Professional Documents
Culture Documents
John Hinch
DAMTP, Cambridge University
22 July 2008
Outline
Micro & macro views
Einstein viscosity
Rotations
Deformations
Interactions
Polymers
Others
Essential
Micro
` L Macro
Micro = particle 1m
`2
1,
L2
can be large
Two-scale problem ` L
1. Macromicro connection
2. Micromacro connection
Macro = continuum = average/smear-out micro details
E.g. average over representative volume V with ` V 1/3 L
Z
1
=
dV
V V
Also ensemble averaging and homogenisation
To be used in averaged = macro momentum equation
u
+ u u = + F
t
NB micro-Reynolds stresses (u)0 u0 small for Re` 1.
Write:
= pI + 2e + +
1
=
V
Z
dV = n
V
dV
particle
types of particle
particle
Z
= pI + 2e + n
n x dA
particle
Homogenisation 2
2 :
k T0 = 0
i.e. T0 = T (x)
Thus T varies only slowly at leading order, with microscale making
small perturbations.
Homogenisation 3
1 :
k T1 = kx T0
Solution T1 is linear in forcing x T0 , details depending on k():
T1 (x, ) = A()x T0
Homogenisation 4
0 :
k T2 = Q x kx T0 kx T1 x k T1
Secularity: hRHSi = 0 else T2 = O( 2 ) which contradicts
asymptoticity. (Periodicity not necessary.) Hence
0 = hQi x hkix T0 x hk
A
ix T0
k T = Q
with
k =
A
k +k
and Q = hQi
Homogenisation 5
NB: Leading order T0 uniform at microlevel, with therefore no
local heat transport
NB: Micro problem forced by T0 . Need to solve
k Tmicro = 0
Tmicro x T0
Solution
Tmicro = (x + A)T0
Hence heat flux
hqi = hkTmicro i = hk + kAi T0
Einstein viscosity
Micro problem
I
Stokes flow
u=0
in r > a
2
0 = p + u
u=V+x
in r > a
on r = a
u U + x U
with
F=
Z
n dA = 0,
r =a
as
V , consts
x n dA = 0
G=
r =a
gives V = U
gives =
Then
a5
5(x E x)
u=U +Ex+xEx 5 x
r
2r 2
3
(x E x)a
p = 5
r5
a3 a5
5
r3
r
= pI + 2E + 5E
4 3
a
3
rod r > 1
disk r < 1
Rotation of particles
r 2 1
r 2 +1
[E p p(p E p)]
Long rods
r 2 1
r 2 +1
+1
Flat disks
r 2 1
r 2 +1
r2
2(ln 2r 23 )
6 ln 2r 11
r2
r 0
10
3r
8
3r
8
3r
rotates to
rotates to
= 1+
r2
3(ln 2r 1.5)
Large at 1 if r 1. Now =
ext
= 1+
4
2
3 ab
and r = ba , so
4na3
9(ln 2r 1.5)
ext = 1 +
9 ln 1/2
10
10nb 3
ext = 1 +
= 1+
3r
9
where for disks b is the largest dimension
(always the largest for Stokes flow).
No semi-dilute theory, yet.
U = (y , 0, 0)
rotates to
( = 1/r with
(r 2 cos2 + sin2 )
r 2 +1
(r 2 1)
sin 2 sin 2
4(r 2 +1)
r
,
r 2 +1
(
0.32r / ln r
= 1+
3.1
rods
disks
Remarks
Alignment gives shear ext
and this material anisotropy will lead to anisotropy of macro flow.
Important to Turbulent Drag Reduction
2 .
3
for ext
r = na
.
r = na2 b for shear
.
= nab 2
for permeability
Rotary diffusivity
Drot =
kT
kT
8a3
8a2
3(ln 2r 1.5)
for spheres,
rods,
kT
8
3 b
disks
+ (pP)
= Drot 2 P
t
p(p)
earlier deterministic.
ext
1/3
X r^2/ln r
Small
strain-hardening
4/15
6
shear
E/D
l Orientation effects
4 r^2/15ln r
Large
shear-thinning
r^2/2ln r (D/) 1/3
Also N1 > 0,
N2 small < 0.
0.32 r/ln r
6
r^3
/ D
Simple shear
Shear viscosity
Anisotropy
Brownian rotations
Macro stress
Viscosities
Closures
Deformations
T
Ea
(normally)
time
time
T
ext Ea
int
ext
Experiments: de Bruijn (1989) (=own), Grace (1982)
Theories: Barthes-Biesel (1972), Rallison (1981), Hinch & Acrivos (1980)
T
0.56
a
higher orders)
T
ext a (k4 A
+ k8 (AA) + . . .
5
2(2+3) ,
5(1)
3(2+3) ,
k2 =
k4 =
40(+1)
(2+3)(19+16)
4
2+3
&
linear in
E &
T
ext a
deformation
xy
angle
N1 , N 2
and T = 0 at ends
Snap straight
H 76
Flexible thread
Double layer
Interactions
Polymers
1
Divergent integral from u 3
r
Need renormalisation: Batchelor or mean-field hierarchy.
= 1 + 2.5 + 6.02
= 1 + 2.5 + 6.02
slope 6.0
Einstein 2.5
Experiments concentrated
a3 /kT
Stokesian Dynamics
(mostly) pairwise additive hydrodynamics
Jamming/locking clusters across the compressive quadrant
Stokesian Dynamics 2
Effective viscosity in shear flow
2 a2 (r 2a)
e
r
r 5
= 1 + 2.5 + 2.82
a
r 5
= velocity r
force distance r
r 3
volume
a
2 coefficient as function of
r
a
Experiments concentrated
Stress as function of shear-rate at different pH.
Suspension of 0.33m aluminium particles at = 0.3
I
I
I
I
I
d
R
, d = 2.3?
Number of particles in floc N =
a
3
R
Volume fraction of flocs floc =
a
Collision between two flocs
Hydro force 6RR = Bond force Fb number of
bonds N Ra
Fb
Hence floc =
6a2
So strong shear-thinning and yields stress Fb /a2 .
Interactions fibres
Interactions drops
deformation
xy
angle
N 1 , N2
into flow
Electrical double-layer
Dilute
Concentrated
3kT
Nb 2
Hence
R = R U
1
2 R
with
3
= pI + 2E + nA
Rheological properties
Shear
I
= constant, N1 2 , N2 = 0.
Distortion xy : a a
Extension
I
0.5
!
1
Distortion e (2E )t
Refinements
1. (boring) Spectrum of internal modes: Rouse 53, Zimm 56
with pre-averaged hydrodynamics
2. (boring) Polydisperse molecular weights
3. (important) Finite extensibility to stop infinite growth
1
e (2E )t
I
1
x
kT
R
Nb 2 1 R 2 /L2
FENE-P closure
RR/(1 R 2 /L2 ) = hRRi /(1 hR 2 i/L2 )
but molecular individualism
L = Nb
1
L2
A= 2
L trace A
a2
A I
3
2
L
= pI + 2E + n 2
A
L trace A
ext
1 + naL2
More refinements
4. Nonlinear bead friction
Hydrodynamic drag increase with size 6(a R)
ext = 1 + nL3
and hysteresis
Couple balance
Afine
A non-affine A
inefficiency of straining
trace A
(AE + E A)
3 + trace A
R2
,
L
number
L2
,
R2
dissipation
R4
L
(trace A)2
L2
A
= pI + 2 1 + n
E + n 2
L
L trace A
Good for contraction flows
arclengths satisfy
si = 41 E (si+1 + 2si si1 )
Kinks model 2
n` = R = Ne
` = e 2Et
H 94
kT ln det M 1
with
M 1 ab =
X
i beads
mi1
g a g b
xi xi
+ x n+1 )f n 12 x n f n
Grassia, Nitsche & H 95
10
(stress-final stress)/n
0.1
0.1
0.01
0.01
0.02
0.03
0.04
0.05
t/n^2
0.06
0.07
0.08
0.09
0.1
10
n^2 t
15
/ 13 N 3 vs N 2 t
/N vs t/N 2 (Rouse)
Stress relaxation: Decay of [1,1] stress component vs t: Intermed. times
(stress-final stress)/n^2
10
Intermediate times
kTN 2 t 1/2
0.1
0.001
0.01
0.1
t
20
1
f (A I )
h
= pI + 2E + G f A
A=
I
I
I
I
Oldroyd B f = 1
FENE-P f = L2 /(L2 trace A)
p
Nonlinear bead friction h = traceA/3
New form of stress
= pI + 2E + 2d (A : E )A + G trace AA
I
I
often reformulated
*
*
*
*
*
*
*
*
*
*
*
*
L2
M3
D = kT /6L
Modulus G = nkT = G D M 3
(expts M 3.4 )
At later time:
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
Stress
Z
(t) = n
0
X 1
2
e p s/D
2
p
p
surving tube
Nsegements
3kT
A u A u
a
ds
a
|A u|2
segment tension
Refinements
1. Chain retraction
deform
retract
1
D
+ u : huui
more shear
Branched polymers
typical in industry
Orientation: S = B/traceB
Stretch: = u : S
1
S (
1)
B = 1O (B I)
while
< max
Blackwell 2002
Polymers
Polymers
Bead-and-spring model
Refinements
FENE-P constitutive equation
Unravelling a polymer chain
Kinks model
Brownian simulations
Entangled polymers
rheology
Refinements
pom-pom
Associating polymers
Surfactants - micells
Aging materials
GENERIC
Outline
Micro & macro views
Einstein viscosity
Rotations
Deformations
Interactions
Polymers
Others