Backward facing step with step wall inclined at 20/15°
 

Hi I'm trying to run a simulation on the BFS with the step angle inclined by about 20° and 15°. However I'm facing a little problem the reattachment is bigger than the experimental data (I'm using the makiola data).

I changed the mesh (refined it and reduced the skewness to about 0.3 for 20° and 0.2 for 15°), the initial boundary condition, the fully developed U-profile I impose at the entry, the schemes and the tolerances at the fvsolution.

For all the other angles I could get good results however for lower angles this proved to be near I'm possible.

Do any of you have any suggestion? I used the komegasst model with the solver simplefoam.

I attached the 0 directory and some other useful files like blockmeshdict, fvsolution, fvschemes, transportproperties. Hope this will help to find out the culprit.

P. S the mesh attached I the coarse mesh used to make a fast estimation the reattachment length.

No one has an idea? I changed everything but still the reattachment length is too high compared to the experimental data (xr/H=6-7 compared to 4).
My last hope is that refining to the extreme the region at the step would solve the problem.

This behavior persist even if I change the expansion ratio and the Reynolds number.

The inlet and outlet distances from the step don't influence the results also the refining outside the side does nothing. I changed the inlet condition and u profile multiple times and I even mapped with the experimental data however the results were still almost the same.

Is it possible that the komegasst doesnt work well with recirculating flow over inclined wall with low angles? Because for angles up 30 the results are really good.