You are on page 1of 1

EXPLICIT VERSUS IMPLICIT NUMERICAL IMPLEMENTATIONS OF THE WALL SLIP BOUNDARY CONDITION

Author* LUIS JORGE LIMA FERRÁS


University of Minho Supervisors: João Miguel Amorim N. C. Nóbrega, Fernando Tavares de Pinho
School of Engineering
Centre of Biological Engineering
* luis.ferras@dep.uminho.pt

Introduction Control volume general equation (implicit vs explicit) Case Study - 4:1 contraction flow
The phenomenon of wall slip has been investigated for several years Imposing the implicit formulation, the equation became less
20H

Xr Normalisation Adopted
by the scientific community, being most of the published work related diagonally dominant. Length 
Length
H2
to experimental measurements performed with the objective to Velocity 
Velocity
U2
evaluate the effect of slip in the velocity distribution, pressure drop or   

Yr
  x V
0

un  up A  A  A  
P
 u = ....... 
  
the onset of flow instabilities.  Slip Intensity
E W S P

un   k       y  t

H1=4H
 f
 U ws
y f 
In the present work the implicit and explicit implementations of the  
50H
U2

Central Coefficient for no-slip

 k / y f     
wall slip boundary condition, in a computational rheology code, are   V
0
1
 un  
 k / y  1  p  = .......
.u A  A  A  u

H2=H
x  P
y

       k/  y  1 y    t


    E W S
 P
 U1 U2
analyzed and compared in terms of performance. The explicit f f f

 Newtonian
x

implementation showed that large relaxation factors must be used in Central Coefficient for slip 
 A P(slip)  A P(no-slip)
the iterative procedure in order to achieve convergence. With the U ws U ws U ws
 0.00  0.63  1.00
implicit implementation, there is no need to use any kind of relaxation Wall slip implementation U2 U2 U2
and the calculations where stable even for high non-linear slip
Explicit Implicit
models.
Slip vs No Slip Initial guess for the
velocity field
Initial guess for
the velocity field

With slip boundary conditions the tangent (to the wall) velocity vector k
is not null. no slip slip  1 Convergence
Solve Navier Stokes
y f
Equations
Solve Navier
k Stokes Equations
 1 divergence Non-Newtonian
h
 y f
u No
Compute velocity at the
wall using
U ws U ws U ws
No Convergence?  0.00  0.033  1.00
u   k
1
n
un0  up0 U2 U2 U2
y f
Wall Slip models
Relaxation Yes
Four slip models implemented in the finite volume code.
 u 0
 u 0

u1n   1      k
n p
   un
0
[1] Navier Linear Slip Law - Navier C. L. M. H., Mem. Acad. R. Sci. Inst. Fr, 6 Compute velocity
 y f
ut   sign  τ   kt [1]  
389-440, 1827. at the wall using
Convergence? (1)

ut  sign  τ  ktn
[2] Navier Nonlinear Slip Law - Schowalter W. R., The Behavior of Complex
[2]

Fluids at Solid Boundaries, J. Non-Newtonian Fluid Mech., 29 25-36, 1988.
k /y 

ut   sign  τ   k sinh k2 t c  [3]
f
n
u  .u
(1)
[3] Hatzikiriakos Slip Law - Hatzikiriakos S., A Slip Model for Linear Polymers
k
 /y  
1
p
1 Based on Adhesive Failure, Intern. Polymer Processing. 8, 135-142, 1993.  f 
ut   sign  τ   k1 ln  1  k2 t  [4]
[4] Asymptotic Slip Law - Polyflow Reference Manual (implementation of
Implicit and explicit comparison (Couette flow) Vortex dimension VS Slip velocity - Newtonian(left) – Non-Newtonian(right)
boundary conditions)
1.8 3.5

1.6
Xr

Normalised Vortex Dimension


Normalised Vortex Dimension
1.4 3
Yr
1.2
Xr
2.5
10000 1
Yr
0.8

y us   k  t Slip Relaxation 2
Slip 0.6
Coefficie Factor -
8000 Level 0.4
tangent stress vector nt Explicit 1.5
Number of Iterations

0.2
slip coefficient
S7 1,0E-07 0 0 1
slip velocity
S6 1,0E-06 0,9 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
6000 Implicit S5 1,0E-05 0,99 Uws/U2 Uws/U2
H S4 1,0E-04 0,999
x 4000
Explicit S3 1,0E-03 0,9999 Conclusions:
S2 1,0E-02 0,999991 (1) Explicit and Implicit implementation of the wall slip boundary conditions were presented and
S1 1,0E-01 divergence compared;
S0 1,0E+00 divergence
2000 (2) The implicit approach gave better results that the explicit one;
(3) In a case study of a 4:1 contraction flow it was possible to achieve the full slip condition with
0
the implicit approach;
S7 S6 S5 S4 S3 S2 S1 S0
(4) For the Newtonian fluid the vortex dimension decreases with the increase of slip velocity;
(5) For PPT fluids the vortex dimensions are almost unaffected by the level of slip imposed,
Slip Level
however its shape was influenced.
ACKNOWLEDGMENTS:
FCT- Portuguese Foundation for Science and Technology

Engenharia para a Qualidade de Vida: MOBILIDADE E ENERGIA – Semana da Escola de Engenharia -11 a 16 de Outubro de 2010 through the PhD grant SFRH/BD/37586/2007
FEDER via FCT, Fundação para a Ciência e Tecnologia,
under the POCI 2010 and Pluriannual programs

You might also like