[Nek5000-users] Periodic BC & Genmap issue

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

In fact, the tolerance was not previously used for the periodic
check, only for the proximity check.  The new code I sent
propagates that tolerance into the periodic check -- diff
against the old version and have a look, you'll see the structure.

Let me know if you have any suggestions...

Thanks!

Paul



On Tue, 22 Jun 2010, nek5000-users at lists.mcs.anl.gov wrote:

>
>
> Hi Paul,
>
>
>
> Is this tolerance used to check the physical coordinates of the periodic faces?  I was afraid that the faces might not be 100 % congruent, but the face cooridantes of the periodic sides should be within at least 0.01 I believe. I could be mistaken though. I am going to try another another geometry where the faces are not so extreme as in this case.
>
>
>
> - Michael
>
>
> ----- Original Message -----
> From: nek5000-users at lists.mcs.anl.gov
> To: "Nekton User List" <nek5000-users at lists.mcs.anl.gov>
> Sent: Tuesday, June 22, 2010 1:47:35 AM GMT -06:00 US/Canada Central
> Subject: Re: [Nek5000-users] Periodic BC & Genmap issue
>
>
> Hi Michael,
>
> It looks like the issue is related to the mesh and to tolerances.
> It appears to me that the periodic faces are not congruent - nearly
> so, but not 100%.
>
> I've changed my version of genmap to make the tolerance in the
> periodicity check reflect the user-prompted value. Â I'm not yet
> commiting this to the repo because I need to think more about the
> consequences. Â However, the attached code does work when you input
> ".1" for the tolerance prompt - and it flags an error when you put
> in .01 or smaller. Â Try diff'g this against your current version
> to see the change.
>
> Paul
>
> PS - rather than attaching, I'm placing in
> www.mcs.anl.gov/~fischer/genmap.f.gz
>
>
> On Mon, 21 Jun 2010, nek5000-users at lists.mcs.anl.gov wrote:
>
>> Hi,
>>
>> I am running into an error when trying to use genmap. I believe it has to do with the periodic boundary conditions I have set up. I have checked the elements, and faces and indeed they are given correctly. This leads me to believe that there may be an issue with genmap. I have attached the rea file. This was to test the periodic faces for this geometry. The output is shown below, thanks for any help with this.
>>
>> - Michael M.
>>
>> Here is the output:
>>
>> test
>> Input mesh tolerance (default 0.2):
>> NOTE: smaller is better, but generous is more forgiving for bad meshes.
>> .2
>> reading .rea file data ...
>> start locglob_lexico: 8 60 480 0.20000000000000001
>> locglob: 1 1 480
>> locglob: 2 27 480
>> locglob: 3 42 480
>> locglob: 1 168 480
>> locglob: 2 168 480
>> locglob: 3 168 480
>> locglob: 1 168 480
>> locglob: 2 168 480
>> locglob: 3 168 480
>> done locglob_lexico: 168 168 480 4
>> start periodic vtx: 60 168
>> 1 3 5 3 3.66325334E-04 1 shift
>> 3 4 4 Matrix: x0
>> 1 x0 1.7389E-03 2.8363E-03 2.8363E-03 1.7389E-03
>> 2 x0 -6.9998E-02 7.0969E-02 6.9026E-02 -6.9998E-02
>> 3 x0 2.0879E-02 2.0879E-02 -2.0937E-02 -2.0821E-02
>> 4 x0 8.3935E-03 8.0685E-03 -2.4856E-02 8.3935E-03
>> 1 4 4 Matrix: x1
>> 1 x1 1.7389E-03 1.7389E-03 2.8351E-03 2.8351E-03
>> 2 x1 -7.0107E-02 -7.0107E-02 6.8930E-02 7.1286E-02
>> 3 x1 2.0876E-02 -2.0824E-02 -2.0927E-02 2.0876E-02
>> 4 x1 8.3867E-03 8.3867E-03 -2.4834E-02 8.0608E-03
>> 3 4 4 Matrix: z0
>> 1 z0 1.7389E-03 2.8363E-03 2.8363E-03 1.7389E-03
>> 2 z0 -1.2500E+00 -1.1090E+00 -1.1110E+00 -1.2500E+00
>> 3 z0 0.0000E+00 0.0000E+00 -4.1816E-02 -4.1700E-02
>> 4 z0 -5.0620E-01 -5.0653E-01 -5.3945E-01 -5.0620E-01
>> 1 4 4 Matrix: z1
>> 1 z1 1.7389E-03 1.7389E-03 2.8351E-03 2.8351E-03
>> 2 z1 -1.2500E+00 -1.2500E+00 -1.1110E+00 -1.1086E+00
>> 3 z1 0.0000E+00 -4.1700E-02 -4.1803E-02 0.0000E+00
>> 4 z1 4.9380E-01 4.9380E-01 4.6058E-01 4.9347E-01
>> 3 4 3 1.00000000E-03 7.09692500E-02 abort: FACE MATCH FAIL
>> 1 3 5 7.09692500E-05 3.66325334E-04 abort: FACE MATCH FAIL
>>
>>
> _______________________________________________
> Nek5000-users mailing list
> Nek5000-users at lists.mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
>