[Pled]

Both sides of an interface have common coordinates. Therefore in a fluid fluid interface a wall shear stress of side 1 and coords x1,y1,z1 is not the same as of side's 2 wall shear stress and the same coords.
I need to use a variable of side 2 that uses the wall shear stresses of side 1 as they have equal coords.
How is it possible?
@REGION: works only when speaking for an areaAve of wall shear stresses.

[ghorrocks]

The difference in wall shear stress across an interface should be negligible. As you refine the mesh to a mesh-independant resolution and differences in wall shear stress should reduce to zero.

Why do you want to access the wall shear stress from the other side of the interface anyway?

And have you done a mesh sensitivity check? What you are proposing should not be necessary if you have a fine enough mesh.

[Pled]

I have modeled the interface as a membrane, where in each side two different fluids are flowing. It is for a biological system where the wall shear stresses of one side cause the flow of the other liquid in the other side of the membrane.
Since in cfd-post I can see the wall shear stresses of each side seperately, I thought it must be a way to consider them in expressions of cfx.

[ghorrocks]

The interface is a membrane? That changes everything. I had to guess what you were talking about in the first post as it was just talking about fluid-fluid interfaces. Please make sure you provide all necessary information in future questions.

First of all, what physical process works like this? How does a wall shear stress drive a flow on the other side of the membrane without moving the membrane? There must be some other physics going on and it might be better to model that process.

Is this a multiphase simulation with a different phase on either side? Or a multicomponent mixture?

[Pled]

I will explain all my problems till here.
First the problem is about an artery and its wall.
In the artery, blood flows along with other components.
Several components pass throught the membrane to the wall (which is porous) and the flux of each one depends on the wall shear stress of the membrane wall shear stress from the side where blood is.
Normally as these components come inside the wall thickness increases.
This is the case. The problems are:
1) porous domain does not support deformation -> thickness variation is impossible
In order to solve this I considered the porous domain as fluid with porous momentum sources .
2) Since components were modelled as additional variables, passing only the components and not the blood was impossible.

In order to solve this I considered a no flux boundary to everything to the membrane in the blood side (which causes no problem since blood flow is fast and the components are considered that do not decrease in the blood flow). In the wall side I considered a source term of the components that depend on the wall shear stress of the other side. That is my problem. I cannot take the WSS of the blood side under account in the wall side.

[ghorrocks]

Can you implement your artery wall model as a additional variable contact resistance, and the contact resistance can be a function of whatever you like (including wall shear stress)?

A second comment: Are you sure the flux through the interface is a function of wall shear stress? This is a very unusual way of considering it. Usually the flux through the interface is function of the concentrations on either side and the resistance of the interface. Wall shear stress influences it by being linked to the concentration on the fluid side. But the actual interface model only requires the concentrations on both sides (and the interface resistance).