From nek5000-users at lists.mcs.anl.gov Fri Dec 3 13:02:12 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 3 Dec 2010 14:02:12 -0500 Subject: [Nek5000-users] Can Prenek read VTK like mesh ? Message-ID: Hello, everyone, I'm using Prenek to create a *.rea file. At the stage of "BUILD INTERACTIVELY ", there is a menu "IMPORT vtk MESH", so I tried to load a vtk mesh generated by GMSH, but unfortunately, I got errors like this: At line 2119 of file build.f (unit = 48, file = 'my_vtk.tmp') Fortran runtime error: Bad integer for item 1 in list input Does prenek support vtk format mesh ? -- Zhicheng Wang -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Sat Dec 4 19:35:40 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sat, 4 Dec 2010 20:35:40 -0500 Subject: [Nek5000-users] genbox issues Message-ID: Hi Neks, After updating my compilation of genbox, it no longer works properly. Has anyone else had this problem? Thanks, David -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Sat Dec 4 20:21:38 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sat, 4 Dec 2010 20:21:38 -0600 (CST) Subject: [Nek5000-users] genbox issues In-Reply-To: Message-ID: <1491536153.169923.1291515698410.JavaMail.root@zimbra.anl.gov> Try changing the sign of ndim and let me know if that fixes the problem. Thanks, Paul ----- Original Message ----- From: nek5000-users at lists.mcs.anl.gov To: nek5000-users at lists.mcs.anl.gov Sent: Saturday, December 4, 2010 7:35:40 PM Subject: [Nek5000-users] genbox issues Hi Neks, After updating my compilation of genbox, it no longer works properly.? Has anyone else had this problem? Thanks, David _______________________________________________ Nek5000-users mailing list Nek5000-users at lists.mcs.anl.gov https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users From nek5000-users at lists.mcs.anl.gov Sat Dec 4 20:50:27 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sat, 4 Dec 2010 21:50:27 -0500 Subject: [Nek5000-users] genbox issues In-Reply-To: <1491536153.169923.1291515698410.JavaMail.root@zimbra.anl.gov> References: <1491536153.169923.1291515698410.JavaMail.root@zimbra.anl.gov> Message-ID: Yes Paul, making ndim positive in the .box file fixed the problem. Thanks! David -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Sun Dec 5 02:17:38 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sun, 5 Dec 2010 02:17:38 -0600 (CST) Subject: [Nek5000-users] genbox issues In-Reply-To: Message-ID: <43806182.170109.1291537058146.JavaMail.root@zimbra.anl.gov> Great! Sorry about that -- We're doing a bit of code re-org / clean-up. For the most part we're trying to avoid disruptions like that one. Paul ----- Original Message ----- From: nek5000-users at lists.mcs.anl.gov To: nek5000-users at lists.mcs.anl.gov Sent: Saturday, December 4, 2010 8:50:27 PM Subject: Re: [Nek5000-users] genbox issues Yes Paul, making ndim positive in the .box file fixed the problem.? Thanks! David _______________________________________________ Nek5000-users mailing list Nek5000-users at lists.mcs.anl.gov https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users From nek5000-users at lists.mcs.anl.gov Mon Dec 6 03:47:34 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 6 Dec 2010 10:47:34 +0100 Subject: [Nek5000-users] Impose localized noise Message-ID: Hi Nek's, I would like to impose noise localized in time and space in my stenotic pipe, pretty much like an impulse. I would be tempted to use the userf subroutine like this: subroutine userf (ix,iy,iz,eg) include 'SIZE' include 'TOTAL' include 'NEKUSE' integer e,f,eg real z_0, sigma_z, istep_0, istep_min, istep_max, Amplitude, Theta z_0 = 0. sigma_z = 1. istept_0 = 100. istep_min = 90. istep_max = 110. sigma_t = 1. if(istep.GE.istep_min.AND.istep.LE.istep_max) Theta = atan2(y,x) ! I would like my forcing to be on u_theta only Amplitude = exp(-((z-z_0)/sigma_z)**2 - ((istep-istep_0)/sigma_t)**2) ffx = - sin(Theta) * Amplitude * Random Noise ffy = cos(Theta) * Amplitude * Random Noise ffz = 0.0 return end The thing is I don't know how I could impose some random noise on the whole domain. Using " call random_number(Random_Noise) " would only give me one single random number. I assume the solution is not to hard but I am however missing the point. Any idea? Kind regards, JC -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Dec 6 07:14:57 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 6 Dec 2010 07:14:57 -0600 (CST) Subject: [Nek5000-users] Impose localized noise In-Reply-To: References: Message-ID: Dear JC, Do you want "random_noise" to be a function of space and time? or of time only ? Paul On Mon, 6 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > Hi Nek's, > > I would like to impose noise localized in time and space in my stenotic > pipe, pretty much like an impulse. > I would be tempted to use the userf subroutine like this: > > subroutine userf (ix,iy,iz,eg) > include 'SIZE' > include 'TOTAL' > include 'NEKUSE' > > integer e,f,eg > real z_0, sigma_z, istep_0, istep_min, istep_max, Amplitude, Theta > > z_0 = 0. > sigma_z = 1. > > istept_0 = 100. > istep_min = 90. > istep_max = 110. > sigma_t = 1. > > if(istep.GE.istep_min.AND.istep.LE.istep_max) > > Theta = atan2(y,x) > ! I would like my forcing to be on u_theta only > Amplitude = exp(-((z-z_0)/sigma_z)**2 - ((istep-istep_0)/sigma_t)**2) > > ffx = - sin(Theta) * Amplitude * Random Noise > ffy = cos(Theta) * Amplitude * Random Noise > ffz = 0.0 > > return > end > > The thing is I don't know how I could impose some random noise on the whole > domain. Using " call random_number(Random_Noise) " would only give me one > single random number. I assume the solution is not to hard but I am however > missing the point. Any idea? > > Kind regards, > JC > From nek5000-users at lists.mcs.anl.gov Mon Dec 6 09:41:01 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 6 Dec 2010 09:41:01 -0600 (CST) Subject: [Nek5000-users] Can Prenek read VTK like mesh ? In-Reply-To: References: Message-ID: Yes - it can read it. The mesh needs to be in ascii and there is usually some modification required either to the source or to the header since the vtk file formats can be varied. You can either edit your prenek source or your file to suit, or if you wish to send me a sample vtk file, I can update prenek to handle this. Paul On Fri, 3 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > Hello, everyone, > > I'm using Prenek to create a *.rea file. > At the stage of "BUILD INTERACTIVELY ", there is a menu "IMPORT vtk MESH", > so I tried to load a vtk mesh generated by GMSH, > but unfortunately, I got errors like this: > > At line 2119 of file build.f (unit = 48, file = 'my_vtk.tmp') > Fortran runtime error: Bad integer for item 1 in list input > > Does prenek support vtk format mesh ? > > > -- > Zhicheng Wang > From nek5000-users at lists.mcs.anl.gov Mon Dec 6 10:01:57 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 6 Dec 2010 11:01:57 -0500 Subject: [Nek5000-users] Can Prenek read VTK like mesh ? In-Reply-To: References: Message-ID: Hi Paul, Thank you very much for your reply. Here I attached an example of vtk like mesh created by Gmsh. It is a square (0,1)*(0,1) with 2*2 elements. zhicheng On Mon, Dec 6, 2010 at 10:41 AM, wrote: > > Yes - it can read it. The mesh needs to be in ascii > and there is usually some modification required either > to the source or to the header since the vtk file formats > can be varied. > > You can either edit your prenek source or your file to > suit, or if you wish to send me a sample vtk file, I can > update prenek to handle this. > > Paul > > > > On Fri, 3 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > > Hello, everyone, >> >> I'm using Prenek to create a *.rea file. >> At the stage of "BUILD INTERACTIVELY ", there is a menu "IMPORT vtk MESH", >> so I tried to load a vtk mesh generated by GMSH, >> but unfortunately, I got errors like this: >> >> At line 2119 of file build.f (unit = 48, file = 'my_vtk.tmp') >> Fortran runtime error: Bad integer for item 1 in list input >> >> Does prenek support vtk format mesh ? >> >> >> -- >> Zhicheng Wang >> >> _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > -- Zhicheng Wang Ph.D. Candidate RM 402, No.11 Beisihuan West Road Institute of Engineering Thermophysics, Chinese Academy of Sciences Beijing, 100190, P.R. China Tel: (86) 10-82543092 E-mail: wangzhicheng09 at gmail.com; wangzhicheng at iet.cn -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Square.vtk Type: application/octet-stream Size: 418 bytes Desc: not available URL: From nek5000-users at lists.mcs.anl.gov Tue Dec 7 03:31:26 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 7 Dec 2010 10:31:26 +0100 Subject: [Nek5000-users] Impose localized noise In-Reply-To: References: Message-ID: Well, I would like to impose random noise in the whole domain (which I then localize in the present calculation through the use of a gaussian) at least for one time step. The aim is just to trigger transition from a steady state to another state. So it is indeed a function of space, and it may be a function of time also if I want zero mean noise but this is not necessary as I'd like to impose it just for a very few time steps. -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Dec 7 04:05:02 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 7 Dec 2010 04:05:02 -0600 (CST) Subject: [Nek5000-users] Impose localized noise In-Reply-To: References: Message-ID: There are two basic approaches: 1) put the function directly into userf(), e.g., ffx = your_function_x(x,y,z,time) ffy = your_function_y(x,y,z,time) ffz = your_function_z(x,y,z,time) 2) Fill arrays your_ffx, your_ffy, your_ffz in userchk and then dereference the arrays in userf(ix,iy,iz,eg) via: include 'SIZE' include 'TOTAL' common /your_arrays/ your_ffx(lx1,ly1,lz1,lelt) $ , your_ffy(lx1,ly1,lz1,lelt) $ , your_ffz(lx1,ly1,lz1,lelt) integer e,eg : : e = gllel(eg) ffx = your_ffx(ix,iy,iz,e) ffy = your_ffy(ix,iy,iz,e) ffz = your_ffz(ix,iy,iz,e) : return end The latter is preferable if evaluation of ffx,ffy,ffz involves nonlocal (i.e., non-pointwise) function evaluations such as derivatives, etc. NOTE: Despite their names, ffx, ffy, and ffz are really _accelerations_ --- not forces. They are later multiplied by rho. (I always use rho=1, so this is not an issue... but beware, and verify...) --Paul On Tue, 7 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > Well, I would like to impose random noise in the whole domain (which I then > localize in the present calculation through the use of a gaussian) at least > for one time step. The aim is just to trigger transition from a steady state > to another state. > > So it is indeed a function of space, and it may be a function of time also > if I want zero mean noise but this is not necessary as I'd like to impose it > just for a very few time steps. > From nek5000-users at lists.mcs.anl.gov Wed Dec 8 04:20:16 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 8 Dec 2010 11:20:16 +0100 Subject: [Nek5000-users] opdiv and multd Message-ID: <4CFF5BE0.4030403@lav.mavt.ethz.ch> Hi all, I am having some problems using the subroutines opdiv and multd. For some reasons they do not give me correct results. Do they need any special treatment or precautions? I was trying to compute the divergence of a 3D field, and i realized that in my case the multd subroutine used by opdiv does not gives good results, while the same derivative computed from gram1 gives no problems. For example, I overwrite an analytical field : ... parameter(lt=lx1*ly1*lz1*lelt) real Res(lt) ... vx(i,j,k,e)= 1.0*sin(x)*cos(y)*cos(z) ... call print_line(vx) ... iflg = 1 call multd (RES(1),vx,rym2,sym2,tym2,2,iflg) call print_line(Res(1) ) call gradm1(work1,Res(1),work2,vx) call print_line(Res(1) ) ... if (lx2.ne.lx1) then ... STOP endif and print output a (xr,:,zr) the result (performed by print_line) . In vxi.eps i compare the result (lines) with the analytical solution( symbols). The result of multd is not correct while all the others match perfectly the analytical solution. In vxy.eps i plot only the result of multd. Note that on y i have 12 elements the same number of peaks of the solution. Does anybody have an idea on how to fix this? Thank you very much for any help or suggestion francesco -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: vxi.eps Type: image/x-eps Size: 24269 bytes Desc: vxi.eps URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: vxy.eps Type: image/x-eps Size: 19571 bytes Desc: vxy.eps URL: From nek5000-users at lists.mcs.anl.gov Wed Dec 8 07:17:19 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 8 Dec 2010 07:17:19 -0600 (CST) Subject: [Nek5000-users] opdiv and multd In-Reply-To: <4CFF5BE0.4030403@lav.mavt.ethz.ch> References: <4CFF5BE0.4030403@lav.mavt.ethz.ch> Message-ID: Hi Francesco, Several of the routines internal to nek are the so-called weak derivatives, implying that they somehow involve the mass matrix or the transpose of the operator, so it takes a bit of investigation to understand precisely what each operator is doing. (Note that the functionality also differs slightly between PN-PN and PN-PN-2 formulations.) As an example, axhelm does not multiply by the (negative) Laplacian. Instead, it creates (for h1==1, h2==0): w = A * u = grad^T B grad u where B is the mass matrix. Similarly, opdiv is most likely returning d = B*grad u (or perhaps the negative of this). Moreover, d is returned on the pressure mesh (mesh 2), which is the same as the velocity mesh (mesh 1) for the PN-PN formulation that you use. In Nek, the mass matrix is diagonal and the mesh 1 inverse is stored in binvm1. It seems probable that you'll get the correct result after call col2(d,binvm1,n) where d holds the output of opdiv and n is the number of points on mesh 1, assuming you're using PN-Pn. Paul On Wed, 8 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > Hi all, > > I am having some problems using the subroutines opdiv and multd. > For some reasons they do not give me correct results. > Do they need any special treatment or precautions? > > I was trying to compute the divergence of a 3D field, and i realized that in > my case the multd subroutine used > by opdiv does not gives good results, while the same derivative computed from > gram1 gives no problems. > > For example, I overwrite an analytical field : > > ... > parameter(lt=lx1*ly1*lz1*lelt) > real Res(lt) > ... > vx(i,j,k,e)= 1.0*sin(x)*cos(y)*cos(z) > ... > call print_line(vx) > ... > iflg = 1 > call multd (RES(1),vx,rym2,sym2,tym2,2,iflg) > call print_line(Res(1) ) > > call gradm1(work1,Res(1),work2,vx) > call print_line(Res(1) ) > ... > if (lx2.ne.lx1) then > ... > STOP > endif > > > and print output a (xr,:,zr) the result (performed by print_line) . > In vxi.eps i compare the result (lines) with the analytical solution( > symbols). The result of multd is not correct while all > the others match perfectly the analytical solution. > In vxy.eps i plot only the result of multd. Note that on y i have 12 elements > the same number of peaks of the solution. > > Does anybody have an idea on how to fix this? > > Thank you very much for any help or suggestion > > francesco > > > > From nek5000-users at lists.mcs.anl.gov Wed Dec 8 08:01:33 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 8 Dec 2010 15:01:33 +0100 Subject: [Nek5000-users] opdiv and multd In-Reply-To: References: <4CFF5BE0.4030403@lav.mavt.ethz.ch> Message-ID: <4CFF8FBD.6010306@lav.mavt.ethz.ch> Thank you Paul, that solved the problem. The correct result is obtained by : call multd (RES(1),vx,rym2,sym2,tym2,2,iflg) call dssum(RES(1)) call col2(RES(1),binvm1,lt) francesco On 12/08/2010 02:17 PM, nek5000-users at lists.mcs.anl.gov wrote: > > Hi Francesco, > > Several of the routines internal to nek are the so-called weak > derivatives, implying that they somehow involve the mass matrix > or the transpose of the operator, so it takes a bit of investigation > to understand precisely what each operator is doing. (Note that the > functionality also differs slightly between PN-PN and PN-PN-2 > formulations.) > > As an example, axhelm does not multiply by the (negative) Laplacian. > Instead, it creates (for h1==1, h2==0): > > w = A * u > > = grad^T B grad u > > where B is the mass matrix. > > Similarly, opdiv is most likely returning d = B*grad u (or perhaps > the negative of this). Moreover, d is returned on the pressure > mesh (mesh 2), which is the same as the velocity mesh (mesh 1) for the > PN-PN formulation that you use. > > In Nek, the mass matrix is diagonal and the mesh 1 inverse is stored > in binvm1. It seems probable that you'll get the correct result > after > > call col2(d,binvm1,n) > > where d holds the output of opdiv and n is the number of points on > mesh 1, assuming you're using PN-Pn. > > Paul > > > > On Wed, 8 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > >> Hi all, >> >> I am having some problems using the subroutines opdiv and multd. >> For some reasons they do not give me correct results. >> Do they need any special treatment or precautions? >> >> I was trying to compute the divergence of a 3D field, and i realized >> that in my case the multd subroutine used >> by opdiv does not gives good results, while the same derivative >> computed from gram1 gives no problems. >> >> For example, I overwrite an analytical field : >> >> ... >> parameter(lt=lx1*ly1*lz1*lelt) >> real Res(lt) >> ... >> vx(i,j,k,e)= 1.0*sin(x)*cos(y)*cos(z) >> ... >> call print_line(vx) >> ... >> iflg = 1 >> call multd (RES(1),vx,rym2,sym2,tym2,2,iflg) >> call print_line(Res(1) ) >> >> call gradm1(work1,Res(1),work2,vx) >> call print_line(Res(1) ) >> ... >> if (lx2.ne.lx1) then >> ... >> STOP >> endif >> >> >> and print output a (xr,:,zr) the result (performed by print_line) . >> In vxi.eps i compare the result (lines) with the analytical solution( >> symbols). The result of multd is not correct while all >> the others match perfectly the analytical solution. >> In vxy.eps i plot only the result of multd. Note that on y i have 12 >> elements the same number of peaks of the solution. >> >> Does anybody have an idea on how to fix this? >> >> Thank you very much for any help or suggestion >> >> francesco >> >> >> >> > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users From nek5000-users at lists.mcs.anl.gov Wed Dec 8 09:02:07 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 8 Dec 2010 09:02:07 -0600 (CST) Subject: [Nek5000-users] opdiv and multd In-Reply-To: <4CFF8FBD.6010306@lav.mavt.ethz.ch> References: <4CFF5BE0.4030403@lav.mavt.ethz.ch> <4CFF8FBD.6010306@lav.mavt.ethz.ch> Message-ID: Thanks Francesco! Paul On Wed, 8 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > Thank you Paul, > that solved the problem. > > The correct result is obtained by : > call multd (RES(1),vx,rym2,sym2,tym2,2,iflg) > call dssum(RES(1)) > call col2(RES(1),binvm1,lt) > > francesco > > > > On 12/08/2010 02:17 PM, nek5000-users at lists.mcs.anl.gov wrote: >> >> Hi Francesco, >> >> Several of the routines internal to nek are the so-called weak >> derivatives, implying that they somehow involve the mass matrix >> or the transpose of the operator, so it takes a bit of investigation >> to understand precisely what each operator is doing. (Note that the >> functionality also differs slightly between PN-PN and PN-PN-2 >> formulations.) >> >> As an example, axhelm does not multiply by the (negative) Laplacian. >> Instead, it creates (for h1==1, h2==0): >> >> w = A * u >> >> = grad^T B grad u >> >> where B is the mass matrix. >> >> Similarly, opdiv is most likely returning d = B*grad u (or perhaps >> the negative of this). Moreover, d is returned on the pressure >> mesh (mesh 2), which is the same as the velocity mesh (mesh 1) for the >> PN-PN formulation that you use. >> >> In Nek, the mass matrix is diagonal and the mesh 1 inverse is stored >> in binvm1. It seems probable that you'll get the correct result >> after >> >> call col2(d,binvm1,n) >> >> where d holds the output of opdiv and n is the number of points on >> mesh 1, assuming you're using PN-Pn. >> >> Paul >> >> >> >> On Wed, 8 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: >> >>> Hi all, >>> >>> I am having some problems using the subroutines opdiv and multd. >>> For some reasons they do not give me correct results. >>> Do they need any special treatment or precautions? >>> >>> I was trying to compute the divergence of a 3D field, and i realized that >>> in my case the multd subroutine used >>> by opdiv does not gives good results, while the same derivative computed >>> from gram1 gives no problems. >>> >>> For example, I overwrite an analytical field : >>> >>> ... >>> parameter(lt=lx1*ly1*lz1*lelt) >>> real Res(lt) >>> ... >>> vx(i,j,k,e)= 1.0*sin(x)*cos(y)*cos(z) >>> ... >>> call print_line(vx) >>> ... >>> iflg = 1 >>> call multd (RES(1),vx,rym2,sym2,tym2,2,iflg) >>> call print_line(Res(1) ) >>> >>> call gradm1(work1,Res(1),work2,vx) >>> call print_line(Res(1) ) >>> ... >>> if (lx2.ne.lx1) then >>> ... >>> STOP >>> endif >>> >>> >>> and print output a (xr,:,zr) the result (performed by print_line) . >>> In vxi.eps i compare the result (lines) with the analytical solution( >>> symbols). The result of multd is not correct while all >>> the others match perfectly the analytical solution. >>> In vxy.eps i plot only the result of multd. Note that on y i have 12 >>> elements the same number of peaks of the solution. >>> >>> Does anybody have an idea on how to fix this? >>> >>> Thank you very much for any help or suggestion >>> >>> francesco >>> >>> >>> >>> >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > From nek5000-users at lists.mcs.anl.gov Thu Dec 9 10:37:54 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 09 Dec 2010 17:37:54 +0100 Subject: [Nek5000-users] Connectivety In-Reply-To: References: <4CFF5BE0.4030403@lav.mavt.ethz.ch> <4CFF8FBD.6010306@lav.mavt.ethz.ch> Message-ID: <4D0105E2.9010509@mech.kth.se> Hello Paul, I have a question about the connectivity of the elements. In the old rea file format (which we convert to the new binary with reatore2) we have listed the Element geometrical date and after that the Boundary conditions and connectivity data. My question is whether the connectivity of the elements stated there is actually used or it is somehow re-computed while the code runs based on the node coordinates. I am talking specifically about the connectivity between the elements, not the boundary conditions with the "outside world". Best regards, Antonios -- --------------------------------------------------------------------------------------------- Antonios Monokrousos Department of Mechanics email: antonios at mech.kth.se Royal Institute of Technology (KTH) phone: +46 8 790 68 76 Osquars Backe 18 telefax: +46 8 796 98 50 SE-100 44 Stockholm, SWEDEN From nek5000-users at lists.mcs.anl.gov Thu Dec 9 11:15:02 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 9 Dec 2010 11:15:02 -0600 (CST) Subject: [Nek5000-users] Connectivety In-Reply-To: <4D0105E2.9010509@mech.kth.se> References: <4CFF5BE0.4030403@lav.mavt.ethz.ch> <4CFF8FBD.6010306@lav.mavt.ethz.ch> <4D0105E2.9010509@mech.kth.se> Message-ID: Hi Antonios, I could be wrong but I think that the only connectivity information that Nek5000 need comes from global vertexi element numbering genereated by .map file which is in turn produced by running genmap on .rea file Best, Aleks On Thu, 9 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > Hello Paul, > I have a question about the connectivity of the elements. > In the old rea file format (which we convert to the new binary with > reatore2) we have listed the Element geometrical date and after that the > Boundary conditions and connectivity data. > My question is whether the connectivity of the elements stated there is > actually used or it is somehow re-computed while the code runs based on > the node coordinates. I am talking specifically about the connectivity > between the elements, not the boundary conditions with the "outside world". > Best regards, > Antonios > > -- > --------------------------------------------------------------------------------------------- > Antonios Monokrousos > Department of Mechanics email: antonios at mech.kth.se > Royal Institute of Technology (KTH) phone: +46 8 790 68 76 > Osquars Backe 18 telefax: +46 8 796 98 50 > SE-100 44 Stockholm, SWEDEN > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > From nek5000-users at lists.mcs.anl.gov Thu Dec 9 11:18:34 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 9 Dec 2010 11:18:34 -0600 (CST) Subject: [Nek5000-users] Connectivety In-Reply-To: Message-ID: <1434703674.1822.1291915114024.JavaMail.root@zimbra.anl.gov> This is correct. -Stefan ----- Original Message ----- From: nek5000-users at lists.mcs.anl.gov To: nek5000-users at lists.mcs.anl.gov Sent: Thursday, December 9, 2010 6:15:02 PM Subject: Re: [Nek5000-users] Connectivety Hi Antonios, I could be wrong but I think that the only connectivity information that Nek5000 need comes from global vertexi element numbering genereated by .map file which is in turn produced by running genmap on .rea file Best, Aleks On Thu, 9 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > Hello Paul, > I have a question about the connectivity of the elements. > In the old rea file format (which we convert to the new binary with > reatore2) we have listed the Element geometrical date and after that the > Boundary conditions and connectivity data. > My question is whether the connectivity of the elements stated there is > actually used or it is somehow re-computed while the code runs based on > the node coordinates. I am talking specifically about the connectivity > between the elements, not the boundary conditions with the "outside world". > Best regards, > Antonios > > -- > --------------------------------------------------------------------------------------------- > Antonios Monokrousos > Department of Mechanics email: antonios at mech.kth.se > Royal Institute of Technology (KTH) phone: +46 8 790 68 76 > Osquars Backe 18 telefax: +46 8 796 98 50 > SE-100 44 Stockholm, SWEDEN > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > _______________________________________________ Nek5000-users mailing list Nek5000-users at lists.mcs.anl.gov https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users From nek5000-users at lists.mcs.anl.gov Thu Dec 9 12:46:36 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 9 Dec 2010 13:46:36 -0500 Subject: [Nek5000-users] rms fluctuations of passive scalar Message-ID: Hi Neks, I'm doing LES of heat transfer in a fully turbulent plane channel flow with uniform thermal boundary condition. Re_tau=150, Pr=0.7. The domain is 6.4/2.0/3.2, 8*8*8 elements using N=8. When I compare my results with Kasagi's, it looks like the mean velocity and temperature distribution are good. AT dimensionless time 165, the rms value of temperature is good, however, as I restart from time 165 to 402, the difference between my results and Kasagi's became larger and larger. I'm using standard eddy viscosity model, the same as example turbChannel. In case some one had this kind of problem, I attached two figures, one is rms fluctuation at dimensionless time 165, the other is that at 402. -- Zhicheng Wang -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: tt_means.eps Type: application/postscript Size: 24302 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: tt_means_bak4.eps Type: application/postscript Size: 24312 bytes Desc: not available URL: From nek5000-users at lists.mcs.anl.gov Fri Dec 10 03:10:16 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 10 Dec 2010 10:10:16 +0100 Subject: [Nek5000-users] Connectivety In-Reply-To: <1434703674.1822.1291915114024.JavaMail.root@zimbra.anl.gov> References: <1434703674.1822.1291915114024.JavaMail.root@zimbra.anl.gov> Message-ID: <4D01EE78.4030307@mech.kth.se> Thank you for the responses. However maybe my question I was clear before. I try again: Ok, the connectivity information comes from the rea file when you run genmap, but is it computed based on the spatial coordinates given in the rea file, or is it just taken from the connectivity information given later, also in the rea file along with the boundary conditions? What we experience is that we have a grid which is created in Matlab and then we write it in "rea file format" which is later read by genmap and reatore2 etc. This rea file, after the spatial coordinates of the element corners, contains a list with all the elements and their connectivity information. Here is what two lines looks like: W 4 1 0.000000 0.000000 0.000000 0.000000 0.000000 E 4 2 5.000000 4.000000 0.000000 0.000000 0.000000 Here I specify Wall condition for element 4 and face 1 and Element condition for face 2 that connects to element 5 with face 4. But then when I removed my connectivity info, that is having every where zeros in the 4th and 5th column: W 4 1 0.000000 0.000000 0.000000 0.000000 0.000000 E 4 2 0.000000 0.000000 0.000000 0.000000 0.000000 the case seemed to run fine, and the flow looked correct. Does this mean that the connectivity specified there does not matter? Best, /Antonios nek5000-users at lists.mcs.anl.gov wrote: > This is correct. > -Stefan > > ----- Original Message ----- > From: nek5000-users at lists.mcs.anl.gov > To: nek5000-users at lists.mcs.anl.gov > Sent: Thursday, December 9, 2010 6:15:02 PM > Subject: Re: [Nek5000-users] Connectivety > > Hi Antonios, > > I could be wrong but I think that the only connectivity information that > Nek5000 need comes from global vertexi element numbering genereated by > .map file which is in turn produced by running genmap on .rea file > > Best, > Aleks > > > On Thu, 9 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > >> Hello Paul, >> I have a question about the connectivity of the elements. >> In the old rea file format (which we convert to the new binary with >> reatore2) we have listed the Element geometrical date and after that the >> Boundary conditions and connectivity data. >> My question is whether the connectivity of the elements stated there is >> actually used or it is somehow re-computed while the code runs based on >> the node coordinates. I am talking specifically about the connectivity >> between the elements, not the boundary conditions with the "outside world". >> Best regards, >> Antonios >> >> -- >> --------------------------------------------------------------------------------------------- >> Antonios Monokrousos >> Department of Mechanics email: antonios at mech.kth.se >> Royal Institute of Technology (KTH) phone: +46 8 790 68 76 >> Osquars Backe 18 telefax: +46 8 796 98 50 >> SE-100 44 Stockholm, SWEDEN >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users -- --------------------------------------------------------------------------------------------- Antonios Monokrousos Department of Mechanics email: antonios at mech.kth.se Royal Institute of Technology (KTH) phone: +46 8 790 68 76 Osquars Backe 18 telefax: +46 8 796 98 50 SE-100 44 Stockholm, SWEDEN From nek5000-users at lists.mcs.anl.gov Fri Dec 10 04:11:26 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 10 Dec 2010 04:11:26 -0600 (CST) Subject: [Nek5000-users] Connectivety In-Reply-To: <4D01EE78.4030307@mech.kth.se> Message-ID: <1360201019.6212.1291975886134.JavaMail.root@zimbra.anl.gov> Antonois: I understand your confusion. Let me backup a little bit. Genmap will try to establish a unique global numbering for all element vertices. Basically, genmap tries to detect (based on the vertex coordinates) the shared vertices and will assign a global id. We compute what you call the connectivity information on the fly (during the setup phase of Nek) based on the global vertex ids read from the .map file. You're right the old ASCII .rea file contains the elemental connectivity but this information is no longer used (we do a formatted dummy read for these lines). The .re2 file does not contain this information anymore. hth, Stefan ----- Original Message ----- From: nek5000-users at lists.mcs.anl.gov To: nek5000-users at lists.mcs.anl.gov Sent: Friday, December 10, 2010 10:10:16 AM Subject: Re: [Nek5000-users] Connectivety Thank you for the responses. However maybe my question I was clear before. I try again: Ok, the connectivity information comes from the rea file when you run genmap, but is it computed based on the spatial coordinates given in the rea file, or is it just taken from the connectivity information given later, also in the rea file along with the boundary conditions? What we experience is that we have a grid which is created in Matlab and then we write it in "rea file format" which is later read by genmap and reatore2 etc. This rea file, after the spatial coordinates of the element corners, contains a list with all the elements and their connectivity information. Here is what two lines looks like: W 4 1 0.000000 0.000000 0.000000 0.000000 0.000000 E 4 2 5.000000 4.000000 0.000000 0.000000 0.000000 Here I specify Wall condition for element 4 and face 1 and Element condition for face 2 that connects to element 5 with face 4. But then when I removed my connectivity info, that is having every where zeros in the 4th and 5th column: W 4 1 0.000000 0.000000 0.000000 0.000000 0.000000 E 4 2 0.000000 0.000000 0.000000 0.000000 0.000000 the case seemed to run fine, and the flow looked correct. Does this mean that the connectivity specified there does not matter? Best, /Antonios nek5000-users at lists.mcs.anl.gov wrote: > This is correct. > -Stefan > > ----- Original Message ----- > From: nek5000-users at lists.mcs.anl.gov > To: nek5000-users at lists.mcs.anl.gov > Sent: Thursday, December 9, 2010 6:15:02 PM > Subject: Re: [Nek5000-users] Connectivety > > Hi Antonios, > > I could be wrong but I think that the only connectivity information that > Nek5000 need comes from global vertexi element numbering genereated by > .map file which is in turn produced by running genmap on .rea file > > Best, > Aleks > > > On Thu, 9 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > >> Hello Paul, >> I have a question about the connectivity of the elements. >> In the old rea file format (which we convert to the new binary with >> reatore2) we have listed the Element geometrical date and after that the >> Boundary conditions and connectivity data. >> My question is whether the connectivity of the elements stated there is >> actually used or it is somehow re-computed while the code runs based on >> the node coordinates. I am talking specifically about the connectivity >> between the elements, not the boundary conditions with the "outside world". >> Best regards, >> Antonios >> >> -- >> --------------------------------------------------------------------------------------------- >> Antonios Monokrousos >> Department of Mechanics email: antonios at mech.kth.se >> Royal Institute of Technology (KTH) phone: +46 8 790 68 76 >> Osquars Backe 18 telefax: +46 8 796 98 50 >> SE-100 44 Stockholm, SWEDEN >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users -- --------------------------------------------------------------------------------------------- Antonios Monokrousos Department of Mechanics email: antonios at mech.kth.se Royal Institute of Technology (KTH) phone: +46 8 790 68 76 Osquars Backe 18 telefax: +46 8 796 98 50 SE-100 44 Stockholm, SWEDEN _______________________________________________ Nek5000-users mailing list Nek5000-users at lists.mcs.anl.gov https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users From nek5000-users at lists.mcs.anl.gov Fri Dec 10 04:27:10 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 10 Dec 2010 11:27:10 +0100 Subject: [Nek5000-users] Connectivety In-Reply-To: <1360201019.6212.1291975886134.JavaMail.root@zimbra.anl.gov> References: <1360201019.6212.1291975886134.JavaMail.root@zimbra.anl.gov> Message-ID: Hi Stefan, It is clear. Thank you very much. Antonios On 10 dec 2010, at 11:11, nek5000-users at lists.mcs.anl.gov wrote: > Antonois: > > I understand your confusion. Let me backup a little bit. > > Genmap will try to establish a unique global numbering for all element vertices. Basically, genmap tries to detect (based on the vertex coordinates) the shared vertices and will assign a global id. We compute what you call the connectivity information on the fly (during the setup phase of Nek) based on the global vertex ids read from the .map file. > > You're right the old ASCII .rea file contains the elemental connectivity but this information is no longer used (we do a formatted dummy read for these lines). The .re2 file does not contain this information anymore. > > > hth, > Stefan > > > ----- Original Message ----- > From: nek5000-users at lists.mcs.anl.gov > To: nek5000-users at lists.mcs.anl.gov > Sent: Friday, December 10, 2010 10:10:16 AM > Subject: Re: [Nek5000-users] Connectivety > > Thank you for the responses. > However maybe my question I was clear before. I try again: Ok, the > connectivity information comes from the rea file when you run genmap, > but is it computed based on the spatial coordinates given in the rea > file, or is it just taken from the connectivity information given later, > also in the rea file along with the boundary conditions? > > What we experience is that we have a grid which is created in Matlab and > then we write it in "rea file format" which is later read by genmap and > reatore2 etc. This rea file, after the spatial coordinates of the > element corners, contains a list with all the elements and their > connectivity information. > Here is what two lines looks like: > > W 4 1 0.000000 0.000000 0.000000 0.000000 0.000000 > E 4 2 5.000000 4.000000 0.000000 0.000000 0.000000 > > Here I specify Wall condition for element 4 and face 1 and Element > condition for face 2 that connects to element 5 with face 4. > But then when I removed my connectivity info, that is having every where > zeros in the 4th and 5th column: > > W 4 1 0.000000 0.000000 0.000000 0.000000 0.000000 > E 4 2 0.000000 0.000000 0.000000 0.000000 0.000000 > > the case seemed to run fine, and the flow looked correct. > Does this mean that the connectivity specified there does not matter? > > Best, > /Antonios > > > > nek5000-users at lists.mcs.anl.gov wrote: >> This is correct. >> -Stefan >> >> ----- Original Message ----- >> From: nek5000-users at lists.mcs.anl.gov >> To: nek5000-users at lists.mcs.anl.gov >> Sent: Thursday, December 9, 2010 6:15:02 PM >> Subject: Re: [Nek5000-users] Connectivety >> >> Hi Antonios, >> >> I could be wrong but I think that the only connectivity information that >> Nek5000 need comes from global vertexi element numbering genereated by >> .map file which is in turn produced by running genmap on .rea file >> >> Best, >> Aleks >> >> >> On Thu, 9 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: >> >>> Hello Paul, >>> I have a question about the connectivity of the elements. >>> In the old rea file format (which we convert to the new binary with >>> reatore2) we have listed the Element geometrical date and after that the >>> Boundary conditions and connectivity data. >>> My question is whether the connectivity of the elements stated there is >>> actually used or it is somehow re-computed while the code runs based on >>> the node coordinates. I am talking specifically about the connectivity >>> between the elements, not the boundary conditions with the "outside world". >>> Best regards, >>> Antonios >>> >>> -- >>> --------------------------------------------------------------------------------------------- >>> Antonios Monokrousos >>> Department of Mechanics email: antonios at mech.kth.se >>> Royal Institute of Technology (KTH) phone: +46 8 790 68 76 >>> Osquars Backe 18 telefax: +46 8 796 98 50 >>> SE-100 44 Stockholm, SWEDEN >>> _______________________________________________ >>> Nek5000-users mailing list >>> Nek5000-users at lists.mcs.anl.gov >>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >>> >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > > > -- > --------------------------------------------------------------------------------------------- > Antonios Monokrousos > Department of Mechanics email: antonios at mech.kth.se > Royal Institute of Technology (KTH) phone: +46 8 790 68 76 > Osquars Backe 18 telefax: +46 8 796 98 50 > SE-100 44 Stockholm, SWEDEN > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users From nek5000-users at lists.mcs.anl.gov Sat Dec 11 12:51:35 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sat, 11 Dec 2010 12:51:35 -0600 (CST) Subject: [Nek5000-users] Connectivety In-Reply-To: <840575852.2204961292093391687.JavaMail.root@neo-mail-3> Message-ID: <903849407.2205351292093495782.JavaMail.root@neo-mail-3> Hi Stefan, Do you recall the repo at which point this started to be the case? We have a similar code to convert gambit files and a majority of the time is spent getting this connectivity information. If we don't need it then that is welcomed news! - Michael M. ----- Original Message ----- From: nek5000-users at lists.mcs.anl.gov To: nek5000-users at lists.mcs.anl.gov Sent: Friday, December 10, 2010 4:11:26 AM GMT -06:00 US/Canada Central Subject: Re: [Nek5000-users] Connectivety Antonois: I understand your confusion. Let me backup a little bit. Genmap will try to establish a unique global numbering for all element vertices. Basically, genmap tries to detect (based on the vertex coordinates) the shared vertices and will assign a global id. We compute what you call the connectivity information on the fly (during the setup phase of Nek) based on the global vertex ids read from the .map file. You're right the old ASCII .rea file contains the elemental connectivity but this information is no longer used (we do a formatted dummy read for these lines). The .re2 file does not contain this information anymore. hth, Stefan ----- Original Message ----- From: nek5000-users at lists.mcs.anl.gov To: nek5000-users at lists.mcs.anl.gov Sent: Friday, December 10, 2010 10:10:16 AM Subject: Re: [Nek5000-users] Connectivety Thank you for the responses. However maybe my question I was clear before. I try again: Ok, the connectivity information comes from the rea file when you run genmap, but is it computed based on the spatial coordinates given in the rea file, or is it just taken from the connectivity information given later, also in the rea file along with the boundary conditions? What we experience is that we have a grid which is created in Matlab and then we write it in "rea file format" which is later read by genmap and reatore2 etc. This rea file, after the spatial coordinates of the element corners, contains a list with all the elements and their connectivity information. Here is what two lines looks like: W 4 1 0.000000 0.000000 0.000000 0.000000 0.000000 E 4 2 5.000000 4.000000 0.000000 0.000000 0.000000 Here I specify Wall condition for element 4 and face 1 and Element condition for face 2 that connects to element 5 with face 4. But then when I removed my connectivity info, that is having every where zeros in the 4th and 5th column: W 4 1 0.000000 0.000000 0.000000 0.000000 0.000000 E 4 2 0.000000 0.000000 0.000000 0.000000 0.000000 the case seemed to run fine, and the flow looked correct. Does this mean that the connectivity specified there does not matter? Best, /Antonios nek5000-users at lists.mcs.anl.gov wrote: > This is correct. > -Stefan > > ----- Original Message ----- > From: nek5000-users at lists.mcs.anl.gov > To: nek5000-users at lists.mcs.anl.gov > Sent: Thursday, December 9, 2010 6:15:02 PM > Subject: Re: [Nek5000-users] Connectivety > > Hi Antonios, > > I could be wrong but I think that the only connectivity information that > Nek5000 need comes from global vertexi element numbering genereated by > .map file which is in turn produced by running genmap on .rea file > > Best, > Aleks > > > On Thu, 9 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > >> Hello Paul, >> I have a question about the connectivity of the elements. >> In the old rea file format (which we convert to the new binary with >> reatore2) we have listed the Element geometrical date and after that the >> Boundary conditions and connectivity data. >> My question is whether the connectivity of the elements stated there is >> actually used or it is somehow re-computed while the code runs based on >> the node coordinates. I am talking specifically about the connectivity >> between the elements, not the boundary conditions with the "outside world". >> Best regards, >> Antonios >> >> -- >> --------------------------------------------------------------------------------------------- >> Antonios Monokrousos >> Department of Mechanics email: antonios at mech.kth.se >> Royal Institute of Technology (KTH) phone: +46 8 790 68 76 >> Osquars Backe 18 telefax: +46 8 796 98 50 >> SE-100 44 Stockholm, SWEDEN >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users -- --------------------------------------------------------------------------------------------- Antonios Monokrousos Department of Mechanics email: antonios at mech.kth.se Royal Institute of Technology (KTH) phone: +46 8 790 68 76 Osquars Backe 18 telefax: +46 8 796 98 50 SE-100 44 Stockholm, SWEDEN _______________________________________________ Nek5000-users mailing list Nek5000-users at lists.mcs.anl.gov https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users _______________________________________________ Nek5000-users mailing list Nek5000-users at lists.mcs.anl.gov https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Sat Dec 11 14:35:11 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sat, 11 Dec 2010 14:35:11 -0600 (CST) Subject: [Nek5000-users] Connectivety In-Reply-To: <903849407.2205351292093495782.JavaMail.root@neo-mail-3> Message-ID: <1199063761.13441.1292099711073.JavaMail.root@zimbra.anl.gov> Hi Michael The e-e bc info has not been needed for many years... p-p is still needed for periodic bcs, however. Paul ----- Original Message ----- From: nek5000-users at lists.mcs.anl.gov To: nek5000-users at lists.mcs.anl.gov Sent: Sat, 11 Dec 2010 12:51:35 -0600 (CST) Subject: Re: [Nek5000-users] Connectivety Hi Stefan, Do you recall the repo at which point this started to be the case? We have a similar code to convert gambit files and a majority of the time is spent getting this connectivity information. If we don't need it then that is welcomed news! - Michael M. ----- Original Message ----- From: nek5000-users at lists.mcs.anl.gov To: nek5000-users at lists.mcs.anl.gov Sent: Friday, December 10, 2010 4:11:26 AM GMT -06:00 US/Canada Central Subject: Re: [Nek5000-users] Connectivety Antonois: I understand your confusion. Let me backup a little bit. Genmap will try to establish a unique global numbering for all element vertices. Basically, genmap tries to detect (based on the vertex coordinates) the shared vertices and will assign a global id. We compute what you call the connectivity information on the fly (during the setup phase of Nek) based on the global vertex ids read from the .map file. You're right the old ASCII .rea file contains the elemental connectivity but this information is no longer used (we do a formatted dummy read for these lines). The .re2 file does not contain this information anymore. hth, Stefan ----- Original Message ----- From: nek5000-users at lists.mcs.anl.gov To: nek5000-users at lists.mcs.anl.gov Sent: Friday, December 10, 2010 10:10:16 AM Subject: Re: [Nek5000-users] Connectivety Thank you for the responses. However maybe my question I was clear before. I try again: Ok, the connectivity information comes from the rea file when you run genmap, but is it computed based on the spatial coordinates given in the rea file, or is it just taken from the connectivity information given later, also in the rea file along with the boundary conditions? What we experience is that we have a grid which is created in Matlab and then we write it in "rea file format" which is later read by genmap and reatore2 etc. This rea file, after the spatial coordinates of the element corners, contains a list with all the elements and their connectivity information. Here is what two lines looks like: W 4 1 0.000000 0.000000 0.000000 0.000000 0.000000 E 4 2 5.000000 4.000000 0.000000 0.000000 0.000000 Here I specify Wall condition for element 4 and face 1 and Element condition for face 2 that connects to element 5 with face 4. But then when I removed my connectivity info, that is having every where zeros in the 4th and 5th column: W 4 1 0.000000 0.000000 0.000000 0.000000 0.000000 E 4 2 0.000000 0.000000 0.000000 0.000000 0.000000 the case seemed to run fine, and the flow looked correct. Does this mean that the connectivity specified there does not matter? Best, /Antonios nek5000-users at lists.mcs.anl.gov wrote: > This is correct. > -Stefan > > ----- Original Message ----- > From: nek5000-users at lists.mcs.anl.gov > To: nek5000-users at lists.mcs.anl.gov > Sent: Thursday, December 9, 2010 6:15:02 PM > Subject: Re: [Nek5000-users] Connectivety > > Hi Antonios, > > I could be wrong but I think that the only connectivity information that > Nek5000 need comes from global vertexi element numbering genereated by > .map file which is in turn produced by running genmap on .rea file > > Best, > Aleks > > > On Thu, 9 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > >> Hello Paul, >> I have a question about the connectivity of the elements. >> In the old rea file format (which we convert to the new binary with >> reatore2) we have listed the Element geometrical date and after that the >> Boundary conditions and connectivity data. >> My question is whether the connectivity of the elements stated there is >> actually used or it is somehow re-computed while the code runs based on >> the node coordinates. I am talking specifically about the connectivity >> between the elements, not the boundary conditions with the "outside world". >> Best regards, >> Antonios >> >> -- >> --------------------------------------------------------------------------------------------- >> Antonios Monokrousos >> Department of Mechanics email: antonios at mech.kth.se >> Royal Institute of Technology (KTH) phone: +46 8 790 68 76 >> Osquars Backe 18 telefax: +46 8 796 98 50 >> SE-100 44 Stockholm, SWEDEN >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users -- --------------------------------------------------------------------------------------------- Antonios Monokrousos Department of Mechanics email: antonios at mech.kth.se Royal Institute of Technology (KTH) phone: +46 8 790 68 76 Osquars Backe 18 telefax: +46 8 796 98 50 SE-100 44 Stockholm, SWEDEN _______________________________________________ Nek5000-users mailing list Nek5000-users at lists.mcs.anl.gov https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users _______________________________________________ Nek5000-users mailing list Nek5000-users at lists.mcs.anl.gov https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users From nek5000-users at lists.mcs.anl.gov Sun Dec 12 21:09:34 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sun, 12 Dec 2010 21:09:34 -0600 Subject: [Nek5000-users] Data Overwritten while using comp_vort3() and hpts() Message-ID: Hello, I am having an issue with hpts data being overwritten while using the comp_vort3() and the hpts() routine. However the data dumps (blah0.f0000) are uncorrupted. Eddy viscosity is computed and stored in temp array, lambda2 and vorticity components are stored in 4 passive scalar arrays. It seems like the last two PS values of previous history points are over-written on vx and vy. I have attached an image of the hpts.out file and also pasted below is a piece of the usr file : common /gaaa/ wo1(lx1,ly1,lz1,lelv) & , wo2(lx1,ly1,lz1,lelv) & , wo3(lx1,ly1,lz1,lelv) if(mod(istep,iostep).eq.0) then ! compute lambda2 vortex call lambda2(t(1,1,1,1,2)) ifpsco(1)=.true. ! compute vorticity if(ldimt.ge.5) call comp_vort3(t(1,1,1,1,3),wo1,wo2,vx,vy,vz) call hpts() ifpsco(2)=.true. ifpsco(3)=.true. ifpsco(4)=.true. endif t(1,1,1,1,1) has eddy viscosity and ldimt is set to 5. Also, I am wondering what ldimt1 and ldimt3 means in the SIZE file ? Thanks. Regards Shriram -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: hpts.png Type: image/png Size: 58034 bytes Desc: not available URL: From nek5000-users at lists.mcs.anl.gov Mon Dec 13 02:00:54 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 13 Dec 2010 02:00:54 -0600 (CST) Subject: [Nek5000-users] Data Overwritten while using comp_vort3() and hpts() In-Reply-To: References: Message-ID: How big is ldimt in SIZE ? For the given usage, it looks like it should be 5? Paul Sun, 12 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > Hello, > > I am having an issue with hpts data being overwritten while using the > comp_vort3() and the hpts() routine. However the data dumps (blah0.f0000) > are uncorrupted. Eddy viscosity is computed and stored in temp array, > lambda2 and vorticity components are stored in 4 passive scalar arrays. It > seems like the last two PS values of previous history points are > over-written on vx and vy. I have attached an image of the hpts.out file and > also pasted below is a piece of the usr file : > > common /gaaa/ wo1(lx1,ly1,lz1,lelv) > & , wo2(lx1,ly1,lz1,lelv) > & , wo3(lx1,ly1,lz1,lelv) > > if(mod(istep,iostep).eq.0) then > ! compute lambda2 vortex > call lambda2(t(1,1,1,1,2)) > ifpsco(1)=.true. > ! compute vorticity > if(ldimt.ge.5) call comp_vort3(t(1,1,1,1,3),wo1,wo2,vx,vy,vz) > call hpts() > ifpsco(2)=.true. > ifpsco(3)=.true. > ifpsco(4)=.true. > endif > > t(1,1,1,1,1) has eddy viscosity and ldimt is set to 5. Also, I am wondering > what ldimt1 and ldimt3 means in the SIZE file ? Thanks. > > Regards > Shriram > From nek5000-users at lists.mcs.anl.gov Mon Dec 13 02:11:51 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 13 Dec 2010 02:11:51 -0600 (CST) Subject: [Nek5000-users] Connectivety In-Reply-To: References: <4CFF5BE0.4030403@lav.mavt.ethz.ch> <4CFF8FBD.6010306@lav.mavt.ethz.ch> <4D0105E2.9010509@mech.kth.se> Message-ID: Hi Antonios, Aleks is correct. The "E-E" connectivity in the .rea BC info is ignored by nek. Connectivity is determined strictly from the mesh topology inferred from the data in the .map file. The same is not true, however for periodic bcs -- this information must come from the "P-P" pairing in the .rea bc info. Does this make sense? Cheers, Paul On Thu, 9 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > Hi Antonios, > > I could be wrong but I think that the only connectivity information that > Nek5000 need comes from global vertexi element numbering genereated by .map > file which is in turn produced by running genmap on .rea file > > Best, > Aleks > > > On Thu, 9 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > >> Hello Paul, >> I have a question about the connectivity of the elements. >> In the old rea file format (which we convert to the new binary with >> reatore2) we have listed the Element geometrical date and after that the >> Boundary conditions and connectivity data. >> My question is whether the connectivity of the elements stated there is >> actually used or it is somehow re-computed while the code runs based on >> the node coordinates. I am talking specifically about the connectivity >> between the elements, not the boundary conditions with the "outside world". >> Best regards, >> Antonios >> >> -- >> >> --------------------------------------------------------------------------------------------- >> Antonios Monokrousos >> Department of Mechanics email: antonios at mech.kth.se >> Royal Institute of Technology (KTH) phone: +46 8 790 68 76 >> Osquars Backe 18 telefax: +46 8 796 98 50 >> SE-100 44 Stockholm, SWEDEN >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > From nek5000-users at lists.mcs.anl.gov Mon Dec 13 03:25:26 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 13 Dec 2010 03:25:26 -0600 (CST) Subject: [Nek5000-users] rms fluctuations of passive scalar In-Reply-To: References: Message-ID: Hi Zhicheng, There are some problems w/ overshoot & undershoot of temperature for passive scalars in under-resolved situations. I'm not certain, however, that that's the difficulty you are encountering in your case. Your early-time results look fine, as you note. I'm trying to recreate the situation here. Paul On Thu, 9 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > Hi Neks, > > I'm doing LES of heat transfer in a fully turbulent plane channel flow with > uniform thermal boundary condition. Re_tau=150, Pr=0.7. > The domain is 6.4/2.0/3.2, 8*8*8 elements using N=8. > > When I compare my results with Kasagi's, it looks like the mean velocity > and temperature distribution are good. AT dimensionless time 165, the rms > value of temperature is good, however, as I restart from time 165 to 402, > the difference between my results and Kasagi's > became larger and larger. > > I'm using standard eddy viscosity model, the same as example turbChannel. > In case some one had this kind of problem, I attached two figures, one is > rms fluctuation at dimensionless time 165, the other is that at 402. > > -- > Zhicheng Wang > From nek5000-users at lists.mcs.anl.gov Mon Dec 13 03:11:47 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 13 Dec 2010 10:11:47 +0100 Subject: [Nek5000-users] Connectivety In-Reply-To: References: <4CFF5BE0.4030403@lav.mavt.ethz.ch> <4CFF8FBD.6010306@lav.mavt.ethz.ch> <4D0105E2.9010509@mech.kth.se> Message-ID: <4D05E353.1000502@mech.kth.se> Hi Paul, Yes, it makes perfect sense. Cheers, Antonios nek5000-users at lists.mcs.anl.gov wrote: > > Hi Antonios, > > Aleks is correct. The "E-E" connectivity in the .rea BC info > is ignored by nek. > > Connectivity is determined strictly from the mesh topology > inferred from the data in the .map file. > > The same is not true, however for periodic bcs -- this information > must come from the "P-P" pairing in the .rea bc info. > > Does this make sense? > > Cheers, > > Paul > > > On Thu, 9 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > >> Hi Antonios, >> >> I could be wrong but I think that the only connectivity information >> that Nek5000 need comes from global vertexi element numbering >> genereated by .map file which is in turn produced by running genmap on >> .rea file >> >> Best, >> Aleks >> >> >> On Thu, 9 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: >> >>> Hello Paul, >>> I have a question about the connectivity of the elements. >>> In the old rea file format (which we convert to the new binary with >>> reatore2) we have listed the Element geometrical date and after that the >>> Boundary conditions and connectivity data. >>> My question is whether the connectivity of the elements stated there is >>> actually used or it is somehow re-computed while the code runs based on >>> the node coordinates. I am talking specifically about the connectivity >>> between the elements, not the boundary conditions with the "outside >>> world". >>> Best regards, >>> Antonios >>> >>> -- >>> >>> --------------------------------------------------------------------------------------------- >>> From nek5000-users at lists.mcs.anl.gov Mon Dec 13 07:42:58 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 13 Dec 2010 07:42:58 -0600 Subject: [Nek5000-users] Data Overwritten while using comp_vort3() and hpts() In-Reply-To: References: Message-ID: Hi Paul, Yes. ldimt is set to 5. Regards Shriram Jagannathan On 13 December 2010 02:00, wrote: > > > How big is ldimt in SIZE ? > > For the given usage, it looks like it should be 5? > > Paul > > > > > Sun, 12 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > > Hello, >> >> I am having an issue with hpts data being overwritten while using the >> comp_vort3() and the hpts() routine. However the data dumps (blah0.f0000) >> are uncorrupted. Eddy viscosity is computed and stored in temp array, >> lambda2 and vorticity components are stored in 4 passive scalar arrays. It >> seems like the last two PS values of previous history points are >> over-written on vx and vy. I have attached an image of the hpts.out file >> and >> also pasted below is a piece of the usr file : >> >> common /gaaa/ wo1(lx1,ly1,lz1,lelv) >> & , wo2(lx1,ly1,lz1,lelv) >> & , wo3(lx1,ly1,lz1,lelv) >> >> if(mod(istep,iostep).eq.0) then >> ! compute lambda2 vortex >> call lambda2(t(1,1,1,1,2)) >> ifpsco(1)=.true. >> ! compute vorticity >> if(ldimt.ge.5) call comp_vort3(t(1,1,1,1,3),wo1,wo2,vx,vy,vz) >> call hpts() >> ifpsco(2)=.true. >> ifpsco(3)=.true. >> ifpsco(4)=.true. >> endif >> >> t(1,1,1,1,1) has eddy viscosity and ldimt is set to 5. Also, I am >> wondering >> what ldimt1 and ldimt3 means in the SIZE file ? Thanks. >> >> Regards >> Shriram >> >> _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Dec 13 08:10:16 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 13 Dec 2010 08:10:16 -0600 (CST) Subject: [Nek5000-users] =?utf-8?q?Data_Overwritten_while_using_comp=5Fvor?= =?utf-8?b?dDMoKSBhbmQJaHB0cygp?= In-Reply-To: References: Message-ID: Hi Shriram, I can't recall if I checked into this before - this might be one of those problems that doesn't show up on our platform, but I'm willing to give it a shot if you can tar up the requisite files and put on our ftp site. How many processors were you using? Paul On Mon, 13 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > Hi Paul, > > Yes. ldimt is set to 5. > > Regards > Shriram Jagannathan > > > On 13 December 2010 02:00, wrote: > >> >> >> How big is ldimt in SIZE ? >> >> For the given usage, it looks like it should be 5? >> >> Paul >> >> >> >> >> Sun, 12 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: >> >> Hello, >>> >>> I am having an issue with hpts data being overwritten while using the >>> comp_vort3() and the hpts() routine. However the data dumps (blah0.f0000) >>> are uncorrupted. Eddy viscosity is computed and stored in temp array, >>> lambda2 and vorticity components are stored in 4 passive scalar arrays. It >>> seems like the last two PS values of previous history points are >>> over-written on vx and vy. I have attached an image of the hpts.out file >>> and >>> also pasted below is a piece of the usr file : >>> >>> common /gaaa/ wo1(lx1,ly1,lz1,lelv) >>> & , wo2(lx1,ly1,lz1,lelv) >>> & , wo3(lx1,ly1,lz1,lelv) >>> >>> if(mod(istep,iostep).eq.0) then >>> ! compute lambda2 vortex >>> call lambda2(t(1,1,1,1,2)) >>> ifpsco(1)=.true. >>> ! compute vorticity >>> if(ldimt.ge.5) call comp_vort3(t(1,1,1,1,3),wo1,wo2,vx,vy,vz) >>> call hpts() >>> ifpsco(2)=.true. >>> ifpsco(3)=.true. >>> ifpsco(4)=.true. >>> endif >>> >>> t(1,1,1,1,1) has eddy viscosity and ldimt is set to 5. Also, I am >>> wondering >>> what ldimt1 and ldimt3 means in the SIZE file ? Thanks. >>> >>> Regards >>> Shriram >>> >>> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> > From nek5000-users at lists.mcs.anl.gov Mon Dec 13 08:15:20 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 13 Dec 2010 08:15:20 -0600 Subject: [Nek5000-users] Data Overwritten while using comp_vort3() and hpts() In-Reply-To: References: Message-ID: Hi Paul, We had discussed this earlier and I had to fix the restart file to get it to work. Now it doesn't seem to come from the restart file though. I am running it on 128 procs (~24k elements) on our cluster. I can try running it on a different machine here to see if I get the same problem. Do you have any instructions on how to upload to the ftp site ? Regards Shriram Jagannathan On 13 December 2010 08:10, wrote: > > Hi Shriram, > > I can't recall if I checked into this before - this might > be one of those problems that doesn't show up on our platform, > but I'm willing to give it a shot if you can tar up the > requisite files and put on our ftp site. How many processors > were you using? > > Paul > > > > On Mon, 13 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > > Hi Paul, >> >> Yes. ldimt is set to 5. >> >> Regards >> Shriram Jagannathan >> >> >> On 13 December 2010 02:00, wrote: >> >> >>> >>> How big is ldimt in SIZE ? >>> >>> For the given usage, it looks like it should be 5? >>> >>> Paul >>> >>> >>> >>> >>> Sun, 12 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: >>> >>> Hello, >>> >>>> >>>> I am having an issue with hpts data being overwritten while using the >>>> comp_vort3() and the hpts() routine. However the data dumps >>>> (blah0.f0000) >>>> are uncorrupted. Eddy viscosity is computed and stored in temp array, >>>> lambda2 and vorticity components are stored in 4 passive scalar arrays. >>>> It >>>> seems like the last two PS values of previous history points are >>>> over-written on vx and vy. I have attached an image of the hpts.out file >>>> and >>>> also pasted below is a piece of the usr file : >>>> >>>> common /gaaa/ wo1(lx1,ly1,lz1,lelv) >>>> & , wo2(lx1,ly1,lz1,lelv) >>>> & , wo3(lx1,ly1,lz1,lelv) >>>> >>>> if(mod(istep,iostep).eq.0) then >>>> ! compute lambda2 vortex >>>> call lambda2(t(1,1,1,1,2)) >>>> ifpsco(1)=.true. >>>> ! compute vorticity >>>> if(ldimt.ge.5) call comp_vort3(t(1,1,1,1,3),wo1,wo2,vx,vy,vz) >>>> call hpts() >>>> ifpsco(2)=.true. >>>> ifpsco(3)=.true. >>>> ifpsco(4)=.true. >>>> endif >>>> >>>> t(1,1,1,1,1) has eddy viscosity and ldimt is set to 5. Also, I am >>>> wondering >>>> what ldimt1 and ldimt3 means in the SIZE file ? Thanks. >>>> >>>> Regards >>>> Shriram >>>> >>>> _______________________________________________ >>>> >>> Nek5000-users mailing list >>> Nek5000-users at lists.mcs.anl.gov >>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >>> >>> >> _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Dec 13 09:53:17 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 13 Dec 2010 16:53:17 +0100 Subject: [Nek5000-users] time derivative Message-ID: <4D06416D.80305@lav.mavt.ethz.ch> hi all, i need to compute time derivatives of the velocity. Is VXLAG(:,:,:,:,1) the vx field of the previous time step? thank you francesco -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Dec 13 09:59:01 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 13 Dec 2010 09:59:01 -0600 (CST) Subject: [Nek5000-users] time derivative In-Reply-To: <4D06416D.80305@lav.mavt.ethz.ch> References: <4D06416D.80305@lav.mavt.ethz.ch> Message-ID: Hi Francesco, Yes ... that should work. Always a good idea to verify by storing your own on a test... The number of values stored varies according to Torder, as I recall. Paul On Mon, 13 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > hi all, > > i need to compute time derivatives of the velocity. > Is VXLAG(:,:,:,:,1) the vx field of the previous time step? > > thank you > francesco > > > From nek5000-users at lists.mcs.anl.gov Tue Dec 14 08:53:43 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Dec 2010 07:53:43 -0700 Subject: [Nek5000-users] 2D axisymmetric problems Message-ID: <8707DE09-5353-4B4C-AA3C-E63A72AC248F@nrel.gov> Hello All. I have a couple of questions regarding simulation of 2D axisymmetric problems where there's a nonzero azimuthal velocity component. First, does the new .rea + .re2 format work with such problems? I tried this with the "axi" problem in the examples directory, and things halted with ERROR READING INITIAL CONDITION/DRIVE FORCE DATA ABORTING IN ROUTINE RDICDF. It worked fine if I stuck with the older .rea format. Second, when doing such problems a while back, we would trick nek for such problems by using temperature to represent the azimuthal velocity. This approach required one to add appropriate terms into the .usr file. Is this still the case, or is Nek's third velocity component now the azimuthal velocity? Thanks. --Mike From nek5000-users at lists.mcs.anl.gov Tue Dec 14 09:13:02 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Dec 2010 16:13:02 +0100 Subject: [Nek5000-users] Impose inflow conditions from another simulation Message-ID: Hi Nek's, I have a question on how to impose inflow boundary conditions that are not analytical. I have for instance a Blasius velocity profile computed thanks to another code. I would like to impose this given profile as my inflow condition. At the moment I am using a polynomial expansion of the velocity profile in the userbc subroutine, however I would like to do things properly. I can interpolate the profile from my previous code onto the wall normal GLL discretization. That would give me a .dat file containing three arrays: first one is the y-discretization, second and third ones are the velocity (u,v) at these points. Question is how could I use these data to impose my blasius profile as inflow boundary condition? I assume it is rather straightforward, however I am not sure how and which indices should vary. Best regards, JC -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Dec 14 09:31:58 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Dec 2010 09:31:58 -0600 (CST) Subject: [Nek5000-users] Impose inflow conditions from another simulation In-Reply-To: References: Message-ID: JC, Are you doing this in serial or parallel? The main thing is to dump out the points at inflow, along with some type of discriminator so that you can subsequently associate function values (from your other code) with those points when you want to read them back in. In serial, this is easy - a bit trickier in parallel. Let me know which case is of interest and we can generate something for you. Paul On Tue, 14 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > Hi Nek's, > > I have a question on how to impose inflow boundary conditions that are not > analytical. > I have for instance a Blasius velocity profile computed thanks to another > code. I would like to impose this given profile as my inflow condition. > At the moment I am using a polynomial expansion of the velocity profile in > the userbc subroutine, however I would like to do things properly. > > I can interpolate the profile from my previous code onto the wall normal GLL > discretization. That would give me a .dat file containing three arrays: > first one is the y-discretization, second and third ones are the velocity > (u,v) at these points. Question is how could I use these data to impose my > blasius profile as inflow boundary condition? > > I assume it is rather straightforward, however I am not sure how and which > indices should vary. > > Best regards, > JC > From nek5000-users at lists.mcs.anl.gov Tue Dec 14 10:01:12 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Dec 2010 17:01:12 +0100 Subject: [Nek5000-users] Impose inflow conditions from another simulation In-Reply-To: References: Message-ID: Well my first code is serial calculation. It actually is nothing more but a matlab script for the Falkner-Skan-Cook boundary layer. The calculation in Nek would be parallel as I'd like to investigate the influence of a roughness element in a boudary layer flow. I can eventually modify my interpolation routine output if you want one particular input format. JC 2010/12/14 > > JC, > > Are you doing this in serial or parallel? > > The main thing is to dump out the points at inflow, along with > some type of discriminator so that you can subsequently > associate function values (from your other code) with those > points when you want to read them back in. > > In serial, this is easy - a bit trickier in parallel. > > Let me know which case is of interest and we can generate > something for you. > > Paul > > > > On Tue, 14 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > > Hi Nek's, >> >> I have a question on how to impose inflow boundary conditions that are not >> analytical. >> I have for instance a Blasius velocity profile computed thanks to another >> code. I would like to impose this given profile as my inflow condition. >> At the moment I am using a polynomial expansion of the velocity profile in >> the userbc subroutine, however I would like to do things properly. >> >> I can interpolate the profile from my previous code onto the wall normal >> GLL >> discretization. That would give me a .dat file containing three arrays: >> first one is the y-discretization, second and third ones are the velocity >> (u,v) at these points. Question is how could I use these data to impose my >> blasius profile as inflow boundary condition? >> >> I assume it is rather straightforward, however I am not sure how and which >> indices should vary. >> >> Best regards, >> JC >> >> _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Dec 14 11:07:41 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Dec 2010 18:07:41 +0100 Subject: [Nek5000-users] time derivative In-Reply-To: References: <4D06416D.80305@lav.mavt.ethz.ch> Message-ID: <4D07A45D.7090802@lav.mavt.ethz.ch> thank you. i do have a follow up question. My case is a DNS of a turbulent channel. I was hoping to use the the characteristics method (IFCHAR=T.) and a target CFL number of 1.5 - 2.0 . To compute time derivatives I am using a normal finite difference [u(t)-u(t-dt)]/dt but the results are suspicious. Given the large time step, do you think it is possible to Have reliable accelerations with this method? Are there any better ways to compute them? thank you a lot. francesco On 12/13/2010 04:59 PM, nek5000-users at lists.mcs.anl.gov wrote: > > Hi Francesco, > > Yes ... that should work. Always a good idea to verify by > storing your own on a test... The number of values stored > varies according to Torder, as I recall. > > Paul > > On Mon, 13 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > >> hi all, >> >> i need to compute time derivatives of the velocity. >> Is VXLAG(:,:,:,:,1) the vx field of the previous time step? >> >> thank you >> francesco >> >> >> > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users From nek5000-users at lists.mcs.anl.gov Tue Dec 14 11:50:15 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Dec 2010 18:50:15 +0100 Subject: [Nek5000-users] opdiv and multd In-Reply-To: References: <4CFF5BE0.4030403@lav.mavt.ethz.ch> <4CFF8FBD.6010306@lav.mavt.ethz.ch> Message-ID: <4D07AE57.5040000@lav.mavt.ethz.ch> Sorry i have to do a correction. A user point it out that the subroutine is defined as dssum(u,nx,ny,nz) in dssum.f . Thus it should be called with 4 arguments : call dssum(RES(1),nx1,ny1,nz1) This did not give me problems and it seems to me that nx,ny,nz are not used. But it is better to be aware of it since i do not know how the compiler will read the wrong call. Sorry about that. francesco On 12/08/2010 04:02 PM, nek5000-users at lists.mcs.anl.gov wrote: > > Thanks Francesco! > > Paul > > > On Wed, 8 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > >> Thank you Paul, >> that solved the problem. >> >> The correct result is obtained by : >> call multd (RES(1),vx,rym2,sym2,tym2,2,iflg) >> call dssum(RES(1)) >> call col2(RES(1),binvm1,lt) >> >> francesco >> >> >> >> On 12/08/2010 02:17 PM, nek5000-users at lists.mcs.anl.gov wrote: >>> >>> Hi Francesco, >>> >>> Several of the routines internal to nek are the so-called weak >>> derivatives, implying that they somehow involve the mass matrix >>> or the transpose of the operator, so it takes a bit of investigation >>> to understand precisely what each operator is doing. (Note that the >>> functionality also differs slightly between PN-PN and PN-PN-2 >>> formulations.) >>> >>> As an example, axhelm does not multiply by the (negative) Laplacian. >>> Instead, it creates (for h1==1, h2==0): >>> >>> w = A * u >>> >>> = grad^T B grad u >>> >>> where B is the mass matrix. >>> >>> Similarly, opdiv is most likely returning d = B*grad u (or perhaps >>> the negative of this). Moreover, d is returned on the pressure >>> mesh (mesh 2), which is the same as the velocity mesh (mesh 1) for >>> the PN-PN formulation that you use. >>> >>> In Nek, the mass matrix is diagonal and the mesh 1 inverse is stored >>> in binvm1. It seems probable that you'll get the correct result >>> after >>> >>> call col2(d,binvm1,n) >>> >>> where d holds the output of opdiv and n is the number of points on >>> mesh 1, assuming you're using PN-Pn. >>> >>> Paul >>> >>> >>> >>> On Wed, 8 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: >>> >>>> Hi all, >>>> >>>> I am having some problems using the subroutines opdiv and multd. >>>> For some reasons they do not give me correct results. >>>> Do they need any special treatment or precautions? >>>> >>>> I was trying to compute the divergence of a 3D field, and i >>>> realized that in my case the multd subroutine used >>>> by opdiv does not gives good results, while the same derivative >>>> computed from gram1 gives no problems. >>>> >>>> For example, I overwrite an analytical field : >>>> >>>> ... >>>> parameter(lt=lx1*ly1*lz1*lelt) >>>> real Res(lt) >>>> ... >>>> vx(i,j,k,e)= 1.0*sin(x)*cos(y)*cos(z) >>>> ... >>>> call print_line(vx) >>>> ... >>>> iflg = 1 >>>> call multd (RES(1),vx,rym2,sym2,tym2,2,iflg) >>>> call print_line(Res(1) ) >>>> >>>> call gradm1(work1,Res(1),work2,vx) >>>> call print_line(Res(1) ) >>>> ... >>>> if (lx2.ne.lx1) then >>>> ... >>>> STOP >>>> endif >>>> >>>> >>>> and print output a (xr,:,zr) the result (performed by print_line) . >>>> In vxi.eps i compare the result (lines) with the analytical >>>> solution( symbols). The result of multd is not correct while all >>>> the others match perfectly the analytical solution. >>>> In vxy.eps i plot only the result of multd. Note that on y i have >>>> 12 elements the same number of peaks of the solution. >>>> >>>> Does anybody have an idea on how to fix this? >>>> >>>> Thank you very much for any help or suggestion >>>> >>>> francesco >>>> >>>> >>>> >>>> >>> _______________________________________________ >>> Nek5000-users mailing list >>> Nek5000-users at lists.mcs.anl.gov >>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > From nek5000-users at lists.mcs.anl.gov Tue Dec 14 12:07:29 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Dec 2010 12:07:29 -0600 (CST) Subject: [Nek5000-users] opdiv and multd In-Reply-To: <4D07AE57.5040000@lav.mavt.ethz.ch> Message-ID: <869885403.25443.1292350049972.JavaMail.root@zimbra.anl.gov> dssum() is just a wrapper for call gs_op() which takes only the input field from the dssum call. But I agree in general it's a good idea to call all subroutines with the correct arguments. -Stefan ----- Original Message ----- From: nek5000-users at lists.mcs.anl.gov To: nek5000-users at lists.mcs.anl.gov Sent: Tuesday, December 14, 2010 6:50:15 PM Subject: Re: [Nek5000-users] opdiv and multd Sorry i have to do a correction. A user point it out that the subroutine is defined as dssum(u,nx,ny,nz) in dssum.f . Thus it should be called with 4 arguments : call dssum(RES(1),nx1,ny1,nz1) This did not give me problems and it seems to me that nx,ny,nz are not used. But it is better to be aware of it since i do not know how the compiler will read the wrong call. Sorry about that. francesco On 12/08/2010 04:02 PM, nek5000-users at lists.mcs.anl.gov wrote: > > Thanks Francesco! > > Paul > > > On Wed, 8 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > >> Thank you Paul, >> that solved the problem. >> >> The correct result is obtained by : >> call multd (RES(1),vx,rym2,sym2,tym2,2,iflg) >> call dssum(RES(1)) >> call col2(RES(1),binvm1,lt) >> >> francesco >> >> >> >> On 12/08/2010 02:17 PM, nek5000-users at lists.mcs.anl.gov wrote: >>> >>> Hi Francesco, >>> >>> Several of the routines internal to nek are the so-called weak >>> derivatives, implying that they somehow involve the mass matrix >>> or the transpose of the operator, so it takes a bit of investigation >>> to understand precisely what each operator is doing. (Note that the >>> functionality also differs slightly between PN-PN and PN-PN-2 >>> formulations.) >>> >>> As an example, axhelm does not multiply by the (negative) Laplacian. >>> Instead, it creates (for h1==1, h2==0): >>> >>> w = A * u >>> >>> = grad^T B grad u >>> >>> where B is the mass matrix. >>> >>> Similarly, opdiv is most likely returning d = B*grad u (or perhaps >>> the negative of this). Moreover, d is returned on the pressure >>> mesh (mesh 2), which is the same as the velocity mesh (mesh 1) for >>> the PN-PN formulation that you use. >>> >>> In Nek, the mass matrix is diagonal and the mesh 1 inverse is stored >>> in binvm1. It seems probable that you'll get the correct result >>> after >>> >>> call col2(d,binvm1,n) >>> >>> where d holds the output of opdiv and n is the number of points on >>> mesh 1, assuming you're using PN-Pn. >>> >>> Paul >>> >>> >>> >>> On Wed, 8 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: >>> >>>> Hi all, >>>> >>>> I am having some problems using the subroutines opdiv and multd. >>>> For some reasons they do not give me correct results. >>>> Do they need any special treatment or precautions? >>>> >>>> I was trying to compute the divergence of a 3D field, and i >>>> realized that in my case the multd subroutine used >>>> by opdiv does not gives good results, while the same derivative >>>> computed from gram1 gives no problems. >>>> >>>> For example, I overwrite an analytical field : >>>> >>>> ... >>>> parameter(lt=lx1*ly1*lz1*lelt) >>>> real Res(lt) >>>> ... >>>> vx(i,j,k,e)= 1.0*sin(x)*cos(y)*cos(z) >>>> ... >>>> call print_line(vx) >>>> ... >>>> iflg = 1 >>>> call multd (RES(1),vx,rym2,sym2,tym2,2,iflg) >>>> call print_line(Res(1) ) >>>> >>>> call gradm1(work1,Res(1),work2,vx) >>>> call print_line(Res(1) ) >>>> ... >>>> if (lx2.ne.lx1) then >>>> ... >>>> STOP >>>> endif >>>> >>>> >>>> and print output a (xr,:,zr) the result (performed by print_line) . >>>> In vxi.eps i compare the result (lines) with the analytical >>>> solution( symbols). The result of multd is not correct while all >>>> the others match perfectly the analytical solution. >>>> In vxy.eps i plot only the result of multd. Note that on y i have >>>> 12 elements the same number of peaks of the solution. >>>> >>>> Does anybody have an idea on how to fix this? >>>> >>>> Thank you very much for any help or suggestion >>>> >>>> francesco >>>> >>>> >>>> >>>> >>> _______________________________________________ >>> Nek5000-users mailing list >>> Nek5000-users at lists.mcs.anl.gov >>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > _______________________________________________ Nek5000-users mailing list Nek5000-users at lists.mcs.anl.gov https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users From nek5000-users at lists.mcs.anl.gov Tue Dec 14 12:38:24 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Dec 2010 12:38:24 -0600 Subject: [Nek5000-users] 2D axisymmetric problems In-Reply-To: <8707DE09-5353-4B4C-AA3C-E63A72AC248F@nrel.gov> References: <8707DE09-5353-4B4C-AA3C-E63A72AC248F@nrel.gov> Message-ID: Hi Mike, In the 'axi' example, genbox produces an unnecessary line in the .rea file when the .re2/.rea files are created. If you delete the following line from axi.rea, it should work: ***** NO THERMAL BOUNDARY CONDITIONS ***** We are working on the azimuthal velocity issue. Thanks, Katie On Tue, Dec 14, 2010 at 8:53 AM, wrote: > Hello All. > > I have a couple of questions regarding simulation of 2D axisymmetric > problems where there's a nonzero azimuthal velocity component. > > First, does the new .rea + .re2 format work with such problems? I tried > this with the "axi" problem in the examples directory, and things halted > with > > ERROR READING INITIAL CONDITION/DRIVE FORCE DATA > ABORTING IN ROUTINE RDICDF. > > It worked fine if I stuck with the older .rea format. > > Second, when doing such problems a while back, we would trick nek for such > problems by using temperature to represent the azimuthal velocity. This > approach required one to add appropriate terms into the .usr file. Is this > still the case, or is Nek's third velocity component now the azimuthal > velocity? > > Thanks. > --Mike > > > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Dec 14 15:34:10 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Dec 2010 15:34:10 -0600 (CST) Subject: [Nek5000-users] time derivative In-Reply-To: <4D07A45D.7090802@lav.mavt.ethz.ch> References: <4D06416D.80305@lav.mavt.ethz.ch> <4D07A45D.7090802@lav.mavt.ethz.ch> Message-ID: Hi Francesco, I would use a higher-order derivative, e.g., (3u^n - 4u^n-1 + u^n-2)/ 2dt would be 2nd order accurate. You could also use centered differences, (u^n+1 - u^n-1) / 2dt to compute du/dt at t^n. The best way to verify your results is to take lim dt--> 0 and see that you realize the expected convergence. There should be no difficulty with the characteristics scheme, as far as I can tell. Paul On Tue, 14 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > thank you. > > i do have a follow up question. > My case is a DNS of a turbulent channel. > I was hoping to use the the characteristics method (IFCHAR=T.) > and a target CFL number of 1.5 - 2.0 . > > To compute time derivatives I am using a normal finite difference > [u(t)-u(t-dt)]/dt > but the results are suspicious. > > Given the large time step, do you think it is possible to > Have reliable accelerations with this method? > Are there any better ways to compute them? > > thank you a lot. > francesco > > > > > On 12/13/2010 04:59 PM, nek5000-users at lists.mcs.anl.gov wrote: >> >> Hi Francesco, >> >> Yes ... that should work. Always a good idea to verify by >> storing your own on a test... The number of values stored >> varies according to Torder, as I recall. >> >> Paul >> >> On Mon, 13 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: >> >>> hi all, >>> >>> i need to compute time derivatives of the velocity. >>> Is VXLAG(:,:,:,:,1) the vx field of the previous time step? >>> >>> thank you >>> francesco >>> >>> >>> >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > From nek5000-users at lists.mcs.anl.gov Wed Dec 15 20:50:05 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 15 Dec 2010 20:50:05 -0600 (CST) Subject: [Nek5000-users] Users Meeting follow-up Message-ID: Dear User Group, A quick note to thank each and every one of you who attended our first users group meeting last week. We were really impressed by the presentations!! For those of you who would like to have your talks posted, you can either send .pdfs directly to me (Paul), or send me a URL and I will post the link on the wiki. This latter option gives you full control over the content from now into the future. I will also follow up with an email regarding some of the conclusions and wish-list items raised during the meeting. Thanks once again. Nek5000 Development Team From nek5000-users at lists.mcs.anl.gov Fri Dec 17 08:22:44 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 17 Dec 2010 15:22:44 +0100 Subject: [Nek5000-users] parameter lmg_rwt too small Message-ID: Hi Nek's, I'm having troubles trying to quickly simulate a boundary layer flow. Here is the message I get as soon as I launch the calculation: gs_setup: 0 unique labels shared setupds time 3.0041E-03 seconds 4 8 4181 300 > *parameter lmg_rwt too small 2400 1 -16* > > call exitt: dying ... > Note however that when I change NSTEP to zero (post-process mode) I do not encounter any problem. Best regards, JC -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Fri Dec 17 09:22:41 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 17 Dec 2010 09:22:41 -0600 (CST) Subject: [Nek5000-users] parameter lmg_rwt too small In-Reply-To: References: Message-ID: JC, Can you send me a tarfile of your SIZE, .usr, and .rea file? Just send directly to my email address... fischer at mcs.anl.gov (make it a gzipd tarfile if the rea file is large...) Thanks, Paul On Fri, 17 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > Hi Nek's, > > I'm having troubles trying to quickly simulate a boundary layer flow. Here > is the message I get as soon as I launch the calculation: > > gs_setup: 0 unique labels shared > > setupds time 3.0041E-03 seconds 4 8 4181 300 >> *parameter lmg_rwt too small 2400 1 -16* >> >> call exitt: dying ... >> > > Note however that when I change NSTEP to zero (post-process mode) I do not > encounter any problem. > > > Best regards, > JC > From nek5000-users at lists.mcs.anl.gov Fri Dec 17 09:23:07 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 17 Dec 2010 09:23:07 -0600 (CST) Subject: [Nek5000-users] 2D axisymmetric problems In-Reply-To: <8707DE09-5353-4B4C-AA3C-E63A72AC248F@nrel.gov> References: <8707DE09-5353-4B4C-AA3C-E63A72AC248F@nrel.gov> Message-ID: Hi Mike, Aleks is looking into this and will get you an answer today. Paul On Tue, 14 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > Hello All. > > I have a couple of questions regarding simulation of 2D axisymmetric problems where there's a nonzero azimuthal velocity component. > > First, does the new .rea + .re2 format work with such problems? I tried this with the "axi" problem in the examples directory, and things halted with > > ERROR READING INITIAL CONDITION/DRIVE FORCE DATA > ABORTING IN ROUTINE RDICDF. > > It worked fine if I stuck with the older .rea format. > > Second, when doing such problems a while back, we would trick nek for such problems by using temperature to represent the azimuthal velocity. This approach required one to add appropriate terms into the .usr file. Is this still the case, or is Nek's third velocity component now the azimuthal velocity? > > Thanks. > --Mike > > > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > From nek5000-users at lists.mcs.anl.gov Fri Dec 17 09:24:47 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 17 Dec 2010 08:24:47 -0700 Subject: [Nek5000-users] 2D axisymmetric problems In-Reply-To: References: <8707DE09-5353-4B4C-AA3C-E63A72AC248F@nrel.gov> Message-ID: <9749D89A-8BFD-484A-BE6A-45D64B244280@nrel.gov> Paul, Great! Katie mentioned a solution to the "ERROR" message -- I think the only remaining thing is the representation of the azimuthal velocity. --Mike On Dec 17, 2010, at 8:23 AM, wrote: > > Hi Mike, > > Aleks is looking into this and will get you an answer today. > > Paul > > > On Tue, 14 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > >> Hello All. >> >> I have a couple of questions regarding simulation of 2D axisymmetric problems where there's a nonzero azimuthal velocity component. >> >> First, does the new .rea + .re2 format work with such problems? I tried this with the "axi" problem in the examples directory, and things halted with >> >> ERROR READING INITIAL CONDITION/DRIVE FORCE DATA >> ABORTING IN ROUTINE RDICDF. >> >> It worked fine if I stuck with the older .rea format. >> >> Second, when doing such problems a while back, we would trick nek for such problems by using temperature to represent the azimuthal velocity. This approach required one to add appropriate terms into the .usr file. Is this still the case, or is Nek's third velocity component now the azimuthal velocity? >> >> Thanks. >> --Mike >> >> >> >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users From nek5000-users at lists.mcs.anl.gov Fri Dec 17 09:30:11 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 17 Dec 2010 09:30:11 -0600 (CST) Subject: [Nek5000-users] 2D axisymmetric problems In-Reply-To: <9749D89A-8BFD-484A-BE6A-45D64B244280@nrel.gov> References: <8707DE09-5353-4B4C-AA3C-E63A72AC248F@nrel.gov> <9749D89A-8BFD-484A-BE6A-45D64B244280@nrel.gov> Message-ID: Generally we store azimuthal velocity as temperature (which I think you've done in the past...), there is a flag that needs to be set, however, for the Helmholtz operator. The issue is how to deal with this in a more general way, so that the user can control it ... and then, what happens if the user also wants heat transfer??? Our current plan is that if you also want heat transfer then temperature will be stored in PS1. As I mentioned, Aleks will get you the details... I think we should be able to have it as a flag from the .usr file so you don't need to touch the source. The caveat will be that the user needs to recognize that a separate set of bcs is required for "temperature", which will be bcs on the azimuthal velocity. Paul On Fri, 17 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > Paul, > > Great! Katie mentioned a solution to the "ERROR" message -- I think the only remaining thing is the representation of the azimuthal velocity. > > --Mike > > On Dec 17, 2010, at 8:23 AM, wrote: > >> >> Hi Mike, >> >> Aleks is looking into this and will get you an answer today. >> >> Paul >> >> >> On Tue, 14 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: >> >>> Hello All. >>> >>> I have a couple of questions regarding simulation of 2D axisymmetric problems where there's a nonzero azimuthal velocity component. >>> >>> First, does the new .rea + .re2 format work with such problems? I tried this with the "axi" problem in the examples directory, and things halted with >>> >>> ERROR READING INITIAL CONDITION/DRIVE FORCE DATA >>> ABORTING IN ROUTINE RDICDF. >>> >>> It worked fine if I stuck with the older .rea format. >>> >>> Second, when doing such problems a while back, we would trick nek for such problems by using temperature to represent the azimuthal velocity. This approach required one to add appropriate terms into the .usr file. Is this still the case, or is Nek's third velocity component now the azimuthal velocity? >>> >>> Thanks. >>> --Mike >>> >>> >>> >>> _______________________________________________ >>> Nek5000-users mailing list >>> Nek5000-users at lists.mcs.anl.gov >>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >>> >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > From nek5000-users at lists.mcs.anl.gov Fri Dec 17 09:49:21 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 17 Dec 2010 08:49:21 -0700 Subject: [Nek5000-users] 2D axisymmetric problems In-Reply-To: References: <8707DE09-5353-4B4C-AA3C-E63A72AC248F@nrel.gov> <9749D89A-8BFD-484A-BE6A-45D64B244280@nrel.gov> Message-ID: <8AEDCDA3-9830-4131-8931-7F0C52B5DEA3@nrel.gov> On Dec 17, 2010, at 8:30 AM, wrote: > > Generally we store azimuthal velocity as temperature (which I think > you've done in the past...), Yep. > Our current plan is that if you also want heat transfer then temperature > will be stored in PS1. I think that's fine, but I have two questions that I'm sure Alexs will answer. First, will the additional forcing terms (for cylindrical coordinates) still need to be added in the .user file? Second, how will axisymmetric problems be handled when the axis of symmetry is included in the underlying problem domain? --Mike From nek5000-users at lists.mcs.anl.gov Fri Dec 17 10:51:38 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 17 Dec 2010 17:51:38 +0100 Subject: [Nek5000-users] parameter lmg_rwt too small In-Reply-To: References: Message-ID: Yes sure, I'll send it to you during the weekend as I've already left my lab. JC 2010/12/17 > > JC, > > Can you send me a tarfile of your SIZE, .usr, and .rea file? > > Just send directly to my email address... fischer at mcs.anl.gov > (make it a gzipd tarfile if the rea file is large...) > > Thanks, > > Paul > > > On Fri, 17 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > > Hi Nek's, >> >> I'm having troubles trying to quickly simulate a boundary layer flow. Here >> is the message I get as soon as I launch the calculation: >> >> gs_setup: 0 unique labels shared >> >> setupds time 3.0041E-03 seconds 4 8 4181 300 >> >>> *parameter lmg_rwt too small 2400 1 -16* >>> >>> call exitt: dying ... >>> >>> >> Note however that when I change NSTEP to zero (post-process mode) I do not >> encounter any problem. >> >> >> Best regards, >> JC >> >> _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Fri Dec 17 21:27:55 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 17 Dec 2010 21:27:55 -0600 (CST) Subject: [Nek5000-users] 2D axisymmetric problems In-Reply-To: References: <8707DE09-5353-4B4C-AA3C-E63A72AC248F@nrel.gov> Message-ID: Hi Mike, We've updated the repo and have added an example "vortex2" that shows axisymmetric plus swirl. I confess I've not tried it with .re2 format yet, but for small problems .rea should be ok. The key is to add a logical "ifaziv" to be true for azimuthal velocity, and then to add the Coriolis force terms to the .usr file, per the example. As always with axisymmetric problems, you must have "A " for the boundary condition on the x-axis if y=0 is part of your domain boundary. Please let us know if this meets your needs, or not...! Cheers, Paul > > On Tue, 14 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > >> Hello All. >> >> I have a couple of questions regarding simulation of 2D axisymmetric >> problems where there's a nonzero azimuthal velocity component. >> >> First, does the new .rea + .re2 format work with such problems? I tried >> this with the "axi" problem in the examples directory, and things halted >> with >> >> ERROR READING INITIAL CONDITION/DRIVE FORCE DATA >> ABORTING IN ROUTINE RDICDF. >> >> It worked fine if I stuck with the older .rea format. >> >> Second, when doing such problems a while back, we would trick nek for such >> problems by using temperature to represent the azimuthal velocity. This >> approach required one to add appropriate terms into the .usr file. Is this >> still the case, or is Nek's third velocity component now the azimuthal >> velocity? >> >> Thanks. >> --Mike >> From nek5000-users at lists.mcs.anl.gov Sun Dec 19 12:53:24 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sun, 19 Dec 2010 12:53:24 -0600 (CST) Subject: [Nek5000-users] 2D axisymmetric problems In-Reply-To: Message-ID: <705047443.12663.1292784804644.JavaMail.root@zimbra.anl.gov> Hi Mike, A little correction to the previous email -- for the current version of axisymmetric setup with azimuthal velocity one needs to set ifaziv = .true. somewhere in .usr routines (including nek5_svn/example/vortex2/v2d.usr) since we do not have a mechanism for reading this flag from .rea file yet -- we will have it in place in one of the next updates... Best, Aleks ----- Original Message ----- From: nek5000-users at lists.mcs.anl.gov To: nek5000-users at lists.mcs.anl.gov Sent: Friday, December 17, 2010 9:27:55 PM Subject: Re: [Nek5000-users] 2D axisymmetric problems Hi Mike, We've updated the repo and have added an example "vortex2" that shows axisymmetric plus swirl. I confess I've not tried it with .re2 format yet, but for small problems .rea should be ok. The key is to add a logical "ifaziv" to be true for azimuthal velocity, and then to add the Coriolis force terms to the .usr file, per the example. As always with axisymmetric problems, you must have "A " for the boundary condition on the x-axis if y=0 is part of your domain boundary. Please let us know if this meets your needs, or not...! Cheers, Paul > > On Tue, 14 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > >> Hello All. >> >> I have a couple of questions regarding simulation of 2D axisymmetric >> problems where there's a nonzero azimuthal velocity component. >> >> First, does the new .rea + .re2 format work with such problems? I tried >> this with the "axi" problem in the examples directory, and things halted >> with >> >> ERROR READING INITIAL CONDITION/DRIVE FORCE DATA >> ABORTING IN ROUTINE RDICDF. >> >> It worked fine if I stuck with the older .rea format. >> >> Second, when doing such problems a while back, we would trick nek for such >> problems by using temperature to represent the azimuthal velocity. This >> approach required one to add appropriate terms into the .usr file. Is this >> still the case, or is Nek's third velocity component now the azimuthal >> velocity? >> >> Thanks. >> --Mike >> _______________________________________________ Nek5000-users mailing list Nek5000-users at lists.mcs.anl.gov https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users From nek5000-users at lists.mcs.anl.gov Sun Dec 19 16:44:48 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sun, 19 Dec 2010 16:44:48 -0600 (CST) Subject: [Nek5000-users] 2D axisymmetric problems In-Reply-To: <705047443.12663.1292784804644.JavaMail.root@zimbra.anl.gov> References: <705047443.12663.1292784804644.JavaMail.root@zimbra.anl.gov> Message-ID: Axisymmetry with azimuthal velocity (swirl) should now work if you add T IFAZIV to the .rea file in addition to ifaxis. You need to also set IFHEAT to be true, since the swirl component is stored in the T() array, and you need boundary condtions. An example is now in examples/vortex2 Hopefully I've properly checked everything into the repo. Paul On Sun, 19 Dec 2010, Aleksandr Obabko wrote: > Hi Mike, > > > A little correction to the previous email -- for the current version of axisymmetric setup with azimuthal velocity one needs to set > > ifaziv = .true. > > somewhere in .usr routines (including nek5_svn/example/vortex2/v2d.usr) since we do not have a mechanism for reading this flag from .rea file yet -- we will have it in place in one of the next updates... > > Best, > Aleks > > > > ----- Original Message ----- > From: nek5000-users at lists.mcs.anl.gov > To: nek5000-users at lists.mcs.anl.gov > Sent: Friday, December 17, 2010 9:27:55 PM > Subject: Re: [Nek5000-users] 2D axisymmetric problems > > > > Hi Mike, > > We've updated the repo and have added an example "vortex2" > that shows axisymmetric plus swirl. I confess I've not > tried it with .re2 format yet, but for small problems .rea > should be ok. > > The key is to add a logical "ifaziv" to be true for azimuthal > velocity, and then to add the Coriolis force terms to the .usr > file, per the example. As always with axisymmetric problems, > you must have "A " for the boundary condition on the x-axis > if y=0 is part of your domain boundary. > > Please let us know if this meets your needs, or not...! > > Cheers, > > Paul > > > >> >> On Tue, 14 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: >> >>> Hello All. >>> >>> I have a couple of questions regarding simulation of 2D axisymmetric >>> problems where there's a nonzero azimuthal velocity component. >>> >>> First, does the new .rea + .re2 format work with such problems? I tried >>> this with the "axi" problem in the examples directory, and things halted >>> with >>> >>> ERROR READING INITIAL CONDITION/DRIVE FORCE DATA >>> ABORTING IN ROUTINE RDICDF. >>> >>> It worked fine if I stuck with the older .rea format. >>> >>> Second, when doing such problems a while back, we would trick nek for such >>> problems by using temperature to represent the azimuthal velocity. This >>> approach required one to add appropriate terms into the .usr file. Is this >>> still the case, or is Nek's third velocity component now the azimuthal >>> velocity? >>> >>> Thanks. >>> --Mike >>> > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > From nek5000-users at lists.mcs.anl.gov Mon Dec 20 13:41:05 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 20 Dec 2010 12:41:05 -0700 Subject: [Nek5000-users] 2D axisymmetric problems In-Reply-To: References: <705047443.12663.1292784804644.JavaMail.root@zimbra.anl.gov> Message-ID: <64AC0195-1DB1-4076-A763-A1D49D1D8AEA@nrel.gov> Paul, Aleks, Thanks! All seems to be working great. Haven't tried with .re2, but .rea works fine. Validated the code against the numerical closed-cell Taylor-Couette results of Czarny et al (Physics of Fluids 15, 2003). --Mike On Dec 19, 2010, at 3:44 PM, wrote: > > Axisymmetry with azimuthal velocity (swirl) should now work if you add > > T IFAZIV > > to the .rea file in addition to ifaxis. You need to also set > IFHEAT to be true, since the swirl component is stored in the T() > array, and you need boundary condtions. > > An example is now in examples/vortex2 > > Hopefully I've properly checked everything into the repo. > > Paul > > > On Sun, 19 Dec 2010, Aleksandr Obabko wrote: > >> Hi Mike, >> >> >> A little correction to the previous email -- for the current version of axisymmetric setup with azimuthal velocity one needs to set >> >> ifaziv = .true. >> >> somewhere in .usr routines (including nek5_svn/example/vortex2/v2d.usr) since we do not have a mechanism for reading this flag from .rea file yet -- we will have it in place in one of the next updates... >> >> Best, >> Aleks >> >> >> >> ----- Original Message ----- >> From: nek5000-users at lists.mcs.anl.gov >> To: nek5000-users at lists.mcs.anl.gov >> Sent: Friday, December 17, 2010 9:27:55 PM >> Subject: Re: [Nek5000-users] 2D axisymmetric problems >> >> >> >> Hi Mike, >> >> We've updated the repo and have added an example "vortex2" >> that shows axisymmetric plus swirl. I confess I've not >> tried it with .re2 format yet, but for small problems .rea >> should be ok. >> >> The key is to add a logical "ifaziv" to be true for azimuthal >> velocity, and then to add the Coriolis force terms to the .usr >> file, per the example. As always with axisymmetric problems, >> you must have "A " for the boundary condition on the x-axis >> if y=0 is part of your domain boundary. >> >> Please let us know if this meets your needs, or not...! >> >> Cheers, >> >> Paul >> >> >> >>> >>> On Tue, 14 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: >>> >>>> Hello All. >>>> >>>> I have a couple of questions regarding simulation of 2D axisymmetric >>>> problems where there's a nonzero azimuthal velocity component. >>>> >>>> First, does the new .rea + .re2 format work with such problems? I tried >>>> this with the "axi" problem in the examples directory, and things halted >>>> with >>>> >>>> ERROR READING INITIAL CONDITION/DRIVE FORCE DATA >>>> ABORTING IN ROUTINE RDICDF. >>>> >>>> It worked fine if I stuck with the older .rea format. >>>> >>>> Second, when doing such problems a while back, we would trick nek for such >>>> problems by using temperature to represent the azimuthal velocity. This >>>> approach required one to add appropriate terms into the .usr file. Is this >>>> still the case, or is Nek's third velocity component now the azimuthal >>>> velocity? >>>> >>>> Thanks. >>>> --Mike >>>> >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users From nek5000-users at lists.mcs.anl.gov Tue Dec 21 20:56:19 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 21 Dec 2010 20:56:19 -0600 (CST) Subject: [Nek5000-users] parameter lmg_rwt too small In-Reply-To: References: Message-ID: Dear JC, Did you ever resolve this issue ? Best regards, Paul On Fri, 17 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > Yes sure, > > I'll send it to you during the weekend as I've already left my lab. > > JC > > 2010/12/17 > >> >> JC, >> >> Can you send me a tarfile of your SIZE, .usr, and .rea file? >> >> Just send directly to my email address... fischer at mcs.anl.gov >> (make it a gzipd tarfile if the rea file is large...) >> >> Thanks, >> >> Paul >> >> >> On Fri, 17 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: >> >> Hi Nek's, >>> >>> I'm having troubles trying to quickly simulate a boundary layer flow. Here >>> is the message I get as soon as I launch the calculation: >>> >>> gs_setup: 0 unique labels shared >>> >>> setupds time 3.0041E-03 seconds 4 8 4181 300 >>> >>>> *parameter lmg_rwt too small 2400 1 -16* >>>> >>>> call exitt: dying ... >>>> >>>> >>> Note however that when I change NSTEP to zero (post-process mode) I do not >>> encounter any problem. >>> >>> >>> Best regards, >>> JC >>> >>> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> > From nek5000-users at lists.mcs.anl.gov Wed Dec 29 14:59:14 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 29 Dec 2010 14:59:14 -0600 (CST) Subject: [Nek5000-users] Question about running error In-Reply-To: References: Message-ID: Dear Albert, It looks like your case ran successfully. (grep tep *.out shows this) You can visualize the results using VisIt. There is also a way to visualize using postx -- please see the (under construction) manual: www.mcs.anl.gov/~fischer/new.pdf on how to build and run postx. At present, you would need to rerun the turbChannel case with subroutine userdat2 : : param(66)=4 param(67)=4 and with these parameters set to 4 in the .rea file. This requirement will go away within a couple of weeks once I convert postx to support the new default output file format (.f00000 vs .fld00). I'm hoping we'll soon have a similar "get started" section about using VisIt in the manual. VisIt it much better than postx for 3D analysis. Paul From nek5000-users at lists.mcs.anl.gov Wed Dec 29 15:01:10 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 29 Dec 2010 15:01:10 -0600 (CST) Subject: [Nek5000-users] Nek5000 example errors In-Reply-To: References: Message-ID: John, Am cc'g this to the users list in case anyone else has seen this problem w/ gfortran / gcc. Paul On Wed, 29 Dec 2010, John Akerley wrote: > Yes, also using gcc (both version 4.4). I did switch IFMPI to false. > Also, I re-downloaded the code yesterday due to other problems, so I > assume it's the latest version. I got a similar error when trying the > turbChannel example. The compilation is successful, but after trying > either nek or nekb, I get an error. Attached is makenek. > > Thanks, > > John > > > > On Wed, Dec 29, 2010 at 12:30 PM, Paul Fischer wrote: >> >> Hi John, >> >> Are you also using gcc ? >> Did you switch IFMPI to false in makenek ? >> >> Did you just recently download the code ? (The test >> that I just made was done by downloading the current >> code after your previous email.) >> >> Confused, but certain we'll get to the bottom of this... >> >> Paul >> >> >> >> >> On Wed, 29 Dec 2010, John Akerley wrote: >> >>> Paul, >>> >>> I tried this and still received "Right-handed check failed for ?256 >>> elements. ?Exiting in routine VERRHE" after using grep -i err logfile >>> | tail command. ?I am using the gfortran compiler (no mpi). ?The >>> machine is a 64 bit with Redhat. >>> >>> On Wed, Dec 29, 2010 at 12:21 PM, Paul Fischer >>> wrote: >>>> >>>> Hi John, >>>> >>>> I just tried this myself and it worked ok. >>>> >>>> Can you try the following: >>>> >>>> makenek clean >>>> makenek eddy_uv >>>> nekb eddy_uv >>>> >>>> (Note, I've posted a new manual online, accessible via the >>>> wiki or directly via: >>>> >>>> www.mcs.anl.gov/~fischer/new.pdf >>>> >>>> ... and this will be continually updated.) >>>> >>>> The "nekb" script is in nek5_svn/trunk/tools/scripts >>>> >>>> >>>> Please let me know if the above works. ? Q: what compiler >>>> are you using ? >>>> >>>> Paul >>>> >>>> >>>> On Wed, 29 Dec 2010, John Akerley wrote: >>>> >>>>> Hi Paul, >>>>> >>>>> I have nek5000 installed and running on our linux machine, but I am >>>>> running into the same error when trying a few of the examples. ?The >>>>> logfile is created, but when checking the error I get "Right-handed >>>>> check failed for ?256 elements. ?Exiting in routine VERRHE". ?Do you >>>>> have a solution? ?I attached the logfile. >>>>> >>>>> Thanks, >>>>> >>>>> John Akerley >>>>> >>>> >> > From nek5000-users at lists.mcs.anl.gov Thu Dec 30 12:12:14 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 30 Dec 2010 12:12:14 -0600 Subject: [Nek5000-users] Post-Processing from avg_all files using Nek / Visit Message-ID: Hi, I would like to post-process quantities from avg_all files (rms, rm2, avg). Particularly I am looking at line plots (mainly) or contours on a plane. I had a question regarding the approach that I might be taking.. If I load the files, say rmsblah.f0000* into nek using the load subroutine and fill the ux,uy,uz arrays with that of , , and call hpts(), how would the interpolation work ? Would it interpolate with the interpolation that is used for velocities ?. The other option would be to do a line plot from Visit and export it to a text file. Also, I am not sure how Visit interpolates. Which is more likely to be accurate ? Regards Shriram -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Thu Dec 30 12:15:54 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 30 Dec 2010 10:15:54 -0800 Subject: [Nek5000-users] Post-Processing from avg_all files using Nek / Visit In-Reply-To: References: Message-ID: Hi Shiriam, It is pretty easy to automate this in VisIt. As we discussed at the user's meeting, VisIt treats each spectral element as a 5x5x5, 7x7x7, etc group of hexahedrons. Within each hexahedron, there is linear interpolation. So: if you want help automating this in VisIt, that should be easy. As far as accuracy, Nek will clearly do the best job, but hopefully you now understand better what VisIt is doing... Best, Hank On Thu, Dec 30, 2010 at 10:12 AM, wrote: > Hi, > > I would like to post-process quantities from avg_all files (rms, rm2, avg). > Particularly I am looking at line plots (mainly) or contours on a plane. I > had a question regarding the approach that I might be taking.. > > If I load the files, say rmsblah.f0000* into nek using the load subroutine > and fill the ux,uy,uz arrays with that of , , and call > hpts(), how would the interpolation work ? Would it interpolate with > the interpolation that is used for velocities ?. The other option would be > to do a line plot from Visit and export it to a text file. Also, I am not > sure how Visit interpolates. Which is more likely to be accurate ? > > Regards > Shriram > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Thu Dec 30 12:20:43 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 30 Dec 2010 12:20:43 -0600 Subject: [Nek5000-users] mixed Velocity BC's Message-ID: Howdy Neks, Has anyone tried or know how to implement a mixed velocity BC? Such as v=w=0 and du/dy=const? Thanks Andrew -- Andrew Duggleby, Ph.D. '01 Assistant Professor Department of Mechanical Engineering Texas A&M University College Station, TX? 77843-3123 109 Engineering Physics Building http://www2.mengr.tamu.edu/FacultyProfiles/facultyinformation.asp?LastName=aduggleby (979) 862-7835 From nek5000-users at lists.mcs.anl.gov Thu Dec 30 12:22:07 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 30 Dec 2010 12:22:07 -0600 Subject: [Nek5000-users] Post-Processing from avg_all files using Nek / Visit In-Reply-To: References: Message-ID: Hi Hank, Thanks. I think I can automate (with python) till I get Visit to do a line plot. Do you know a way I can just export the data from line plot to a text file ? I might have about 10 line plots (900 time steps) which I would like to automate. Regards Shriram Jagannathan On 30 December 2010 12:15, wrote: > > Hi Shiriam, > > It is pretty easy to automate this in VisIt. As we discussed at the user's > meeting, VisIt treats each spectral element as a 5x5x5, 7x7x7, etc group of > hexahedrons. Within each hexahedron, there is linear interpolation. > > So: if you want help automating this in VisIt, that should be easy. As far > as accuracy, Nek will clearly do the best job, but hopefully you now > understand better what VisIt is doing... > > Best, > Hank > > On Thu, Dec 30, 2010 at 10:12 AM, wrote: > >> Hi, >> >> I would like to post-process quantities from avg_all files (rms, rm2, >> avg). Particularly I am looking at line plots (mainly) or contours on a >> plane. I had a question regarding the approach that I might be taking.. >> >> If I load the files, say rmsblah.f0000* into nek using the load subroutine >> and fill the ux,uy,uz arrays with that of , , and call >> hpts(), how would the interpolation work ? Would it interpolate with >> the interpolation that is used for velocities ?. The other option would be >> to do a line plot from Visit and export it to a text file. Also, I am not >> sure how Visit interpolates. Which is more likely to be accurate ? >> >> Regards >> Shriram >> >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> >> > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Thu Dec 30 12:30:30 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 30 Dec 2010 10:30:30 -0800 Subject: [Nek5000-users] Post-Processing from avg_all files using Nek / Visit In-Reply-To: References: Message-ID: Hi Shririam, I would do: OpenDatabase("file.nek5000") # use real filename AddPlot("Pseudocolor", "pressure") # or whatever variable you want DrawPlots() Lineout((1,1,1),(2,2,2)) # takes lineout from point (1,1,1) to point (2,2,2) SetActiveWindow(2) sw = SaveWindowAttributes() sw.format = sw.CURVE sw.family = 0 sw.fileName = "myname_line1" SetSaveWindowAttributes(sw) # saves out myname_line1.curve SaveWindow() DeleteAllPlots() SetActiveWindow(1) Lineout((-1, 1, -1), (2, 4, 5)) SetActiveWindow(2) sw.fileName = "myname_line2" SetSaveWindowAttributes(sw) SaveWindow() DeleteAllPlots() SetActiveWindow(1) .... In VisIt 2.2, this will be much easier ... you won't have to make a Pseudocolor plot and do a Lineout query ... you will be able to add the Lineout directly via a curve plot. Best, Hank On Thu, Dec 30, 2010 at 10:22 AM, wrote: > Hi Hank, > > Thanks. I think I can automate (with python) till I get Visit to do a line > plot. Do you know a way I can just export the data from line plot to a text > file ? I might have about 10 line plots (900 time steps) which I would like > to automate. > > Regards > Shriram Jagannathan > > > > On 30 December 2010 12:15, wrote: > >> >> Hi Shiriam, >> >> It is pretty easy to automate this in VisIt. As we discussed at the >> user's meeting, VisIt treats each spectral element as a 5x5x5, 7x7x7, etc >> group of hexahedrons. Within each hexahedron, there is linear >> interpolation. >> >> So: if you want help automating this in VisIt, that should be easy. As >> far as accuracy, Nek will clearly do the best job, but hopefully you now >> understand better what VisIt is doing... >> >> Best, >> Hank >> >> On Thu, Dec 30, 2010 at 10:12 AM, wrote: >> >>> Hi, >>> >>> I would like to post-process quantities from avg_all files (rms, rm2, >>> avg). Particularly I am looking at line plots (mainly) or contours on a >>> plane. I had a question regarding the approach that I might be taking.. >>> >>> If I load the files, say rmsblah.f0000* into nek using the load >>> subroutine and fill the ux,uy,uz arrays with that of , , and >>> call hpts(), how would the interpolation work ? Would it interpolate >>> with the interpolation that is used for velocities ?. The other option would >>> be to do a line plot from Visit and export it to a text file. Also, I am not >>> sure how Visit interpolates. Which is more likely to be accurate ? >>> >>> Regards >>> Shriram >>> >>> _______________________________________________ >>> Nek5000-users mailing list >>> Nek5000-users at lists.mcs.anl.gov >>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >>> >>> >> >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> >> > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Thu Dec 30 12:44:17 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 30 Dec 2010 12:44:17 -0600 Subject: [Nek5000-users] Post-Processing from avg_all files using Nek / Visit In-Reply-To: References: Message-ID: Hi Hank, Thank you very much for the script. I will try this now. We have VisIt 2.1 on local machines but the cluster is powered with VisIt 1.12 only. Thanks again ! Regards Shriram Jagannathan On 30 December 2010 12:30, wrote: > Hi Shririam, > > I would do: > OpenDatabase("file.nek5000") # use real filename > AddPlot("Pseudocolor", "pressure") # or whatever variable you want > DrawPlots() > Lineout((1,1,1),(2,2,2)) # takes lineout from point (1,1,1) to point > (2,2,2) > SetActiveWindow(2) > sw = SaveWindowAttributes() > sw.format = sw.CURVE > sw.family = 0 > sw.fileName = "myname_line1" > SetSaveWindowAttributes(sw) # saves out myname_line1.curve > SaveWindow() > DeleteAllPlots() > SetActiveWindow(1) > Lineout((-1, 1, -1), (2, 4, 5)) > SetActiveWindow(2) > sw.fileName = "myname_line2" > SetSaveWindowAttributes(sw) > SaveWindow() > DeleteAllPlots() > SetActiveWindow(1) > .... > > In VisIt 2.2, this will be much easier ... you won't have to make a > Pseudocolor plot and do a Lineout query ... you will be able to add the > Lineout directly via a curve plot. > > Best, > Hank > > > On Thu, Dec 30, 2010 at 10:22 AM, wrote: > >> Hi Hank, >> >> Thanks. I think I can automate (with python) till I get Visit to do a line >> plot. Do you know a way I can just export the data from line plot to a text >> file ? I might have about 10 line plots (900 time steps) which I would like >> to automate. >> >> Regards >> Shriram Jagannathan >> >> >> >> On 30 December 2010 12:15, wrote: >> >>> >>> Hi Shiriam, >>> >>> It is pretty easy to automate this in VisIt. As we discussed at the >>> user's meeting, VisIt treats each spectral element as a 5x5x5, 7x7x7, etc >>> group of hexahedrons. Within each hexahedron, there is linear >>> interpolation. >>> >>> So: if you want help automating this in VisIt, that should be easy. As >>> far as accuracy, Nek will clearly do the best job, but hopefully you now >>> understand better what VisIt is doing... >>> >>> Best, >>> Hank >>> >>> On Thu, Dec 30, 2010 at 10:12 AM, wrote: >>> >>>> Hi, >>>> >>>> I would like to post-process quantities from avg_all files (rms, rm2, >>>> avg). Particularly I am looking at line plots (mainly) or contours on a >>>> plane. I had a question regarding the approach that I might be taking.. >>>> >>>> If I load the files, say rmsblah.f0000* into nek using the load >>>> subroutine and fill the ux,uy,uz arrays with that of , , and >>>> call hpts(), how would the interpolation work ? Would it interpolate >>>> with the interpolation that is used for velocities ?. The other option would >>>> be to do a line plot from Visit and export it to a text file. Also, I am not >>>> sure how Visit interpolates. Which is more likely to be accurate ? >>>> >>>> Regards >>>> Shriram >>>> >>>> _______________________________________________ >>>> Nek5000-users mailing list >>>> Nek5000-users at lists.mcs.anl.gov >>>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >>>> >>>> >>> >>> _______________________________________________ >>> Nek5000-users mailing list >>> Nek5000-users at lists.mcs.anl.gov >>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >>> >>> >> >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> >> > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Thu Dec 30 13:03:09 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 30 Dec 2010 13:03:09 -0600 Subject: [Nek5000-users] Post-Processing from avg_all files using Nek / Visit In-Reply-To: References: Message-ID: Hi Hank, It works w/o any issues. Thanks. Regards Shriram Jagannathan On 30 December 2010 12:44, Shriram Jagannathan wrote: > Hi Hank, > > Thank you very much for the script. I will try this now. We have VisIt 2.1 > on local machines but the cluster is powered with VisIt 1.12 only. Thanks > again ! > > Regards > Shriram Jagannathan > > > > On 30 December 2010 12:30, wrote: > >> Hi Shririam, >> >> I would do: >> OpenDatabase("file.nek5000") # use real filename >> AddPlot("Pseudocolor", "pressure") # or whatever variable you want >> DrawPlots() >> Lineout((1,1,1),(2,2,2)) # takes lineout from point (1,1,1) to point >> (2,2,2) >> SetActiveWindow(2) >> sw = SaveWindowAttributes() >> sw.format = sw.CURVE >> sw.family = 0 >> sw.fileName = "myname_line1" >> SetSaveWindowAttributes(sw) # saves out myname_line1.curve >> SaveWindow() >> DeleteAllPlots() >> SetActiveWindow(1) >> Lineout((-1, 1, -1), (2, 4, 5)) >> SetActiveWindow(2) >> sw.fileName = "myname_line2" >> SetSaveWindowAttributes(sw) >> SaveWindow() >> DeleteAllPlots() >> SetActiveWindow(1) >> .... >> >> In VisIt 2.2, this will be much easier ... you won't have to make a >> Pseudocolor plot and do a Lineout query ... you will be able to add the >> Lineout directly via a curve plot. >> >> Best, >> Hank >> >> >> On Thu, Dec 30, 2010 at 10:22 AM, wrote: >> >>> Hi Hank, >>> >>> Thanks. I think I can automate (with python) till I get Visit to do a >>> line plot. Do you know a way I can just export the data from line plot to a >>> text file ? I might have about 10 line plots (900 time steps) which I would >>> like to automate. >>> >>> Regards >>> Shriram Jagannathan >>> >>> >>> >>> On 30 December 2010 12:15, wrote: >>> >>>> >>>> Hi Shiriam, >>>> >>>> It is pretty easy to automate this in VisIt. As we discussed at the >>>> user's meeting, VisIt treats each spectral element as a 5x5x5, 7x7x7, etc >>>> group of hexahedrons. Within each hexahedron, there is linear >>>> interpolation. >>>> >>>> So: if you want help automating this in VisIt, that should be easy. As >>>> far as accuracy, Nek will clearly do the best job, but hopefully you now >>>> understand better what VisIt is doing... >>>> >>>> Best, >>>> Hank >>>> >>>> On Thu, Dec 30, 2010 at 10:12 AM, wrote: >>>> >>>>> Hi, >>>>> >>>>> I would like to post-process quantities from avg_all files (rms, rm2, >>>>> avg). Particularly I am looking at line plots (mainly) or contours on a >>>>> plane. I had a question regarding the approach that I might be taking.. >>>>> >>>>> If I load the files, say rmsblah.f0000* into nek using the load >>>>> subroutine and fill the ux,uy,uz arrays with that of , , and >>>>> call hpts(), how would the interpolation work ? Would it interpolate >>>>> with the interpolation that is used for velocities ?. The other option would >>>>> be to do a line plot from Visit and export it to a text file. Also, I am not >>>>> sure how Visit interpolates. Which is more likely to be accurate ? >>>>> >>>>> Regards >>>>> Shriram >>>>> >>>>> _______________________________________________ >>>>> Nek5000-users mailing list >>>>> Nek5000-users at lists.mcs.anl.gov >>>>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >>>>> >>>>> >>>> >>>> _______________________________________________ >>>> Nek5000-users mailing list >>>> Nek5000-users at lists.mcs.anl.gov >>>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >>>> >>>> >>> >>> _______________________________________________ >>> Nek5000-users mailing list >>> Nek5000-users at lists.mcs.anl.gov >>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >>> >>> >> >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> >> > -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Thu Dec 30 13:12:33 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 30 Dec 2010 11:12:33 -0800 Subject: [Nek5000-users] Post-Processing from avg_all files using Nek / Visit In-Reply-To: References: Message-ID: Cool! Keep the questions coming and I'll help out with whatever I can. It's very little effort for me to put together little scripts like that... Best, Hank On Thu, Dec 30, 2010 at 11:03 AM, wrote: > Hi Hank, > > It works w/o any issues. Thanks. > > Regards > Shriram Jagannathan > > > > On 30 December 2010 12:44, Shriram Jagannathan wrote: > >> Hi Hank, >> >> Thank you very much for the script. I will try this now. We have VisIt 2.1 >> on local machines but the cluster is powered with VisIt 1.12 only. Thanks >> again ! >> >> Regards >> Shriram Jagannathan >> >> >> >> On 30 December 2010 12:30, wrote: >> >>> Hi Shririam, >>> >>> I would do: >>> OpenDatabase("file.nek5000") # use real filename >>> AddPlot("Pseudocolor", "pressure") # or whatever variable you want >>> DrawPlots() >>> Lineout((1,1,1),(2,2,2)) # takes lineout from point (1,1,1) to point >>> (2,2,2) >>> SetActiveWindow(2) >>> sw = SaveWindowAttributes() >>> sw.format = sw.CURVE >>> sw.family = 0 >>> sw.fileName = "myname_line1" >>> SetSaveWindowAttributes(sw) # saves out myname_line1.curve >>> SaveWindow() >>> DeleteAllPlots() >>> SetActiveWindow(1) >>> Lineout((-1, 1, -1), (2, 4, 5)) >>> SetActiveWindow(2) >>> sw.fileName = "myname_line2" >>> SetSaveWindowAttributes(sw) >>> SaveWindow() >>> DeleteAllPlots() >>> SetActiveWindow(1) >>> .... >>> >>> In VisIt 2.2, this will be much easier ... you won't have to make a >>> Pseudocolor plot and do a Lineout query ... you will be able to add the >>> Lineout directly via a curve plot. >>> >>> Best, >>> Hank >>> >>> >>> On Thu, Dec 30, 2010 at 10:22 AM, wrote: >>> >>>> Hi Hank, >>>> >>>> Thanks. I think I can automate (with python) till I get Visit to do a >>>> line plot. Do you know a way I can just export the data from line plot to a >>>> text file ? I might have about 10 line plots (900 time steps) which I would >>>> like to automate. >>>> >>>> Regards >>>> Shriram Jagannathan >>>> >>>> >>>> >>>> On 30 December 2010 12:15, wrote: >>>> >>>>> >>>>> Hi Shiriam, >>>>> >>>>> It is pretty easy to automate this in VisIt. As we discussed at the >>>>> user's meeting, VisIt treats each spectral element as a 5x5x5, 7x7x7, etc >>>>> group of hexahedrons. Within each hexahedron, there is linear >>>>> interpolation. >>>>> >>>>> So: if you want help automating this in VisIt, that should be easy. As >>>>> far as accuracy, Nek will clearly do the best job, but hopefully you now >>>>> understand better what VisIt is doing... >>>>> >>>>> Best, >>>>> Hank >>>>> >>>>> On Thu, Dec 30, 2010 at 10:12 AM, wrote: >>>>> >>>>>> Hi, >>>>>> >>>>>> I would like to post-process quantities from avg_all files (rms, rm2, >>>>>> avg). Particularly I am looking at line plots (mainly) or contours on a >>>>>> plane. I had a question regarding the approach that I might be taking.. >>>>>> >>>>>> If I load the files, say rmsblah.f0000* into nek using the load >>>>>> subroutine and fill the ux,uy,uz arrays with that of , , and >>>>>> call hpts(), how would the interpolation work ? Would it interpolate >>>>>> with the interpolation that is used for velocities ?. The other option would >>>>>> be to do a line plot from Visit and export it to a text file. Also, I am not >>>>>> sure how Visit interpolates. Which is more likely to be accurate ? >>>>>> >>>>>> Regards >>>>>> Shriram >>>>>> >>>>>> _______________________________________________ >>>>>> Nek5000-users mailing list >>>>>> Nek5000-users at lists.mcs.anl.gov >>>>>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >>>>>> >>>>>> >>>>> >>>>> _______________________________________________ >>>>> Nek5000-users mailing list >>>>> Nek5000-users at lists.mcs.anl.gov >>>>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >>>>> >>>>> >>>> >>>> _______________________________________________ >>>> Nek5000-users mailing list >>>> Nek5000-users at lists.mcs.anl.gov >>>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >>>> >>>> >>> >>> _______________________________________________ >>> Nek5000-users mailing list >>> Nek5000-users at lists.mcs.anl.gov >>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >>> >>> >> > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Fri Dec 31 09:27:15 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 31 Dec 2010 09:27:15 -0600 (CST) Subject: [Nek5000-users] Post-Processing from avg_all files using Nek / Visit In-Reply-To: References: Message-ID: Hi Shriram, I see that Hank answered some of your questions already. If we define u = + u' then we have ^2 = - ^2 with stored in rms and stored in avg. (Similar forms exist for the cross terms.) postx can load the rms and avg files to give you ^2 -- but not yet with the .f0000 format. Another common approach is to manipulate these by loading all the avg files inside the solver itself. We can set up an example that illustrates this. At that point, you could also interpolate the output onto a finer mesh before passing to visit. Paul On Thu, 30 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > Hi, > > I would like to post-process quantities from avg_all files (rms, rm2, avg). > Particularly I am looking at line plots (mainly) or contours on a plane. I > had a question regarding the approach that I might be taking.. > > If I load the files, say rmsblah.f0000* into nek using the load subroutine > and fill the ux,uy,uz arrays with that of , , and call > hpts(), how would the interpolation work ? Would it interpolate with > the interpolation that is used for velocities ?. The other option would be > to do a line plot from Visit and export it to a text file. Also, I am not > sure how Visit interpolates. Which is more likely to be accurate ? > > Regards > Shriram > Z_ kk kk http://cuireadhchunceoil.ie/index.php?option=com_content&view=article&id=58 From nek5000-users at lists.mcs.anl.gov Fri Dec 31 09:29:15 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 31 Dec 2010 09:29:15 -0600 (CST) Subject: [Nek5000-users] mixed Velocity BC's In-Reply-To: References: Message-ID: Hi Andrew, On what type of surface do you want this? (z plane?, y plane?...?) Also, if y-plane, wouldn't it be: v=0 w_y = 0 u_y = const ? (Or am I totally missing what you're trying to do ?) Cheers, Paul On Thu, 30 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > Howdy Neks, > > Has anyone tried or know how to implement a mixed velocity BC? Such > as v=w=0 and du/dy=const? > > Thanks > Andrew > > -- > Andrew Duggleby, Ph.D. '01 > Assistant Professor > Department of Mechanical Engineering > Texas A&M University > College Station, TX? 77843-3123 > > 109 Engineering Physics Building > http://www2.mengr.tamu.edu/FacultyProfiles/facultyinformation.asp?LastName=aduggleby > (979) 862-7835 > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > From nek5000-users at lists.mcs.anl.gov Fri Dec 31 09:40:22 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 31 Dec 2010 09:40:22 -0600 Subject: [Nek5000-users] mixed Velocity BC's In-Reply-To: References: Message-ID: Ultimately I'd like to be able to do this on any general surface. Basically, impose a slip velocity in the streamwise direction with a specified gradient (so w_y=0 and u_y=const in a y-plane), no penetration in the wall-normal (v=0). Thanks, Andrew On Dec 31, 2010, at 9:29 AM, nek5000-users at lists.mcs.anl.gov wrote: > > Hi Andrew, > > On what type of surface do you want this? (z plane?, y plane?...?) > > Also, if y-plane, wouldn't it be: > > v=0 > w_y = 0 > u_y = const > > ? > > (Or am I totally missing what you're trying to do ?) > > Cheers, > > Paul > > > On Thu, 30 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > >> Howdy Neks, >> >> Has anyone tried or know how to implement a mixed velocity BC? Such >> as v=w=0 and du/dy=const? >> >> Thanks >> Andrew >> >> -- >> Andrew Duggleby, Ph.D. '01 >> Assistant Professor >> Department of Mechanical Engineering >> Texas A&M University >> College Station, TX 77843-3123 >> >> 109 Engineering Physics Building >> http://www2.mengr.tamu.edu/FacultyProfiles/facultyinformation.asp?LastName=aduggleby >> (979) 862-7835 >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users From nek5000-users at lists.mcs.anl.gov Fri Dec 31 09:57:01 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 31 Dec 2010 09:57:01 -0600 Subject: [Nek5000-users] Post-Processing from avg_all files using Nek / Visit In-Reply-To: References: Message-ID: Hi Paul, Thanks. I did set up a python script (thanks to Hank) in VisIt to do the interpolation and o/p as txt file. Does rmsblah.f0000 file have the other components of Reynolds stress viz. , , ? Yes, we have a routine for getting the fluctuating components from nek and outside of it as well. Regards Shriram On 31 December 2010 09:27, wrote: > > Hi Shriram, > > I see that Hank answered some of your questions already. > > If we define u = + u' then we have > > ^2 = - ^2 > > with stored in rms and stored in avg. (Similar > forms exist for the cross terms.) postx can load the > rms and avg files to give you ^2 -- but not yet with the > .f0000 format. > > Another common approach is to manipulate these by loading all the avg files > inside the solver itself. We can set up an > example that illustrates this. At that point, you could > also interpolate the output onto a finer mesh before passing > to visit. > > Paul > > > > > On Thu, 30 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > > Hi, >> >> I would like to post-process quantities from avg_all files (rms, rm2, >> avg). >> Particularly I am looking at line plots (mainly) or contours on a plane. I >> had a question regarding the approach that I might be taking.. >> >> If I load the files, say rmsblah.f0000* into nek using the load subroutine >> and fill the ux,uy,uz arrays with that of , , and call >> hpts(), how would the interpolation work ? Would it interpolate with >> the interpolation that is used for velocities ?. The other option would be >> to do a line plot from Visit and export it to a text file. Also, I am not >> sure how Visit interpolates. Which is more likely to be accurate ? >> >> Regards >> Shriram >> >> > > Z_ > > > kk > > kk > http://cuireadhchunceoil.ie/index.php?option=com_content&view=article&id=58 > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Fri Dec 31 10:00:06 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 31 Dec 2010 10:00:06 -0600 (CST) Subject: [Nek5000-users] mixed Velocity BC's In-Reply-To: References: Message-ID: Hi Andrew, At present, this can be done with Pn-Pn-2, which allows specification of traction boundary conditions. One issue is that an arbitrary surface is fundamentally different than an aligned one because the constraints U.n = 0 and U.t_i = tau_i couple the components of U=(u,v,w), meaning that these velocity components must be solved in a single Helmholtz solve instead of as 3 independent ways. The stress formulation takes care of this. It should be possible to do this in the Pn-Pn formulation as well, but we're not there yet. One possibility might be to use a high-order explicit prescription of the boundary condition. (I'd have to think about that a bit...) Paul On Fri, 31 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > Ultimately I'd like to be able to do this on any general surface. Basically, impose a slip velocity in the streamwise direction with a specified gradient (so w_y=0 and u_y=const in a y-plane), no penetration in the wall-normal (v=0). > > Thanks, > Andrew > On Dec 31, 2010, at 9:29 AM, nek5000-users at lists.mcs.anl.gov wrote: > >> >> Hi Andrew, >> >> On what type of surface do you want this? (z plane?, y plane?...?) >> >> Also, if y-plane, wouldn't it be: >> >> v=0 >> w_y = 0 >> u_y = const >> >> ? >> >> (Or am I totally missing what you're trying to do ?) >> >> Cheers, >> >> Paul >> >> >> On Thu, 30 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: >> >>> Howdy Neks, >>> >>> Has anyone tried or know how to implement a mixed velocity BC? Such >>> as v=w=0 and du/dy=const? >>> >>> Thanks >>> Andrew >>> >>> -- >>> Andrew Duggleby, Ph.D. '01 >>> Assistant Professor >>> Department of Mechanical Engineering >>> Texas A&M University >>> College Station, TX 77843-3123 >>> >>> 109 Engineering Physics Building >>> http://www2.mengr.tamu.edu/FacultyProfiles/facultyinformation.asp?LastName=aduggleby >>> (979) 862-7835 >>> _______________________________________________ >>> Nek5000-users mailing list >>> Nek5000-users at lists.mcs.anl.gov >>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > From nek5000-users at lists.mcs.anl.gov Fri Dec 31 10:02:35 2010 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 31 Dec 2010 10:02:35 -0600 (CST) Subject: [Nek5000-users] =?utf-8?q?Post-Processing_from_avg=5Fall_files_us?= =?utf-8?q?ing_Nek_/=09Visit?= In-Reply-To: References: Message-ID: Hi Shriram, the rm2 file has uv vw wu Look for "outpost" in navier5.f, which is where the avg_all() routine is found. Note that, because of space limitations, not _all_ of the fields are computed/stored.... but these clearly could be added for someone interested in turbulent Nu computations. Cheers, Paul On Fri, 31 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: > Hi Paul, > > Thanks. I did set up a python script (thanks to Hank) in VisIt to do the > interpolation and o/p as txt file. Does rmsblah.f0000 file have the other > components of Reynolds stress viz. , , ? Yes, we have a routine > for getting the fluctuating components from nek and outside of it as well. > > Regards > Shriram > > > On 31 December 2010 09:27, wrote: > >> >> Hi Shriram, >> >> I see that Hank answered some of your questions already. >> >> If we define u = + u' then we have >> >> ^2 = - ^2 >> >> with stored in rms and stored in avg. (Similar >> forms exist for the cross terms.) postx can load the >> rms and avg files to give you ^2 -- but not yet with the >> .f0000 format. >> >> Another common approach is to manipulate these by loading all the avg files >> inside the solver itself. We can set up an >> example that illustrates this. At that point, you could >> also interpolate the output onto a finer mesh before passing >> to visit. >> >> Paul >> >> >> >> >> On Thu, 30 Dec 2010, nek5000-users at lists.mcs.anl.gov wrote: >> >> Hi, >>> >>> I would like to post-process quantities from avg_all files (rms, rm2, >>> avg). >>> Particularly I am looking at line plots (mainly) or contours on a plane. I >>> had a question regarding the approach that I might be taking.. >>> >>> If I load the files, say rmsblah.f0000* into nek using the load subroutine >>> and fill the ux,uy,uz arrays with that of , , and call >>> hpts(), how would the interpolation work ? Would it interpolate with >>> the interpolation that is used for velocities ?. The other option would be >>> to do a line plot from Visit and export it to a text file. Also, I am not >>> sure how Visit interpolates. Which is more likely to be accurate ? >>> >>> Regards >>> Shriram >>> >>> >> >> Z_ >> >> >> kk >> >> kk >> http://cuireadhchunceoil.ie/index.php?option=com_content&view=article&id=58 >> >> _______________________________________________ >> Nek5000-users mailing list >> Nek5000-users at lists.mcs.anl.gov >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> >