[Nek5000-users] Periodicity with curvature
nek5000-users at lists.mcs.anl.gov
nek5000-users at lists.mcs.anl.gov
Sat Jan 13 18:31:14 CST 2018
Hi Jon,
You raise two separate issues ---
1. periodic-auto is simply the way that prenek assigns the P-P matching between
a pair of faces. It requires the faces to be parallel, aligned, and with unit normals
point in opposite directions. Generally under these circumstances prenek/pretex will
find the matched pairs and thus generate the correct .rea file. From there, genmap
and Nek5000 will establish the correct connectivity to execute the code (primarily,
direct stiffness summation).
2. If the faces are not aligned, you need to have a way to assign the P-P association.
Assuming you can do this then there are two possible cases:
a) The unit normals on P-P face pairs are aligned and in opposite direction.
This case will go through genmap/Nek5000 without any issue.
b) The unit normals on P-P faces are not aligned, i.e., you're running the CYCLIC
case (ifcyc=.true.). Here, there's going to be an issues in setting up the .map
file. genmap likely will not work. You can recover this inside nek because
we allow the user to prescribe some connectivity information, but it's clear
that this will have to be done on a case-by-case basis.
One of things we often do is to have a fake mesh with the same topology but
with a geometry such that the normals are aligned (Case 2.a). Then generate
the .map file and use it with Case 2.b.
If this isn't possible, I'd suggest you contact me off-list and I can try to give
some guidance.
In the meantime, for your model problem, I'd suggest proceeding with one of
the approaches outlined above, just so you can familiarize yourself with the
overall process for the cyclic bcs.
hth,
Paul
a) Unit normals are parallel and in
________________________________
From: Nek5000-users <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: Saturday, January 13, 2018 2:34:58 PM
To: nek5000-users at lists.mcs.anl.gov
Subject: Re: [Nek5000-users] Periodicity with curvature
Hi Paul,
In the future, I will need a more complex geometry where cyclic BCs are not in-line. This simple example I presented here is simply a proof of concept for another mesh.
What would be the best method do this, when the geometry isn't simple enough to use auto-periodic?
Thank you,
Jon
On Sat, Jan 13, 2018 at 2:28 PM, <nek5000-users at lists.mcs.anl.gov<mailto:nek5000-users at lists.mcs.anl.gov>> wrote:
Hi Jon,
It's been awhile since I've run the cyclic BCs that you're trying to use.
(Here: cyclic = periodic + turning through an angle.)
I can't recall the full workflow at the moment (and am not in a position
to check), but I think you'll have better luck if you build a straight mesh,
run genmap, and then morph the geometry into the desired curved
shape in usrdat2().
The reason for doing so is that you'll then be guaranteed to have
the correct mesh connectivity.
hth,
Paul
________________________________
From: Nek5000-users <nek5000-users-bounces at lists.mcs.anl.gov<mailto:nek5000-users-bounces at lists.mcs.anl.gov>> on behalf of nek5000-users at lists.mcs.anl.gov<mailto:nek5000-users at lists.mcs.anl.gov> <nek5000-users at lists.mcs.anl.gov<mailto:nek5000-users at lists.mcs.anl.gov>>
Sent: Saturday, January 13, 2018 1:00:48 PM
To: nek5000-users at lists.mcs.anl.gov<mailto:nek5000-users at lists.mcs.anl.gov>
Subject: Re: [Nek5000-users] Periodicity with curvature
My apologies, can you clarify what you mean?
I made global refinements one more time, and re-assigned the periodic BCs within prex. For this 2D case (attached), I have linked the 8 element sides with the right 8 sides. All other sides are walls.
Thank you,
Jon
On Sat, Jan 13, 2018 at 12:49 PM, <nek5000-users at lists.mcs.anl.gov<mailto:nek5000-users at lists.mcs.anl.gov>> wrote:
can you try again with at least 3 elements in other direction.
On 13 Jan 2018, at 19:42, "nek5000-users at lists.mcs.anl.gov<mailto:nek5000-users at lists.mcs.anl.gov>" <nek5000-users at lists.mcs.anl.gov<mailto:nek5000-users at lists.mcs.anl.gov>> wrote:
Thanks for the quick response. I made some global refinements but it seems I'm still getting the same error with genmap.
Attached is the new .rea file with 4 elements in the periodic direction.
On Sat, Jan 13, 2018 at 12:32 PM, <nek5000-users at lists.mcs.anl.gov<mailto:nek5000-users at lists.mcs.anl.gov>> wrote:
You need to have at least three elements in the periodic direction.
On 13 Jan 2018, at 19:19, "nek5000-users at lists.mcs.anl.gov<mailto:nek5000-users at lists.mcs.anl.gov>" <nek5000-users at lists.mcs.anl.gov<mailto:nek5000-users at lists.mcs.anl.gov>> wrote:
Dear Nek Users,
I am trying to get periodicity to work in the lastest version of Nek. I've made an extremely simple mesh within prenek (prex) in polar coordinates and cannot seem to run this case. When I run genmap, I get the error below. Did I do something wrong in setting up this mesh?
>> genmap
reading a.rea
Input mesh tolerance (default 0.2):
NOTE: smaller is better, but generous is more forgiving for bad meshes.
0.01
reading mesh data ...
start locglob_lexico: 4 2 8 1.0000000000000000E-002
locglob: 1 1 8
locglob: 2 5 8
locglob: 1 6 8
locglob: 2 6 8
done locglob_lexico: 6 6 8 2
start periodic vtx: 2 6
1 4 4 Matrix: x0
1 x0 1.0000E+00 1.0000E+00 5.4347-323 4.9407-324
2 x0 -3.5355E-01 3.5355E-01 6.9423-310 6.9423-310
3 x0 3.5355E-01 -3.5355E-01 6.9529-310 6.9529-310
4 x0 6.9423-310 6.9529-310 2.3477-313 3.1632-317
2 4 4 Matrix: x1
1 x1 1.0000E+00 1.0000E+00 6.3240-322 0.0000E+00
2 x1 -3.5355E-01 3.5355E-01 0.0000E+00 6.9423-310
3 x1 -3.5355E-01 3.5355E-01 6.9423-310 6.9529-310
4 x1 6.9423-310 6.9423-310 6.9423-310 6.9529-310
1 4 4 Matrix: z0
1 z0 1.0000E+00 1.0000E+00 5.4347-323 4.9407-324
2 z0 -2.1213E+00 -1.4142E+00 6.9423-310 6.9423-310
3 z0 2.1213E+00 1.4142E+00 6.9529-310 6.9529-310
4 z0 6.9423-310 6.9529-310 2.3477-313 3.1632-317
2 4 4 Matrix: z1
1 z1 1.0000E+00 1.0000E+00 6.3240-322 0.0000E+00
2 z1 1.4142E+00 2.1213E+00 0.0000E+00 6.9423-310
3 z1 1.4142E+00 2.1213E+00 6.9423-310 6.9529-310
4 z1 6.9423-310 6.9423-310 6.9423-310 6.9529-310
1 3 0 1.00000000E-03 2.12132000E+00 abort: FACE MATCH FAIL
2 4 3 2.12132000E-03 1.00000031E+00 abort: FACE MATCH FAIL
0 quit
The mesh within my .rea file is as follows:
20.0000 20.0000 -8.00000 -4.00000 XFAC,YFAC,XZERO,YZERO
**MESH DATA** 1st line is X of corner 1,2,3,4. 2nd line is Y.
2 2 2 NEL,NDIM,NELV
ELEMENT 1 [ 1A] GROUP 0
-2.121320 -1.414213 -0.8742276E-07 -0.1311341E-06
2.121320 1.414213 2.000000 3.000000
ELEMENT 2 [ 1B] GROUP 0
-0.1311341E-06 -0.8742276E-07 1.414213 2.121320
3.000000 2.000000 1.414213 2.121320
***** CURVED SIDE DATA *****
4 Curved sides follow IEDGE,IEL,CURVE(I),I=1,5, CCURVE
2 1 -2.00000 0.00000 0.00000 0.00000 0.00000 C
4 1 3.00000 0.00000 0.00000 0.00000 0.00000 C
2 2 -2.00000 0.00000 0.00000 0.00000 0.00000 C
4 2 3.00000 0.00000 0.00000 0.00000 0.00000 C
***** BOUNDARY CONDITIONS *****
***** FLUID BOUNDARY CONDITIONS *****
P 1 1 2.00000 3.00000 0.00000 0.00000 0.00000
W 1 2 0.00000 0.00000 0.00000 0.00000 0.00000
E 1 3 2.00000 1.00000 0.00000 0.00000 0.00000
W 1 4 0.00000 0.00000 0.00000 0.00000 0.00000
E 2 1 1.00000 3.00000 0.00000 0.00000 0.00000
W 2 2 0.00000 0.00000 0.00000 0.00000 0.00000
P 2 3 1.00000 1.00000 0.00000 0.00000 0.00000
W 2 4 0.00000 0.00000 0.00000 0.00000 0.00000
***** NO THERMAL BOUNDARY CONDITIONS *****
<mime-attachment.png>
_______________________________________________
Nek5000-users mailing list
Nek5000-users at lists.mcs.anl.gov<mailto:Nek5000-users at lists.mcs.anl.gov>
https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users<https://urldefense.proofpoint.com/v2/url?u=https-3A__lists.mcs.anl.gov_mailman_listinfo_nek5000-2Dusers&d=DwMFaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=gP2CkfxBOr64FG89YpFokg&m=ncKzoPY2dwLIbAc1FITAEJsCsTeH2zRgDAqkjXYZe5A&s=nLM4v-0kL5jW9dMNEgiVqdVGDFSs6TFHQxJK53ePw04&e=>
_______________________________________________
Nek5000-users mailing list
Nek5000-users at lists.mcs.anl.gov<mailto:Nek5000-users at lists.mcs.anl.gov>
https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users<https://urldefense.proofpoint.com/v2/url?u=https-3A__lists.mcs.anl.gov_mailman_listinfo_nek5000-2Dusers&d=DwMFaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=gP2CkfxBOr64FG89YpFokg&m=BAFsct9orEUodBvhKrXssNMVdu-OcqgveZCrA11i8M8&s=DBQt47t5SSmsVmTy3GCoo7VqaiTms-bCOKWC0nrs9xo&e=>
_______________________________________________
Nek5000-users mailing list
Nek5000-users at lists.mcs.anl.gov<mailto:Nek5000-users at lists.mcs.anl.gov>
https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users<https://urldefense.proofpoint.com/v2/url?u=https-3A__lists.mcs.anl.gov_mailman_listinfo_nek5000-2Dusers&d=DwMFaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=gP2CkfxBOr64FG89YpFokg&m=BAFsct9orEUodBvhKrXssNMVdu-OcqgveZCrA11i8M8&s=DBQt47t5SSmsVmTy3GCoo7VqaiTms-bCOKWC0nrs9xo&e=>
<a.rea>
_______________________________________________
Nek5000-users mailing list
Nek5000-users at lists.mcs.anl.gov<mailto:Nek5000-users at lists.mcs.anl.gov>
https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users<https://urldefense.proofpoint.com/v2/url?u=https-3A__lists.mcs.anl.gov_mailman_listinfo_nek5000-2Dusers&d=DwMFAg&c=ODFT-G5SujMiGrKuoJJjVg&r=gP2CkfxBOr64FG89YpFokg&m=sAFtxxjU2oId9i7igsGJ8tVajh1Q5GHP-CQMMPawlcs&s=NLpzCQaEmDTGEe8B503wySfumBz2BxWtkf6eEvhyGAU&e=>
_______________________________________________
Nek5000-users mailing list
Nek5000-users at lists.mcs.anl.gov<mailto:Nek5000-users at lists.mcs.anl.gov>
https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/nek5000-users/attachments/20180114/c0b8c35c/attachment-0001.html>
More information about the Nek5000-users
mailing list