Ask Your Question

NBQ and OBCs

asked 2020-06-22 21:49:16 +0100

this post is marked as community wiki

This post is a wiki. Anyone with karma >5 is welcome to improve it.


This is a new question regarding the NBQ mode and boundary conditions. I am building my way up to imposing a time-dependent stratified flow at the western boundary in the non-hydrostatic mode. As in my earlier post I am using a simple rotating, non-hydrostatic geostrophic adjustment problem (define NBQ and UV_COR) to learn about the model. The flow is nominally d/dy=0 (NS_PERIODIC with Mm=3). In this case the calculation works as expected when the x-direction boundary conditions are either EW_PERIODIC or EW-walls. However, if I define OBC_EAST and OBC_WEST (attempting to get a simple radiation condition) the calculation fails almost immediately. The velocities immediately adjacent to the two x-boundaries become very large. I suspect that this could be related to the acoustic mode, but am uncertain whether this is the case, and how to handle it. There are no problems with OBCs in the hydrostatic mode. Any help is appreciated.


edit retag flag offensive close merge delete

4 answers

Sort by » oldest newest most voted

answered 2020-06-26 20:31:13 +0100

nguc gravatar image

Thanks Patrick, that fixed the problem for Lucie's S2DV section too. Cyril

edit flag offensive delete link more

answered 2020-06-26 19:07:10 +0100

krh gravatar image

Hi Patrick, That fixed the problem for me also, so many thanks for the quick help. -Karl

edit flag offensive delete link more

answered 2020-06-26 18:20:27 +0100

OK, I could reproduced the blow-up with the KH_INST test case and found the problem. I commited the fix on the Master branch on Gitlab. You can do a git pull of the Master branch or just comment the call to rnbq_bc in step3d_fast.F. Please tell us if it fixes your problem as well ... Patrick

edit flag offensive delete link more

answered 2020-06-26 16:06:10 +0100

Hi Karl

I cannot reproduce the blow-up. I tried the IGW test case with the same options (NBQ, UV_COR, NS_PERIODIC, OBC_EAST, OBC_WEST), except that in IGW tides are forced through the boundaries. I'll try the KH_INST test case with open boundaries ... Patrick

edit flag offensive delete link more
Login/Signup to Answer

Question Tools

1 follower


Asked: 2020-06-22 21:49:16 +0100

Seen: 201,594 times

Last updated: Jun 26 '20