[sjbae]

I'm currently simulating flow inside a reverse-osmosis membrane channel (feed side) with spacers included. The problem is, I want to quantify recirculation zones occurring behind the spacers as the zones with low-velocity are accumulated with solute and are prone to fouling. These spacer configurations are to be compared so that I can see which configuration is the best.

I have read other posts regarding post-processing recirculation regions, but in my case, I don't believe any of those apply to my case because

i) my simulation is carried out in 3D; making a cross sectional area and computing the flow that passes it won't capture the flow rate accurately.

ii) my simulation conditions have low Reynold's number. I don't see much recirculation with velocity vector profiles. Why aren't the recirculation zones visible?

I have tried viewing shear stress and shear stress parallel to the flow direction. I'm not sure if integrating it gives me an accurate result.

Any suggestions?

[FMDenaro]

For a 3D flow, I suggest to detect the vortical structures using a methods such as the lambda-2.

[sjbae]

I've just read the Lambda2 method on detecting recirculation regions. As far as I know, there aren't any built-in methods for doing this in CFX post. Do I have to define a variable via expressions and variable to detect the vortices?

[FMDenaro]

Originally Posted by sjbae

I've just read the Lambda2 method on detecting recirculation regions. As far as I know, there aren't any built-in methods for doing this in CFX post. Do I have to define a variable via expressions and variable to detect the vortices?

You have to code the method, then store the 3d matrix of the lambda-2 function and load it in Tecplot to represent the isosurface.
However, Tecplot has also the internal lambda-2 representation but I don’t like how it works.