[shayan_mv]

Hi,

I am dealing with a multi phase problem ; a 2D domain filled with air as the primary phase and a fluid flow with variable viscosity in y-direction moving in x-direction on the bottom wall as the secondary phase.

Wondering if you could help with a UDF code by which I can define a function that varies the viscosity values in y-direction?

Thanks

Shayan

[ComputerGuy]

What about something like this, hooked to the viscosity of whatever phase in your simulation you're trying to change?

DEFINE_PROPERTY(cell_viscosity, c, t)
{
  real mu;
  real x[ND_ND];
  real Y_coord;
  C_CENTROID(x,c,t);
  Y_coord=x[1];
  mu= /*YOUR FUNCTION OF VISCOSITY HERE */
  return mu;
}

Can you elaborate why the viscosity would be spatially dependent? Is it because you don't want to simulate the effects of one phase diffusing into another and affecting the viscosity?

ComputerGuy

[shayan_mv]

Thanks a lot for the code;it was really helpful. I did some minor changes to get what I want, but there's still an issue:

#include "udf.h"

DEFINE_PROPERTY(cell_viscosity, c, t)
{
real mu;
real x[ND_ND];
real Y_coord;
C_CENTROID(x,c,t);
Y_coord=x[1];

if(Y_coord<=/ H) /*H: thickness of the secondary phase)*/
mu=(a*Y_coord)+A; /*viscosity profile & a and b are constants*/
else
mu=A;
return mu;
}

Well, I'm actually dealing with two layers of fluids with various viscosities. Instead of defining the 3rd phase, I decided to change the viscosity in y-direction! nevertheless, there's also a variation in "mu" in y-dir for each layer.

Assuming "A(@Y_coord=0 or wall)"<=mu<="B(@ Y_coord=H or interface)", I couldn't get the B value for the viscosity at the interface, and its contour shows me a less value. Like, B=5.7 Pa.S instead of 6! Is there any chance to fix this issue??

Cheers