UDF in parallel computing
 

Hi all,

Hope everyone is fine. My problem is that I want to use a UDF for variation of Unsteady Mach. The problem is big enough that I have to run it in parallel. But Fluent says that you require separate UDFs for host and nodes. I am not getting with this Kindly help me in this regard if anybody had used UDF in parallel before!

Regards

Re: UDF in parallel computing
 

i think fluent will find it automaticly. Or else just look in the src folder or a udfconfig-host.h

Re: UDF in parallel computing
 

Thanks lurk

Actually when I run UDF for Mach on single PC it runs very well. But when I read it on host computer (in parallel) it gives a long error msg that it couldnt find some cortex file or... etc ( i didnt remember exact words ! sorry!).

So should I have to make 5 separate UDFs for 5 clustered computers? What is the remedy?

Re: UDF in parallel computing
 

In general for parallel use UDF has to be parallelised. Communication between different calculation nodes has to be ensured. I have never done this in FLUENT before but it should be described in the UDF guide. Otherwise, ask support for more help.

Regards

Re: UDF in parallel computing
 

Yeah I know that but for that should i have to write as many UDFs as tha no PCs in parallel?