[Nek5000-users] Flow Rate Oscillation on 16384 cores

[nek5000-users at lists.mcs.anl.gov]

Dear JP,
We have been using the constant flow rate feature in a number of cases 
as well, but we have not seen your problem.

Maybe a more general question; why does the flow rate vary at all in 
your case? In all our cases, due to the implicit superposition of the 
Stokes solution, we get the flow rate exactly constant without fluctuations.

Best regards,
Philipp

On 2015-03-26 10:40, nek5000-users at lists.mcs.anl.gov wrote:
> Dear All,
>
> I have setup a simulation with Nek's constant flow rate option. The
> simulation consisted of a number of restarts, in particular one restart
> was carried out on 16384 cores whilst all the other runs where on 8192
> cores. In the plot of flow rate against dump in the following link,
> which I carry out with my post-processor, the flow rate oscillates
> strangely in the region where the run was on 16384 cores, i.e. between
> 220 and 343.
>
> https://www.dropbox.com/s/fnfuqhj2nxbh5g5/FlowRate_restart_section.png?dl=0
>
> Any idea on what might be causing this oscillation? All parameters,
> geometry, machine settings etc. have been left the same for the run on
> 16384 cores (except for changing the max no. of processors and max no.
> of elements per processor in SIZE).
>
> Regards,
> JP
>
>
> _______________________________________________
> Nek5000-users mailing list
> Nek5000-users at lists.mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
>