Tangential velocity on OPEN WIND boundaries with SEM

[ericvmueller]

Here is a case with OPEN boundaries and WIND that is not aligned with the grid. Typically, the tangential velocity at the boundaries is set properly.

However, if you add synthetic turbulence to a boundary (even a negligible amount such as with a single eddy), the tangential velocity is not being set, except on the perimeter of the VENT.

I think this can be traced to this part of the code where the ghost cell values for open boundaries are only being set with the WIND velocity if there is no SEM. And the eddy velocities are not picked up either.

fds/Source/velo.f90

Lines 2098 to 2099 in f177050

	OPEN_AND_WIND_BC: IF ((IWM==0 .OR. WALL(IWM)%BOUNDARY_TYPE==OPEN_BOUNDARY) .AND. &
	(IWP==0 .OR. WALL(IWP)%BOUNDARY_TYPE==OPEN_BOUNDARY) .AND. .NOT.SYNTHETIC_EDDY_METHOD) THEN

OW_SEM.txt

[rmcdermo]

Try with PR #15625 and let me know. Hopefully this is not a whack-a-mole.

[ericvmueller]

Yup, looks like I'm getting the correct flow in this case now. Thanks!