CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   obtaining paraller run in cfx (http://www.cfd-online.com/Forums/cfx/80788-obtaining-paraller-run-cfx.html)

alinik October 6, 2010 11:15

obtaining paraller run in cfx
 
Hi everyone

I have the most wierd problem ever!

I was able to obtain solution in parallel mode in ansys-cfx, but since yesterday I can not.

the machine can not partition the problem and solve it in parallel mode, though it has no problem solving in serial mode.

I even re-installed ansys, but it didn't work

looking forward for your suggestions.

michael_owen October 6, 2010 14:46

What is the error you receive?

alinik October 7, 2010 00:41

I receive no errors. The solution simply does not start.

CFX solver can not partition the problem and start the solution. I didn't have any problem with parallel mode solution till the day before yesterday

michael_owen October 7, 2010 15:48

Quote:

Originally Posted by alinik (Post 278190)
I receive no errors. The solution simply does not start.

CFX solver can not partition the problem and start the solution. I didn't have any problem with parallel mode solution till the day before yesterday

Have you contacted technical support?

hosseinhgf October 8, 2010 18:50

which version of cfx do you work?

feizaghaee October 9, 2010 05:17

Quote:

Originally Posted by alinik (Post 278113)
Hi everyone

I have the most wierd problem ever!

I was able to obtain solution in parallel mode in ansys-cfx, but since yesterday I can not.

the machine can not partition the problem and solve it in parallel mode, though it has no problem solving in serial mode.

I even re-installed ansys, but it didn't work

looking forward for your suggestions.

have you increased the mesh?

RossFS October 14, 2010 20:25

I think I read in the CFX documentation that the problem you have can come about from the simulation you are running not being all that robust (not quite the right word). i.e. your simulation is a bit tricky and ANSYS doesn't like it.

One suggestion is to get the simulation going in serial if you can and then try starting it from the simulation data you have in parallel once the residuals have dropped down somewhat.

For me, I can't run my RNG k-e simulations in parallel almost ever seemingly, but I can with my SSTk-w simulations. Weird.


All times are GMT -4. The time now is 05:36.