprising of several elements is
represented by the interconnection of the blocks for
individual elements.
R(s) C( s )
G(s)
C(s)
G( s ) 
Input Output R( s )
• Signals entering the block represents the input
• Signals leaving the block represents the output
• The block stands for the transfer function.
• Information flow in one direction, from input to
output.
• A summery point is where signals are algebraically
+
added together. -
+
- 21
• Take-off Point is represented by junction.

• Block is written with transfer function written

inside it.

• Forward Path is the path that the signal passes

the direction input-to-output for the system as a
whole

• Feedback Path is for those elements through

which a signal passes when being fed back from
the output towards the input.

22
 Blocks in Series
G (s)

θi(s) θo(s)
G1(s) G2(s) G3(s)
Input θ1(s) θ2(s) Output

• The overall transfer function be G(s).

1 ( s ) 2 ( s ) o ( s )
G1 ( s )  ;G2 ( s )  ;G3 ( s )  ;
i ( s ) 1 ( s ) 2 ( s )
 o ( s ) 1 ( s )  2 ( s )  o ( s )
G( s )   . .  G1 ( s )G2 ( s )G3 ( s )
 i ( s )  i ( s ) 1 ( s )  2 ( s )
• For n-number of blocks, the overall transfer
function is given by the product of the transfer
function of each block. 23
Block with feedback Loop
θi(s) + e θo(s)
G1(s) θi(s) G1 ( s ) θo(s)
-
1  G1 ( s )H( s )
H(s)

Block in Parallel θi(s) + θo(s)

G1(s)
+

G2(s)

o ( s )  i ( s )G1( s )  i ( s )G2 ( s )
 o ( s )  i ( s )[ G1( s )  G2 ( s )]  i ( s )G( s )
o ( s )
 G( S )   G1 ( s )  G2 ( s )
i ( s ) 24
Multiple input, Multiple output system
• For the case of a linear system, each input can be
treated independently and the complete output
of the system will be obtained by superposition.

U(s)
R(s) + e C(s)
G1(s) G2(s)
-
H(s)

Step 1: Assume U(s)=0, hence the system will look like:

R(s) + e CR(s)
G1(s) G2(s)
-

H(s) 25
• The output will then be:
G1G2
CR ( s )  R( s )
1  G1G2 H
Step 2: Assume R(s)=0 and the system will look like:
U(s) + e Cu(s)
G2(s)
-
G1(s) H(s)

• The output will then be:

G2
Cu ( s )  .U ( s )
1  G1G2 H
26
Step 3: Super position of both outputs
C( s )  CR ( s )  Cu ( s )
G1G2 G2
 C( s )  .R( s )  .U ( s )
1  G1G2 H 1  G1G2 H
G2
 C( s )  G1 R( s )  U( s )
1  G1G2 H
For open-loop system, if it has r inputs and m outputs as
shown below, the ith output is given by:
R1(S) C1(S)
R2(S) Gij(s) C2(S)

Rr(S) Cm(S)
r
Ci ( s )   Gij ( s )R j ( s ), i  1,2,...,m 27
j 1
Transfer Function for Liquid Level Control System
Q

T-2
T-1 H1 Q1 Q2
H2
P-1 P-2
• Because of interaction of the tanks, the complete
transfer function can not be obtained by
multiplying individual transfer functions of the
tanks.
 Q1 and Q2 : Steady state outflow rates
 R1 and R2 : Resistance of pipes 1 and 2
 C1 and C2 : Liquid Capacitance of tank 1 and 2
 H1 and H2 : Heads in tank 1 and 2
 ΔQ : Small deviation in the flow rate Q
 ΔH1, ΔH2, ΔQ1 and ΔQ2 : Small deviations of all
parameters from their steady state values 28
Signal Flow Graph (SFG)
• For complicated systems, BD reduction process
will be tedious and time consuming .
• In such cases SFG is an alternative method and is
developed by S.J. Mason.
• SFG does not require any reduction process due
to availability of a flow graph gain formulas.
• The formulas relate the input and output system
variables.
R 1 E G C R 1 E G C 1 C
Input Input Output
Node -1 H Node -1 H Node
B B
29
Signal flow terms
• Node: represents a system variable which is equal
to the sum of all incoming signals at the nodes.
 Outgoing signals from the node don’t affect the value
of the node variable.
 R, E and C are nodes.

• Branch: a line along which a signal travels from

one node to another in the direction of arrow.
 In this process the signals get multiplied by the gain or
transmittance of the branch.
 Signals from node E to node C = GE, where G is the
branch transmittance or gain.

30
• Input Node (source): node only with outgoing
branches.
 R is an input node
• Output Node (Sink): node only with incoming
branches.
 Achieved by introducing a branch with unit gain.
• Path: it is the traversal of connected branches in
the direction of the branch arrows such that no
node is traversed more than once.
• Forward Path: from input node to the output
node, R-E-C is an example
• Loop: a path which originates and terminates at
the same node, E-C-B-E is an example.

31
• Non touching Loops: loops which don’t posses
any common node.
• Forward Path Gain (FPG): is the product of
branch gains encountered in traversing a forward
path.
 FPG of path REC is G.
• Loop Gain(LG): is the product of the branch gains
encountered in traversing the loop
 LG of loop ECBE is -GH

R 1 E G C R 1 E G C 1 C
Input Input Output
Node -1 H Node -1 H Node
B B
32
 Construction of Signal Flow Graph
• It is constructed from its describing equations
• Example:
– Let a system is described by the following set of
equations
x2  a12 x1  a32 x3  a42 x4  a52 x5
x3  a23 x2
x4  a34 x3  a44 x4
x5  a35 x3  a45 x4
where x1 is the input variable and
x5 is the output variable 33
Steps:
1-
x1 x2 x3 x4 x5
a42

a12 a32
2- x2
x1 x3 x4 x5

a52
a23
3- x2
x1 x3 x4 x5

a44
a34
4- x2
x1 x3 x4 x5

a45
5-
x1 x2 x3 x4 x5

a35 34
• Combining the above five relations
a42 a44
a12 a23 a34
6- x2
x1 x3 x4 a45 x5
a32
a35
a52

• The overall gain from input to output may be

obtained by Mason’s gain formula.
• Definition:
– The relationship between an input variable
and an output variable of a signal flow graph
is given by the net gain between the input and
output nodes and is known as the OVERALL
GAIN of the system. 35
• Mason’s gain formula for the determination of
the overall system gain is given by:
1
T

 P
k
k k

where Pk is path gain of k th forward path

 is determinant of the graph
 = 1-(sum of loop gains of all individual loops)+
(sum of gain products of all possible combinations
of two non-touching loops)-(sum of gain products
of all possible combinations of three non touching
loops)+...
36
• That is:
  1   Pm1   Pm2  Pm3  ...
m m m

where Pmr  Gain product of mth possible combination of

r non-toughing loops
k = The value of  for that part of the graph not
touching the k th forward path
T = Overall gain of the system

• Exercise:
– Find the overall gain of the above example using
Mason’s gain formula

37
Application:
• Consider the following model of an electro-
mechanical system
• The system objective is to move the load at a
desired speed.

Ra
Amplifier
if

TD
er e kA ea Ia TM Load
ω J, C
DC Motor
Tachometer
et (Conventional
DC Generator)

38
• DC tachometer is a conventional DC generator
with permanent magnetic excitation (PME)
• The DC tachometer gives a voltage et proportional
to the output speed ω,
• er-et=e is the difference signal used after
amplification to control ia of the motor such that
the motor acquires the desired speed ωo.
• The open circuit voltage will then be :
et  k1  et  
 volt.s 
 et  kt , kt   ...........................( 1 )
 rad 
where kt is called the tachometer constant 39
• The voltage at the armature terminals of the motor
is given by:
ea  k Ae
where k A is the transfer function of the amplifier
 ea  k A ( er  et ).....................(2)
• For the armature circuit (applying KVL):
Ra I a  kb  ea .................(3)
where kb is the back EMF of the motor
when there is constant field current
i.e kb  ka 40
• For constant field current, the torque
developed will be:
60
TM  kT ia [ kT  k M  ,k M  ka ]
2
Hence, the torque equation of the system
is:
d
J  C  TD  TM  kT ia ...............( 4 )
dt
• Taking the Laplace transform of the above 4
equations we will get:
41
( 1 ) Et ( s )  kt( s )
( 2 ) Ea ( s )  k A [ Er ( s )  Et ( s )]
 Ea ( s )  k A [ Er ( s )  kt( s )]
( 3 ) Ra I a ( s )  kb( s )  Ea ( s )
1
 Ia ( s )  [ Ea ( s )  kb( s )]
Ra
( 4 ) kT I a ( s )  ( Js  C )( s )  TD ( s )
 ( s )( Js  C )  TM ( s )  TD ( s )
1
 ( s )  [ kT I a ( s )  TD ( s )]
Js  C 42
Taking the nodes:
1. Er ( s ),2. Ea ( s ),3. I a ( s ),4. ( s ) and 5. TD ( s )
• The SFG representation of the above LT equations
will then be :
TD(s)
1 1
Er(s) 1 KA Ea(s) Ra Ia(s) K
T
-1
Js  C 1 ( s )
( s )
-Kb

-Kt

• This is a multiple (2) input system and should be

treated by a method of superposition. 43
End of Chapter Two

44