Triply periodic boundary conditions #556

makrandak · 2024-05-07T17:25:15Z

makrandak
May 7, 2024

Hello,

I am trying to prescribe periodic boundary conditions for all pairs of faces on a box (periodic in x, y, z). I generated the box using genbox. Below is the script for the generation of the box

base.rea
-3                     spatial dimension  ( < 0 --> generate .rea/.re2 pair)
1                      number of fields
#=======================================================================
#
#    Example of .box file for Taylor-Green
#
#    If nelx (y or z) < 0, then genbox automatically generates the
#                          grid spacing in the x (y or z) direction
#                          with a geometric ratio given by "ratio".
#                          ( ratio=1 implies uniform spacing )
#
#    Note that the character bcs _must_ have 3 spaces.
#
#=======================================================================
#
Box
-16  -16  -16                                        nelx,nely,nelz for Box
-1.570796327 1.570796327 1.                          x0,x1,gain  (rescaled in usrdat)
-1.570796327 1.570796327 1.                          y0,y1,gain  (rescaled in usrdat)
-1.570796327 1.570796327 1.                          z0,z1,gain
P  ,P  ,P  ,P  ,P  ,P                                bc's  (3 chars each!)

I also prescribe the following in the .usr file to make sure the boundaryIDs are zero for all faces (following the documentation), I skipped the exo2nek step because I generated the mesh using genbox.

usrdat2 is as follows.

c-----------------------------------------------------------------------
      subroutine usrdat2()  ! This routine to modify mesh coordinates
      include 'SIZE'
      include 'TOTAL'
      
      integer e,f

      n = lx1*ly1*lz1*nelv
      nxz = nx1*nz1
      nface = 2*ldim

      do iel=1,nelt
         cbc(1,iel,1) = 'P  '
         cbc(2,iel,1) = 'P  '
         cbc(3,iel,1) = 'P  '
         cbc(4,iel,1) = 'P  '
         cbc(5,iel,1) = 'P  '
         cbc(6,iel,1) = 'P  '
      do ifc=1,2*ndim
         cbc(ifc,iel,2)  = cbc(ifc,iel,1) 
         if (cbc(ifc,iel,1) .eq. 'P  ') boundaryID(ifc,iel) = 0
         if (cbc(ifc,iel,2) .eq. 'P  ') boundaryID(ifc,iel) = 0
         if (cbc(ifc,iel,3) .eq. 'P  ') boundaryID(ifc,iel) = 0
         if (cbc(ifc,iel,4) .eq. 'P  ') boundaryID(ifc,iel) = 0
         if (cbc(ifc,iel,5) .eq. 'P  ') boundaryID(ifc,iel) = 0
         if (cbc(ifc,iel,6) .eq. 'P  ') boundaryID(ifc,iel) = 0
      enddo
      enddo

      return
      end
c-----------------------------------------------------------------------

When I try to run a Taylor-Green vortex case with the setup above, I see non-periodic behaviour for x-velocity in x direction, y-velocity in y direction and so on. Is there something else I have to do to prescribe periodicity?
Also, I found some this routine in the ktauChannel setup saying kludge for periodic in z. I am not sure if this is required for periodicity

subroutine usrsetvert(glo_num,nel,nx,ny,nz) ! to modify glo_num
     integer*8 glo_num(1)

     ! kludge for periodic bc in z
     nxy  = nx*ny
     nxyz = nx*ny*nz
     do iel = 1,nel
        ioff = nxyz*(iel-1)
        do ixy = 1,nxy
           glo_num(ioff + nxy*(nz-1) + ixy) = glo_num(ioff + ixy)
        enddo
     enddo

     return
     end

I am new to nekRS and I have not used nek5000 in the past. I am not sure what I am doing wrong in the setup. I would really appreciate your help. I would be happy to share all case files if needed to reproduce the problem.

With sincere regards

yslan · 2024-05-07T17:55:44Z

yslan
May 7, 2024

Hi,

Can you check ldimt in .cache/nek5000/SIZE ?
cbc is of size cbc(6,lelt,0:ldimt1) I think 6 (from (cbc(ifc,iel,6)) is uncommonly large and you could access the memory for other variables.

You don't need to add usersetvert since the periodicity should be directly coming from ma2/parCON.
In fact, I don't think you need to set boundaryID at all. NekRS should read cbc as-is.

I suggest to comment all those out.
Or, it should at least work if you simply replace the mesh in the tgv example.

Hope this helps,
Yu-Hsiang

0 replies

makrandak · 2024-05-07T18:07:53Z

makrandak
May 7, 2024
Author

Hello Yu-Hsiang, I commented all of it out and added boundaryTypeMap = none in my .par file. My box file is pretty much the same as the one in tgv example except for fewer elements (domain is rescled in `.usr` file in the tgv) example. I still get non-periodic results. How is the mesh in the tgv example generated? from the box file? With sincere regards, Makrand Makrand A. Khanwale ------------------------------------------- Postdoctoral Fellow, Center for Turbulence Research, Department of Mechanical Engineering, Stanford University Website: makrandak.github.io<http://makrandak.github.io/> Email: ***@***.******@***.***>

…

________________________________ From: Yu-Hsiang Lan ***@***.***> Sent: Tuesday, May 7, 2024 10:56 AM To: Nek5000/nekRS ***@***.***> Cc: Makrand Khanwale ***@***.***>; Author ***@***.***> Subject: Re: [Nek5000/nekRS] Triply periodic boundary conditions (Discussion #556) Hi, Can you check ldimt in .cache/nek5000/SIZE ? cbc is of size cbc(6,lelt,0:ldimt1) I think 6 (from (cbc(ifc,iel,6)) is uncommonly large and you could access the memory for other variables. You don't need to add usersetvert since the periodicity should be directly coming from ma2/parCON. In fact, I don't think you need to set boundaryID at all. NekRS should read cbc as-is. I suggest to comment all those out. Or, it should at least work if you simply replace the mesh in the tgv example. Hope this helps, Yu-Hsiang — Reply to this email directly, view it on GitHub<#556 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ALCCMN4QTIN52VDGXQYCJYTZBEITNAVCNFSM6AAAAABHLNWFTGVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4TGNBVGE4TC>. You are receiving this because you authored the thread.Message ID: ***@***.***>

1 reply

yslan May 7, 2024

Yes, there is a box file here
https://github.com/Nek5000/nekRS/blob/master/examples/tgv/input.box

Your box file looks fine though. It should work.
I recommend to rescale the domain in usrdat2 since box file only deal with ascii with a fewer digits.

I wonder if you are indeed using the re2 from genbox.
The output from genbox should be a box.re2 file and you typically need to rename is to match your .par

makrandak · 2024-05-07T18:17:56Z

makrandak
May 7, 2024
Author

Hello Yu-Hsiang, I am doing exactly that. I am using genbox and then renaming the file to match the .par name. I can see the output in visit and the mesh matches my settings in the box file. I have commented all of that stuff from the usr file and the udf file is something I wrote. The original effort was to setup a periodic HIT simulation with linear forcing, but I wanted to try Taylor-Green first as I was getting discontinuous solutions when I prescribed the energy spectrum based initial conditions. Right now the only thing in udf that is interesting is the initial condition, nothing to do with boundary conditions. I am attaching my case files to the email. Thanks again for your help. With sincere regards, Makrand Makrand A. Khanwale ------------------------------------------- Postdoctoral Fellow, Center for Turbulence Research, Department of Mechanical Engineering, Stanford University Website: makrandak.github.io<http://makrandak.github.io/> Email: ***@***.******@***.***>

…

________________________________ From: Yu-Hsiang Lan ***@***.***> Sent: Tuesday, May 7, 2024 11:10 AM To: Nek5000/nekRS ***@***.***> Cc: Makrand Khanwale ***@***.***>; Author ***@***.***> Subject: Re: [Nek5000/nekRS] Triply periodic boundary conditions (Discussion #556) Yes, there is a box file here https://github.com/Nek5000/nekRS/blob/master/examples/tgv/input.box I wonder if you are indeed using the re2 from genbox. The output from genbox should be a box.re2 file and you typically need to rename is to match your .par — Reply to this email directly, view it on GitHub<#556 (reply in thread)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ALCCMN5GQYDRWQYM4WKC7ULZBEKJNAVCNFSM6AAAAABHLNWFTGVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4TGNBVGQYTQ>. You are receiving this because you authored the thread.Message ID: ***@***.***>

0 replies

makrandak · 2024-05-07T20:39:18Z

makrandak
May 7, 2024
Author

I think I found a bug, my domain was from -pi/2 to pi/2 instead of pi to pi. I will test further and let you know. Thank you very much for your help and time. With sincere regards, Makrand Makrand A. Khanwale ------------------------------------------- Postdoctoral Fellow, Center for Turbulence Research, Department of Mechanical Engineering, Stanford University Website: makrandak.github.io<http://makrandak.github.io/> Email: ***@***.******@***.***>

…

________________________________ From: Yu-Hsiang Lan ***@***.***> Sent: Tuesday, May 7, 2024 11:10 AM To: Nek5000/nekRS ***@***.***> Cc: Makrand Khanwale ***@***.***>; Author ***@***.***> Subject: Re: [Nek5000/nekRS] Triply periodic boundary conditions (Discussion #556) Yes, there is a box file here https://github.com/Nek5000/nekRS/blob/master/examples/tgv/input.box I wonder if you are indeed using the re2 from genbox. The output from genbox should be a box.re2 file and you typically need to rename is to match your .par — Reply to this email directly, view it on GitHub<#556 (reply in thread)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ALCCMN5GQYDRWQYM4WKC7ULZBEKJNAVCNFSM6AAAAABHLNWFTGVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4TGNBVGQYTQ>. You are receiving this because you authored the thread.Message ID: ***@***.***>

0 replies

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Triply periodic boundary conditions #556

{{title}}

Replies: 4 comments 1 reply

{{title}}

{{editor}}'s edit

{{editor}}'s edit

{{title}}

{{title}}

{{editor}}'s edit

{{editor}}'s edit

{{title}}

{{title}}

Select a reply

Triply periodic boundary conditions #556

makrandak May 7, 2024

Replies: 4 comments · 1 reply

yslan May 7, 2024

makrandak May 7, 2024 Author

yslan May 7, 2024

makrandak May 7, 2024 Author

makrandak May 7, 2024 Author

makrandak
May 7, 2024

Replies: 4 comments 1 reply

yslan
May 7, 2024

makrandak
May 7, 2024
Author

makrandak
May 7, 2024
Author

makrandak
May 7, 2024
Author