superposing uniform flow
14 views
Skip to first unread message
Hans Hornung
Feb 19, 2013, 6:11:11 PM2/19/13
to amrita...@googlegroups.com
This is a new question.
I am actually having a lot of fun running amrita scripts on an 8 processor machine in parallel. So fast!
My question is this. I am studying the interaction of a shock with a compressible vortex. After the shock passes through the vortex, the vortex gets convected along. If I want to follow it for a long time, I need to make a very long domain. So I would like to (at some point after the shock has hit the vortex) superimpose a uniform velocity on the whole flow field, so as to effectively go into the frame of the vortex and avoid computing a long domain. There is the trick that is used in ShockWave.amr which one could use:
W'state2 ::= W'state1<RHO*=r2[],U+=u2[],P*=p2[]>
if only one could input, instead of W'state1, the current flow field (which is effectively available through flowin e.g.)
I would not be surprised if one of you has done this. If so, please let me in on it.
I am actually having a lot of fun running amrita scripts on an 8 processor machine in parallel. So fast!
My question is this. I am studying the interaction of a shock with a compressible vortex. After the shock passes through the vortex, the vortex gets convected along. If I want to follow it for a long time, I need to make a very long domain. So I would like to (at some point after the shock has hit the vortex) superimpose a uniform velocity on the whole flow field, so as to effectively go into the frame of the vortex and avoid computing a long domain. There is the trick that is used in ShockWave.amr which one could use:
W'state2 ::= W'state1<RHO*=r2[],U+=u2[],P*=p2[]>
if only one could input, instead of W'state1, the current flow field (which is effectively available through flowin e.g.)
I would not be surprised if one of you has done this. If so, please let me in on it.
Reply all
Reply to author
Forward
0 new messages