CE30240 Advanced Principles of Chemical Engineering Non-Newtonian Fluids

1. NON-NEWTONIAN FLUIDS

1.1 Classification of Fluid Behaviour

In all the work we have done in Years 1 and 2 Fluid Mechanics, we have assumed that the
viscosity of the fluid is independent of its velocity, shear rate, shear stress, time of shearing
and kinematic history, i.e. it is a Newtonian fluid.

Definition of a Newtonian Fluid (e.g. water, molten metals and low Mw polymers)

We recall that for a Newtonian fluid

 yx
Shear stress [Pa] = Viscosity [kg/m s] × Shear rate [1/s]
plane direction
du
 yx =  =   yx
dy
For a Newtonian fluid the viscosity can depend upon three factors, namely, temperature,
pressure and chemical composition of the fluid.   fn (  and  )

Definition of Non-Newtonian Fluids

There are many fluids known as non-Newtonian fluids for which the viscosity depends on
flow conditions such as flow geometry, shear rate (or stress) developed within the fluid, time
of shearing, kinematic history of the sample. Non-Newtonian fluids are divided into different
categories (though somewhat arbitrary):

1. Substances for which the rate of shear,  , (or viscosity) is dependent only on the current
value of the shear stress,  , or vice versa; this class of materials is variously known as
purely viscous, time independent, or generalised Newtonian fluids.  = fn (  and  )

2. More complex materials for which the relation between the shear stress and the shear
rate also depends upon the duration of shearing, the previous kinematic history, etc;
these are known as time-dependent fluids.  = fn (  , and t )

3. Materials exhibiting combined characteristics of both an elastic solid and a viscous fluid
and showing partial elastic and recoil recovery after deformation, the so-called
viscoelastic fluids.  = fn ( elastic + viscous components )
___________________________________________________________________________
9 JC
CE30240 Advanced Principles of Chemical Engineering Non-Newtonian Fluids

Purely viscous fluids (Inelastic or Plastic)

• Pseudoplastic fluids: shear thinning (e.g. yogurt) - majority of NNF
• Dilatant fluids: shear thickening (e.g. paint, starch) - high solids content
• Bingham (plastic) fluid: solid until yield stress is reached (e.g. toothpaste)

Time dependent fluids

• Thixotropic fluids: thins with time of shearing (e.g. gels, clay) - popular
• Rheopectic fluids: thickens with time of shearing (e.g. lubricants)

Viscoelastic fluids
• Viscoelastic fluids: viscous and elastic behaviour (e.g. paste extrusion) - complex

This classification is quite arbitrary in that most real materials often display a combination of
two or even all the three types of non-Newtonian characteristics.

For each group of fluids, we will attempt to find a constitutive equation which adequately
models the behaviour of the fluids.

 = fn (  )

Notes:
(i) the aim of studies of non-Newtonian fluids is to develop expressions for the velocity
profile and volumetric flow rate when these fluids flow through pipes and other
equipment. Pressure drop can be determined using these results;

(ii) Note that in some texts the shear stress is given the opposite sign. This depends
whether one looks at the shear stress on the fluid or on the wall.

___________________________________________________________________________
10 JC
CE30240 Advanced Principles of Chemical Engineering Non-Newtonian Fluids

There are a number methods of presenting  = fn ( ) data, two of which are:

(a) shear stress in terms of shear rate

Shear thickening
such as starch Newtonian such as
water and most
organic solvents

Shear thinning such as a

Shear polymer or colloidal
suspension (90% of fluids)
stress, 

Shear rate,

(b) Apparent Viscosity, app

Using the Newtonian definition, we can define the “apparent viscosity” for a non-Newtonian
fluid as


 = app   app =


Shear thickening

log [app] Newtonian

Shear thinning

log [  ]

Apparent viscosity is also referred to “viscosity” or “effective viscosity” by some authors.

Care must be taken with how this term is defined.

___________________________________________________________________________
11 JC
CE30240 Advanced Principles of Chemical Engineering Non-Newtonian Fluids

The relationship between app and  can also be expressed as analytical functions or a set of
experimental data. The table below summarises some two- and three-parameter flow
models for describing shear rate,  , versus shear stress,  .

[DO NOT MEMORISE]

 =  Newtonian model
 =kn Power law model used extensively in handling applications.
Bingham model for fluids that features Newtonian flow after
 =  y +  B
the yield stress is reached.
Herschel-Bulkley model for fluids that follows a power law
 =  y + k n
behaviour after the yield stress is reached.
 Ellis model for low-shear rate data containing  y .
=
1 (1 nE ) −1  where  y is the yield stress, k E is the consistency index, nE is
 + k E ( ) 
 y  the flow index.
Sisko model for high shear rate data containing  .
 =   + k S  n S
where  is the infinite shear viscosity, kS is consistency
index, nS is the flow index.
Cross model for data over a wide range of shear rates.
0 −  where  is the infinite-shear viscosity,  is the zero-shear
app =  +
1 + ( c  )
m
viscosity,  c is the relaxation time of polymer and m is a
dimensionless exponent.
Carreau model for data over a wide range of shear rates.
0 − 
app =  + N where c is the relaxation time of polymer and N is a
1 + ( c )2 
 
dimensionless exponent.
 =  y + B  Casson model used especially in treating data on chocolates.

 n =  y n +   n
1 1 2
Generalised model.
 = A − B log ( t ) Weltman model (time dependent)

 =   y + k n 
d Tiu and Boger model
= k1 (  − e )
2

dt

___________________________________________________________________________
12 JC
CE30240 Advanced Principles of Chemical Engineering Non-Newtonian Fluids

1.1.1 Purely Viscous Fluids (Inelastic)

1.1.1.1 Pseudoplastic Fluids

The viscosity of shear-thinning or pseudoplastic fluids decrease as the shear rate is

increased. They are the most commonly encountered type of time-independent fluid
behaviour.

decreasing n
Pseudoplastic, 0 < n < 1
Apparent
viscosity,
app

Shear rate, 
There are various explanations for pseudoplastic behaviour. For some fluids it is thought that
at low shear, particles or molecules within the fluid are more randomly dispersed or
entangled, but at high shear they align in the direction of the shear, thus reducing the
viscosity. Pseudoplastic fluids often contain polymers or large molecules that could behave
in this manner. Examples include polymeric solutions, paints, mayonnaise, paper pulp etc..

The behaviour can often be modelled using the Ostwald-de Waele Power Law (normally just
known as power law). The constitutive equation can be written as

 yx = k  yx
n

where k = consistency index [Pa·sn] (  k  depends on the value of n)  yx = k yx n

  Pa  =  k   s −1 
n
n = flow behaviour index (0  n  1)

When n = 1, the fluid is Newtonian and k equals the viscosity.   k  =  Pa   s n 

When n  1 , k is not the viscosity.

For pseudoplastic fluids 0 < n < 1, thus if the shear rate doubles, the shear stress will increase
by a lower ratio.
app = k  yx n −1
In terms of apparent viscosity,  log  app  = log  k  + ( n − 1) log  yx 
 yx k  yx n
 Plot log  app  vs log  yx 
Definition: app = = = k  yx n −1
 yx  yx  slope = n − 1
 intercept = log  k 
___________________________________________________________________________
13 JC
CE30240 Advanced Principles of Chemical Engineering Non-Newtonian Fluids

1.1.1.2 Dilatant Fluids ( n  1)

The viscosity increases as the shear rate is increased. The power law model can also be used
but n > 1 for dilatant fluids. Examples include high solids content solutions of starch and
sand. At low shear rates water acts as a lubricant, but at higher rates, water moves out of
the solid structure and there is less lubrication so the viscosity increases. Dilatant fluids are
much less common than pseudoplastic fluids. denser particle
packing

1.1.1.3 Bingham (Plastic) Fluids (solid + liquid like behaviour)

This class of materials is characterised by the existence of a yield stress that must be
exceeded before the fluid will deform or flow. In other words, Bingham plastic fluids behave
as a solid until a critical shear rate is achieved. Bingham plastics often have three
dimensional structures sufficient to resist flow until the yield stress is achieved. Examples
include drilling mud, toothpaste and grain. They can be modelled by
yield coefficient of
stress rigidity
 yx =  y +  B yx for  yx   y Important information for
piping, agitator design,
 yx = 0 for  yx   y pump selections etc.
zero
deformation
where B is the coefficient of rigidity or coefficient of plastic viscosity and y is the critical
shear stress (also called yield stress) of the fluid. Note that B is not equal to apparent
viscosity.
fluid behaviour,
 yx =  y +  B yx
slope =  B

 app NB: similar

behaviour to
 yx pseudoplastic
y solid app =  B (power law)
 yx
behaviour,
 yx = 0

This model assumes that the fluid will have a constant viscosity, i.e. Newtonian, once the
critical shear stress has been exceeded. In practice this is not always the case and a better
model is a combination of the model given and the power law. This is known as the
Herschel-Bulkley model.
___________________________________________________________________________
14 JC
CE30240 Advanced Principles of Chemical Engineering Non-Newtonian Fluids

Herschel-Bulkley

  yx =  y + k yx n for  yx   y

 yx = 0 for  yx   y



From these two constitutive equations we get a procedure for modelling a plastic fluid

(i) Plot  vs  on a linear graph

(ii) If the intercept is not zero then the fluid exhibits plastic behaviour
(iii) If the slope is constant the classical Bingham model can be used, otherwise HB
(iv) If the slope is not constant plot ( –  y) vs  on a log-log graph
(v) The slope is n and the intercept at  = 1 is k  yx =  y + k yx n
(vi) Quote all results with reference to the equation used   yx −  y = k yx n
 log  yx −  y  = log  k  + n log  yx 

1.1.2 Time Dependent Fluids

1.1.2.1 Thixotropic Fluids

For many industrially important materials, the shear flow properties depend on both the
rate of shear as well as on the time of shearing. For instance, when the aqueous suspensions
of red mud and of bentonite, crude oils and certain food-stuffs, etc. are sheared at a
constant rate of shear following a period of rest, their apparent viscosity gradually decreases
as their internal “structure” is progressively broken down. In these fluids the structure is
broken down or molecules or particles are aligned as the fluid is subjected to shear.

hysteresis Exponential
 app

 time

___________________________________________________________________________
15 JC
CE30240 Advanced Principles of Chemical Engineering Non-Newtonian Fluids

1.1.2.2 Rheopectic Fluids

Fluid movement can, under certain critical conditions, lead to re-formation of structure
which results in an increased resistance to relative motion. Thus apparent viscosity increases
with length of shearing. This behaviour is known as rheopexy. These fluids are rare.
Examples are such as printer ink and gypsum pastes.

1.1.3 Viscous and Elastic Fluids

1.1.3.1 Viscoelastic Fluids

In addition to viscous (cause energy dissipation) behaviour, these fluids exhibit some elastic
(or rubbery) behaviour. They generally contain elastic particles which are stretched during
flow but which recoil when the stress is reduced. Viscoelastic behaviour is important in the
flow of hot plastics into extrusion moulds. Other examples include jelly and some polymers.
These fluids are usually modelled using the Kelvin Model and Maxwell Model.

Viscoelastic fluids display the Weissenberg Effect and Die Swell

Weissenberg effect
As a rod is rotated, the fluid wraps around it. The elastic
component causes it to behave like elastic. The elastic effects are
greatest where the shear rate is the greatest. This causes
compression on the fluid “within” the elastic wrapping. The fluid
will naturally respond (or recoil) by being squeezed upwards as it
has nowhere else to travel.
e.g. mixing of cake batter

Die Swell effect

When a viscoelastic fluid is extruded from a die or flows from the
exit of a tube, it swells to a diameter up to 3 times the diameter of
the tube as it relaxes. As the flow is increased the swelling usually
increases.
e.g. contributes to the challenges of designing extruder dies to
produce the desired shape of food or snack

___________________________________________________________________________
16 JC
CE30240 Advanced Principles of Chemical Engineering Non-Newtonian Fluids

Example: CE30119, 2018-19 Exam, Question 2

China clay is an essential ingredient in the manufacture of china and porcelain and is widely
used in the making of paper, rubber, paint etc. The flow behaviour of a china clay suspension
in a circular pipe can be approximated by either a power law or a Bingham plastic model
over the shear rate range of 10 to 100 s-1.

(a) Assuming the suspension is a Bingham plastic fluid, determine the rheological
properties by using Figure A shown below. (Ans:  y = 16 Pa,  B = 0.18 Pa s)

(b) What will be the approximate values of the consistency index and flow behaviour
index if the flow is modelled using power law? Express your answers to 2 decimal
points. Sketch a plot of shear stress versus shear rate. (Ans: n = 0.28, k = 9.34 Pa sn)

40

35

30
Shear stress [Pa]

25

20

15

10

0
0 20 40 60 80 100 120
-1
Shear rate [s ]
Figure A

___________________________________________________________________________
17 JC
CE30240 Advanced Principles of Chemical Engineering Non-Newtonian Fluids

1.2 Laminar Flow of Non-Newtonian Fluids

1.2.1 Newtonian Flow in Circular Pipe

For Newtonian fluids we balance the forces on a tube of fluid under laminar flow conditions.
Please ensure that you know the derivations for Newtonian fluids; this is the fundamental
knowledge that you need for non-Newtonian fluid derivations and is also potentially
examinable.

 rz
R r
z
u

dz
p p + dp

From a force balance of an element of radius r inside a circular tube

p r 2 = ( p + dp ) r 2 +  rz 2 rdz

where rz is the shear stress applied by the cylinder of fluid on its surrounding at a distance r
from the centre. Hence

r dp
 rz = − No difficulty here!
2 dz

Note: This is the generalised expression for fluid flows in a circular pipe regardless of
Newtonian or non-Newtonian fluids. The main assumptions being  is independent of time
(only on shear rate) and no wall slip i.e. u = 0 at wall.

For Newtonian fluids:

Note:  rz  0 as drawn (definition of  = retarding the flow)
du
0 velocity decreases as r increases
dr
 du  du
 +  rz =   −    rz = − 
 dr  dr
___________________________________________________________________________
18 JC
CE30240 Advanced Principles of Chemical Engineering Non-Newtonian Fluids

Summary of results for Newtonian fluid:

R 2 − r 2  dp 
(i) Velocity profile: u= −  parabolic
4  dz 

 R 4  dp 
(ii) Volumetric flow rate: Q= − 
8  dz 

R 2  dp 
(iii) Centreline velocity: umax = − 
4  dz 
  r 2 
 u = 1 −  2  umax
  R 

R 2  dp  1
(iv) Mean velocity: um =  −  = umax
8  dz  2

32 um L
(v) Pressure drop: p = Hagen-Poiseuille Equation
4R2

___________________________________________________________________________
19 JC
CE30240 Advanced Principles of Chemical Engineering Non-Newtonian Fluids

1.2.2 Power Law Fluids in Circular Pipe

n
Force balance for a pipe flow  du 
For power law:  rz = k   
 dr 
 rz = −
r dp Note:  rz  0 as drawn
2 dz
du
n
 0 velocity decreases as r increases
 du  r dp dr
 k −  = −
 dr  2 dz rearrangement  du 
n

 +  rz =   − 
 dr 
1
du  r dp  n
− = − 
dr  2k dz 

Velocity profile:
Integration w.r.t. r:
1

 1 dp  n  n  nn+1
u = −−   r + C
 2k dz   n + 1 

  fn ( r )

Boundary condition: r = R, u = 0 i.e. no-slip at wall

 n  n+1
C =   R
n

 n + 1 
 n 
 u=
n +1

  R −r
 n +1 
n
n+1
n
( )
Volumetric flow rate:

Q =  2 r u (r ) dr
R

 2 n 
( rR ) dr
R

n +1 2 n+1
=   −r substitute in u ( r ) and   fn ( r )
n n

 n +1  0

R
 2 n   r 2 nn+1 n 3 n+1 
=    R − r n  integrate with respect to r
 n +1   2 3n + 1 0



 2 n   n + 1  3 nn+1
1
n  1 dp  n 3 nn+1
=    R Q=  −  R Phew!
 n +1   2 ( 3n + 1)  3n + 1  2k dz 
___________________________________________________________________________
20 JC
CE30240 Advanced Principles of Chemical Engineering Non-Newtonian Fluids

Mean velocity:

Q n n+1
 um = =  R n

 R 2 3n + 1

Ratio of velocity to mean velocity:

u (r ) 3n + 1   r  n 
n+1

 = 1 −   
um n + 1   R  

1
u  r n
(i) n ~ small  = 1−   → flat profile (pseudoplastic)
um R

u   r 2 
(ii) n = 1  = 2 1 −    → parabolic (Newtonian)
um   R  

u  r
(iii) n ~ large  = 3 1 −  → linear, sharp (dilatant)
um  R

pseudoplastic

parabolic

dilatant

Pressure drop:
Integrating w.r.t. z:
n
 3n + 1  n
p = 2 Lk   Q
  nR 3 nn+1 
 

When n = 1,

 R 4  dp 
Q= −  same as Newtonian fluid when k = 
8k  dz 

32kuL
 p = same as Newtonian fluid when k = 
4R 2
___________________________________________________________________________
21 JC
CE30240 Advanced Principles of Chemical Engineering Non-Newtonian Fluids

1.2.3 Bingham Fluids in Circular Pipe

cross-over point, c, is not a constant.
It depends on rheological properties
of fluids and the flow conditions
rz > y , deformation
R
Rc rz < y , solid plug

Constitutive relationship:

 rz =  y +  Brz for  rz   y

rz = 0 for  rz   y

As the shear stress for flow in the pipe varies from zero at the centreline to a maximum at
the wall, i.e. deformation, genuine flow of a Bingham plastic fluid occurs only in the part of
cross section where the shear stress is greater than the yield stress, y.

In the part where  rz   y , the material remains as a solid plug and is transported by the
genuinely flowing outer material.

Boundary conditions: r = Rc (cross-over) r = R (pipe wall)

u = uc u=0
du
rz = y  rz =  y −  B NB:  rz  0
dr

At cross-over: rz = y From force balance of pipe flow:

r dp r dp
y = − at Rc   rz = −
2 dz 2 dz
2 y
 Rc = − At the cross-over:  rz =  y
dp dz
Rc dp
2 y Rc  y = −
If we define:  =− = 2 dz
R ( dp dz ) R
 Rc = R

___________________________________________________________________________
22 JC
CE30240 Advanced Principles of Chemical Engineering Non-Newtonian Fluids

Fluid Section: Rc  r  R
 R  r  R
From force balance of pipe flow:
r dp
 du  r dp   rz = −
 rz =  y +  B  − =− 2 dz
 dr  2 dz
rearrangement
du y r dp
− =− −
dr  B 2 B dz

R dp
Now, y = − (from above)
2 dz
eliminate  y
du R dp r dp
− = −
dr 2 B dz 2 B dz

no  y , no Rc
Velocity profile:
Integrating w.r.t. r:
1  dp 
0 R
 −  du =   (R − r ) dr reverse limits of
ur
2  B  dz  r
integration
1  dp 
r
 − ( 0 − ur ) =  −  ( R − r ) dr
2  B  dz  R

r
 r2 
 ur = A   Rr − 
 2 R no Rc , no Rc

 r2 R2 
 ur = A   Rr − −  R +
2

 2 2 

no Rc , no Rc

 ur =
1  dp   2 
 −  R (r − R ) + R − r 
2 B  dz  
1 2
( ) or
2 
1  dp  2 2  y
 ur = (
−  R −r +
4  B  dz  B
(r − R) )
with full rheological data
Rc and  B

___________________________________________________________________________
23 JC
CE30240 Advanced Principles of Chemical Engineering Non-Newtonian Fluids

Volumetric flow rate:

R
Qf =

 2 r u
R = Rc
r dr

 1 2 2 
R
= 2 A  r  R ( r − R ) +
 2
(R − r  dr

) substitute in u ( r )
R

1  dp 
where A= − 
2 B  dz 

 
R
Q f = 2 A  (
 R r − Rr
2
) +
1 2
2
( )
R r − r 3  dr

 R integrate with
respect to r
R
  r 3 Rr 2  1  R2r 2 r 4  
= 2 A   R  −  +  − 
  3 2  2 2 4
R

.
.
.
.
.

2 R 4  dp   5 4 3 2  1 
=  −   − + − − + 
2 B  dz   24 2 4 6 8 





 R 4  dp   5 4  2  1
 Qf =  −  − + 3 − − + 
2 B  dz   12 2 3 4

___________________________________________________________________________
24 JC
CE30240 Advanced Principles of Chemical Engineering Non-Newtonian Fluids

Solid plug section: 0  r  Rc

 u = uc i.e. uniform velocity from r = 0 to Rc

Velocity profile:
Note that the velocity in the “fluid section” and “solid plug section” coincide at the cross
over, thus (using the expression for velocity profile)

u (R ) = uc =
1  dp  
( 2 
 −  R (R − R ) + R − (R ) 
2 B  dz  
1 2
) from uc = u f ( r =  R )
2 
1  dp   2 R 2 R2 
 uc = −   − R 2 + 
2 B  dz   2 2

Volumetric flow rate:

uc is constant between r = 0 and r =  R
Qs =  Rc 2  uc =  ( R ) uc
2

 R 4  dp   4 2 
 Qs = −   − + 
3

2 B  dz   2 2 

Overall volumetric flow rate:

QT = Q f + Qs

 R 4  dp   5 4 2  1 4 2 
QT =  −   − +  3
− − + + −  3
+ 
2 B  dz   12 2 3 4 2 2 





 R 4  dp   1   4 
 QT = −  − +
2 B  dz   4 3 12 

Now check that when y = 0 and  = 0

 R 4  dp 
 QT =  −  same as that for Newtonian fluid.
8 B  dz 

___________________________________________________________________________
25 JC
CE30240 Advanced Principles of Chemical Engineering Non-Newtonian Fluids

Example: Flow down a plane under gravity

liquid

yz z

wall air
y
y


u = fn(y)

Derive the expression for velocity profile and volumetric flow assuming the fluid is:

(a) Newtonian fluid

(b) Power-law fluid

(c) Bingham-plastic fluid (Hint: You may assume that the cross-over point is at
y = yc =  when  =  y )

Answers:
(a) Newtonian fluid
g  y2 
Velocity profile: u=  y − 
  2 

 g 3
Volumetric flow rate: q =
3

(b) Power law fluid

1

  g  n n  nn+1
 − ( − y ) n 
n +1
Velocity profile: u = 
 k  n +1  
1

  g n n 2 n +1
Volumetric flow rate: q =    n
 k  2n + 1

___________________________________________________________________________
26 JC
CE30240 Advanced Principles of Chemical Engineering Non-Newtonian Fluids

(c) Bingham-plastic fluid

g  y2 
Fluid section: uf =   y − 
B  2 

 g 2 2
Solid section: us =
2 B

 g 3  2  3 
Volumetric flow rate per unit depth: qT =  − 
2 B  3

___________________________________________________________________________
27 JC