You are on page 1of 34

The effect of interfacial slip on the rheology of a dilute emulsion of

drops for small capillary numbers


Arun Ramachandran and L. Gary Leal

Citation: Journal of Rheology 56, 1555 (2012); doi: 10.1122/1.4749836


View online: https://doi.org/10.1122/1.4749836
View Table of Contents: https://sor.scitation.org/toc/jor/56/6
Published by the The Society of Rheology

ARTICLES YOU MAY BE INTERESTED IN

The effect of interfacial slip on the dynamics of a drop in flow: Part I. Stretching, relaxation,
and breakup
Journal of Rheology 56, 45 (2012); https://doi.org/10.1122/1.3663379

Effect of interfacial slip on the deformation of a viscoelastic drop in uniaxial extensional flow
field
Physics of Fluids 29, 032105 (2017); https://doi.org/10.1063/1.4977949

Slip at polymer–polymer interfaces: Rheological measurements on coextruded multilayers


Journal of Rheology 46, 145 (2002); https://doi.org/10.1122/1.1427912

Deformation of a surfactant-laden viscoelastic droplet in a uniaxial extensional flow


Physics of Fluids 30, 122108 (2018); https://doi.org/10.1063/1.5064278

Polymer-polymer interfacial slip by direct visualization and by stress reduction


Journal of Rheology 54, 1207 (2010); https://doi.org/10.1122/1.3479389

Interfacial slip between polymer melts studied by confocal microscopy and rheological
measurements
Journal of Rheology 47, 795 (2003); https://doi.org/10.1122/1.1566035
The effect of interfacial slip on the rheology of a dilute
emulsion of drops for small capillary numbers

Arun Ramachandrana)

Department of Chemical Engineering and Applied Chemistry,


University of Toronto, Toronto, Ontario M5S3E5, Canada

L. Gary Leal

Department of Chemical Engineering, University of California at Santa Barbara,


Santa Barbara, California 93106

(Received 4 April 2012; final revision received 12 August 2012;


published 25 September 2012)

Synopsis
We present the constitutive equation for the volume-averaged hydrodynamic stress for a dilute
emulsion in a linear ambient flow, when there is slip at the liquid-liquid interface between the
Newtonian drop and the suspending fluids. Slip is modeled using a simple Navier slip boundary
condition. We provide analytical solutions in the limit of small capillary numbers for the shape
deformation, viscosity, and normal stresses. Slip moderates these quantities, with changes from the
no-slip case being more pronounced for large viscosity ratios of the drop relative to the suspending
fluid. It has been suggested in the past that slip can explain the anomalously low viscosities of
certain polymeric blends. Our analysis indicates that slip can only partially account for these
deviations, and that other mechanisms should be explored to explain the residual discrepancy.
C 2012 The Society of Rheology. [http://dx.doi.org/10.1122/1.4749836]
V

I. INTRODUCTION
Over the past few decades, blending of immiscible polymers has become an increas-
ingly popular method for producing composites with hybrid properties [e.g. see Macosko
(2000)]. A tremendous effort has been devoted in the literature, therefore, to establish the
connections between flow conditions in the blending process, drop size distribution, the
rheology of the blend, and the properties of the composite. Blend rheology, which is
the focus of this paper, is important to understand because it governs the processability of
the mixture, and is required to design the mixing and pumping requirements. It also pro-
vides information on the morphology of the blend [Tucker and Moldenaers (2002)],
which critically governs the mechanical properties of the composite. The extensive stud-
ies of rheology of blends have uncovered some interesting and counterintuitive trends,
the most prominent one among these being related to the blend viscosity. It is known, for
example, that the viscosity of a blend can be “anomalously” low, sometimes less than the

a)
Author to whom correspondence should be addressed; electronic mail: Arun.Ramchandran@utoronto.ca

C 2012 by The Society of Rheology, Inc.


V
J. Rheol. 56(6), 1555-1587 November/December (2012) 0148-6055/2012/56(6)/1555/33/$30.00 1555
1556 A. RAMACHANDRAN AND L. G. LEAL

viscosities of the individual polymers. Such negative-deviation blends have been


reviewed by Utracki [Utracki and Kamal (1982); Utracki (1983)].
In the literature, the negative deviations in blend viscosities have often been attributed
to the existence of slip at the interface between the two blended polymers [Lin (1979);
Lyngaae-Jorgensen and Thomsen (1988); Bousmina et al. (1999); Zhao and Macosko
(2002)]. The primary origin of interfacial slip is that two different polymers do not mix,
as a result of which the degree of entanglement in the diffuse region at the interface
between the polymeric fluids is greatly reduced. The diffuse region thus has a viscosity
that is much lower than the viscosities of either of the bulk fluids [Furukawa et al.
(1989); De Gennes (1979); Brochard-Wyart and de Gennes (1993); Ajdari (1993);
Goveas and Fredrickson (1998); Adhikari and Goveas (2004)] and, thus, from a macro-
scopic point of view this thin region appears as a slip layer. Recently, Macosko and co-
workers [Zhao and Macosko (2002); Lee et al. (2009); Park et al. (2010)] and Lam et al.
(2003) have furnished definitive experimental evidence for the existence of slip.
A natural question that arises, therefore, is, how interfacial slip affects the rheology of
emulsions. In this work, we address this question by analyzing the simplest case of the
rheology of a dilute emulsion of equal-sized, surfactant-free, Newtonian drops, with inter-
facial slip. In the process, we explore how drop deformation and other material functions,
such as the normal stress differences in simple shear flow, the viscosity and normal
stresses in extensional flow, and the complex viscosity in an oscillatory shear flow, are
affected by interfacial slip. The procedure that we follow for performing this calculation is
a regular, domain perturbation expansion for small capillary numbers, which has been pre-
viously implemented for drops for the no-slip case [Cox (1969); Schowalter et al. (1968);
Frankel and Acrivos (1970); Barthès-Biesel and Acrivos (1973); Rallison (1980)]. We
model slip at the interface using the Navier slip boundary condition. We then use these
results as a basis to discuss the problem of negative deviations in the viscosities of blends.
In Sec. II, we define the problem mathematically, carry out the perturbation expansion
at the zeroth and first orders in the capillary number Ca, and calculate the drop shape and
the total emulsion stress. In Sec. III, we study the effect of slip on drop deformation and
rheology for uniaxial extensional flow, simple shear flow, and oscillatory shear flow, and
compare our results with prior work in this area that has employed the no-slip assump-
tion. In Sec. IV, we present a critique of slip as an explanation for the anomalously low
viscosities observed for polymer blends. In particular, we make quantitative comparisons
between the theory and the experimental measurements to examine if slip can account
for these deviations. We summarize our findings in Sec. V.

II. THEORY
Consider a spherical drop of unit radius placed in a suspending fluid undergoing a lin-
ear flow u1 (see Fig. 1), given by
1
u1 ¼ C  x ¼ ðE þ XÞ  x ¼ E  x þ ðx Ù xÞ: (1)
2

Here, x represents the position vector measured from the origin. The tensor E is the
dimensionless rate of strain tensor, while x is the dimensionless vorticity.
Under creeping flow conditions, the governing equations for a Newtonian suspending
fluid and drop (drop variables denoted by the circonflex) are

r2 u  rp ¼ 0; r  u ¼ 0;
(2)
r2 u
^  r^
p ¼ 0; ru ^ ¼ 0;
RHEOLOGY OF DILUTE EMULSIONS WITH SLIP 1557

FIG. 1. An initially spherical drop (dotted line) of unit radius deformed by a linear ambient flow u1 (a simple
shear flow is shown as an example). The vectors n and t are the local normal and tangent vectors on the surface
of the deformed drop.

with the boundary conditions,


u ! u1 as x ! 1;
uu ^ ¼ a½I  nn  ðr  nÞ for x 2 S; (3)
1
ðr  k^
rÞ  n ¼ ðr  nÞn for x 2 S:
Ca
The variables appearing in (2) and (3) have been rendered dimensionless according to

x0 ¼ ax; t0 ¼ tla=c; u0 ; u^ 0 ; u0 1 ¼ Gau; Ga^


u ; Gau1
0 0 0 0 (4)
p ; r ¼ lGp; lGr; p^ ; r
^ ¼l ^ G^ p; l
^ G^
r:

Here, the primed quantities are the dimensional variables. We denote the interfacial tension as
c: The dimensional radius of the undeformed drop is a. The scalar G represents the magnitude
of the velocity gradient tensor ru0 1 . The suspending fluid has a viscosity l, while the drop
^ ¼ kl. In addition to the viscosity ratio k, the dimensionless parameters that
viscosity is l
appear in the governing equations (2) and (3) are the capillary number, Ca  lGa=c; and a
dimensionless slip coefficient, a ¼ a0 l=a: We shall define the dimensional slip coefficient a0
shortly.
It is convenient to describe the shape of the drop in terms of the points x, for which a
function F satisfies

Fðx; tÞ ¼ 0: (5)

The normal vector n is then defined in terms of F as

rF
n¼ : (6)
jrFj

This leads to the dimensionless kinematic condition [Leal (2007)]

1 @F
þ Ca ðu  nÞ ¼ 0 (7)
jrFj @t

with the velocity being evaluated at the interface.


The problem set up so far is nonlinear, even though the creeping flow equations are lin-
ear. The nonlinearity arises from the fact that the shape of drop is not known; its calculation
1558 A. RAMACHANDRAN AND L. G. LEAL

requires knowledge of the pressure and stress fields, which, depend, in turn, on the flow
field. The solution of the full problem for arbitrary values of the parameters requires a nu-
merical method. To make analytical progress, in this paper, we consider the limiting case
of small drop deformations, so that we can solve via a domain perturbation expansion in
terms of a small deformation parameter, e: As pointed out in earlier work [Cox (1969);
Frankel and Acrivos (1970); Barthès-Biesel and Acrivos (1973); Hakimi and Schowalter
(1980); Rallison (1980)], small drop deformations can arise in two situations: For small
capillary numbers ðCa  1Þ, when the flow is not strong enough to stretch the drop signifi-
cantly; and for large viscosity ratios ðk  1Þ, when the drop is too viscous to be stretched
out by the relatively inviscid flow. In this paper, we only address the case e ¼ Ca.
The slip coefficient corresponding to the diffuse interface between two polymeric fluids
can be approximated [Ramachandran et al. (2012)] as a0 ¼ dI =lI , where dI is the thickness
of the diffuse interface and lI is the viscosity of the interfacial region.1 Using the theories
of Helfand and Tagami (1972) and Goveas and Fredrickson (1998), we can evaluate dI and
lI for the combination of the polymers PDMS and PBd [see Park et al. (2003) for details]
as 0.5 nm and 0.02 Pa s, respectively. The slip coefficient a0 for the PDMS-PBd combina-
tion is, therefore, 2:5  108 SI units. If we consider suspending fluid viscosities ranging
from 1 to 100 Pa s, and drop radii ranging from 1 to 100 lm [e.g., see Ramachandran et al.
(2012)], the dimensionless slip coefficient in Eq. (3) can vary from about 104 to 1. To
encompass this wide range of values, we will consider in this paper three different, nonzero
values of a: 0 (no-slip), 0.1, and 1, to examine the effect of slip.
Since we are considering only small deviations from the spherical shape of the drop,
we can write the shape function, F, as
h x i
Fðx; tÞ ¼ r  1 þ Ca f ; t ¼ 0 (8)
r

so that when Ca ¼ 0, the drop is a sphere. The definition of the normal vector n becomes

rF 1 x 
n¼ ¼  CarF ; (9)
jrFj jrFj r

where

jrFj ¼ ð1 þ Ca2 jrf j2 Þ1=2 : (10)

The curvature r  n, then, is


 
1 2 1 hx i
rn¼  Car2 f  Ca2 3
 Carf  ðrðrf Þ  rf Þ: (11)
jrFj r jrFj r
pffiffiffiffiffiffiffiffiffi
Here, we have used r ¼ x  x and x  rf ¼ rð@f =@rÞ ¼ 0. The kinematic condition (7)
can also be simplified as follows:

@f  1=2
¼ ujr¼1þCaf  n 1 þ Ca2 ðrf Þ2 : (12)
@t

1
In this model, we assume that the viscosity is constant across the diffuse layer. Other representations of the
interfacial viscosity were considered by Ramachandran et al. (2012).
RHEOLOGY OF DILUTE EMULSIONS WITH SLIP 1559

We can now expand the shape function f and the velocity, pressure, and stress fields
as regular perturbation expansions in Ca:

f ¼ f ð0Þ þ Ca f ð1Þ þ Ca2 f ð2Þ þ    ;


p ¼ pð0Þ þ Ca pð1Þ þ Ca2 pð2Þ þ    ;
u ¼ uð0Þ þ Ca uð1Þ þ Ca2 uð2Þ þ    ; (13)
p^ ¼ p^ð0Þ þ Ca p^ð1Þ þ Ca2 p^ð2Þ þ    ;
^¼u
u ^ ð0Þ þ Ca u^ ð1Þ þ Ca2 u
^ ð2Þ þ    :

The interfacial quantities at r ¼ 1 þ Caf in a domain perturbation framework can be eval-


uated approximately in terms of quantities at the surface of a sphere (r ¼ 1) using a Taylor
series expansion. If we then substitute (13), we obtain the full expansions for p; u; p^; and u^
at r ¼ 1 þ Caf in terms of their values at the spherical surface, r ¼ 1. For example,
 
pjr¼1þCaf ¼ pð0Þ jr¼1 þ Ca pð1Þ jr¼1 þ rpð0Þ  xjr¼1 f ð0Þ þ    ; (14)

and the other variables can be expanded in the same way.

A. Zeroth order solution


At the leading order in Ca, the governing equations are

r2 uð0Þ  rpð0Þ ¼ 0; r  uð0Þ ¼ 0;


(15)
^ ð0Þ  r^
r2 u p ð0Þ ¼ 0; ru ^ ð0Þ ¼ 0;

with the boundary conditions

uð0Þ ! ðE þ XÞ  x as r ! 1; (16)
 
nð0Þ  uð0Þ  u^ ð0Þ ¼ 0;
     r¼1  (17)
I  nð0Þ nð0Þ  uð0Þ  u ^ ð0Þ jr¼1 ¼ a I  nð0Þ nð0Þ  ðrð0Þ  nð0Þ Þjr¼1 ;

and
  1
nð0Þ  rð0Þ  k^ r ð0Þ jr¼1 ¼ r  nð0Þ þ r  nð1Þ ;
   Ca   (18)
I  nð0Þ nð0Þ  rð0Þ  k^ r ð0Þ jr¼1  nð0Þ ¼ 0:

In the above equation, we have split up the boundary conditions at the interface into nor-
mal and tangential components.
We use the method of vector harmonics [Leal (2007)] to write the solutions to flow
equations. Recognizing that disturbance velocity and pressure fields must be linear in E
and X at this order, we can write

pð0Þ ¼ p1 þ 2c1 Dð2Þ : E;


xpð0Þ (19)
uð0Þ ¼ ðE þ XÞ  x þ þ c2 Dð1Þ  E þ c3 Dð3Þ : E
2
¼ ðE þ XÞ  x þ c1 xðDð2Þ : EÞ þ c2 Dð1Þ  E þ c3 Dð3Þ : E:
1560 A. RAMACHANDRAN AND L. G. LEAL

Here, p1 is the ambient pressure, while DðnÞ is the decaying spherical harmonic of order
n, defined as
     
1 1 1
Dð1Þ ¼ r ; Dð2Þ ¼ rr ; Dð3Þ ¼ rrr ; etc: (20)
r r r

Similarly, we can write the solution to drop phase pressure and velocity fields as

p^ð0Þ ¼ p^0 þ 2d1 Gð2Þ : E;


p ð0Þ
x^
u^ ð0Þ ¼ þ X  x þ d2 Gð1Þ  E þ d3 Gð3Þ : E (21)
2  
¼ X  x þ d1 x Gð2Þ : E þ d2 Gð1Þ  E þ d3 Gð3Þ : E:

GðnÞ is the growing spherical harmonic of order n and is defined as

GðnÞ ¼ r 2nþ1 DðnÞ : (22)

Note that, at this order, the vorticity tensor X appears in the velocity expressions only
in the form of a rigid body rotation X  x for both the suspending and drop fluid phases.
Because f is a scalar and expected to be linearly related to the rate of strain tensor E,
we can write the geometry of the drop to leading order as
xx
f ð0Þ ¼ 3b1 E : : (23)
r2

The normal vector to leading order from Eq. (9) is, therefore, just the radial unit vector:
x
nð0Þ ¼ ¼ ir : (24)
r
We now need to determine the six constants, c1 through c3 and d1 through d3 . First, we
apply the two continuity equations in the drop and suspending fluids, which yield the two
equations:

c2 ¼ 0;
5 (25)
d3 ¼ d1 :
21

Before we implement the boundary conditions to get the remaining four constants, we
need the interfacial velocities and stresses for the drop and suspending fluids. For the sus-
pending fluid, we have

uð0Þ r¼1 ¼ ð1 þ 6c3 ÞE  x þ X  x þ ð3c1  15c3 Þx ðE : xxÞ;
(26)
rð0Þ r¼1  nð0Þ ¼ ð2 þ 6c1  48c3 ÞE  x þ ð24c1 þ 120c3 Þx ðE : xxÞ;

while for the drop fluid,


   
ð0Þ 10 4
u^ jr¼1 ¼ X  x þ d2 þ d1 E  x þ  d1 x ðE : xxÞ;
 7  7  (27)
ð0Þ ð0Þ 32 38
^ jr¼1  n ¼ ^
r p0x þ d1  2d2 E  x þ  d1 x ðE : xxÞ:
7 7
RHEOLOGY OF DILUTE EMULSIONS WITH SLIP 1561

Finally, we need an expression for the force jump across the interface. Examining the
RHS of Eq. (18) and using (11), we get
1     2
r  nð0Þ þ r  nð1Þ ¼ þ 12b1 E : xx: (28)
Ca Ca
We can now apply the boundary conditions at the interface using the relations (26)–
(28). The velocity restrictions in Eq. (17) yield
 
6
ð1 þ 3c1  9c3 Þ  d2 þ d1 ¼ 0; (29)
7
 
10
ð1 þ 6c3 Þ  d2 þ d1 ¼ að2 þ 6c1  48c3 Þ; (30)
7

d~2 ¼ 1; (31)

while the condition connecting the interfacial stresses of the drop and suspending fluids
in Eq. (18) gives
2
p^0  p1 ¼ ; (32)
Ca
 
6
ð2  18c1 þ 72c3 Þ  k  d1  2d2 ¼ 12b1 ; (33)
7
 
32
ð2 þ 6c1  48c3 Þ  k d1  2d2 ¼ 0; (34)
7

c~2 ¼ 0: (35)

We can solve Eqs. (29), (30), (33), and (34) to get the four constants as

4b1 ½3kð80a þ 19Þ þ 48 þ 95k2 ð2a þ 1Þ  5kð80a þ 3Þ  80


c1 ¼  ;
3½38k2 ð5a þ 1Þ þ kð240a þ 89Þ þ 48
12b1 ½kð10a þ 3Þ þ 2 þ 19k2  kð80a þ 3Þ  16
c3 ¼  ;
3½38k2 ð5a þ 1Þ þ kð240a þ 89Þ þ 48
14f3b1 ½2kð5a þ 1Þ þ 3  20akg (36)
d1 ¼ ;
½38k2 ð5a þ 1Þ þ kð240a þ 89Þ þ 48
 
14½4b1 48kð5a þ 1Þ þ 57  5kð128a þ 19Þ  80
d2 ¼ :
½38k2 ð5a þ 1Þ þ kð240a þ 89Þ þ 48

The evolution of the drop shape at this order can now be determined from the kinematic
condition (7) as

@½b1 E f24b1 ½kð5a þ 1Þ þ 1  ½kð80a þ 19Þ þ 16g


3 ¼ 5 E: (37)
@t ½38k2 ð5a þ 1Þ þ kð240a þ 89Þ þ 48

B. First order solution


At the first order in Ca, the governing equations are
1562 A. RAMACHANDRAN AND L. G. LEAL

r2 uð1Þ  rpð1Þ ¼ 0; r  uð1Þ ¼ 0;


(38)
^ ð1Þ  r^
r2 u p ð1Þ ¼ 0; ru ^ ð1Þ ¼ 0;

with the boundary conditions

uð1Þ ! 0 as r ! 1; (39)

^ Þjr¼1þCa f  nð1Þ ¼ 0;
½ðu  u
(40)
^ Þjr¼1þCa f ð1Þ ¼ a½ðI  nnÞ  ðrjr¼1þCa f  nÞð1Þ ;
½ðI  nnÞ  ðu  u

and

r Þjr¼1þCa f : nnð1Þ ¼ ðr  nÞð2Þ ;


½ðr  k^
 (41)
½ðI  nnÞ  ðr  k^ r Þjr¼1þCa f  nÞð1Þ ¼ 0:

Since the velocity fields satisfy the creeping flow equations, we can again write the
general forms of these fields using vector harmonics. We realize that at this order of
expansion in Ca, the velocity fields must depend, at the most, quadratically on the rate of
strain tensor E. Hence,

pð1Þ ¼ 2c4 Dð2Þ : E þ 2c5 Dð2Þ : ðE  EÞ þ 2c6 ðDð4Þ : EÞ : E þ 2~


c 5 Dð2Þ : ðX  EÞ þ 2
c 5 D : ðX  XÞ;
xpð1Þ
uð1Þ ¼ þ c7 Dð1Þ  E þ c8 Dð3Þ : E þ c9 ðE  EÞ  Dð1Þ þ c10 E  ðDð3Þ : EÞ þ c11 ðDð5Þ : EÞ : E
2
_
þ c12 Dð1Þ E : E þ c13 Dð3Þ : ðE  EÞ þ c~7 X  Dð1Þ þ c~9 ðX  EÞ  Dð1Þ þ c 9 ðE  XÞ  Dð1Þ
þ c~10 X  ðDð3Þ : EÞ þ c9 ðX  XÞ  Dð1Þ þ c12 Dð1Þ ðX : XÞ; (42)

and

p^ð1Þ ¼ 2d4 Gð2Þ : E þ 2d5 Gð2Þ : ðE  EÞ þ 2d6 ðGð4Þ : EÞ : E þ 2d14 E : E


þ 2d~5 ðX  EÞ : Gð2Þ þ 2d5 ðX  XÞ : Gð2Þ þ 2d14 X : X;
x p^ð1Þ
^ ð1Þ ¼
u þ d7 E  Gð1Þ þ d8 E : Gð3Þ þ d9 ðE  EÞ  Gð1Þ þ d10 E  ðGð3Þ : EÞ
2l (43)
þ d11 ðGð5Þ : EÞ : E þ d12 Gð1Þ E : E þ d13 ðGð3Þ  EÞ : E þ d~7 X  Gð1Þ
_
þ d~9 ðX  EÞ  Gð1Þ þ d 9 ðE  XÞ  Gð1Þ þ d~10 X  ðGð3Þ : EÞ
þ d9 ðX  XÞ  Gð1Þ þ d12 Gð1Þ X : X:

Also, at this order of the expansion, the shape function f ð1Þ is a scalar and linearly
related to quadratic combinations of E and X composed of the rate of strain tensor. We
can write this shape function as

xEx x  ðE  EÞ  x ðx  E  xÞ2
f ð1Þ ¼ 3b2 þ b3 E : E þ b4 þ b5
r2 r2 r4
x  ðX  EÞ  x  x  ðX  XÞ  x
þ b~4 þ b 3 X : X þ b4 : (44)
r2 r2
RHEOLOGY OF DILUTE EMULSIONS WITH SLIP 1563

Since the drop volume must remain constant, the coefficients b2 through b5 cannot be in-
dependent; they must be related. To determine this relationship, we write the drop volume as
ð ð 3
1 1 4p
V¼ r 3 dX ¼ 1 þ ef ð0Þ þ e2 f ð1Þ dX ¼ : (45)
3 S 3 S 3

After integrating, simplifying, and collecting terms of the same order in Ca, we get

b4 2 6
b3 þ þ b5 þ b21 ¼ 0: (46)
3 15 5

b4
b3 þ ¼ 0: (47)
3

From Eq. (9), the normal vector at the first order is


 
nð1Þ ¼ rf ð0Þ ¼ 6b1 xðx  E  xÞ  E  x jr¼1 : (48)

We now have to determine the unknown coefficients in the above equations. The con-
^ ð1Þ ¼ 0 at
tinuity equations in the drop and suspending fluid phases r  uð1Þ ¼ 0 and r  u
this order provide 12 equations:

c10 ¼ 2c6 ; c9 ¼ 0; c7 ¼ 0;
_
c 9 ¼ c~9 ; c9 ¼ 0;

and
5 6 21
d8 ¼ d4 ; d5 ¼ d10 þ d13 ; (49)
21 5 5
55
d6 ¼ d11 ; d9 ¼ 3ðd14  d12 Þ;
7
14 21
d 5 ¼  d~10 þ d~13 ;
~
5 5
21
d5 ¼ d13 ; d9 ¼ 3d12 þ 3d14 :
5

As before, prior to the application of the boundary conditions to get the remaining con-
stants, we need to set up the interfacial velocities and stresses for the drop and suspending flu-
ids at first order in e. The interfacial velocity and stress at this order for the suspending fluid are

ðujr¼1þCa f Þð1Þ ¼ uð1Þ jr¼1 þ f ð0Þ ruð0Þ  xjr¼1 ;


(50)
ðrjr¼1þCa f  nÞð1Þ ¼ rð1Þ jr¼1  nð0Þ þ rð0Þ jr¼1  nð1Þ þ f ð0Þ ðrrð0Þ  xjr¼1 Þ  nð0Þ :

Similar expressions can be written down for these quantities in the drop fluid. To imple-
ment the boundary conditions conveniently, we write the normal and tangential components
of the interfacial velocities and stresses separately. At this order, the components of these
fields along the interfacial normal and tangential directions are, for the suspending fluid,

½ujr¼1þCaf  nð1Þ ¼ ½ujr¼1þCaf ð1Þ  nð0Þ þ ½ujr¼1þCaf ð0Þ  nð1Þ ; (51)


1564 A. RAMACHANDRAN AND L. G. LEAL

½ðI  nnÞ  ujr¼1þCa f ð1Þ ¼ ðujr¼1þCaf Þð1Þ  nð0Þ ðujr¼1þCaf  nÞð1Þ  nð1Þ ðujr¼1þCaf  nÞð0Þ ;
(52)

ðrjr¼1þCa f : nnÞð1Þ ¼ ðrjr¼1þCa f  nÞð1Þ  nð0Þ þ ðrjr¼1þCa f  nÞð0Þ  nð1Þ ; (53)


  ð1Þ
ðI  nnÞ  rjr¼1þCa f n ¼ ðrjr¼1þCaf  nÞð1Þ  nð1Þ ðrjr¼1þCaf : nnÞð0Þ

nð0Þ ðrjr¼1þCaf : nnÞð1Þ : (54)

We can use analogous expansions to evaluate the corresponding drop fluid fields.
Finally, we need an expression for the force jump across the interface. Examining the
RHS of the first equation in (40) and using (11), we get the dimensionless curvature of
the drop at this order to be
 2   
ðr  nÞð2Þ ¼ 2 f ð0Þ  f ð1Þ  ðr2 f Þð1Þ  ðrf ð0Þ Þ2  x  rf ð0Þ  rrf ð0Þ
¼ ð90b21 þ 18b5 Þðx  E  xÞ2 þ ð12b2 Þx  E  x þ ð2b3  2b4 ÞE : E
þð4b4  8b5 ÞðE  xÞ2 þ 4b~4 x  ðX  EÞ  x þ ð2b3  2b4 ÞX : X
þð4b4  8b5 Þx  ðX  XÞ  x: (55)

We now have all the expressions required to apply the boundary conditions in Eqs.
(40) and (41). The resulting equations are presented in Appendix A [note that the conti-
nuity equation results in Eq. (49) have already been applied in writing these equations].
On closer examination of the equations in Appendix A, it appears that there are more
equations than unknowns and the system is overspecified. However, it turns out that by
applying the coefficient relationships (29)–(35) developed at the zeroth order, this system
of equations has a unique solution, and the constants can be evaluated in terms of the
shape constants b2 ; b4 ; b5 ; b~4 , and b4 . We have solved these equations in MATLAB, but are
unable to present them here due to the length of the solution variables.
We are now ready to apply the kinematic equation at the first order:

@f ð1Þ
¼ ðujr¼1þCaf  nÞð1Þ : (56)
@t

Substitution of the normal component of the interfacial velocity ðujr¼1þCaf  nÞð1Þ


2   3
b4 2 6 2 b4
 þ b5 þ b1 E : E  X : X
@6 6 3 15 5 3 7
7
6 7
@t 4 þð3b2 E þ b4 E  E þ b~4 X  E þ b4 X  XÞ : xx 5
þb5 ðx  E  xÞ2
¼ ðc5 þ 6c6  30c11  c12 þ 3c13 ÞE : E þ ð c 5  c12 þ 3 c 13 ÞX : X


ð3c4  9c8 ÞE þ ½3c5  48c6 þ 300c11  9c13  6b1 ð1 þ 6c3 ÞE : E
þ : xx
þð3~c 5 þ 6~
c 10  9~
c 13 þ 6b1 c~2 ÞX  E þ ð3
c 5  9c 13 ÞX  X
þ½3b1 ð3  6c1 þ 48c3 Þ þ ð75c6  525c11 Þðx  E  xÞ2 : (57)

There are six shape parameters remaining to be determined: b1 ; b2 ; b4 ; b5 ; b~4 , and b4 .
RHEOLOGY OF DILUTE EMULSIONS WITH SLIP 1565

For steady ambient flows (i.e., when Eij and Xij are not functions of time), the kine-
matic conditions (37) and (57) decompose to six ordinary differential equations for the
unknown shape parameters:

@b1 1
¼ þ c1  3c3 ;
@t 3
@b2
¼ c4  3c8 ;
@t
@b4
¼ 3c5  48c6 þ 300c11  9c13  6b1 ð1 þ 6c3 Þ;
@t
@b5 (58)
¼ 3b1 ð3  6c1 þ 48c3 Þ þ ð75c6  525c11 Þ;
@t
@ b~4
¼ 3~c 5 þ 6~
c 10  9~
c 13 þ 6b1 c~2 ;
@t
@ b4
¼ 3c 5  9
c 13 :
@t

To gain a qualitative insight into the rate and degree of drop deformation, let us con-
sider the first effect of the flow on the drop shape by examining the differential equation
for the shape parameter b1 .

@b1 5 f24b1 ½kð5a þ 1Þ þ 1  ½kð80a þ 19Þ þ 16g ðb1  b 1 Þ


¼ 2
¼ : (59)
@t 3 ½38k ð5a þ 1Þ þ kð240a þ 89Þ þ 48 tc

We see that the determination of the drop shape at this order is reduced to the determi-
nation of only one scalar—b1 , which decays exponentially with time to the value

½kð80a þ 19Þ þ 16


b 1 ¼ : (60)
24½kð5a þ 1Þ þ 1

The dimensional time constant for the decay is

t0c ½38k2 ð5a þ 1Þ þ kð240a þ 89Þ þ 48


tc ¼ ¼ : (61)
la=c 40½kð5a þ 1Þ þ 1

This time constant is shown in Fig. 2(a). One can see that if a is zero, i.e., if the no-slip
condition is employed, the time scale for drop stretching is la=c for inviscid drops, and l ^ a=c
for highly viscous drops. But surprisingly, with the introduction of slip, there is little change
in this time scale; furthermore, this weak effect is only discernible at intermediate viscosity
ratios. The dimensionless shape constant b 1 [see Fig. 2(b)], which is a measure of the defor-
mation of the drop, is quite insensitive to the viscosity ratio in the absence of slip, varying
only from 2/3 to 19/24 between the inviscid and the highly viscous limits. Slip reduces the
weak disparity in the two limits even further by decreasing deformation of the drop for higher
drop viscosities, simply because the tangential stress acting on the drop is less effective at
stretching it in the presence of slip. We anticipate, therefore, that slip will not affect the shape
of the deformed drop strongly, except at high viscosity ratios.
Now, if we assume that the drop has attained a steady shape, and set the time deriva-
tives db1 =dt through db5 =dt to zero, the shape parameters attain the following asymptotic
values:
1566 A. RAMACHANDRAN AND L. G. LEAL

½kð80a þ 19Þ þ 16


b1 ¼ ;
24½kð5a þ 1Þ þ 1
b2 ¼ 0;
2 3
k3 ð79200a3 þ 460250a2 þ 128903a þ 8717Þ
6 7
½kð80a þ 19Þ þ 164 þk2 ð40480a2 þ 14439a þ 18758Þ 5
þkð6688a þ 10393Þ þ 352
b4 ¼ ;
30240½kð9a þ 1Þ þ 1½kð5a þ 1Þ þ 13
" #
k2 ð29520a2 þ 9974a þ 751Þ (62)
½kð80a þ 19Þ þ 16
þkð9184a þ 1407Þ þ 656
b5 ¼ ;
1728½kð9a þ 1Þ þ 1½kð5a þ 1Þ þ 12
6 b4 2
b3 ¼  b21   b5 ;
5 3 15
½kð80a þ 19Þ þ 16½38k2 ð5a þ 1Þ þ kð240a þ 89Þ þ 48
b~4 ¼ ;
160½kð5a þ 1Þ þ 12
b3 ¼ b4 ¼ 0:
These equations will be used subsequently to evaluate the steady-state drop shape and
rheology.

C. Rheology of a dilute emulsion of drops with interfacial slip


For a dilute suspension, in the general linear flow given by (1), we can use the analysis
of Batchelor (1970) to write the volume averaged suspension stress as
/ 0
hr0 i ¼ hp0 iI þ 2l0 GE þ S; (63)
Vd
where the stresslet S0 is a symmetric, traceless tensor that signifies the contribution made
by drop to the total stress due to the modification of the velocity and stress fields by its
presence. It is defined as

ð    T  1  ð 
0 3 1 T
S ¼ lGa ðr  nÞx þ ðr  nÞx  Iððr  nÞ  xÞ dS  l unþðunÞ dS :
S 2 3 S
(64)
As shown in Appendix B, we can extract the drop contribution to the total stress by
examining the leading order disturbance terms in the velocity field of the suspending
fluid. These terms decay as 1=r2 and are associated with the stresslet arising from the
presence of the drop in the ambient flow. The particulate stress is simply proportional to
the stresslet [Batchelor (1970); Kim and Karrila (2005)].
2  3
1
6 ½1  3/ðc1 þ Ca c4 ÞE þ Ca c5 E  E  3 IðE : EÞ 7
hr0 i ¼ hp0 iI þ 2/lG6 4
7
5
1
þCa~ c 5 ðE  X  X  EÞ
2
2 2
þ Oð/Ca lG; / lGÞ: (65)

Since we already know the constants c1 ; c4 ; c5 , and c~5 in terms of the shape parame-
ters, whose evolution is governed by Eqs. (37) and (57), we can easily deduce the total
stress.
RHEOLOGY OF DILUTE EMULSIONS WITH SLIP 1567

FIG. 2. (a) The time constant t0c =ðla=cÞ in Eq. (61) and (b) the dimensionless drop deformation b 1 in Eq. (60)
as functions of the dimensionless slip coefficient a and the viscosity ratio k.

For the steady shapes described by the shape parameters in Eq. (62), the constitutive
equation assumes the rather simple form

 
5kð2a þ 1Þ þ 2
hri ¼ hpiI þ 2l G 1 þ / E
2kð5a þ 1Þ þ 2
2 3
f ðk; aÞðE  X  X  EÞ
 
þ / Ca l G4 E : E 5 þ Oð/Ca2 lG; /2 lGÞ; (66)
þgðk; aÞ E  E I
3
1568 A. RAMACHANDRAN AND L. G. LEAL

where
 
1 kð80a þ 19Þ þ 16 2
f ðk; aÞ ¼ ; (67)
40 kð5a þ 1Þ þ 1
and

3½kð80a þ 19Þ þ 16½5k2 ð20a2 þ 4a þ 5Þ þ kð40a þ 41Þ þ 4


gðk; aÞ ¼ : (68)
140½kð5a þ 1Þ þ 13

Note that the tensor E  X  X  E is symmetric and traceless, as is required for the
stresslet.
Based on the work of Frankel and Acrivos (1970), we can extend the above work to
weakly time-dependent linear flows, with the vorticity appearing in the constitutive equa-
tion only in the form of a Jaumann derivative, as follows:
 
5kð2a þ 1Þ þ 2
hr0 i ¼ hp0 iI þ 2lGE þ 2lG/ E
2kð5a þ 1Þ þ 2
  
DE E:E
þlG/Ca f ðk; aÞ þgðk; aÞ E  E I : (69)
Dt 3

Here, DE=Dt is the usual Jaumann derivative, defined as

DE @E
¼ þ u1  rE þ E  X  X  E: (70)
Dt @t

In Sec. III, we will consider specific examples of the ambient flow to evaluate the results
of drop deformation and rheology developed in this section.

III. DROP DEFORMATION AND EMULSION RHEOLOGY FOR SOME


COMMON AMBIENT FLOWS
A. Uniaxial extensional flow
Consider the extensional flow along the x3 axis described by the following ambient
rate-of-strain and vorticity tensors:
2 3
1 0 0
14
E¼ 0 1 0 5; X ¼ 0: (71)
2
0 0 2

To compute the shape of the drop for this flow, we first evaluate the following quantities:
xEx 1 ðx21 þ x22 Þ x23
¼  þ 2;
r2 2 r2 r
3
E:E¼ ;
2 (72)
x  ðE  EÞ  x 1 ðx21 þ x22 Þ x23
¼ þ 2;
r2 4 r2 r
1
E  E ¼ ðdi1 dj1 þ di2 dj2 Þ þ di3 dj3 :
4
RHEOLOGY OF DILUTE EMULSIONS WITH SLIP 1569

where the definitions of x1 , x2 , and x3 are

x1 ¼ r cos / sinh;
x2 ¼ r sin / sinh; (73)
x3 ¼ r cos h:

Using Eqs. (8), (13), (44), (62) and (72), the shape is described by
 
1 ðx21 þ x22 Þ x23
r ¼ 1 þ 3b1 Ca  þ
2 r2 r2
2   3
1 ðx21 þ x22 Þ x23 3
3b
6 2  þ þ b3 7
6 2 r2 r2 2 7
þ Ca2 6    2 7: (74)
4 1 ðx21 þ x22 Þ x23 1 ðx21 þ x22 Þ x23 5
þb4 þ 2 þ b5  þ 2
4 r2 r 2 r2 r

We will use the deformation parameter D originally employed by Taylor to describe


the deformation:
  
rðh ¼ 0Þ  rðh ¼ p=2Þ 9b1 9b2 3b4 3b5 27
D¼ ¼ Ca þ Ca2 þ þ  b21 þ OðCa3 Þ:
rðh ¼ 0Þ  rðh ¼ p=2Þ 4 4 8 8 16
(75)

The deformation parameter is shown in Figs. 3(a)–3(c) as a function of the capillary


number for different slip coefficients and viscosity ratios. We have also shown predictions
of the linear theory (dotted lines), and it can be seen that the quadratic corrections (solid
lines) are a significant change from the linear theory especially at higher capillary numbers
shown in these subfigures. As expected from our discussion in Sec. II, we see that the effect
of slip on deformation is most evident at high viscosity ratios. The shape of the drop
becomes less and less ellipsoidal [see Fig. 3(d)] as the slip coefficient is increased.
The Trouton or extensional viscosity l  for the dilute emulsion is given by

 hr033 i  hr011 i hr033 i  hr022 i


l
¼ ¼
l lG lG

 
5kð2a þ 1Þ þ 2 gðk; aÞ
¼3 1þ/ þ Ca/ þ Oð/Ca2 ; /2 Þ: (76)
2kð5a þ 1Þ þ 2 4

We can see that in the limit of the drop retaining the spherical shape, the Trouton viscos-
ity for the pure suspending fluid, 3l, is only appended by the Einstein-type contribution.
The leading order contribution of the drop deformation to the Trouton viscosity is
3Ca/gðk; aÞ=4, and is shown in Fig. 4 as a function of the viscosity ratio and the slip
coefficient for a fixed capillary number of 0.05. Again, we see that slip mitigates the
effect of drop deformation on the Trouton viscosity, especially at high viscosity ratios,
where the reduction can be by as much as a factor of approximately 7.5 for large slip
coefficients.

B. Simple shear flow


We will now consider the quantities calculated in Sec. II for the simple shear flow
u1i ¼ Gdi1 dj2 xj , i.e., for
1570 A. RAMACHANDRAN AND L. G. LEAL

FIG. 3. The deformation parameter D as a function of the capillary number Ca for different values of the slip
coefficient a and for k ¼ 0:1 (a), k ¼ 1 (b), and k ¼ 10 (c). The square and brace brackets at the right edges of
(a), (b), and (c) denote the linear ½OðCaÞ and quadratic ½OðCa2 Þ results, respectively. The solid blue, dashed
green, and dash-dotted red lines represent the results for a ¼ 0, a ¼ 0:1, and a ¼ 1, respectively. In (d), we have
shown the steady-state shape of the drop for Ca ¼ 0:1 and k ¼ 10 shapes for different slip coefficients.

2 3 2 3
0 1 0 0 1 0
14 14
E¼ 1 0 0 5; X¼ 1 0 0 5: (77)
2 2
0 0 0 0 0 0

For this ambient flow,

x  E  x x1 x2 1 x  ðE  EÞ  x 1 ðx21 þ x22 Þ
2
¼ 2 ; E:E¼ ; ¼ ;
r r 2 r2 4 r2
ðx  E  xÞ2 x1 x2 2 1 x  ðX  EÞ  x 1 ðx21  x22 Þ
¼ ; E  E ¼ ðd i1 dj1 þ d i2 dj2 Þ; ¼ ;
r4 r2 4 r2 4 r2
1 x  ðX  XÞ  x 1 ðx21 þ x22 Þ
XX¼ ; 2
¼ ;
2 r 4 r2
1
E  X þ E  X ¼ ðdi2 dj2  di1 dj1 Þ; (78)
2

where the definitions of x1 , x2 , and x3 are slightly different from the extensional flow
case:
RHEOLOGY OF DILUTE EMULSIONS WITH SLIP 1571

FIG. 4. The Trouton viscosity l


 as a function of the viscosity ratio k for different slip coefficients.

x1 ¼ r cos h;
x2 ¼ r cos / sinh; (79)
x3 ¼ r sin / sinh:

The shape of the drop is



2    3
x1 x2 b3  b3 b4  b4 ðx21 þ x22 Þ
  6 3b2 þ þ 7
x1 x2 r2 2 4 r2
rðh; /Þ ¼ 1 þ Ca 3b1 2 þ Ca2 6
4
7
5
r  
x1 x2 2 b~4 ðx1  x2 Þ
2 2
þb5 þ
r2 4 r2
3
þ OðCa Þ: (80)

In Fig. 5, we have shown the steady-state drop shapes in the 1-2 (flow-gradient) and 1-
3 (flow-vorticity) planes passing through the origin for Ca ¼ 0:1 and k ¼ 10 for different
slip coefficients. It may be seen that deformation from the spherical shape is stronger in
the flow-gradient plane than in the flow-vorticity plane. Slip, however, does not appear to
influence the shapes very strongly even for this high viscosity ratio.
As we did for extensional flows, we have also examined the deformation parameter D,
defined for simple shear flow as

max rðh; / ¼ 0Þ  min rðh; / ¼ 0Þ


h h
D¼ : (81)
max rðh; / ¼ 0Þ  min rðh; / ¼ 0Þ
h h

D provides a measure of the deformation of the drop in the plane of shear, and is shown
in Fig. 6 for different viscosity ratios and slip coefficients. As before, it may be seen that
slip only affects the deformation at higher viscosity ratios. A consistent feature in all the
1572 A. RAMACHANDRAN AND L. G. LEAL

FIG. 5. Drop shapes in the flow-gradient ðx1 ; x2 Þ (a), and flow-vorticity ðx1 ; x3 Þ (b) planes passing through the
origin for Ca ¼ 0:1 and k ¼ 10. The solid blue lines, the dashed green lines, and the dash-dotted red lines are
the results for a ¼0, 0.1, and 1 respectively.

subfigures in Fig. 6, when compared to subfigures (a)–(c) in Fig. 3, is that the deforma-
tion in simple shear flow is, in general, weaker than the deformation in extensional flow.
Thus, extensional flow is more effective at stretching the drops. However, slip appears to
decrease the deformation less strongly in simple shear flow as compared to extensional
flow.
The emulsion viscosity, leff , corresponding to the steady shapes is

 
hr012 i 5kð2a þ 1Þ þ 2
leff ¼ ¼l 1þ/ þ Oð/2 ; /Ca2 Þ; (82)
G 2kð5a þ 1Þ þ 2

while the first and second normal stress differences, N1 and N2 , respectively, are

N1 ¼ hr011 i  hr022 i ¼ lG/Caf ðk; aÞ þ Oð/2 ; /Ca 2


 Þ;
f ðk; aÞ gðk; aÞ (83)
N2 ¼ hr022 i  hr033 i ¼ lG/Ca  þ þ Oð/2 ; /Ca2 Þ:
2 4

Oldroyd (1953) derived the viscosity of a dilute emulsion for arbitrarily time-dependent
linear ambient flows. The viscosity in Eq. (82) is consistent with his result applied to a

FIG. 6. The deformation in the flow-velocity gradient plane defined in Eq. (81) for different values of the slip
coefficient a and for k ¼ 0:1 (a), k ¼ 1 (b), and k ¼ 10 (c). The color code is the same as for prior figures. For
the case in (c), we have chosen a smaller range of Ca. This is because the range of Ca for which our perturbation
expansion is valid reduces for large viscosity ratios. In this case, with k ¼ 10 and Ca > 0:1, we observe cusps
and kinks in the shape of the drop, which confirm the breakdown of our shape approximations. In each subfig-
ure, the solid blue lines, the dashed green lines, and the dash-dotted red lines are the results for a ¼0, 0.1, and 1,
respectively.
RHEOLOGY OF DILUTE EMULSIONS WITH SLIP 1573

simple shear flow. It is also consistent with the result derived by Luo and Pozrikidis
(2008) for slipping solid spheres, in the limit k ! 1. The leading order expression for
the effective emulsion viscosity is independent of the shear rate, but the normal stress dif-
ferences are, to leading order, proportional to the square of the shear rate G (since the
definition of the capillary number also contains G, Ca ¼ lGR=c), as would be expected
for a second-order fluid. The expressions for the normal stresses are consistent with the
results of Schowalter et al. (1968) and Frankel and Acrivos (1970) in the absence of slip.
The effective viscosity leff and the normalized first and second normal stress differences,
N1 =lG/Ca and N2 =lG/Ca, are presented in Figs. 7(b) and 7(c), respectively. An impor-
tant feature of difference between the results for simple shear flow and extensional flows
is that for a fixed volume fraction, the magnitude of the difference from the no-slip result
is much larger for extensional flows [compare Figs. 4 and 7(a)]. This suggests that exten-
sional rheology is likely a better probe of the presence of interfacial slip than simple
shear flow.
The viscosity and the first normal stress difference show similar trends—These quanti-
ties are relatively unaffected by slip for small viscosity ratios, but diminish for higher vis-
cosity ratios, with the decrease being greater for higher slip coefficients. We have also

FIG. 7. (a) The effective emulsion viscosity leff =l; (b) first normal stress difference, N1 , normalized by
lG/Ca; (c) second normal stress difference, N2 , normalized by lG/Ca; and (d) the ratio of the second normal
stress difference to the first, N2 =N1 , as functions of the viscosity ratio, k, and the slip coefficient, a.
1574 A. RAMACHANDRAN AND L. G. LEAL

shown the ratio N2 =N1 as a function of the viscosity ratio k and the slip coefficient a in
Fig. 7(d). We see that in the absence of slip, the magnitude of N2 relative to N1 ranges
from 0.45 in the inviscid limit to 0.22 for large viscosity ratios. With the introduction of
slip, the ratio is unaffected at low viscosity ratios, but increases for higher viscosity
ratios. It is notable that disparity in the first and second normal stress differences is less
pronounced in emulsions, particularly with slip, than for a single-component entangled
polymer, for which N2 =N1 ranges from 0 to around 0.25 [Larson (1999)].
To conclude this subsection, and as an aside to our general approach of considering
Ca  1, we consider the case when the drops are spherical (b1 through b5 are 0), and the
interfacial tension is small (so that capillary contributions to the stress are negligible).
This case will also be useful in our discussion of oscillatory flows at high frequencies in
Sec. II C. In this limit, we can use (69) to determine the viscosity at the instant shearing
begins with the Eq. (36) used to calculate c1 . The result is

 
ð19k þ 16Þðk  1Þ þ 2akð19k  40Þ
leff 1 ¼l 1þ5 / : (84)
ð19k þ 16Þð2k þ 3Þ þ 10akð19k þ 24Þ

We observe that when a is 0, leff reduces to the analytical expression derived by


Batchelor and Green (1972). Also, in the limit of large k, the result is identical to that of
Luo and Pozrikidis (2008).
The first effect of the volume fraction on the effective viscosity in this limiting case,
ðleff 1  lÞ=l/, is shown in Fig. 8 as a function of interfacial slip and viscosity ratio. It
can be seen in Fig. 8 that the effective viscosity for spherical drops is lower at this initial
instant than the suspending fluid viscosity for small k, and higher than the suspending
fluid viscosity for large k. The viscosity ratio, kc , at which the transition between these
two behaviors occurs depends on the slip coefficient; the larger the slip coefficient, the
larger the value of k for the transition

FIG. 8. The emulsion effective viscosity as a function of the viscosity ratio and slip coefficient for spherical
drops at the instant of initiating shear when the contribution of interfacial to the stress is negligible. The black
horizontal line denotes leff 1 ¼ l.
RHEOLOGY OF DILUTE EMULSIONS WITH SLIP 1575

pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
3 þ 80a þ 6400a2 þ 2912a þ 1225
kc ¼ : (85)
76a þ 38

An interesting result that can be seen from Eq. (84) is that, unlike the case of small
capillary numbers where results in the limits a ! 1 and k ! 0 are identical, the viscos-
ity at the initiation of shear for spherical drops in these two limits is not the same. The
results are


ð19k  40Þ
leff ja!1 ¼l 1þ / ; (86)
ð19k þ 24Þ

and


5
leff jk!0 ¼l 1 / : (87)
3

We see that there is still a residual dependence on the viscosity ratio for large slip
limit. This can be explained as follows: At small capillary numbers, both the bubble and
the perfectly slipping drop assumed steady shapes. For a bubble, the contribution of the
bubble to the total viscous dissipation is identically zero due to its negligible viscosity.
For a perfectly slipping drop, the contribution of the drop to the viscous dissipation rate
is also zero for steady drop shapes (Appendix C), because the rate of strain within the
drop vanishes. Thus, the viscous energy dissipation rate in the suspending fluid is exactly
the same for bubbles and slipping drops with the same steady shape. In the special case
discussed above, the viscous dissipation within the bubble is still zero, because of the
negligible viscosity. However, for a perfectly slipping drop, although tangential forces
applied by the suspending fluid are ineffective in producing motion within the drop, nor-
mal forces do produce motion, which will of course cause the drop to deform after the
initial instant where we have assumed it to be spherical. The motion of the drop fluid pro-
duced by these normal forces makes a positive contribution to the total instantaneous vis-
cous dissipation, and therefore to the emulsion viscosity.

C. Small-amplitude oscillatory shear flow


Finally, let us now examine the linear viscoelastic response of the emulsion. To do
this, we consider the characteristic shear rate G in Eq. (1) to be

G ¼ G0 expðix0 t0 Þ ¼ G0 expðixtÞ; (88)

where

x0 al
x¼ ; (89)
c

x0 being the frequency of oscillation.


We are interested in the linear response of the stress to this oscillatory flow,

lG0
hr012 i ¼ G; (90)
x

where
1576 A. RAMACHANDRAN AND L. G. LEAL

G ¼G0 þ ixG00 : (91)

For small capillary numbers G0 al=r  1, we only need to consider the zeroth order
solution presented in Sec. II B to determine the viscoelastic response. Comparing (90)
with the leading order terms in the general expression for the volume averaged emulsion
stress (65), we find

G ¼ xð1  3/c1 Þ: (92)

Hence, to determine G* from the above equation, we need c1 from Eq. (36), which, in
turn, requires b1 . It can be shown that for the oscillatory simple shear flow in Eq. (88),
the expression (B3) for b1 is simply modified to

@b1 ðb1  b 1 Þ
þ ixnb1 ¼  ; (93)
@t tc

where b 1 and tc are defined in Eqs. (60) and (61), respectively. The asymptotic long time
solution to this differential equation is

b 1
b1 ¼ : (94)
ixtc þ 1

This yields c1 as

5fix½19k2 ð2a þ 1Þ  kð80a þ 3Þ  16 þ 4½5kð2a þ 1Þ þ 2g


c1 ¼  ; (95)
3fix½38k2 ð5a þ 1Þ þ kð240a þ 89Þ þ 48 þ 40½kð5a þ 1Þ þ 1g

which means that G* is



 
c ix½19k2 ð2a þ 1Þ  kð80a þ 3Þ  16 þ 4½5kð2a þ 1Þ þ 2
G ¼ ix 1 þ 5/ :
a ix½38k2 ð5a þ 1Þ þ kð240a þ 89Þ þ 48 þ 40½kð5a þ 1Þ þ 1
(96)

Since the suspending fluid is taken to be Newtonian in our analysis, the complex modu-
lus shows, to leading order, a purely viscous response. The leading order elastic response
arises due to deformation of the drops and is proportional to the volume fraction. It is in-
structive to examine the moduli G0 and G00 in the low and high frequency limits

ðc=aÞx2 /½kð80a þ 19Þ þ 162


G0 jx1 ¼ ;
80½kð5a þ 1Þ þ 12
20ðc=aÞ/½kð80a þ 19Þ þ 162
G0 jx1 ¼ ;
½38k2 ð5a þ 1Þ þ kð240a þ 89Þ þ 482

(97)
½5kð2a þ 1Þ þ 2
G00 jx1 ¼ ðc=aÞx 1 þ / ;
½2kð5a þ 1Þ þ 2


½19k2 ð2a þ 1Þ  kð80a þ 3Þ  16
G00 jx1 ¼ ðc=aÞx 1 þ 5/ :
½38k2 ð5a þ 1Þ þ kð240a þ 89Þ þ 48
RHEOLOGY OF DILUTE EMULSIONS WITH SLIP 1577

FIG. 9. The contribution of the drops to the storage modulus (a) and loss modulus (b) of the dilute emulsion for
k ¼ 10 in an oscillatory shear flow experiment.

At low frequencies, the storage modulus varies as x2 , while at high frequencies it asymp-
totes to a constant, as may be seen in Fig. 9(a). One can see that slip affects the elastic
response only at high frequencies, and the reduction in the elastic modulus is significant at
high viscosities, being about 10% for a slip coefficient of 0.1 and k ¼ 10. The contribution
of the drops to the viscous response, which we quantify as g ¼ ½G00 ðc=aÞx=½ðc=aÞx/ is
shown in Fig. 9(b). It is notable from the expressions in Eq. (97) that in the low frequency
limit, the viscous response is proportional to the effective viscosity for the drop deformation
at steady state [see Eq. (82)], while in the high frequency limit, it is proportional to the effec-
tive viscosity with zero surface tension or high capillary numbers [see Eq. (84)]. The latter
effective viscosity is also the viscosity of an emulsion of drops with zero interfacial tension.

IV. DISCUSSION
In this section, we speculate on the implications of our calculations of the rheology of
a dilute emulsion for the idea in the literature that slip is responsible for the anomalously
low viscosities of uncompatibilized polymer blends.
1578 A. RAMACHANDRAN AND L. G. LEAL

Before we proceed with this discussion, we digress briefly to discuss what constitutes
anomalousness in the behavior of blend viscosity. The “irregularities” in the measured
viscosities of polymer blends have been quantified by defining a deviation of the experi-
mental viscosity data with respect to a predicted or expected value. This predicted value
has often been obtained from an empirical expression such as the log-additive mixing
rule [e.g., see Utracki and Kamal (1982)], or from expressions derived on the basis of
assuming a simple morphology for the blend [Lin (1979); Bousmina et al. (1999)]. But as
explained by Han (2009), these empirical or theoretical expressions account neither for
the correct morphology of the blend nor for the history of the flow field to which the
blend is subjected. It is questionable, therefore, that these expressions represent the
expected viscosity. For example, the log-additive mixing rule defines the viscosity, leff ,
of a blend of two immiscible polymeric fluids A and B as

1/ /
leff ¼ lA lB ; (98)

where lA and lB are viscosities of A and B, respectively, and / is the volume fraction of
the phase B. This empirical equation predicts that the viscosity varies monotonically
from lA to lB as / is varied from 0 to 1. So, if lA ¼ lB , the viscosity should not be a
function of the volume fraction at all, a prediction that would be correct only if interfacial
tension effects were absent; this is rarely the case in the uncompatibilized blends.
Unfortunately, the exact relationships between the rheology of an emulsion, the viscosity
ratio, the flow field, the interfacial tension, and the volume fraction of the dispersed phase
still elude us. It is difficult, therefore, to formally define/quantify a deviation or irregular-
ity when we do not know what the exact behavior should be. However, there is one type
of behavior that can be truly classified (qualitatively) as irregular and that has attracted
the most attention in the literature—the addition of drops with a higher viscosity than the
suspending fluid leading to a decrease in the viscosity relative to the suspending fluid.
We choose to focus on this case in our discussion.
A reduction in the viscosity upon the inclusion of a more viscous drop phase has been
observed in literature for polymer blends. Some “negative-deviation” blends presented
by Utracki and Kamal (1982) and Utracki (1983) show this trend. For example, Utracki
and Kamal (1982) reported a summary of experiments of Carley, in which a relative vis-
cosity of 0.35 upon the addition of 10% by weight of polyamide-12 ðk ¼ 3:3Þ to polysty-
rene at 210
C and a shear rate of 10 s1 was observed. Shih (1976) reported capillary
rheometry data for Viton/ethylene propylene diene monomer rubber (EPDM) blends, in
which the introduction of a small concentration (<2.5 wt. %) of the more viscous EPDM
(viscosity: 25 000 Pa s) to Viton (viscosity 7500 poise) leads to the blend viscosity
dropping precipitously to about 3000 poise, a relative viscosity of 0.4. In another study,
Chuang and Han (1984) examined polyvinylidene difluoride/polymethyl methacrylate
(PVDF/PMMA) blends in a cone and plate viscometer at 230
C, and observed that even
at shear rates where the two constituent polymers were Newtonian (0.018 s1), a 20%
mixture by weight of the more viscous (3105 Pa s) PMMA in the less viscous PVDF
(2875 Pa s) has a viscosity of 2560 Pa s ðlrel ¼ 0:89Þ.
It is known that when a blend prepared with a low volume fraction of the dispersed
phase is subjected to a simple shear flow, the steady-state morphology of the mixture is
simply a suspension of drops of one polymer in the other, with the drop size distribution
determined by drop breakup and coalescence rates [Tucker and Moldenaers (2002)].
Generally speaking, the addition of drops to a suspending fluid contributes in two ways
to the changes in the effective viscosity of the overall blend. First, some of the suspend-
ing fluid is replaced by fluid with a different viscosity. Second, the drop creates a local
RHEOLOGY OF DILUTE EMULSIONS WITH SLIP 1579

perturbation to the flow field in its neighborhood, and this changes the rate of dissipation
compared to that associated with the mean motion of the suspending fluid in the absence
of drops. The first effect produces either an increase or a decrease in the viscosity of
Oð/Þ; depending on whether the viscosity ratio k is greater than or less than 1.
The question is whether the trends in the variation of the viscosity with volume frac-
tion from inclusion of slip in our dilute emulsion theory may provide some insight as to
whether the measured viscosities of polymer blends should be considered as anomalous.
We consider this point with the further caveat that the majority of the experimental sys-
tems was studied at disperse phase concentrations above that where we expect the dilute
theory to apply. Dilute suspension theories in the absence of slip and at low capillary
number lead to the conclusion that the addition of drops that take a steady shape always
leads to an increase of the effective viscosity, for any viscosity ratio, i.e.,
ðdleff =d/Þj/¼0 > 0 [cf. Eq. (82) when a ¼ 0]. Hence, when the effect of interfacial ten-
sion is strong enough to maintain the drop in a near-sphere configuration, the additional
dissipation in the suspending fluid due to the presence of the drop dominates, and the vis-
cosity increases, irrespective of the viscosity ratio. Indeed, in the absence of slip, prior
work on dilute emulsions shows that a decrease in viscosity can only be achieved for
drops that attain a steady shape if the viscosity ratio is low, and the capillary number is
large enough to allow for finite deformation of the drop shape. In this limit, the change in
viscosity is primarily due to the reduction in the viscous dissipation arising from the
replacement of the more viscous suspending fluid with the less viscous drop fluid, which
leads to a decrease in the viscosity of the emulsion relative to the suspending fluid [see
Eq. (84)]. As an example of this limit, Manga and Loewenberg (2001) demonstrated
numerically that a dilute suspension of bubbles achieves a steady shear viscosity less
than that of the suspending liquid at capillary numbers less than 1 (the bubbles reach
steady shapes at these capillary numbers).
The trend of a decrease in viscosity via the addition of more viscous drops (i.e., of a
blend with a viscosity lower than the viscosity of other of the bulk fluids) is thus not
anticipated from dilute emulsion theory, unless we introduce slip at the liquid-liquid
interface, and this is the reason offered frequently in the literature for these low blend vis-
cosities. For example, Macosko and co-workers [Zhao and Macosko (2002); Lee et al.
(2009)] co-extruded alternating layers of polystyrene/polypropylene into a slit die and
demonstrated that the measured pressure drops were smaller than those expected from
the pure fluid viscosities alone. They also showed that the pressure drop decreased as the
number of layers and therefore the interfacial area increased, which suggests an underly-
ing interfacial-slip mechanism. Since the slip layer at the interface between the two poly-
mers can have a viscosity that is orders of magnitude smaller than the viscosities in the
bulk regions of the polymers [e.g., see Park et al. (2003)], this can lead to low emulsion
effective viscosities, lower even than the viscosities of the individual polymers.
For this multiple-alternating layer blend morphology considered experimentally by
Macosko and coworkers [Zhao and Macosko (2002); Lee et al. (2009)] and Lam et al.
(2003), and theoretically by Lin (1979), Lyngaae-Jorgensen and Thomsen (1988), and
Bousmina et al. (1999), hydrodynamic stresses are transmitted through all the alternating
layers. However, in a droplet emulsion, which is the likely morphology at low volume
fractions, hydrodynamic stresses are transmitted through the drops and the suspending
fluid, and also purely through the suspending fluid. Therefore, the magnitude of the impact
of slip on blend viscosity when the morphology is an emulsion is likely to be smaller.
The present theory shows that for drops that are near spherical, slip cannot account for
the observed decrease in viscosity. Indeed, if the drop has a steady shape, (82) shows that
the viscosity increases at all viscosity ratios, even in the limit a ! 1: Furthermore, even
1580 A. RAMACHANDRAN AND L. G. LEAL

the instantaneous viscosity of a spherical drop that is deforming increases unless the vis-
cosity ratio is small [cf. Eq. (84)]. One additional result that have already shown in
Sec. III B is that when perfectly slipping drops achieve steady shapes, the viscosity of the
emulsion is identical to the viscosity of a suspension of bubbles (or any low viscosity
fluid) of the same size/size distribution.2 This suggests that we can therefore invoke all
the theoretical work developed for bubbles to explore the low viscosities in an emulsion
with slipping drops. We particularly refer to the work of Manga and Loewenberg (2001).
Using a boundary integral technique, they numerically calculated the viscosity of an
emulsion of bubbles and found that for Ca < 1, when the bubbles shapes are steady, the
relative viscosity of the suspension is less than 1. For example, the emulsion relative vis-
cosity for Ca ¼ 1 is about 1  0:25/ for low volume fractions; for a volume fraction of
0.1, the relative viscosity would be 0.975. When Manga and Loewenberg performed their
simulations at a higher volume fraction of 0.3 at Ca ¼ 1, the emulsion relative viscosity
was found to be 0.93, but this is still not as low as the experimentally observed relative
viscosities.
We thus see that although the decrease in the emulsion viscosity upon the introduction
of slip in the theory at moderate capillary numbers is qualitatively consistent with experi-
ment, the magnitude of the change at low volume fractions is not nearly as strong as
observed experimentally for blends, even when the drop interfaces are assumed to be per-
fectly slipping. From these results, we suggest that slip alone is incapable of explaining
the anomalous behavior of lowering the viscosity upon the addition of a more viscous
phase at low volume fractions.
An obvious question is what other possibilities may exist to explain the residual dis-
crepancy. Perhaps, one is the appearance of highly elongated bubbles in polymers
sheared at very high stresses (order of MPa) arising from cavitation due to dissolved
gases and impurities. This cavitation, which was observed by Archer et al. (1997) for
“pure” polystyrene, coincided with a sharp decrease in the measured shear stress. Pre-
sumably, drops of a second phase in a pure polymer could provide the sites for cavitation
in a sheared polymer blend in an analogous manner. However, these bubbles are only
observed after long periods of shearing, and the residence times in capillary rheometer
and cone and plate measurements referred to in this paper were short. Nevertheless, this
possibility still needs to be eliminated by a direct experiment.
A second possibility is suggested by the fact that the stresses imposed in the experi-
ments are extremely large. For example, the shear rate in the work of Shih (1976) was
14 s1, which corresponds to a shear stress of about 100 kPa for a suspending fluid viscos-
ity of 7500 poise. At such high stresses, assuming a critical capillary number of unity3
and an interfacial tension of 5 mN/m, the drop sizes one should obtain should be less than
50 nm. A similar conclusion may be drawn from the experiments of Carley reported by
Utracki and Kamal (1982), in which the shear stresses applied were of the order of
100 kPa. The upshot of these observations is that the stresses in these experiments were
large enough to produce drop sizes in the range of tens of nanometers, with even smaller
satellite drops.

2
The problem definition for the flow field, and therefore, the viscous dissipation, in the suspending fluid for a
perfectly slipping drop at steady state is identical to that for a bubble, and the viscous dissipation within a per-
fectly slipping drop at steady state is identically zero (see Appendix C).
3
The viscosity ratio in these experiments was just above 3, and therefore, the critical capillary number is
expected to be O(1) [Tucker and Moldenaers (2002)].
RHEOLOGY OF DILUTE EMULSIONS WITH SLIP 1581

The presence of nanometer-sized droplets in the emulsion suggests a re-examination


of the continuum assumption that is built into our flow models for the drop and suspend-
ing fluids. The breakdown of the continuum ideas may, thus, not be restricted simply to
the interface (which leads to slip), but may extend into the bulk of the two fluids. This
breakdown could occur in two ways. First, since the molecular weight of the drop fluid is
typically high, the radius of gyration of the polymer may be large, of the order of a few
tens of nanometers. The droplet size may, thus, become comparable to radius of gyration
of the droplet polymeric fluid. Second, since the suspending fluid employed in the experi-
ments displaying the low emulsion effective viscosities is also typically a high molecular
weight polymeric fluid, it is possible that the droplets produced in the experiments are of
sizes are comparable to length scales characteristic of the structure of the suspending
fluid. Evidence of such effects has already been reported in the literature. It has been
observed, for example, that the addition of small rigid nanoparticles (<10 nm) can lower
the viscosity of a polymeric liquid [Tuteja et al. (2007)] significantly. Although the nano-
droplet sizes calculated above appear to be larger than the nanoparticle diameters in the
study of Tuteja et al. mentioned above, they may be sufficiently small to produce a reduc-
tion in the viscosity for the polymers employed in the experiments. A detailed analysis of
all of the above possibilities would, however, need to be the subject of a new
investigation.

V. CONCLUSIONS
In this paper, we have presented the constitutive equation for the rheology of a dilute of
emulsion of drops of one Newtonian fluid suspended in a second Newtonian fluid, with the
essential difference from prior literature that the interface between the two Newtonian flu-
ids shows slip. We assume that the drop undergoes small deformations (low Capillary num-
bers), and interfacial slip is modeled using a simple Navier slip boundary condition. Slip
leads to moderation of all the material functions: Viscosity and normal stress differences in
simple shear flow, the Trouton viscosity in extensional flows, and the complex viscosity in
linear viscoelasticity measurements. The effect of slip on droplet deformation and rheology
is, however, most pronounced for extensional flows, and experimentation in such flows is a
likely avenue to characterize the effect of interfacial slip.
One observation that formed the basis of this work was the data in the literature sug-
gesting that the viscosity of a polymer blend can be much lower than the suspending fluid
viscosity even when only a small weight fraction of a more viscous polymer is introduced
into the suspending fluid, with emulsion relative viscosities being as low as 0.4 in some
experiments. Interfacial slip has often been purported to be the reason for this discrep-
ancy in the literature. However, our analysis suggests that although slip can reduce the
viscosity of an emulsion below that of a suspending fluid, it cannot explain the observed
magnitudes of the decrease in the viscosity, even if the interface is assumed to be per-
fectly slipping. This indicates that there probably are other mechanisms influencing the
decrease in emulsion viscosity in addition to interfacial slip.

ACKNOWLEDGMENTS
Dr. Ramachandran is grateful to the Department of Chemical Engineering and
Applied Chemistry at University of Toronto for funding. The authors thank Professor
Ronald Larson for providing the reference to his work describing cavitation phenomena
in polymers sheared at high stresses.
1582 A. RAMACHANDRAN AND L. G. LEAL

APPENDIX A: EQUATIONS FOR THE CONSTANTS AT THE FIRST ORDER


OF EXPANSION

1. Equality of normal component of interfacial velocity

 
6
ð3c4  9c8 Þ  d7 þ d4 ¼ 0;
7
 
6 120 6
ðc5 þ 6c6  30c11  c12 þ 3c13 Þ   d10 þ d11  d12 þ d14  d13 ¼ 0;
5 7 5
2 3
ð3c5  48c6 þ 300c11  9c13 Þ   
4 48 1200 18
5 10
 6b1 ð1 þ 6c3 Þ  d2 þ d1 ¼ 0;
 d10  d11 þ 3d12  3d14 þ d13 7
5 7 5
  
38
½ð75c6  525c11 Þ  ð15d10 þ 300d11 Þ þ 3b1 ð3  6c1 þ 48c3 Þ  3d2 þ d1 ¼ 0;
7
 
_ 12 18
c 5 þ 6~
3~ c 13 þ 6b1  d~9 þ d 9  d~10 þ d~13  6b1 d~2 ¼ 0;
c 10  9~
5 5
 
 6  
c 5  c12 þ 3c 13  d 12  d 13 þ d 14 ¼ 0;
5
 
 18  
c 5  9
3 c 13  3d 12 þ d 13  3d 14 ¼ 0:
5

2. Jump in the tangential component of interfacial velocity


( ) ( )
3b1 ð3  42c3 þ 6c1 Þ  3b1 ð6d1  3d2 Þ 6b1 ð4  27c1 þ 234c3 Þ
¼a ;
þ½ð30c6 þ 420c11 Þ  ð15d10 þ 420d11 Þ þð450c6  5040c11 Þ
8  9
> ( )
< 3b ð6c þ 30c  5Þ  3b  62 d þ 5d > = 6b1 ð6 þ 21c1  186c3 Þ
1 1 3 1 1 2
7 ¼a ;
>
: >
; þð450c6 þ 5040c11 Þ
þ½ð30c6  420c11 Þ  ð420d11 þ 15d10 Þ
2    3
10
þ6b1 ð1 þ 6c3 Þ  d2 þ d1 ( )
6 7 7 6b1 ð4 þ 6c1  36c3 Þ
6 7
6 7¼a
4 ð12c6 þ 120c11  6c13 Þ 5 þð6c5 þ 156c6  1440c11 þ 48c13 Þ
ð6d10 þ 120d11  3d12 þ 3d14  6d13 Þ
" # ( )
ð12c6  120c11 þ 6c13 Þ 6b1 ð2  12c3 Þ
¼a
ð6d10  120d11 þ 3d12  3d14 þ 6d13 Þ þð6c5  156c6 þ 1440c11  48c13 Þ
_
ð6~ c 13  6b1 Þ  ðd~9  d 9  6d~10  6d~13 þ 6b1 d~2 Þ ¼ að6~
c 10  6~ c 5  18~
c 10 þ 48~
c 13 Þ
ð~ ~
c 7 Þ  ðd 7 Þ ¼ a3~
c7
ð~
c 9 þ 6~ c 13 Þ  ðd~9 þ 6d~10 þ 3d~13 Þ ¼ að3~
c 10 þ 3~ c 5 þ 3~
c 9  24~c 10  24~
c 13 Þ
ð~ c 13 Þ  ðd9  3d~13 Þ ¼ að3~
c 9  3~ c 5 þ 3~
c 9  6~
c 10 þ 24~c 13 Þ
ð15~ c 10 Þ  ð15d~10 Þ ¼ að75~c 10 Þ
c 13  ð3d12 þ 6d13  3d14 Þ ¼ að6
6 c 5  48c 13 Þ:
RHEOLOGY OF DILUTE EMULSIONS WITH SLIP 1583

3. Jump in the interfacial normal stresses


  
10
6c8  d7 þ d4 ¼að6c4 48c8 Þ;
7 
6
ð18c4 þ72c8 Þk  d4 2d7 ¼12b2;
 7  
58
6b1 ð4þ39c1 276c3 Þk d1 4d2
7
þ½ð810c6 þ6300c11 Þkð90d10 þ1350d11 Þ¼90b21 þ18b5;
8    9
> 64 >
>
< 6b1 ð412c1 þ96c3 Þk  d1 þ4d2 >
=
 7  
>
> 252 5400 18 >
: þ ð18c5 þ504c6 3600c11 þ72c13 Þk d10  d11 þ6d12 6d14  d13 > ;
5 7 5
¼4b4 8b5 ;  
24 540 6
ð6c5 60c6 þ360c11 þ4c12 24c13 Þk  d10 þ d11 2d12 þ d13 ¼2b3 2b4 ;
 5 7 5
12 18
ð18~c 5 48~ c 10 þ72~ c 13 Þk 2d~9 þ2d9 þ d~10  d~13 ¼4b~4 ;
 5 5
6
c 5 þ4
6 c 12 24 c 13 k d 13 2d12 ¼2ðb3 þ b4 Þ
 5 
18 
18 c 5 þ72 c 13 k  d 13 þ6d 12 6d 14 ¼4b4 :
 
5

4. Continuity of interfacial tangential stress


 
32
ð6c4  48c8 Þ  k d4  2d7 ¼ 0;
7
  
96
6b1 ð4  27c1 þ 234c3 Þ  k d1  4d2
7
þ ½ð450c6  5040c11 Þ  kð90d10 þ 2880d11 Þ ¼ 0;
  
128
6b1 ð6 þ 21c1  186c3 Þ  k  d1 þ 6d2
7
þ½ð450c6 þ 5040c11 Þ  kð90d10  2880d11 Þ ¼ 0;
  
52
6b1 ð4 þ 6c1  36c3 Þ  k d1  4d2
7
2 3
ð6c5 þ 156c6  1440c11 þ 48c13 Þ
6  7
þ4 156 5760 96 5 ¼ 0;
k  d10 þ d11  6d12 þ 6d14  d13
5 7 5
  
20
6b1 ð2  12c3 Þ  k 2d2  d1
7
2 3
ð6c5  156c6 þ 1440c11  48c13 Þ
6  7
þ4 156 5760 96 5 ¼ 0;
k d10  d11 þ 6d12  6d14 þ d13
5 7 5
c~7 ¼ 0;
1584 A. RAMACHANDRAN AND L. G. LEAL

 
~
_ 48 ~ 48 ~
ð3~
c 5 þ 3~ c 9  24~ c 10  24~ c 13 Þ  k d 9 þ d 9 þ d 10 þ d 13 ¼ 0;
 5 5 
_ 12 48
ð3~ c 5 þ 3~ c 9  6~ c 10 þ 24~ c 13 Þ  k d~9  d 9 þ d~10  d~13 ¼ 0;
5 5
ð75~ c 10 Þ kð30d~10 Þ ¼ 0; 
_ 36 96
ð6~ c 5  18~ c 10 þ 48~ c 13 Þ  k 2d~9  2d 9  d~10  d~13 ¼ 0;
 5  5
96
6c5  48c13  k d13 þ 6d12  6d14 ¼ 0:
5

APPENDIX B: DERIVATION OF THE STRESSLET FROM THE MULTIPOLE


EXPANSION OF THE VELOCITY FIELD
For a dilute suspension, the volume averaged stress is
/
hrij i ¼ hpidij þ 2lhe1
ij i þ Sij ; (B1)
Vd
where Sij is the stresslet due to the presence of a drop in the ambient field. The stresslet can
be obtained by simply examining the leading order disturbance velocity in the ambient field.
The suspending fluid velocity field to leading order for this problem is (preserving
only the 1=r 2 terms)
2    c 3
3c4 5
  þ 5 xi Ejk xj xk þ  3 xi Ejk Ekj
3c1 6 r r 7
ui ¼ Eij xj þ 5
xi Ejk xj xk þ Ca6
4     7; (B2)
5
r 3c5 3~c5
þ 5 xi Ejk xk Ejl xl þ x i x j X jk E kl x l
r r5

which may be written compactly as


2  3
1
6 ðc1 þ Ca c4 ÞEjk þ Ca c5 Ejl Ekl  djk Eml Eml 7x d 
ui ¼ Eij xj  6 3 7 i jk  3xi xj xk :
4 Ca 5 r3 r5
þ c~5 ðXjl Elk þ Xkl Elj Þ
2
(B3)
In the above manipulation, we have used Eii ¼ 0 and Xij Eji ¼ 0.
We recall from the boundary integral equations for creeping flow [e.g., see Leal
(2007)] that the leading order 1=r 2 disturbance terms arise from both the single layer and
the double layer terms. In particular,

ð ð
1 @Gij ðxÞ 1
ui ¼ Eij xj þ fj xk dS  Rijk ðxÞ uj nk dS
8pl @xk 8p
ðS  S
 
1 @Gij ðxÞ 1 l @Gij ðxÞ @Gik ðxÞ
¼ Eij xj þ fj xk dS  Pi ðxÞdjk þ þ
8pl @xk S 8pl 2 @xk @xj
ð
 uj nk dS ðBy definitionÞ
S (B4)
ð   
1 @Gij ðxÞ 1 1 @Gij ðxÞ @Gik ðxÞ
¼ Eij xj þ fj xk dS  þ
8pl @xk S 8p 2 @xk @xj
ð

 uj nk dS From uj nj ¼ 0 at the interface :
S
RHEOLOGY OF DILUTE EMULSIONS WITH SLIP 1585

Taking the symmetric and antisymmetric parts of the single layer term, we get
 ð  
1 1 @Gij ðxÞ @Gik ðxÞ fj xk þ fk xj 1 ðuj nk þ uk nj Þ
ui ¼ Eij xj þ þ  djk fl xl  l dS
8pl 2 @xk @xj S 2 3 2
 ð  
1 1 @Gij ðxÞ @Gik ðxÞ fj xk  fk xj
þ  dS
8pl 2 @xk @xj S 2
   
1 1 @Gij ðxÞ @Gik ðxÞ 1 1 @Gij ðxÞ @Gik ðxÞ
¼ Eij xj þ þ Sjk þ  Tjk :
8pl 2 @xk @xj 8pl 2 @xk @xj
(B5)

In the above equation, the disturbance components are the stresslet and the rotlet parts for
the drop.
We know that only the stresslet will contribute to the rheology. Now, noting that
   
1 @Gij ðxÞ @Gik ðxÞ xi djk 3xi xj xk
þ ¼  ; (B6)
2 @xk @xj r3 r5

we can get the stresslet by simply locating terms of the above form in the velocity field
(B3). Thus,
   
1 Ca
Sjk ¼ 8plVd ðc1 þ Ca c4 ÞEjk þ Ca c5 Ejl Ekl  djk Elm Eml þ c~5 ðXjl Elk þ Xkl Elj Þ :
3 2
(B7)

This gives us the rheology according to (B1). Therefore, we can simply get the rheology
for such types of problems by examining the disturbance velocity field in the suspending
fluid.

APPENDIX C: THE VISCOUS ENERGY DISSIPATION RATE WITHIN A


PERFECTLY SLIPPING DROP
For a drop occupying a volume V and bounded by a surface S, the rate of viscous dis-
sipation within the drop is [Kim and Karrila (2005)]
ð ð
U¼ r ^ ij e^ij dV ¼ u^i r
^ ij nj dS: (C1)
V S

Since the drop is perfectly slipping a ! 1, the tangential shear stress on the drop surface
is zero, and therefore
^ ij nj ¼ h ni ;
r (C2)

where h is a scalar function of the interfacial position. Hence, we have


ð
U ¼ h^ u i ni dS: (C3)
S

When the drop assumes a steady shape in the flow, the normal component of the velocity is
zero, i.e., u^i ni ¼ 0, as required by the kinematic condition. Equation (C3) yields, therefore,

U  0: (C4)
1586 A. RAMACHANDRAN AND L. G. LEAL

Thus, the energy dissipation


Ð within a perfectly slipping drop at steady state is identically
zero. Since U ¼ 2k V e^ij e^ij dV, and e^ij e^ij 0, U ¼ 0, implies that e^ij ¼ 0. The velocity
field u^ can thus be at the most, a rigid body rotation, but since u^i ni on the drop surface is
zero, u^  0 (unless the drop is a sphere). This result is independent of the volume fraction
of the perfectly slipping drops.

References
Adhikari, N. P., and J. L. Goveas, “Effects of slip on the viscosity of polymer melts,” J. Polym. Sci., Part B:
Polym. Phys. 42, 1888–1904 (2004).
Ajdari, A. “Slippage at a polymer/polymer interface: Entanglements and associated friction,” C. R. Acad. Sci.,
Ser. II: Mec., Phys., Chim., Sci. Terre Univers 317, 1159–1163 (1993).
Archer, L. A., D. Ternet, and R. G. Larson, “‘Fracture’ phenomena in shearing flow of viscous liquids,” Rheol.
Acta 36, 579–584 (1997).
Batchelor, G. K., “The stress system in a suspension of force free particles,” J. Fluid Mech. 41, 545–570 (1970).
Batchelor, G. K., and J. T. Green, “The determination of the bulk stress in a suspension of spherical particles to
order c2,” J. Fluid Mech. 56, 401–427 (1972).
Barthès-Biesel, D., and A. Acrivos, “Deformation and burst of a liquid droplet freely suspended in a linear shear
field,” J. Fluid Mech. 61, 1–22 (1973).
Bousmina, M., J. F. Palierne, and L. A. Utracki, “Modeling of structured polyblend flow in a laminar shear
field,” Polym. Eng. Sci. 39(6), 1049–1059 (1999).
Brochard-Wyart, F., and P. G. de Gennes, “Sliding molecules at a polymer/polymer interface,” C. R. Acad. Sci.,
Ser. II: Mec., Phys., Chim., Sci. Terre Univers 317, 13–17 (1993).
Chuang, H., and C. D. Han, “Rheological behavior of polymer blends,” J. Appl. Polym. Sci. 29, 2205–2229
(1984).
Cox, R. G., “The deformation of a drop in a general time-dependent fluid flow,” J. Fluid Mech. 37, 601–623
(1969).
De Gennes, P. G., “Viscometric flows of tangled polymers,” C.R. Acad. Sci. Paris B 288, 219–220 (1979).
Frankel, N. A., and A. Acrivos, “The constitutive equation for a dilute emulsion,” J. Fluid Mech. 44, 65–78
(1970).
Furukawa, H., “Sliding along the interface of strongly segregated polymer melts,” Phys. Rev. A 40, 6403–6406
(1989).
Goveas, J. L., and G. H. Fredrickson, “Apparent slip at a polymer-polymer interface,” Eur. Phys. J. B 2, 79–92
(1998).
Hakimi, F. S., and W. R. Schowalter, “The effects of shear and vorticity on deformation of a drop,” J. Fluid
Mech. 98, 635–645 (1980).
Han, C. D., “On modeling the composition dependence of the bulk viscosity of immiscible polymer blends,”
Polym. Eng. Sci. 49, 1671–1687 (2009).
Helfand, E., and Y. Tagami, “Theory of the interface between immiscible polymers,” J. Chem. Phys. 57, 1812–
1813 (1972).
Kim, S., and S. J. Karrila, Microhydrodynamics: Principles and Selected Applications (Dover, New York,
2005).
Lam, Y. C., L. Jiang, C. Y. Yue, K. C. Tam, and L. Li, “Interfacial slip between polymer melts studied by con-
focal microscopy and rheological measurements,” J. Rheol. 47(3), 795–807 (2003).
Larson, R. G., The Structure and Rheology of Complex Fluids (Oxford University Press, New York, 1999).
Leal, L. G., Advanced Transport Phenomena: Fluid Mechanics and Convective Transport Processes (Cam-
bridge University Press, New York, 2007).
Lee, P. C., H. E. Park, D. C. Morse, and C. W. Macosko, “Polymer-polymer interfacial slip in multilayered
films,” J. Rheol. 53, 893–915 (2009).
Lin, C. C., “A mathematical model for viscosity in capillary extrusion of two-component polyblends,” Polym. J.
11, 185–192 (1979).
RHEOLOGY OF DILUTE EMULSIONS WITH SLIP 1587

Luo, H., and C. Pozrikidis, “Effect of surface slip on stokes flow past a spherical particle in an infinite fluid and
near a plane wall,” J. Eng. Math. 62, 1–21 (2008).
Lyngaae-Jïrgensen, J., L. D. Thomsen, K. Rasmussen, K. Sïndergaard, and F. E. Andersen, “On the influence
of interfacial slip on melt flow properties of polymer blends,” Int. Polym. Process. 2, 123–130 (1988).
Macosko, C. W., “Morphology development and control in immiscible polymer blends,” Macromol. Symp. 149,
171–184 (2000).
Manga, M., and M. Loewenberg, “Viscosity of magmas containing highly deformable bubbles,” J. Volcanol.
Geotherm. Res. 105, 19–24 (2001).
Oldroyd, J. G., “The elastic and viscous properties of emulsions and suspensions,” Proc. R. Soc. London Ser. A
218, 122–132 (1953).
Park, C. C., F. Baldessari, and L. G. Leal, “Study of molecular weight effects on coalescence: Interface slip
layer,” J. Rheol. 47, 911–942 (2003).
Park, H. E., P. C. Lee, and C. W. Macosko, “Polymer-polymer interfacial slip by direct visualization and by
stress reduction,” J. Rheol. 54, 1207–1218 (2010).
Rallison, J. M., “Note on the time-dependent deformation of a viscous drop which is almost spherical,” J. Fluid
Mech. 98, 625–633 (1980).
Ramachandran, A., K. Tsigklifis, A. Roy, and L. G. Leal, “The effect of interfacial slip on the dynamics of a
drop in flow: Part I. Stretching, relaxation and breakup,” J. Rheol. 56, 45–97 (2012).
Schowalter, W. R., C. E. Chaffey, and H. Brenner, “Rheological behavior of a dilute emulsion,” J. Colloid Inter-
face Sci. 26, 152–160 (1968).
Shih, C. K., “Rheological properties of blends of two elastomers,” Polym. Eng. Sci. 16, 742–746 (1976).
Tucker, C. L., and P. Moldenaers, “Microstructural evolution in polymer blends,” Annu. Rev. Fluid Mech. 34,
177–210 (2002).
Tuteja, A., M. E. Mackay, S. Narayanan, S. Asokan, and M. S. Wong, “Breakdown of the continuum Stokes-
Einstein relation for nanoparticle diffusion,” Nano Lett. 7, 1276–1281 (2007).
Utracki, L. A., “Melt flow of polymer blends,” Polym. Eng. Sci. 23, 602–609 (1983).
Utracki, L. A., and M. R. Kamal, “Melt rheology of polymer blends,” Polym. Eng. Sci. 22, 63–78 (1982).
Zhao, R., and C. W. Macosko, “Slip at polymer-polymer interfaces: Rheological measurements on coextruded
multilayers,” J. Rheol. 46, 145–167 (2002).

You might also like