[ubeneej]

Hello, using ANSYS CFX I am trying to simulate fluid flow through a small well that contains a membrane (small outline in the first image before the bottom region). The well narrows in the middle as you will see in the images due to a washer that is placed in it for experimental reasons. We have experimentally determined the permeability of the membrane.

I have defined a subdomain at the membrane region as shown and set an isotropic loss model with permeability coefficient. The problem is that whenever we try to set the permeability coefficient too low the model will not converge. The velocity streamlines make no sense as shown in the figures. I have even tried setting it higher to begin with and then using those results as initial values for the next runs. It still will not converge. This is shown in the convergence plot where you can see how the RMS values increase for each trial.

If anyone has any input as to why this low permeability would cause the convergence issues that would be great.

Thank you

[ubeneej]

This is the image showing the isotropic loss model for the membrane region (in green)

[Rajaero]

Hello there,
My suggestions for you,
1.try to find out the inerital and viscous coefficient value by darcy or forchiemer equation and set the value.
2. try to reduce the relaxation factor and enable higher order discretixation and try again.
3.if it is pressure-velocity coupling problem then use couples solver and for the pressure you use presto or standard method.
4. if it is the transient analysis then try to reduce your time step size and increse the no of iteration to solve one time step.
5.your time step size should be less than your element size.
6.check the mesh quality and try to use different mesh size.
7.check the inlet and outlet pressure setting.

if you get any error then let me know i will try to find the solution for you...

best regards..