October 2010 - Nek5000-users - mailman3.cels.anl.gov

problem with closed internal boundaries
by nek5000-users＠lists.mcs.anl.gov 23 Jul '13

23 Jul '13

Hello Neks, I'm trying to model a flow past an obstacle in a channel. As a first step I would like to do a 2D simulation. I'm encountering some problems during the geometry data generation in the nek5000 execution : it looks like the program can only generate NaN values although genmap seems to work well on the .rea file. I can also reopen the rea file with prenek without any problems. Do you have any idea what is going wrong here? Is it problematic to have a closed internal boundary condition? Thanks in advance for any help. Best, Michael B

1 9

Periodic B.C and Genmap Issue
by nek5000-users＠lists.mcs.anl.gov 09 Nov '12

09 Nov '12

Hi Paul, This post is in reference to the previous post by Michael (dated Jun 22). When I try to get a map file for a rea file that has periodic vertices, I get the following error. The rea file was generated the same way as before (Gambit -> Nek through Matlab). I checked on Matlab by plotting the periodic vertices and confirmed for a few faces that they are indeed periodic. Also, as you had earlier suggested (with the new genmap), I tried changing the tolerances fro 0.09 to 0.2 and I get the same periodic mismatch error. Could you provide any leads on where things are going wrong ? Please find below the error message. Thanks. <error> Input (.rea) file name: domain Input mesh tolerance (default 0.2): NOTE: smaller is better, but generous is more forgiving for bad meshes. 0.09 reading .rea file data ... start locglob_lexico: 8 3912 31296 8.99999999999999967E-002 locglob: 1 1 31296 locglob: 2 36 31296 locglob: 3 1224 31296 locglob: 1 4851 31296 locglob: 2 4851 31296 locglob: 3 4851 31296 locglob: 1 4851 31296 locglob: 2 4851 31296 locglob: 3 4851 31296 done locglob_lexico: 4851 4851 31296 10 start periodic vtx: 3912 4851 abort: PERIODIC MISMATCH 2: 1093 1 P ie 0 1 je 4.14470000000000029E-003 0.0000000000000000 bc 8 quit </error> Regards Shriram

1 10

Re: [Nek5000-users] Vanishing Jacobian error description
by nek5000-users＠lists.mcs.anl.gov 10 Mar '11

10 Mar '11

Hi Paul, Here is the rea, usr, & SIZE file I am using. I ran it again without the curved side information and it worked, so I think the problem is somewhere in the curved side section of the rea. The geometry is a square base lofted into a circle in the +z direction. The mesh was generated in gambit. I've developed a matlab code that will take a mesh created in gambit and generate the required REA formatting. I have created some basic shapes to test the matlab portion and the curved side portion, and it worked in most cases (not this one yet). I'm not sure what is going on with the curved side data in this case however. Thanks for looking at this, Michael ----- Original Message ----- From: nek5000-users(a)lists.mcs.anl.gov To: "Nekton User List" <nek5000-users(a)lists.mcs.anl.gov> Sent: Tuesday, April 20, 2010 3:36:44 PM GMT -06:00 US/Canada Central Subject: Re: [Nek5000-users] Vanishing Jacobian error description Michael, The code is trying to check to ensure that the mesh is valid. Sometimes, however, there are other causes for the problem. Can you send a gzip'd rea/usr file, and I can take a look later tonight. Paul On Tue, 20 Apr 2010, nek5000-users(a)lists.mcs.anl.gov wrote: > Hi, > > I am running into an issue with 'Vanishing Jacobian'. To help understand what the problem > is, could someone explain what is meant in the error? Specifically, what it means by "Xth node" of the element. > Also, I ran into this issue when I was deforming geometries in the usr file, where one of the elements > would overlap another by mistake. However in this case I know there are no overlaps. > > Here is a portion: > > 0 ERROR: Vanishing Jacobian near 91th node of element 69. > 1.40692804605498586E-002 -7.79619559165757245E-003 > 0 xyz: 2.65874E-01 3.54498E-01 2.87050E+00 > 0 xyz: 2.07855E-01 1.77249E-01 2.87050E+00 > > > 0 ERROR: Vanishing Jacobian near 76th node of element 70. > 2.59472856644434001E-002 -2.12442422718029170E-002 > 0 xyz: 2.11655E-01 4.23310E-01 2.62500E+00 > 0 xyz: 3.54498E-01 1.77249E-01 2.87050E+00 > > Thanks for any help deciphering this error! > > - Michael > > > > > > _______________________________________________ Nek5000-users mailing list Nek5000-users(a)lists.mcs.anl.gov https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users

1 4

lambda2
by nek5000-users＠lists.mcs.anl.gov 28 Jan '11

28 Jan '11

dear neks, in the source file postpro.f i found the following subroutine: subroutine lambda2(l2) c c Generate Lambda-2 vortex of Jeong & Hussein, JFM '95 c ... but it is not used in the code and i wanted calculate it for some visualizations. may you can explain me how i can use it. because i tried just to call it in the usrchk subroutine but then i get an error message (segmentation fault...). best, fred

1 4

Visit and NEK 3D data
by nek5000-users＠lists.mcs.anl.gov 06 Jan '11

06 Jan '11

Hello all, I am running into a problem with Visit and NEK 3D data files. Visit (version 1.12.2) reads them in, but the mesh is all jumbled up. If anyone could try Visit on this file, I'd be grateful, perhaps there is some setting in Visit that I overlooked. The data file (4MB) is: http://tetra.fluid.tuwien.ac.at/fmuldoo/engineering-marangoni-flows/sim2/si… I use the following meta data file: NEK5000 version: 1.0 filetemplate: c203d.fld%02d firsttimestep: 2 numtimesteps: 1 Cheers, Frank -- Frank Herbert Muldoon, Ph.D. Mechanical Engineering Technische Universität Wien (Technical University of Vienna) Inst. f. Strömungsmechanik und Wärmeübertragung (Institute of Fluid Mechanics and Heat Transfer) Resselgasse 3 1040 Wien Tel: +4315880132232 Fax: +4315880132299 Cell:+436765203470 fmuldoo (skype) http://tetra.fluid.tuwien.ac.at/fmuldoo/public_html/webpage/frank-muldoon.h…

1 29

Re: [Nek5000-users] Compute the derivatives
by nek5000-users＠lists.mcs.anl.gov 26 Nov '10

26 Nov '10

Hi Michael, iostep dump works only for the dump of quantities specified as T in your .rea like ***** OUTPUT FIELD SPECIFICATION ***** 8 SPECIFICATIONS FOLLOW T COORDINATES T VELOCITY T PRESSURE T TEMPERATURE F TEMPERATURE GRADIENT 2 PASSIVE SCALARS F F Since userchk is called every timestep so you'll need something like if (mod(istep,iostep).eq.0) then ... Best, Aleks ----- Original Message ----- From: nek5000-users(a)lists.mcs.anl.gov To: nek5000-users(a)lists.mcs.anl.gov Sent: Thursday, July 1, 2010 7:57:07 PM GMT -06:00 US/Canada Central Subject: Re: [Nek5000-users] Compute the derivatives Hi Paul and Aleks, I have a question regarding this. I have implemented this and it is working nicely, however it is dumping out a .f file every step, rather than every iotime step I have specified in the rea. Do you know how to fix this? - Michael ----- Original Message ----- From: nek5000-users(a)lists.mcs.anl.gov To: nek5000-users(a)lists.mcs.anl.gov Sent: Wednesday, June 2, 2010 10:57:41 AM GMT -06:00 US/Canada Central Subject: Re: [Nek5000-users] Compute the derivatives Hi JC, To dump the computed derivatives into separate field files you can use in userchk ifxyo = .true. ! for cooordinates ifpo = .false. ! turn off pressure output ifto = .false. ! turn off pressure output call outpost(vxx,vxy,vxz,pr,t,'dvx') call outpost(vyx,vyy,vyz,pr,t,'dvy') ... Best, Aleks On Wed, 2 Jun 2010, nek5000-users(a)lists.mcs.anl.gov wrote: > > dudxyz is somewhat dated and slow... > > I recommend somthing like: > > parameter (lt=lx1*ly1*lz1*lelv) > common /mygrad/ vxx(lt),vxy(lt),vxz(lt) > $ , vyx(lt),vyy(lt),vyz(lt) > $ , vzx(lt),vzy(lt),vzz(lt) > > call gradm1(vxx,vxy,vxz,vx) > call gradm1(vyx,vyy,vyz,vy) > call gradm1(vzx,vzy,vzz,vz) > > Note that putting these arrays into a common block, as > shown, is a good idea for several reasons, one of which > is that it guarantees that the data will be byte-aligned with the cache line. > On some machines (e.g., BG/P) this > is crucial for proper functionality - for this reason we > don't mix variable types in a given common block. > > If you subsequently wish to access elements of vxx etc. > on an element-by-element basis you can of course do something > like > > parameter (lt=lx1*ly1*lz1) > common /mygrad/ vxx(lt,lelt),vxy(lt,lelt),vxz(lt,lelt) > : > : > > Paul > > On Wed, 2 Jun 2010, nek5000-users(a)lists.mcs.anl.gov wrote: > >> Hi Nek's, >> >> I have a home-made global stability code, but in order to use it I first >> need to compute the derivatives of my base flow: dUdX, dUdY, dVdX, dVdY. >> How could one compute these derivatives using Nek 5k and drop them along >> with X Y U V in the blah.f**** file (or perhaps another .f**** file) ? I >> assume I have to make use of the subroutine *dudxyz >> (du,u,rm1,sm1,tm1,jm1,imsh,isd)*, but I'm not sure what to put as inputs. >> >> Regards, >> JC >> > _______________________________________________ > Nek5000-users mailing list > Nek5000-users(a)lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > _______________________________________________ Nek5000-users mailing list Nek5000-users(a)lists.mcs.anl.gov https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users _______________________________________________ Nek5000-users mailing list Nek5000-users(a)lists.mcs.anl.gov https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users

1 9

Re: [Nek5000-users] Grid 2 Grid Interpolation
by nek5000-users＠lists.mcs.anl.gov 30 Oct '10

30 Oct '10

Neks, We are trying to get the grid to grid interpolation routine working. I am using a simple box as a test case, with a coarse grid of 64 as the "old geometry" that I've let go through several flow throughs, and a finer grid of 700 elements that we want to interpolate on to. I went through the steps on the website given for grid to grid interpolation, but on the last step when I run old geometry with g2gi called in usrchk, I get the following error message in the output file. call usrchk grid-to-grid interpolation ABORT: nelr>lelr on rank 0 ABORT: nelr>lelr on rank 1 call exitt: dying ... This occurred when I ran the old geometry on two processors. When I ran it on one processor, it gave the same message, except it didn't have the 2nd ABORT message. Any help would be much appreciated, as grid2grid should help us tremendously on the full case we are trying to get through. Thanks, Josh

1 0

Finding the maximum
by nek5000-users＠lists.mcs.anl.gov 29 Oct '10

29 Oct '10

Hi, For a computation with eddy viscosity, I would like to find the maximum eddy viscosity within a particular region of my domain and also the location of the max/min value. Is there a routine in nek to give limits on x,y,z and then find max of a particular variable ? glmax() finds the maximum in the entire domain . . . Regards Shriram

1 3

scaling of nek5000 on a Cray
by nek5000-users＠lists.mcs.anl.gov 25 Oct '10

25 Oct '10

Hi nek5000-users & developers, I wonder if anyone has experience with the performance of nek5000 on a Cray ? In particular what scaling could one expect compared to e.g. a Blue Gene? We have a new Cray XT6m system, based on the AMD Opteron 12-core “Magny-Cours” <http://www.cray.com/Products/XT/Specifications.aspx> (2.1 GHz) processors and the Cray SeaStar2 interconnect technology. It consists of 11040 compute cores and 10 service cores. The scaling is not too encouraging (see attached pdf-files). The attached files show three different runs, and the scaling is apparently highly dependent on where in the torus the job happens to be. This is maybe ok, but is the leveling off after 2048 cores consistent with what people observe on a Cray ? Thanks for your suggestions and experience! Best regards, Johan -- Johan Malm Department of Mechanics, KTH SE-100 44, Stockholm, Sweden Phone: +46 8 7906876 E-mail: johan(a)mech.kth.se

1 5

Steady state computation
by nek5000-users＠lists.mcs.anl.gov 25 Oct '10

25 Oct '10

Hi Nek's :-) I would like to compute steady state solutions of the Navier-Stokes equations and use them as a base flow for stability calculations. Being base flows, these solutions are not necessarily the time-asymptotic steady solutions (if they exists). As a matter of fact, I was wondering if any steady-state solver was currently available in Nek or if I would have implement my own one? Sincerely, JC

1 2