Bug
In src/pre_process/m_perturbation.fpp lines 86-87, the x-velocity is modified in place before the y-velocity reads it:
q_prim_vf(mom_idx%beg)%sf(i,j,k) = (1 + r) * vx ! line 86: x-vel overwritten
q_prim_vf(mom_idx%end)%sf(i,j,k) = r * vx_modified ! line 87: reads (1+r)*vx instead of vx
The y-velocity perturbation ends up as r * (1+r) * vx_original instead of the intended r * vx_original, introducing an unintended (1+r) coupling factor. The fix is to swap the two lines so y-velocity is computed from the original x-velocity before x-velocity is modified.
Fix
PR #1216