MODEL HIERARCHY AND

ITS IMPORTANCE
IN ANALYSIS

Mathematical Modeling in Chemical Engineering

 Summary of Model Building Process
2

1. Draw a sketch of the system to be modeled and

label/define various geometric, physical & chemical
quantities.
2. Carefully select the important dependent (response)
variables.
3. Select the possible independent variables (e.g., z, t),
changes in which must necessarily affect the
dependent variables.
4. List the parameters (physical constants, physical size
and shape) that are expected to be important; also
note the possibility of non-constant parameters [e.g.
viscosity changing with temperature, (T)].

Mathematical Modeling in Chemical Engineering

 Summary of Model Building Process
3

5. Draw a sketch of the expected behavior of the

dependent variable(s), such as the “expected”
temperature profile.
6. Establish a “control volume” for the system to be
modeled; sketch the element and indicate all inflow-
outflow paths.
7. Write the “conservation law” for the volume
element: Express flux and reaction rate terms using
general symbols.
8. After rearrangement to the proper differential
format, invoke the fundamental lemma of calculus
to produce a differential equation.

Mathematical Modeling in Chemical Engineering

 Summary of Model Building Process
4

9. Introduce specific forms of flux (e.g., Nr=-DC/r)

and rate (RA=-kCA); note, the opposite of generation
is depletion, so when a species is depleted then its
loss rate must be entered with the appropriate sign
in the conservation law (i.e., replace “ + generation”
with “ - depletion”).
10. Write out all possibilities for boundary values of the
dependent variables; the choice among these will be
made in conjunction with the solution method
selected for the defining (differential) equation.
11. Search out solution methods, and consider possible
approximations for: (i) the defining equation, (ii) an
acceptable final solution.

Mathematical Modeling in Chemical Engineering

 Model Hierarchy and
Its Importance in Analysis
5
Depending on the real purposes of the modeling effort,
the scope and depth of these decisions will determine
the complexity of the mathematical description of a
process.
Hierarchy of models

Mathematical complexity
lowest
Black box level
All possible transport
processes and all other highest
concepts are taken into level
account

Hierarchy is dictated by the number of rules (transport, thermo., etc.)

Mathematical Modeling in Chemical Engineering
 Heat removal from a solvent bath

Heat is removed from a bath of hot solvent by

immersing steel rod into the bath and allowing the
heat to dissipate from the hot solvent bath through
the rod and hence to the atmosphere.
Mathematical Modeling in Chemical Engineering 6
 Level 1: Simplest model
Let us assume:
• The rod temperature is uniform, that is, from the
bath to the atmosphere.
• Ignore heat transfer at the two flat ends of the rod.
• Overall heat transfer coefficients are known and
constant.
• No solvent evaporates from the solvent-air
interface.
• Steady-state

Mathematical Modeling in Chemical Engineering 7

 Steady-state heat balance:
hL (2RL1 )(T1  T)  hG (2RL2 )(T  T0 )  0
Solution is:
(T0  T1 ) hLL1
T where 
(1  ) hGL 2

If  >> 1 then T is very near T1.

The rate of heat transfer is:

hL 2RL 1 1
Q (T1  T0 )  2R(T1  T0 )
(1   )  1 1 
  
 hLL1 hGL 2 
If  >> 1  Q  2Rh GL 2 (T1  T0 )

Mathematical Modeling in Chemical Engineering 8

 Level 2: Relax the first assumption

Notice
Assume that the rod
location temperature below the
of
origin solvent liquid surface
is uniform at a value
T1.

Is this reasonable?
Yes. The liquid has a
much higher thermal
conductivity than air.

Mathematical Modeling in Chemical Engineering 9

 Heat Balance around a differential volume element
of thickness x:
R 2 q x  R 2 q x  x  2RxhG (T  T0 )  0

Divide by R2x and take limit as x  0

dq 2
 d2T 2hG
 hG (T  T0 )  0 k 2  (T  T0 )
dx R dx R

Boundary conditions:
@ x  0, T  T1
dT
@ x  L2 , 0 (Flux from the top surface is small)
dx

Mathematical Modeling in Chemical Engineering 10

 coshm(L2  x )
Solution is: T  T0  (T1 - T0 )
cosh(mL2 )
2hG
where m
Rk

dT
The rate of heat loss is: Q  R k2
dx x 0

tanh(mL2 )
Q  2RhGL2(T1  T0 ) where 
mL2

Mathematical Modeling in Chemical Engineering 11

A plot of  versus mL2 If mL2 << 1, then   1

When is this
physically achieved?

• The rod k is large

• L2 is short

Then Q  2Rh GL 2 (T1  T0 )  Identical to first model

Therefore, first model is valid only when mL2 << 1

Mathematical Modeling in Chemical Engineering 12

 Level 3: Relax the first assumption completely

We now allow temperature gradients in the rod for

segments above and below the solvent-air
interface.
Let TI = temperature of the rod below the
interface
TII = temperature of the rod above the
interface

1-D heat balances for the two rod segments:

d2TI 2hL I d2TII 2hG II

 (T  T1 ) 2
 (T  T0 )
dx 2
Rk dx Rk
Mathematical Modeling in Chemical Engineering 13
 Boundary conditions:
dT I
@ x  L1 0
dx
Zero flux at the flat
dT II ends of the rod
@ x  L2 0
dx

@x  0 T I  T II Continuity of
temperature

dT I dT II
@x  0  Continuity of flux
dx dx

Mathematical Modeling in Chemical Engineering 14

 Solution is: TI  T1  A cosh[ n( x  L1 )]
TII  T0  B cosh[ m(L 2  x )]

2hG 2hL
where m and n
Rk Rk

(T1  T0 )
A
n sinh( nL 1 )
[cosh( mL 1 )  cosh( mL 2 )]
m sinh( mL 2 )

(T1  T0 )
B
m sinh( mL 2 )
[cosh( mL 2 )  cosh( nL 1 )]
n sinh( nL 1 )

Mathematical Modeling in Chemical Engineering 15

 I
dT
Rate of heat transfer is: Q   R 2k
dx x 0
(T1  T0 )
Q  2RhGL 2
m tanh(mL2 ) tanh(mL2 )
[1  ] where  
n tanh(nL1 ) mL2

In the Level 2 model, we have obtained:

Q  2RhGL2(T1  T0 )

m tanh( mL 2 )
Level 2 model is valid if  1
n tanh( nL 1 )

Mainly controlled by m/n = (hG/hL)1/2 (always less than 1)

Mathematical Modeling in Chemical Engineering 16
 Higher levels of modeling yield more information
about the system and hence provide needed
criteria to validate the model one level lower.

What is the next level of

modeling?

At what point and under what

conditions do radial gradients
become significant?

Mathematical Modeling in Chemical Engineering 17

 Level 4: Include radial heat conduction

When is this important?

If the rod diameter is

large relative to length

Assume that there is no

resistance to heat flow
below solvent interface
x  0  T = T1
Schematic diagram of
shell for heat balance

Mathematical Modeling in Chemical Engineering 18

 Heat balance around a cylindrical differential volume in
radial and axial directions:
(2rxq r ) r  (2rxq r ) r r  (2rrq x ) x  (2rrq x ) x x  0

Divide by 2rx and take limit as r0 and x0

 qx
 (rqr )  r 0
r x

Insert the two forms of Fourier’s laws:

T T
qr  k qx  k
r x

1   T   2T  Here we have assumed

k r  2 
0 that k of the rod is
 r r  r  x  isotropic and constant

Mathematical Modeling in Chemical Engineering 19

 Boundary conditions:
T Symmetry at the center of
@r  0 0
r the rod

T
@r  R -k  hG (T - T0 ) Newton’s cooling condition
r

@x  0 T  T1 Constant temperature

T
@ x  L2 0 No heat flow at the flat end
x

Mathematical Modeling in Chemical Engineering 20

 Partial differential equation
 need to simplify the notation and to
deduce the parameters
Introduce non-dimensional variables :
T  To r x
u ,  , 
T1  To R L2

Two dimensionless parameters will arise :

R hGR
    , Bi  (Biot number)
 L2  k

Remember Biot Number?

Mathematical Modeling in Chemical Engineering 21

 Dimensionless equations are:

1   u  2  u 2
     0
      2

u u
@   0; 0 @   1;  Biu
 

u
@   0; u 1 @   1; 0


Mathematical Modeling in Chemical Engineering 22

 Solution is:
 n 
cosh  (1   )
T  To 
 1, K n   
u  K n ()
T1  To n1  K n , K n   n 
cosh  


where K n ()  Jo (n) & nJ1(n )  BiJo (n )

J1(n )
 1, K n  J0 and J1 are
n tabulated
 relations called
 n  
2
J (n ) 2
Bessel functions
 K n , K n  1    
1

2   Bi  

Mathematical Modeling in Chemical Engineering 23

 Rate of heat transfer can be calculated using heat
flux at x=0 but also accounting for radial variation
of temperature so that elemental area is 2rdr:
R  T 
Q  k x  2rdr
0
  x 0

In dimensionless form :

2R 2k 1 u(0) 
Q (T1  T0 )   d
L2 0
  

2R 2k(T1  T0 )  n  1, K n  2  n 
Q
L2
n 1  K n , K n 
tanh  


Mathematical Modeling in Chemical Engineering 24

 For small Biot number (Bi << 1) :
nJ1(n )  BiJo (n )  1  (2Bi)1/ 2

smallest eigenvalue

Then the solution for Q becomes :

Level 2
Q  2Rh GL 2(T1  T0 )
solution

Radial heat conduction can be neglected when

Bi  1
Mathematical Modeling in Chemical Engineering 25
 When is a model of a process good enough?

It can only be answered fully when

economics of design and practicality
are taken into account.

The user must decide

when tractability is
more important than
precision.

Mathematical Modeling in Chemical Engineering 26