June 2018 - Nek5000-users - mailman3.cels.anl.gov

Re: [Nek5000-users] Stream function
by nek5000-users＠lists.mcs.anl.gov 10 Jul '18

10 Jul '18

Hi Phillip, I am using the “Blasius” example test outputting the streamfunction as a scalar. I included my code below. I am not sure what I am doing wrong. In userchk, I added “compute streamfunction from vorticity” from the psi_omega example, but it does not compute the psi properly… (max_psi = min_psi = 0 in every step) c----------------------------------------------------------------------- subroutine userchk include 'SIZE' include 'TOTAL' common /myblas/ ub(lx1,ly1,lz1,lelt),vb(lx1,ly1,lz1,lelt) common /blasiusr/ u(lx1,ly1,lz1,lelv),v(lx1,ly1,lz1,lelv) common /blasiusg/ bin(lx1,ly1,lz1,lelv) common /scrns/ dum(lx1*ly1*lz1*lelv) common /xtream/ psi(lx1*ly1*lz1*lelt) $ , rhs(lx1*ly1*lz1*lelt) $ , h1 (lx1*ly1*lz1*lelt) $ , h2 (lx1*ly1*lz1*lelt) integer bin real normu(50),normv(50) real xmax,xmin,dx save xmax,xmin,dx n = nx1*ny1*nz1*nelv c. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . c Compute streamfunction from vorticity call col2(rhs,bm1,n) call rone (h1 ,n) call rzero(h2 ,n) tol = param(22) imsh = 1 call hmholtz('psi ',psi,rhs,h1,h2,v1mask,vmult,imsh,tol,200,1) c. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . if (istep.eq.0) call outpost(vx,vy,vz,pr,psi,' ') c Compute delta* do i=1,n dum(i) = 1.-vx(i,1,1,1) enddo tot_deficit = glsc2(dum,bm1,n) xmin = glmin(xm1,n) xmax = glmax(xm1,n) delta_star = tot_deficit/(xmax-xmin) psimin = glmin(psi,n) psimax = glmax(psi,n) if (nid.eq.0) write(6,1) istep,time,delta_star,psimin,psimax 1 format(i8,1p4e13.5,' blasius: delta*') return end

1 6

Re: [Nek5000-users] Measure resolution for simulations with low-diffusivity scalars
by nek5000-users＠lists.mcs.anl.gov 29 Jun '18

29 Jun '18

What are you looking for? How can I get a bounded scalar solution for a given (low) diffusivity and resolution? Stefan -----Original message----- > From:[email protected] <nek5000-users(a)lists.mcs.anl.gov> > Sent: Friday 29th June 2018 17:44 > To: nek5000-users(a)lists.mcs.anl.gov > Subject: [Nek5000-users] Measure resolution for simulations with low-diffusivity scalars > > Hi, > > I want to evolve a scalar field \phi in my simulation with as low diffusivity (conductivity/(rhocp)) as possible. > I have performed simulations with sequentially reduced diffusivity D: > > Simulation 1: D = 1*2.17*10^-5 m^2/s, polynomial order 7 > Simulation 2: D = (1/4)*10^-5 m^2/s, polynomial order 7 > Simulation 3: D = (1/4)*2.17*10^-5 m^2/s, polynomial order 11 > Simulation 4: D = (1/8)*2.17*10^-5 m^2/s, polynomial order 11 > > Reducing the diffusivity, I keep the ratio D/dt constant, where dt is the time step. > At t = 0 \phi is bounded to 0 <= \phi <= 1, and it has no sources or sinks. > Currently, I check the resolution of the simulation after 25 s, by observing the range of \phi. > I find these values: > > Simulation 1: -0.01 < \phi < 0.06 > Simulation 2: -0.23 < \phi < 0.61 > Simulation 3: -0.044 < \phi < 0.16 > Simulation 4: -0.4 < \phi < 0.91 > > Indeed, the smaller the diffusivity D becomes, the worse resolution of \phi becomes. > Also, it seems as increasing the polynomial order improves the resolution of \phi. > However, the Courant numbers in my simulations are (on average): > > Simulation 1: C = 0.31 > Simulation 2: C = 0.06 > Simulation 3: C = 0.14 > Simulation 4: C = 0.06 > > That the Courant number goes down with increasing diffusivity is (I think) just a consequence of reducing the time step (keeping the ratio D/dt constant). > According to the Courant number, I have really great resolution in Simulation > 4, but according to the range of \phi I have not. Also, the Helmholtz solver seem to reach the tolerance in fewer iterations for low values of D. > > Can somehow Nek output a measure of the resolution of a scalar field? > How can I improve the resolution at a low D/allow for lower D and keep a good resolution? > > Best, > > Johan > > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users(a)lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users

1 1

Measure resolution for simulations with low-diffusivity scalars
by nek5000-users＠lists.mcs.anl.gov 29 Jun '18

29 Jun '18

Hi, I want to evolve a scalar field \phi in my simulation with as low diffusivity (conductivity/(rhocp)) as possible. I have performed simulations with sequentially reduced diffusivity D: Simulation 1: D = 1*2.17*10^-5 m^2/s, polynomial order 7 Simulation 2: D = (1/4)*10^-5 m^2/s, polynomial order 7 Simulation 3: D = (1/4)*2.17*10^-5 m^2/s, polynomial order 11 Simulation 4: D = (1/8)*2.17*10^-5 m^2/s, polynomial order 11 Reducing the diffusivity, I keep the ratio D/dt constant, where dt is the time step. At t = 0 \phi is bounded to 0 <= \phi <= 1, and it has no sources or sinks. Currently, I check the resolution of the simulation after 25 s, by observing the range of \phi. I find these values: Simulation 1: -0.01 < \phi < 0.06 Simulation 2: -0.23 < \phi < 0.61 Simulation 3: -0.044 < \phi < 0.16 Simulation 4: -0.4 < \phi < 0.91 Indeed, the smaller the diffusivity D becomes, the worse resolution of \phi becomes. Also, it seems as increasing the polynomial order improves the resolution of \phi. However, the Courant numbers in my simulations are (on average): Simulation 1: C = 0.31 Simulation 2: C = 0.06 Simulation 3: C = 0.14 Simulation 4: C = 0.06 That the Courant number goes down with increasing diffusivity is (I think) just a consequence of reducing the time step (keeping the ratio D/dt constant). According to the Courant number, I have really great resolution in Simulation 4, but according to the range of \phi I have not. Also, the Helmholtz solver seem to reach the tolerance in fewer iterations for low values of D. Can somehow Nek output a measure of the resolution of a scalar field? How can I improve the resolution at a low D/allow for lower D and keep a good resolution? Best, Johan

1 0

Wall models
by nek5000-users＠lists.mcs.anl.gov 29 Jun '18

29 Jun '18

Dear Neks, Do any of you have experience in turbulent wall modeling for smooth or rough regime in Nek ? One would have to fix the velocities of a determined number of points above the boundary at each time step. Is that feasible ? Sincerely, Armand.

1 0

Re: [Nek5000-users] total volume for velocity mesh
by nek5000-users＠lists.mcs.anl.gov 28 Jun '18

28 Jun '18

What are you looking for if you say cell coordinates? It's trivial to find the coordinates of the say 8 (in 3D) element vortices as findpts returns iel and nid for a given position X. -----Original message----- > From:[email protected] <nek5000-users(a)lists.mcs.anl.gov> > Sent: Thursday 28th June 2018 15:38 > To: nek5000-users(a)lists.mcs.anl.gov > Subject: Re: [Nek5000-users] total volume for velocity mesh > > Hi Stefan, > > I used findpts() in interp_v and interp_v3v to find the velocities in particles positions. Can you please tell me how I can return the cell coordinates that particles belong to because I didnt find any subroutine in the hemi example. > > Kind regards, > > Jian > _______________________________________________ > Nek5000-users mailing list > Nek5000-users(a)lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users

1 0

Re: [Nek5000-users] total volume for velocity mesh
by nek5000-users＠lists.mcs.anl.gov 28 Jun '18

28 Jun '18

Hi Stefan, I used findpts() in interp_v and interp_v3v to find the velocities in particles' positions. Can you please tell me how I can return the cell coordinates that particles belong to because I didn't find any subroutine in the hemi example. Kind regards, Jian

1 0

Running genmap
by nek5000-users＠lists.mcs.anl.gov 27 Jun '18

27 Jun '18

Good morning to all users, I've created a mesh on GAMBIT and then converted it to Nek using a Matlab software that I found in this community. The conversion seems to have worked fine, but right now I want to launch the simulation with several processors. Therefore, I need to run the "genmap" function and I am having trouble with it. Could anyone explain me how to create the .map file? I've tried writing the command "genmap" in the terminal with the corresponding directories but it's not working. The .rea/.usr/SIZE, etc files are in directory /home3/Desktop/R0025 While the genmap file is in the directory /home3/Programmi/nek5_svn/trunk/tools/genmap. What should I exactly write in the terminal as a command? Thank you very much, Roger

1 2

Genmap, Periodic Condition and Matlab converter
by nek5000-users＠lists.mcs.anl.gov 27 Jun '18

27 Jun '18

Good afternoon community, I am having a problem running a simulation on Nek5000. I will try to summarize the steps so that maybe you can easily identify a solution to the problem: 1. Mesh creation using Gambit with Inlet, Outlet, Wall and Periodic Boundary conditions. The mesh has a total of about 120,000 elements. 2. Mesh converstion to Nek format using a Matlab converter. More precisely, the one I found in this link: https://lists.mcs.anl.gov/pipermail/nek5000-users/2010-June/000645.html 3. Running genmap to create the partitions. At this point, I get the following error: PERIODIC MISMATCH 2. I've tried eliminating the boundary conditions and the map file is done correctly, which means that the error is probably on the boundary conditions. I've also checked mesh and every periodic boundary condition seems to be applied properly. The only thing that makes me think is wrong is the matlab conversion from Gambit to Nek5000. In the output mesh, the Periodic Condition is defined as follows (the following is an example): P 5470 2 0 0.00000 0.00000 0.00000 0.00000 Which esentially means that face 2 from element 5470 is periodic with face 0 of element 0. The 3rd and 4th column should represent the face and number of element which is periodic with the first one. In the output mesh after the matlab conversion, these two elements are always 0 and I believe this is a mistake. Has any of you had any experience with this matlab conversor and periodic boundary conditions? Could any of you guide me towards a solution for this problem? Thank you very much, Roger

1 0

Re: [Nek5000-users] total volume for velocity mesh
by nek5000-users＠lists.mcs.anl.gov 27 Jun '18

27 Jun '18

Yes, the code works for arbitrary meshes! findpts() returns (iel,nid,r,s,t) for a given position X. Check the hemi example. Stefan -----Original message----- > From:[email protected] <nek5000-users(a)lists.mcs.anl.gov> > Sent: Wednesday 27th June 2018 11:40 > To: nek5000-users(a)lists.mcs.anl.gov > Subject: Re: [Nek5000-users] total volume for velocity mesh > > Dear Paul, > > Thanks for your reply. In that way, can I say that the volume in channel can be calculated as Lx (length in x direction set in box file) * Ly * Lz? For a pipe, do you think the code is still working to obtain the volume? > > Besides, I am stuck with finding pipe cells where particles exist in lagragian particle tracking. Do you have any suggestion, I can give a limit of pipe cell to determine which particle belongs to? > > Kind regards, > > Jian > > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users(a)lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users

1 0

Re: [Nek5000-users] cell coordinates in the pipe mesh
by nek5000-users＠lists.mcs.anl.gov 27 Jun '18

27 Jun '18

Not sure what you are looking for. You'll find the GLL gridpoints of all (local) elements in XM1,YM1,ZM1. There is nothing special when your elements are deformed like in pipe. Stefan -----Original message----- > From:[email protected] <nek5000-users(a)lists.mcs.anl.gov> > Sent: Wednesday 27th June 2018 11:53 > To: nek5000-users(a)lists.mcs.anl.gov > Subject: [Nek5000-users] cell coordinates in the pipe mesh > > Hi nek, > > Is there anyone know what the cell coordinates refer to in the pipe mesh? In the channel, I can understand the cell coordinates are the ones of rectangual cell centre. For the pipe which has a > squished cell in the mesh, what does the cooridate refer to? > > kind regards, > > Jian > _______________________________________________ > Nek5000-users mailing list > Nek5000-users(a)lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users

1 0