Learning Forum

[benjamin.arnold]

Dear Ansys community

I am performing a simple simulation of a single phase fluid flow in a rod bundle geometry. I use only walls (no slip, no roughness) as well as periodic conditions on top, bottom, front, back and left and right side. I set all fluid properties to constants, such that I achieve a Reynolds number of 5000. Now I simulate both, a crossflow and a streamwise flow, by setting the periodic mass flow rate to the propper direction. While for corssflow, I get reasonable results, I do not get so for the streamwise flow. In the streamwise case, I get an asymmetric velocity and (if turned on) temperature profile. This happens for different turbulence models and for different meshes,

I tried very crude meshes without any wall inflations, furthermore I tried very fine meshes, the results are always pretty much the same. 

I tried to run transient solver instead of steady, the results are literally the same.Regarding the convergence, there are no problems, the case converges arbitrarily low, depending on the number of iterations. Also, the extracted data do not change reasonably, no matter how small the residual gets, i.e. there are no oscillating results.

The following picture shows the streamwise velocity, I changed the colorscale a bit, such that the difference between the left and right half of the symetric geometry can be seen. 

It would be great if someone could help me out with this issue.

Best regards

 

Benjamin

[Federico]

Hello, 

which is the "streamwise" direction?

[Dato]

Hello, it looks like it's a fuel rod assembly geometry. Is the picture shows whole domain of the model? I guess you use translational periodic condition. It seems that the asymmetry of velocity contour is not obvious if a default colormap is used. Regards