[Nek5000-users] symmetry boundary conditions and genmap

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

Hi Marco,

genmap uses recursive spectral bisection to partition the graph across processors.

RSB does not guarantee connected subgraphs except in the continuous case (i.e., infinite
resolution).

We have several heuristics to repair disconnected subgraphs, if they all fail you get the 
warning below.   For small integers, like the one shown, there is now worry.  However,
if the integers were in the 100s or higher it could be detrimental to parallel performance,
particularly when using the (default) XXt-based coarse-grid solver.  (AMG-based coarse
grid solver is not so sensitive...)

Paul


________________________________________
From: nek5000-users-bounces at lists.mcs.anl.gov [nek5000-users-bounces at lists.mcs.anl.gov] on behalf of nek5000-users at lists.mcs.anl.gov [nek5000-users at lists.mcs.anl.gov]
Sent: Friday, July 01, 2016 3:18 PM
To: nek5000-users at lists.mcs.anl.gov
Subject: Re: [Nek5000-users] symmetry boundary conditions and genmap

Paul,

I fixed my python script and was able to generate the map file. I noticed, however, in the output the following message:
not connected           1           3           9           0
Could you please explain what means this message?

Thanks,
Marco
> On Jul 1, 2016, at 3:03 PM, nek5000-users at lists.mcs.anl.gov wrote:
>
>
> Hi Marco,
>
> It looks maybe like you replaced "W" by "SYM"  ?
>
> In the .rea file, if a line has a character that is read by nek, then the line has to be formatted,
> which means that spaces in the line are important.
>
> Thus, you should replace "W  "  ('W-space-space) by "SYM"  , i.e., all bcs are read as 3 character
> strings.
>
> Now, concerning SYM, there is another constraint --- namely the following:
>
> Any SYM boundary must be aligned with either X , Y, or Z unit normals, unless you're using the
> stress formulation.
>
> The reason is that, say, for a SYM boundary that has unit normal in the Z direction, the velocity
> bcs are:
>
>   du/dz = 0 ,  dv/dz = 0,  w=0
>
> That is, Dirichlet for w and Neumann for u and v.
>
> If the SYM bc is not aligned, it ends up being a mixed bc for the 3 velocity components, which
> in Nek can at present be handled only with the stress formulation, which means PN-PN-2, and
> not PN-PN.
>
> Best, Paul
>
> ________________________________________
> From: nek5000-users-bounces at lists.mcs.anl.gov [nek5000-users-bounces at lists.mcs.anl.gov] on behalf of nek5000-users at lists.mcs.anl.gov [nek5000-users at lists.mcs.anl.gov]
> Sent: Thursday, June 30, 2016 7:46 AM
> To: nek5000-users at lists.mcs.anl.gov
> Subject: [Nek5000-users] symmetry boundary conditions and genmap
>
> Hello,
>
> I have two rea files that are exactly identical but for the boundary conditions: one has ‘W’ and the other one has ‘SYM’. If I am not mistaken, the ‘SYM’ boundary type is supported by Nek5000 but is causing problem when I try to generate a map file with genmap. It returns the following error message:
> reading .rea file data ...
>  ERROR: error reading    1       79200      299357
>  aborting 530 in routine rdbdry.
>
>           1  quit
>
> The rea file with ‘W’ boundary type successfully goes through the genmap script. Is there anything I should be aware when setting SYM bc type in the rea file?
>
> I could not find any related post in the archive of the mailing list.
>
> Thanks,
> Marco
>
> _______________________________________________
> Nek5000-users mailing list
> Nek5000-users at lists.mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
> _______________________________________________
> Nek5000-users mailing list
> Nek5000-users at lists.mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
>

_______________________________________________
Nek5000-users mailing list
Nek5000-users at lists.mcs.anl.gov
https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users