From nek5000-users at lists.mcs.anl.gov Mon Nov 6 06:13:11 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 6 Nov 2017 13:13:11 +0100 Subject: [Nek5000-users] IFCHAR option in rea file Message-ID: Hello, I have been using IFCHAR = TRUE in my simulations. The CFL number is around 1.2 and the computations appear to go on nicely. I would like to know if the flow field computed using IFCHAR=TRUE option is different than the one obtained with IFCHAR=FALSE (albeit with smaller dt)? If yes, can the difference be significant? Many thanks, NN From nek5000-users at lists.mcs.anl.gov Mon Nov 6 06:30:47 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 6 Nov 2017 12:30:47 +0000 Subject: [Nek5000-users] IFCHAR option in rea file In-Reply-To: References: Message-ID: Dear NN, If your flow is turbulent the answer is definitely Yes (they will be different) because of the standard sensitivity to initial condition concerns. Despite this, both schemes are O(dt^k) accurate, meaning that as dt-->0, the error goes to _zero_ as dt^k, where k=Torder is the prescribed temporal order. Since the error goes to zero, both will deliver the same answer provided that the physical solution is stable and has an attractor. [Understanding when such conditions prevail is often a challenge.] Paul ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Monday, November 6, 2017 6:13:11 AM To: nek5000-users at lists.mcs.anl.gov Subject: [Nek5000-users] IFCHAR option in rea file Hello, I have been using IFCHAR = TRUE in my simulations. The CFL number is around 1.2 and the computations appear to go on nicely. I would like to know if the flow field computed using IFCHAR=TRUE option is different than the one obtained with IFCHAR=FALSE (albeit with smaller dt)? If yes, can the difference be significant? Many thanks, NN _______________________________________________ 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 Nov 6 06:34:48 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 6 Nov 2017 12:34:48 +0000 Subject: [Nek5000-users] Hackathon Message-ID: Dear All, Just a reminder to sign up for the upcoming Hackathon at UIUC, this coming Sunday--Tuesday (11/12--11/14), if you plan to attend. Here is the link: https://doodle.com/poll/yxxtw49ewxtzzruk In addition to helping attendees with their challenging problems, we also are asking attendees to develop an example case (with a known solution, if possible) appropriate for posting to the Examples suite. Our goal is to have the examples posted by Tuesday. Looking forward to seeing you next Sunday! The Nek development team. -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Nov 6 06:40:28 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 06 Nov 2017 15:40:28 +0300 Subject: [Nek5000-users] =?utf-8?q?Axis_in_spectral_elements?= Message-ID: Hi, Neks! I am working with a complex geometry and want to change positions of some points in spectral elements. I've read in documentation that i,j,k,e in xm1(i,j,k,e), for example, are changed from 1 to nx1,ny1,nz1 and nelv respectively. But during my tests it's seemed that in different elements x,y,z axis not always correspond to the global coordinate system. Am I right? And how could I change only the lowest slice of an element, for example? Best regards, Elizabeth -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Nov 6 07:34:18 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 6 Nov 2017 14:34:18 +0100 Subject: [Nek5000-users] IFCHAR option in rea file In-Reply-To: References: Message-ID: Thanks very much Paul. Sincerely, NN. On Monday 06 November 2017 01:30 PM, nek5000-users at lists.mcs.anl.gov wrote: > > > Dear NN, > > > If your flow is turbulent the answer is definitely Yes (they will be > different) because of the standard sensitivity to initial condition > concerns. > > > Despite this, both schemes are O(dt^k) accurate, meaning that as > dt-->0, the error goes to _zero_ as dt^k, where k=Torder is the > prescribed temporal order. Since the error goes to zero, both will > deliver the same answer provided that the physical solution is stable > and has an attractor. [Understanding when such conditions prevail is > often a challenge.] > > > Paul > > > ------------------------------------------------------------------------ > *From:* Nek5000-users on > behalf of nek5000-users at lists.mcs.anl.gov > > *Sent:* Monday, November 6, 2017 6:13:11 AM > *To:* nek5000-users at lists.mcs.anl.gov > *Subject:* [Nek5000-users] IFCHAR option in rea file > Hello, > > I have been using IFCHAR = TRUE in my simulations. The CFL number is > around > 1.2 and the computations appear to go on nicely. I would like to know > if the flow field computed using IFCHAR=TRUE option is different than the > one obtained with IFCHAR=FALSE (albeit with smaller dt)? If yes, can the > difference > be significant? > > Many thanks, > NN > > _______________________________________________ > 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 Wed Nov 8 03:55:44 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 08 Nov 2017 10:55:44 +0100 Subject: [Nek5000-users] Dissipation rate In-Reply-To: References: <07d9c63f18562f417c187df4ce221889@vki.ac.be> Message-ID: Dear Neks, I would like to compute the turbulent dissipation rate, in order to calculate the Kolmogorov microscales. Does a subroutine already exist or should I compute it from userchk ? Thank you very much for your answers, Samuel From nek5000-users at lists.mcs.anl.gov Wed Nov 8 06:35:42 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 8 Nov 2017 13:35:42 +0100 Subject: [Nek5000-users] Problem with drag coefficient of ext_cyl Message-ID: Hello, I just started my PhD work in FSI of NACA profile so I am using NEK5000 for the simulations. So, in order to start with NEK5000, first I try to validate a very simple test case which is the flow around a 2D cylinder. This test case is already in the example folder of the solver kit. So, I try to simply run the test case and tried to compare the drag coefficient with the given Re number. It was giving a wrong value. for example it gives 21 instead of 3.5 for Re 100. I had changed the domain a bit according the research papper I have. I used torque calc function to compute the drag. Could you please help me to validate this test case? So that I can continue with lmy moving cylinder test case and I will be more confident in NEK5000 Thanks SIjo George -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Nov 8 14:46:34 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 8 Nov 2017 20:46:34 +0000 Subject: [Nek5000-users] Problem with drag coefficient of ext_cyl Message-ID: Hi SIjo George, Can you tell how have you changed the domain? Also, are you running the latest version of the repo? When I use the current case in NekExamples repo - ext_cyl - and run the calculation, I get a drag coefficient of about 1.35 for Re=100, which I think is about right for this Re. Below is a plot of drag coefficient evolution over time that I got. Ketan [cid:B3193C5A-4983-41E9-B997-E058759EA923] -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Screen Shot 2017-11-08 at 2.43.30 PM.png Type: image/png Size: 18616 bytes Desc: Screen Shot 2017-11-08 at 2.43.30 PM.png URL: From nek5000-users at lists.mcs.anl.gov Thu Nov 9 17:29:33 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 09 Nov 2017 23:29:33 +0000 Subject: [Nek5000-users] genmap error Message-ID: Hey neks, I'm having a wierd genmap error while trying to convert from a gmsh format to nek format using cpraveen's script. The .rea file generates correctly but I cannot run genmap after that. trying to compile the case without running genmap also gives an error. The error message from genmap is ERROR: error reading 1 1 891 aborting 510 in routine rdbdry. I can email the .geo and .msh files if required Sincerely, -- *Amitvikram Dutta* Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Fri Nov 10 00:46:04 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 10 Nov 2017 07:46:04 +0100 Subject: [Nek5000-users] genmap error In-Reply-To: References: Message-ID: I guess you convert does somehow not produce the expected format. -----Original message----- > From:nek5000-users at lists.mcs.anl.gov > Sent: Friday 10th November 2017 5:48 > To: nek5000-users at lists.mcs.anl.gov > Subject: [Nek5000-users] genmap error > > Hey neks, > > Im having a wierd genmap error while trying to convert from a gmsh format to nek format using cpraveens script. The .rea file generates correctly but I cannot run genmap after that. trying to compile the case without running genmap also gives an error. > > The error message from genmap is > ERROR: error reading??? 1?????????? 1???????? 891 > ? aborting 510 in routine rdbdry. > > I can email the .geo and .msh files if required > > Sincerely, > -- > Amitvikram Dutta > Graduate Research Assistant > Fluid Mechanics Research Lab > University of Waterloo > _______________________________________________ > 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 Nov 10 07:08:06 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 10 Nov 2017 13:08:06 +0000 Subject: [Nek5000-users] Hackathon Update Message-ID: Dear All, There are currently about 20 registrants for the upcoming Hackathon, which starts Sunday, November 12 at 1 PM in 4405 Siebel Hall at the University of Illinois, Urbana Champaign. (This is the UIUC computer science building, 4th floor.) It's still not too late to sign up. If you wish to join, please register so that we can have an accurate head count for the meals. Parking is free on Sunday, so you can park most anywhere near Siebel. Sunday evening dinner will be at Pizza M in Urbana - a 13 minute walk from Siebel. Monday and Tuesday we will meet at 9 AM in the first floor of the NCSA building, which is next door to Siebel. As mentioned in an earlier post, we'd like to start off with people taking a two or three minutes to stand up and describe the problem they'd like to tackle during the hackathon (e.g., particle tracking, turbulence in a given domain, etc.) so that we can self organize into small focus groups. Also, we'll have a few requests from the development team for assistance with certain tasks aimed toward getting out the latest release and with developing new examples to be posted. Looking forward to seeing you all on Sunday! The Nek development team. -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Fri Nov 10 08:56:09 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 10 Nov 2017 15:56:09 +0100 Subject: [Nek5000-users] Dissipation rate In-Reply-To: <662f888a2aaf0ff2f3a4b3f4ebea4036@vki.ac.be> References: <07d9c63f18562f417c187df4ce221889@vki.ac.be> <662f888a2aaf0ff2f3a4b3f4ebea4036@vki.ac.be> Message-ID: Dear Nek experts, I'm still working on the turbulent dissipation rate. I would like to compute epsilon=2*nu* with Sij=0.5*(grad(U)+grad(U)^T) I found in the forum a proposition to calculate the scalar dissipation rate as follow. If I understood well, this code is basically doing "chi := D * |grad Z|^2" while I would like to compute chi := D * D * |grad Z + grad^T Z|^2 Does anyone have a suggestion ? Many thanks, Samuel c----------------------------------------------------------------------- subroutine magSqr(a,b1,b2,b3,n) include 'SIZE' real a(1) real b1(1),b2(1),b3(1) do i=1,n a(i) = b1(i)*b1(i) + b2(i)*b2(i) enddo return end c----------------------------------------------------------------------- subroutine scalDisp(chi,Z,D) c c compute scalar dissipation rate c chi := D * |grad Z|^2 c include 'SIZE' include 'TOTAL' real chi(lx1,ly1,lz1,1) real Z (lx1,ly1,lz1,1) real D (lx1,ly1,lz1,1) common /scrns/ w1(lx1,ly1,lz1,lelt) $ ,w2(lx1,ly1,lz1,lelt) $ ,w3(lx1,ly1,lz1,lelt) ntot = nx1*ny1*nz1*nelv call opgrad (w1,w2,w3,Z) call opdssum(w1,w2,w3) call opcolv (w1,w2,w3,binvm1) call magsqr (chi,w1,w2,w3,ntot) call col2 (chi,D,ntot) return end c----------------------------------------------------------------------- Le 2017-11-08 10:55, Samuel Ahizi a ?crit?: > Dear Neks, > > I would like to compute the turbulent dissipation rate, in order to > calculate the Kolmogorov microscales. Does a subroutine already exist > or should I compute it from userchk ? > > Thank you very much for your answers, > Samuel From nek5000-users at lists.mcs.anl.gov Fri Nov 10 09:12:08 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 10 Nov 2017 15:12:08 +0000 Subject: [Nek5000-users] Dissipation rate In-Reply-To: References: <07d9c63f18562f417c187df4ce221889@vki.ac.be> <662f888a2aaf0ff2f3a4b3f4ebea4036@vki.ac.be>, Message-ID: Dear Samuel, Several of these routines are in the turbchannel example .usr file (and nek5000 source). Please have a look at the .usr file in that example and you'll see the below. hth, Paul call comp_gije(sij,vx(1,1,1,e),vy(1,1,1,e),vz(1,1,1,e),e) call comp_sije(sij) call mag_tensor_e(snrm(1,e),sij) call cmult(snrm(1,e),2.0,ntot) ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Friday, November 10, 2017 8:56:09 AM To: nek5000-users at lists.mcs.anl.gov Subject: Re: [Nek5000-users] Dissipation rate Dear Nek experts, I'm still working on the turbulent dissipation rate. I would like to compute epsilon=2*nu* with Sij=0.5*(grad(U)+grad(U)^T) I found in the forum a proposition to calculate the scalar dissipation rate as follow. If I understood well, this code is basically doing "chi := D * |grad Z|^2" while I would like to compute chi := D * D * |grad Z + grad^T Z|^2 Does anyone have a suggestion ? Many thanks, Samuel c----------------------------------------------------------------------- subroutine magSqr(a,b1,b2,b3,n) include 'SIZE' real a(1) real b1(1),b2(1),b3(1) do i=1,n a(i) = b1(i)*b1(i) + b2(i)*b2(i) enddo return end c----------------------------------------------------------------------- subroutine scalDisp(chi,Z,D) c c compute scalar dissipation rate c chi := D * |grad Z|^2 c include 'SIZE' include 'TOTAL' real chi(lx1,ly1,lz1,1) real Z (lx1,ly1,lz1,1) real D (lx1,ly1,lz1,1) common /scrns/ w1(lx1,ly1,lz1,lelt) $ ,w2(lx1,ly1,lz1,lelt) $ ,w3(lx1,ly1,lz1,lelt) ntot = nx1*ny1*nz1*nelv call opgrad (w1,w2,w3,Z) call opdssum(w1,w2,w3) call opcolv (w1,w2,w3,binvm1) call magsqr (chi,w1,w2,w3,ntot) call col2 (chi,D,ntot) return end c----------------------------------------------------------------------- Le 2017-11-08 10:55, Samuel Ahizi a ?crit : > Dear Neks, > > I would like to compute the turbulent dissipation rate, in order to > calculate the Kolmogorov microscales. Does a subroutine already exist > or should I compute it from userchk ? > > Thank you very much for your answers, > Samuel _______________________________________________ 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 Nov 13 09:39:12 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 13 Nov 2017 15:39:12 +0000 Subject: [Nek5000-users] Defining 2 different drichlet velocities in 2 different zones Message-ID: Hi all, I'm using a geometry where I need to define 2 different velocities at 2 inlets. How do I differentiate the zones in the .usr file where i set up my bcs? Whats the format for this? Sincerely, -- *Amitvikram Dutta* Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Nov 13 14:29:14 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 13 Nov 2017 20:29:14 +0000 Subject: [Nek5000-users] Defining 2 different drichlet velocities in 2 different zones In-Reply-To: References: Message-ID: Hi Amitvikram, One way to do it by geometry since userbc() has x,y,z info for each GLL point. Another way is during the mesh construction: use a specific boundary character codes say 'v01' and 'v02' instead of default 'v ' and then in usrdat2() loop over cbc array to switch them back to 'v ' remembering the local element numbers for a future use in userbc(). Aleks ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Monday, November 13, 2017 9:39:12 AM To: nek5000-users at lists.mcs.anl.gov Subject: [Nek5000-users] Defining 2 different drichlet velocities in 2 different zones Hi all, I'm using a geometry where I need to define 2 different velocities at 2 inlets. How do I differentiate the zones in the .usr file where i set up my bcs? Whats the format for this? Sincerely, -- Amitvikram Dutta Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Nov 13 20:51:00 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 13 Nov 2017 23:51:00 -0300 Subject: [Nek5000-users] Converting Mesh using exo2nek Message-ID: Dear Neks, The latest version of Nek doesn't suport MOAB for mesh convertion, so the only way to import a complex mesh (e.g. from Trelis/CUBIT) is using exo2nek. Am I right? Is there any example case to understand how does it work? The old examples to understand MOAB cases showed how to set the BC's using the sidesets in the rea file, but now I don't know how to set the BC's in the usr file, and I haven't found any guide or example for this. And, finally, if I'm using an old Nek version, (.rea instead of .par for parameters) would it work having the .rea and the .re2 file, for the parameters and for the mesh respectively? Thanks in advance, Juan Pablo. -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Nov 13 21:23:04 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Nov 2017 04:23:04 +0100 Subject: [Nek5000-users] Converting Mesh using exo2nek Message-ID: - Yes, this is correct - Check usrdat2() of short_test/ethier.usr - exo2nek stores the side set ID in bc(5,ifc,iel,1) - Yes! -----Original message----- > From:nek5000-users at lists.mcs.anl.gov > Sent: Tuesday 14th November 2017 3:52 > To: nek5000-users at lists.mcs.anl.gov > Subject: Re: [Nek5000-users] Converting Mesh using exo2nek > > Dear Neks,? > > The latest version of Nek doesnt suport MOAB for mesh convertion, so the only way to import a complex mesh (e.g. from Trelis/CUBIT) is using exo2nek. Am I right? > > Is there any example case to understand how does it work? The old examples to understand MOAB cases showed how to set the BCs using the sidesets in the rea file, but now I dont know how to set the BCs in the usr file, and I havent found any guide or example for this. > > And, finally, if Im using an old Nek version, (.rea instead of .par for parameters) would it work having the .rea and the .re2 file, for the parameters and for the mesh respectively? > > Thanks in advance, > > Juan Pablo. > _______________________________________________ > 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 Nov 13 14:56:39 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 13 Nov 2017 20:56:39 +0000 Subject: [Nek5000-users] n2to3 mesh size error Message-ID: Hi all, I keep running into an error whenever I want to use n2to3 to extrude my mesh from 2 dimensions to 3. My workflow is as follows: gmsh .geo file -> gmsh .msh file -> .rea file -> n2to3 > .rea (3D) file. Now if the original 2D .rea file obtained from gmsh contains less than 1000 elements, n2to3 fails with the error At line 651 of file n2to3.f (unit = 10, file = 'msh_test.rea') Fortran runtime error: Bad value during integer read However, should this not be the case (mesh size > 1000) its all smooth sailing. The trouble is with a mesh of that size my processor and memory requirements quickly blow up. Incidentally, genmap fails under almost exactly the same conditions. Is there a lower limit to mesh size issue for nek5000 that I'm missing here? Sincerely, -- *Amitvikram Dutta* Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Nov 13 21:52:37 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Nov 2017 04:52:37 +0100 Subject: [Nek5000-users] n2to3 mesh size error In-Reply-To: References: Message-ID: Looks like the boundary conditions in your .rea file are somehow messed up. How do you generate you .rea? -----Original message----- > From:nek5000-users at lists.mcs.anl.gov > Sent: Tuesday 14th November 2017 4:47 > To: nek5000-users at lists.mcs.anl.gov > Subject: [Nek5000-users] n2to3 mesh size error > > Hi all, > > I keep running into an error whenever I want to use n2to3 to extrude my mesh from 2 dimensions to 3. My workflow is as follows: > > gmsh .geo file -> gmsh .msh file -> .rea file -> n2to3 > .rea (3D) file. > > Now if the original 2D .rea file obtained from gmsh contains less than 1000 elements, n2to3 fails with the error > > At line 651 of file n2to3.f (unit = 10, file = msh_test.rea) > Fortran runtime error: Bad value during integer read > > However, should this not be the case (mesh size > 1000) its all smooth sailing. The trouble is with a mesh of that size my processor and memory requirements quickly blow up. > > Incidentally, genmap fails under almost exactly the same conditions. > > Is there a lower limit to mesh size issue for nek5000 that Im missing here? > > Sincerely, > -- > Amitvikram Dutta > Graduate Research Assistant > Fluid Mechanics Research Lab > University of Waterloo > _______________________________________________ > 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 Nov 13 22:40:16 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Nov 2017 04:40:16 +0000 Subject: [Nek5000-users] n2to3 mesh size error In-Reply-To: References: Message-ID: Gmsh to .rea using CPraveen's python mesh converter. On Mon, Nov 13, 2017, 10:52 PM wrote: > Looks like the boundary conditions in your .rea file are somehow messed > up. How do you generate you .rea? > > -----Original message----- > > From:nek5000-users at lists.mcs.anl.gov > > Sent: Tuesday 14th November 2017 4:47 > > To: nek5000-users at lists.mcs.anl.gov > > Subject: [Nek5000-users] n2to3 mesh size error > > > > Hi all, > > > > I keep running into an error whenever I want to use n2to3 to extrude my > mesh from 2 dimensions to 3. My workflow is as follows: > > > > gmsh .geo file -> gmsh .msh file -> .rea file -> n2to3 > .rea (3D) file. > > > > Now if the original 2D .rea file obtained from gmsh contains less than > 1000 elements, n2to3 fails with the error > > > > At line 651 of file n2to3.f (unit = 10, file = msh_test.rea) > > Fortran runtime error: Bad value during integer read > > > > However, should this not be the case (mesh size > 1000) its all smooth > sailing. The trouble is with a mesh of that size my processor and memory > requirements quickly blow up. > > > > Incidentally, genmap fails under almost exactly the same conditions. > > > > Is there a lower limit to mesh size issue for nek5000 that Im missing > here? > > > > Sincerely, > > -- > > Amitvikram Dutta > > Graduate Research Assistant > > Fluid Mechanics Research Lab > > University of Waterloo > > _______________________________________________ > > 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 > -- *Amitvikram Dutta* Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Nov 14 07:13:21 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Nov 2017 13:13:21 +0000 Subject: [Nek5000-users] n2to3 mesh size error In-Reply-To: References: Message-ID: What is the number of elements in your mesh? On Nov 13, 2017, at 11:40 PM, nek5000-users at lists.mcs.anl.gov wrote: Gmsh to .rea using CPraveen's python mesh converter. On Mon, Nov 13, 2017, 10:52 PM > wrote: Looks like the boundary conditions in your .rea file are somehow messed up. How do you generate you .rea? -----Original message----- > From:nek5000-users at lists.mcs.anl.gov > > Sent: Tuesday 14th November 2017 4:47 > To: nek5000-users at lists.mcs.anl.gov > Subject: [Nek5000-users] n2to3 mesh size error > > Hi all, > > I keep running into an error whenever I want to use n2to3 to extrude my mesh from 2 dimensions to 3. My workflow is as follows: > > gmsh .geo file -> gmsh .msh file -> .rea file -> n2to3 > .rea (3D) file. > > Now if the original 2D .rea file obtained from gmsh contains less than 1000 elements, n2to3 fails with the error > > At line 651 of file n2to3.f (unit = 10, file = msh_test.rea) > Fortran runtime error: Bad value during integer read > > However, should this not be the case (mesh size > 1000) its all smooth sailing. The trouble is with a mesh of that size my processor and memory requirements quickly blow up. > > Incidentally, genmap fails under almost exactly the same conditions. > > Is there a lower limit to mesh size issue for nek5000 that Im missing here? > > Sincerely, > -- > Amitvikram Dutta > Graduate Research Assistant > Fluid Mechanics Research Lab > University of Waterloo > _______________________________________________ > 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 -- Amitvikram Dutta Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo _______________________________________________ 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 Nov 14 08:02:47 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Nov 2017 14:02:47 +0000 Subject: [Nek5000-users] n2to3 mesh size error In-Reply-To: References: Message-ID: Its a 25X25 square grid in 2D. Im trying to extrude this with 10 elements in the z direction. On Tue, Nov 14, 2017 at 8:13 AM wrote: > What is the number of elements in your mesh? > > On Nov 13, 2017, at 11:40 PM, nek5000-users at lists.mcs.anl.gov wrote: > > Gmsh to .rea using CPraveen's python mesh converter. > > On Mon, Nov 13, 2017, 10:52 PM wrote: > >> Looks like the boundary conditions in your .rea file are somehow messed >> up. How do you generate you .rea? >> >> -----Original message----- >> > From:nek5000-users at lists.mcs.anl.gov >> > Sent: Tuesday 14th November 2017 4:47 >> > To: nek5000-users at lists.mcs.anl.gov >> > Subject: [Nek5000-users] n2to3 mesh size error >> > >> > Hi all, >> > >> > I keep running into an error whenever I want to use n2to3 to extrude my >> mesh from 2 dimensions to 3. My workflow is as follows: >> > >> > gmsh .geo file -> gmsh .msh file -> .rea file -> n2to3 > .rea (3D) file. >> > >> > Now if the original 2D .rea file obtained from gmsh contains less than >> 1000 elements, n2to3 fails with the error >> > >> > At line 651 of file n2to3.f (unit = 10, file = msh_test.rea) >> > Fortran runtime error: Bad value during integer read >> > >> > However, should this not be the case (mesh size > 1000) its all smooth >> sailing. The trouble is with a mesh of that size my processor and memory >> requirements quickly blow up. >> > >> > Incidentally, genmap fails under almost exactly the same conditions. >> > >> > Is there a lower limit to mesh size issue for nek5000 that Im missing >> here? >> > >> > Sincerely, >> > -- >> > Amitvikram Dutta >> > Graduate Research Assistant >> > Fluid Mechanics Research Lab >> > University of Waterloo >> > _______________________________________________ >> > 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 >> > -- > > *Amitvikram Dutta* > > Graduate Research Assistant > > Fluid Mechanics Research Lab > > University of Waterloo > _______________________________________________ > 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 > -- *Amitvikram Dutta* Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Nov 14 08:08:42 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Nov 2017 14:08:42 +0000 Subject: [Nek5000-users] n2to3 mesh size error In-Reply-To: References: Message-ID: So your mesh has more than 1000 elements. I think that the way the python script is written the boundary conditions in a rea file is not correct. I had this issue in the past. On Nov 14, 2017, at 9:02 AM, nek5000-users at lists.mcs.anl.gov wrote: Its a 25X25 square grid in 2D. Im trying to extrude this with 10 elements in the z direction. On Tue, Nov 14, 2017 at 8:13 AM > wrote: What is the number of elements in your mesh? On Nov 13, 2017, at 11:40 PM, nek5000-users at lists.mcs.anl.gov wrote: Gmsh to .rea using CPraveen's python mesh converter. On Mon, Nov 13, 2017, 10:52 PM > wrote: Looks like the boundary conditions in your .rea file are somehow messed up. How do you generate you .rea? -----Original message----- > From:nek5000-users at lists.mcs.anl.gov > > Sent: Tuesday 14th November 2017 4:47 > To: nek5000-users at lists.mcs.anl.gov > Subject: [Nek5000-users] n2to3 mesh size error > > Hi all, > > I keep running into an error whenever I want to use n2to3 to extrude my mesh from 2 dimensions to 3. My workflow is as follows: > > gmsh .geo file -> gmsh .msh file -> .rea file -> n2to3 > .rea (3D) file. > > Now if the original 2D .rea file obtained from gmsh contains less than 1000 elements, n2to3 fails with the error > > At line 651 of file n2to3.f (unit = 10, file = msh_test.rea) > Fortran runtime error: Bad value during integer read > > However, should this not be the case (mesh size > 1000) its all smooth sailing. The trouble is with a mesh of that size my processor and memory requirements quickly blow up. > > Incidentally, genmap fails under almost exactly the same conditions. > > Is there a lower limit to mesh size issue for nek5000 that Im missing here? > > Sincerely, > -- > Amitvikram Dutta > Graduate Research Assistant > Fluid Mechanics Research Lab > University of Waterloo > _______________________________________________ > 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 -- Amitvikram Dutta Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo _______________________________________________ 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 -- Amitvikram Dutta Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo _______________________________________________ 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 Nov 14 08:36:38 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Nov 2017 15:36:38 +0100 Subject: [Nek5000-users] n2to3 mesh size error In-Reply-To: References: Message-ID: Note, this converter is not part of Nek5000 and not support by us. Yes this converter does not output the correct BC.? On 14 Nov 2017, at 08:13, "nek5000-users at lists.mcs.anl.gov " > wrote: So your mesh has more than 1000 elements. I think that the way the python script is written the boundary conditions in a rea file is not correct. I had this issue in the past. On Nov 14, 2017, at 9:02 AM, nek5000-users at lists.mcs.anl.gov wrote: Its a 25X25 square grid in 2D. Im trying to extrude this with 10 elements in the z direction. On Tue, Nov 14, 2017 at 8:13 AM > wrote: What is the number of elements in your mesh? On Nov 13, 2017, at 11:40 PM, nek5000-users at lists.mcs.anl.gov wrote: Gmsh to .rea using CPraveen's python mesh converter. On Mon, Nov 13, 2017, 10:52 PM > wrote: Looks like the boundary conditions in your .rea file are somehow messed up. How do you generate you .rea? -----Original message----- > From:nek5000-users at lists.mcs.anl.gov > > Sent: Tuesday 14th November 2017 4:47 > To: nek5000-users at lists.mcs.anl.gov > Subject: [Nek5000-users] n2to3 mesh size error > > Hi all, > > I keep running into an error whenever I want to use n2to3 to extrude my mesh from 2 dimensions to 3. My workflow is as follows: > > gmsh .geo file -> gmsh .msh file -> .rea file -> n2to3 > .rea (3D) file. > > Now if the original 2D .rea file obtained from gmsh contains less than 1000 elements, n2to3 fails with the error > > At line 651 of file n2to3.f (unit = 10, file = msh_test.rea) > Fortran runtime error: Bad value during integer read > > However, should this not be the case (mesh size > 1000) its all smooth sailing. The trouble is with a mesh of that size my processor and memory requirements quickly blow up. > > Incidentally, genmap fails under almost exactly the same conditions. > > Is there a lower limit to mesh size issue for nek5000 that Im missing here? > > Sincerely, > -- > Amitvikram Dutta > Graduate Research Assistant > Fluid Mechanics Research Lab > University of Waterloo > _______________________________________________ > 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 -- Amitvikram Dutta Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo _______________________________________________ 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 -- Amitvikram Dutta Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo _______________________________________________ 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 Tue Nov 14 08:38:58 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Nov 2017 14:38:58 +0000 Subject: [Nek5000-users] n2to3 mesh size error In-Reply-To: References: Message-ID: can you generate the same mesh in a exodus format? If you can, use the exo2nek script to convert the mesh in a re2 file. Marco On Nov 14, 2017, at 9:36 AM, nek5000-users at lists.mcs.anl.gov wrote: Note, this converter is not part of Nek5000 and not support by us. Yes this converter does not output the correct BC. On 14 Nov 2017, at 08:13, "nek5000-users at lists.mcs.anl.gov" > wrote: So your mesh has more than 1000 elements. I think that the way the python script is written the boundary conditions in a rea file is not correct. I had this issue in the past. On Nov 14, 2017, at 9:02 AM, nek5000-users at lists.mcs.anl.gov wrote: Its a 25X25 square grid in 2D. Im trying to extrude this with 10 elements in the z direction. On Tue, Nov 14, 2017 at 8:13 AM > wrote: What is the number of elements in your mesh? On Nov 13, 2017, at 11:40 PM, nek5000-users at lists.mcs.anl.gov wrote: Gmsh to .rea using CPraveen's python mesh converter. On Mon, Nov 13, 2017, 10:52 PM > wrote: Looks like the boundary conditions in your .rea file are somehow messed up. How do you generate you .rea? -----Original message----- > From:nek5000-users at lists.mcs.anl.gov > > Sent: Tuesday 14th November 2017 4:47 > To: nek5000-users at lists.mcs.anl.gov > Subject: [Nek5000-users] n2to3 mesh size error > > Hi all, > > I keep running into an error whenever I want to use n2to3 to extrude my mesh from 2 dimensions to 3. My workflow is as follows: > > gmsh .geo file -> gmsh .msh file -> .rea file -> n2to3 > .rea (3D) file. > > Now if the original 2D .rea file obtained from gmsh contains less than 1000 elements, n2to3 fails with the error > > At line 651 of file n2to3.f (unit = 10, file = msh_test.rea) > Fortran runtime error: Bad value during integer read > > However, should this not be the case (mesh size > 1000) its all smooth sailing. The trouble is with a mesh of that size my processor and memory requirements quickly blow up. > > Incidentally, genmap fails under almost exactly the same conditions. > > Is there a lower limit to mesh size issue for nek5000 that Im missing here? > > Sincerely, > -- > Amitvikram Dutta > Graduate Research Assistant > Fluid Mechanics Research Lab > University of Waterloo > _______________________________________________ > 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 -- Amitvikram Dutta Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo _______________________________________________ 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 -- Amitvikram Dutta Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo _______________________________________________ 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 Tue Nov 14 09:15:20 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Nov 2017 15:15:20 +0000 Subject: [Nek5000-users] n2to3 mesh size error In-Reply-To: References: Message-ID: Does exo2nek support 3D conversion? Also, someone should consider updating the README for the exo2nek tools. Its not quite clear for someone who is starting out with nek5000, what exactly sideset ids are and how fluidic bcs are actually assigned. I will however give it a go. Thanks for your help everyone. On Tue, Nov 14, 2017 at 9:39 AM wrote: > can you generate the same mesh in a exodus format? If you can, use the > exo2nek script to convert the mesh in a re2 file. > > Marco > > On Nov 14, 2017, at 9:36 AM, nek5000-users at lists.mcs.anl.gov wrote: > > Note, this converter is not part of Nek5000 and not support by us. Yes > this converter does not output the correct BC. > > On 14 Nov 2017, at 08:13, "nek5000-users at lists.mcs.anl.gov" < > nek5000-users at lists.mcs.anl.gov> wrote: > > So your mesh has more than 1000 elements. I think that the way the python > script is written the boundary conditions in a rea file is not correct. I > had this issue in the past. > > On Nov 14, 2017, at 9:02 AM, nek5000-users at lists.mcs.anl.gov wrote: > > Its a 25X25 square grid in 2D. Im trying to extrude this with 10 elements > in the z direction. > > On Tue, Nov 14, 2017 at 8:13 AM wrote: > >> What is the number of elements in your mesh? >> >> On Nov 13, 2017, at 11:40 PM, nek5000-users at lists.mcs.anl.gov wrote: >> >> Gmsh to .rea using CPraveen's python mesh converter. >> >> On Mon, Nov 13, 2017, 10:52 PM wrote: >> >>> Looks like the boundary conditions in your .rea file are somehow messed >>> up. How do you generate you .rea? >>> >>> -----Original message----- >>> > From:nek5000-users at lists.mcs.anl.gov >>> > Sent: Tuesday 14th November 2017 4:47 >>> > To: nek5000-users at lists.mcs.anl.gov >>> > Subject: [Nek5000-users] n2to3 mesh size error >>> > >>> > Hi all, >>> > >>> > I keep running into an error whenever I want to use n2to3 to extrude >>> my mesh from 2 dimensions to 3. My workflow is as follows: >>> > >>> > gmsh .geo file -> gmsh .msh file -> .rea file -> n2to3 > .rea (3D) >>> file. >>> > >>> > Now if the original 2D .rea file obtained from gmsh contains less than >>> 1000 elements, n2to3 fails with the error >>> > >>> > At line 651 of file n2to3.f (unit = 10, file = msh_test.rea) >>> > Fortran runtime error: Bad value during integer read >>> > >>> > However, should this not be the case (mesh size > 1000) its all smooth >>> sailing. The trouble is with a mesh of that size my processor and memory >>> requirements quickly blow up. >>> > >>> > Incidentally, genmap fails under almost exactly the same conditions. >>> > >>> > Is there a lower limit to mesh size issue for nek5000 that Im missing >>> here? >>> > >>> > Sincerely, >>> > -- >>> > Amitvikram Dutta >>> > Graduate Research Assistant >>> > Fluid Mechanics Research Lab >>> > University of Waterloo >>> > _______________________________________________ >>> > 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 >>> >> -- >> >> *Amitvikram Dutta* >> >> Graduate Research Assistant >> >> Fluid Mechanics Research Lab >> >> University of Waterloo >> _______________________________________________ >> 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 >> > -- > > *Amitvikram Dutta* > > Graduate Research Assistant > > Fluid Mechanics Research Lab > > University of Waterloo > _______________________________________________ > 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 > -- *Amitvikram Dutta* Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Nov 14 09:45:14 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Nov 2017 12:45:14 -0300 Subject: [Nek5000-users] Converting Mesh using exo2nek Message-ID: Thanks for the answer. One last question: userdat2 looks like: do iel=1,nelt do ifc=1,2*ndim idss = bc(5,ifc,iel,1) if (idss.eq.1) cbc(ifc,iel,1)='v ' enddo enddo the last value "1" in bc array, and cbc array refers to velocity field, right? So for temperature BC I should do the same but specifying "2"? e.g do iel=1,nelt do ifc=1,2*ndim idss = bc(5,ifc,iel,2) if (idss.eq.1) cbc(ifc,iel,2)='t ' enddo enddo JP. -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Nov 14 10:27:13 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Nov 2017 16:27:13 +0000 Subject: [Nek5000-users] n2to3 mesh size error In-Reply-To: References: Message-ID: It works for 2D and 3D meshes. Sideset ids should be set in the package you use to create the mesh. Then in the subroutine usrdat (inside the *.usr file), you will have to loop over the faces of each element and set the boundary conditions: do i=1,nelt do j=1,2*ndim !write(6,*) bc(5,j,i,1) if(bc(5,j,i,1).eq.1) then cbc(j,i,1)='mv ' !write(6,*) 'mv bc' elseif(bc(5,j,i,1).eq.2) then cbc(j,i,1)='W ' elseif(bc(5,j,i,1).eq.3) then cbc(j,i,1)='W ' elseif(bc(5,j,i,1).eq.4) then cbc(j,i,1)='W ' elseif(bc(5,j,i,1).eq.5) then cbc(j,i,1)='W ' elseif(bc(5,j,i,1).eq.6) then cbc(j,i,1)='W ' elseif(bc(5,j,i,1).eq.7) then cbc(j,i,1)='W ' elseif(bc(5,j,i,1).eq.8) then cbc(j,i,1)='W ' elseif(bc(5,j,i,1).eq.9) then cbc(j,i,1)='W ' else cbc(j,i,1)='E ' endif enddo enddo The installation of the exo2nek script is not that complicated if you follow the instructions in the README file. Marco On Nov 14, 2017, at 10:15 AM, nek5000-users at lists.mcs.anl.gov wrote: Does exo2nek support 3D conversion? Also, someone should consider updating the README for the exo2nek tools. Its not quite clear for someone who is starting out with nek5000, what exactly sideset ids are and how fluidic bcs are actually assigned. I will however give it a go. Thanks for your help everyone. On Tue, Nov 14, 2017 at 9:39 AM > wrote: can you generate the same mesh in a exodus format? If you can, use the exo2nek script to convert the mesh in a re2 file. Marco On Nov 14, 2017, at 9:36 AM, nek5000-users at lists.mcs.anl.gov wrote: Note, this converter is not part of Nek5000 and not support by us. Yes this converter does not output the correct BC. On 14 Nov 2017, at 08:13, "nek5000-users at lists.mcs.anl.gov" > wrote: So your mesh has more than 1000 elements. I think that the way the python script is written the boundary conditions in a rea file is not correct. I had this issue in the past. On Nov 14, 2017, at 9:02 AM, nek5000-users at lists.mcs.anl.gov wrote: Its a 25X25 square grid in 2D. Im trying to extrude this with 10 elements in the z direction. On Tue, Nov 14, 2017 at 8:13 AM > wrote: What is the number of elements in your mesh? On Nov 13, 2017, at 11:40 PM, nek5000-users at lists.mcs.anl.gov wrote: Gmsh to .rea using CPraveen's python mesh converter. On Mon, Nov 13, 2017, 10:52 PM > wrote: Looks like the boundary conditions in your .rea file are somehow messed up. How do you generate you .rea? -----Original message----- > From:nek5000-users at lists.mcs.anl.gov > > Sent: Tuesday 14th November 2017 4:47 > To: nek5000-users at lists.mcs.anl.gov > Subject: [Nek5000-users] n2to3 mesh size error > > Hi all, > > I keep running into an error whenever I want to use n2to3 to extrude my mesh from 2 dimensions to 3. My workflow is as follows: > > gmsh .geo file -> gmsh .msh file -> .rea file -> n2to3 > .rea (3D) file. > > Now if the original 2D .rea file obtained from gmsh contains less than 1000 elements, n2to3 fails with the error > > At line 651 of file n2to3.f (unit = 10, file = msh_test.rea) > Fortran runtime error: Bad value during integer read > > However, should this not be the case (mesh size > 1000) its all smooth sailing. The trouble is with a mesh of that size my processor and memory requirements quickly blow up. > > Incidentally, genmap fails under almost exactly the same conditions. > > Is there a lower limit to mesh size issue for nek5000 that Im missing here? > > Sincerely, > -- > Amitvikram Dutta > Graduate Research Assistant > Fluid Mechanics Research Lab > University of Waterloo > _______________________________________________ > 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 -- Amitvikram Dutta Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo _______________________________________________ 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 -- Amitvikram Dutta Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo _______________________________________________ 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 -- Amitvikram Dutta Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo _______________________________________________ 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 Nov 14 10:30:05 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Nov 2017 16:30:05 +0000 Subject: [Nek5000-users] n2to3 mesh size error In-Reply-To: References: Message-ID: Thanks a lot Marco! On Tue, Nov 14, 2017 at 11:27 AM wrote: > It works for 2D and 3D meshes. Sideset ids should be set in the package > you use to create the mesh. Then in the subroutine usrdat (inside the *.usr > file), you will have to loop over the faces of each element and set the > boundary conditions: > > do i=1,nelt > do j=1,2*ndim > !write(6,*) bc(5,j,i,1) > if(bc(5,j,i,1).eq.1) then > cbc(j,i,1)='mv ' > !write(6,*) 'mv bc' > elseif(bc(5,j,i,1).eq.2) then > cbc(j,i,1)='W ' > elseif(bc(5,j,i,1).eq.3) then > cbc(j,i,1)='W ' > elseif(bc(5,j,i,1).eq.4) then > cbc(j,i,1)='W ' > elseif(bc(5,j,i,1).eq.5) then > cbc(j,i,1)='W ' > elseif(bc(5,j,i,1).eq.6) then > cbc(j,i,1)='W ' > elseif(bc(5,j,i,1).eq.7) then > cbc(j,i,1)='W ' > elseif(bc(5,j,i,1).eq.8) then > cbc(j,i,1)='W ' > elseif(bc(5,j,i,1).eq.9) then > cbc(j,i,1)='W ' > else > cbc(j,i,1)='E ' > endif > enddo > enddo > > The installation of the exo2nek script is not that complicated if you > follow the instructions in the README file. > > Marco > > > On Nov 14, 2017, at 10:15 AM, nek5000-users at lists.mcs.anl.gov wrote: > > Does exo2nek support 3D conversion? Also, someone should consider updating > the README for the exo2nek tools. Its not quite clear for someone who is > starting out with nek5000, what exactly sideset ids are and how fluidic bcs > are actually assigned. I will however give it a go. > > Thanks for your help everyone. > > On Tue, Nov 14, 2017 at 9:39 AM wrote: > >> can you generate the same mesh in a exodus format? If you can, use the >> exo2nek script to convert the mesh in a re2 file. >> >> Marco >> >> On Nov 14, 2017, at 9:36 AM, nek5000-users at lists.mcs.anl.gov wrote: >> >> Note, this converter is not part of Nek5000 and not support by us. Yes >> this converter does not output the correct BC. >> >> On 14 Nov 2017, at 08:13, "nek5000-users at lists.mcs.anl.gov" < >> nek5000-users at lists.mcs.anl.gov> wrote: >> >> So your mesh has more than 1000 elements. I think that the way the python >> script is written the boundary conditions in a rea file is not correct. I >> had this issue in the past. >> >> On Nov 14, 2017, at 9:02 AM, nek5000-users at lists.mcs.anl.gov wrote: >> >> Its a 25X25 square grid in 2D. Im trying to extrude this with 10 elements >> in the z direction. >> >> On Tue, Nov 14, 2017 at 8:13 AM wrote: >> >>> What is the number of elements in your mesh? >>> >>> On Nov 13, 2017, at 11:40 PM, nek5000-users at lists.mcs.anl.gov wrote: >>> >>> Gmsh to .rea using CPraveen's python mesh converter. >>> >>> On Mon, Nov 13, 2017, 10:52 PM wrote: >>> >>>> Looks like the boundary conditions in your .rea file are somehow messed >>>> up. How do you generate you .rea? >>>> >>>> -----Original message----- >>>> > From:nek5000-users at lists.mcs.anl.gov >>> > >>>> > Sent: Tuesday 14th November 2017 4:47 >>>> > To: nek5000-users at lists.mcs.anl.gov >>>> > Subject: [Nek5000-users] n2to3 mesh size error >>>> > >>>> > Hi all, >>>> > >>>> > I keep running into an error whenever I want to use n2to3 to extrude >>>> my mesh from 2 dimensions to 3. My workflow is as follows: >>>> > >>>> > gmsh .geo file -> gmsh .msh file -> .rea file -> n2to3 > .rea (3D) >>>> file. >>>> > >>>> > Now if the original 2D .rea file obtained from gmsh contains less >>>> than 1000 elements, n2to3 fails with the error >>>> > >>>> > At line 651 of file n2to3.f (unit = 10, file = msh_test.rea) >>>> > Fortran runtime error: Bad value during integer read >>>> > >>>> > However, should this not be the case (mesh size > 1000) its all >>>> smooth sailing. The trouble is with a mesh of that size my processor and >>>> memory requirements quickly blow up. >>>> > >>>> > Incidentally, genmap fails under almost exactly the same conditions. >>>> > >>>> > Is there a lower limit to mesh size issue for nek5000 that Im missing >>>> here? >>>> > >>>> > Sincerely, >>>> > -- >>>> > Amitvikram Dutta >>>> > Graduate Research Assistant >>>> > Fluid Mechanics Research Lab >>>> > University of Waterloo >>>> > _______________________________________________ >>>> > 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 >>>> >>> -- >>> >>> *Amitvikram Dutta* >>> >>> Graduate Research Assistant >>> >>> Fluid Mechanics Research Lab >>> >>> University of Waterloo >>> _______________________________________________ >>> 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 >>> >> -- >> >> *Amitvikram Dutta* >> >> Graduate Research Assistant >> >> Fluid Mechanics Research Lab >> >> University of Waterloo >> _______________________________________________ >> 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 >> > -- > > *Amitvikram Dutta* > > Graduate Research Assistant > > Fluid Mechanics Research Lab > > University of Waterloo > _______________________________________________ > 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 > -- *Amitvikram Dutta* Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Nov 14 11:13:51 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 14 Nov 2017 18:13:51 +0100 Subject: [Nek5000-users] fix projection for multiple passive scalars Message-ID: Hi Neks, thanks for dealing with the projection of passive scalars. I saw that there were 2 commits enabling projection for passive scalars (586e675926e14ca89982adf7dfc3a4c682c6d8c7) and disabling them again (71573b1904e27a9fa15bff13752292545d3d2943) because of too much memory consumption. If I understand the code correctly, there is no simple flag to enable projection for passive scalars in the rea file. Can you tell what needs to be changed in the code in order to enable them? Best, Steffen Straub Message: 5 Date: Thu, 19 Oct 2017 14:13:13 +0000 From:nek5000-users at lists.mcs.anl.gov To:"nek5000-users at lists.mcs.anl.gov" Subject: Re: [Nek5000-users] fix projection for multiple passive scalars Message-ID: Content-Type: text/plain; charset="windows-1252" Yes, this is a bug. Sorry - Because of another issue (now resolved), I hadn't been attending to this part of projection yet (since it was essentially broken before). I'll try to get this fixed asap. Paul ________________________________ From: Nek5000-users on behalf ofnek5000-users at lists.mcs.anl.gov Sent: Thursday, October 19, 2017 3:50:23 AM To:nek5000-users at lists.mcs.anl.gov Subject: [Nek5000-users] fix projection for multiple passive scalars Hi Neks, in order to use projection (p94!=0) for one temperature field and at least one additional passive scalar, I believe that the variables napproxt and approxt need to be extended by the size of ldimt. See the attached git-diff.txt for my changes to the code. When I adjust the code like that, projection for veloctity, temperature and all passive scalars works, otherwise it fails with: ... 22 Error Hmholtz TEMP 100 NaN NaN 1.0000E-08 22proj_ortho: 1 2 TEMP Detect rank deficiency: NaN NaN 22 Project PS 1 NaN NaN NaN 1 0 22 Error Hmholtz PS 1 100 NaN NaN 1.0000E-08 22proj_ortho: 1 2 PS 1 Detect rank deficiency: NaN NaN 22 Project PS 2 NaN NaN NaN 1 0 22 Error Hmholtz PS 2 100 NaN NaN 1.0000E-08 22proj_ortho: 1 2 PS 2 Detect rank deficiency: NaN NaN ... Can one of the developers please confirm and if confirmed include it in the master branch? Best, Steffen Straub -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Nov 15 02:21:36 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 15 Nov 2017 09:21:36 +0100 Subject: [Nek5000-users] Dissipation rate In-Reply-To: References: <07d9c63f18562f417c187df4ce221889@vki.ac.be> <662f888a2aaf0ff2f3a4b3f4ebea4036@vki.ac.be> Message-ID: Thank you Paul, your answer was really helpful Le 2017-11-10 15:56, Samuel Ahizi a ?crit?: > Dear Nek experts, > > I'm still working on the turbulent dissipation rate. > I would like to compute epsilon=2*nu* with > Sij=0.5*(grad(U)+grad(U)^T) > > I found in the forum a proposition to calculate the scalar dissipation > rate as follow. > > If I understood well, this code is basically doing "chi := D * |grad > Z|^2" while I would like to compute chi := D * D * |grad Z + grad^T > Z|^2 > > Does anyone have a suggestion ? > > Many thanks, > Samuel > > c----------------------------------------------------------------------- > subroutine magSqr(a,b1,b2,b3,n) > include 'SIZE' > real a(1) > real b1(1),b2(1),b3(1) > > do i=1,n > a(i) = b1(i)*b1(i) + b2(i)*b2(i) > enddo > > return > end > c----------------------------------------------------------------------- > subroutine scalDisp(chi,Z,D) > c > c compute scalar dissipation rate > c chi := D * |grad Z|^2 > c > include 'SIZE' > include 'TOTAL' > > real chi(lx1,ly1,lz1,1) > real Z (lx1,ly1,lz1,1) > real D (lx1,ly1,lz1,1) > > common /scrns/ w1(lx1,ly1,lz1,lelt) > $ ,w2(lx1,ly1,lz1,lelt) > $ ,w3(lx1,ly1,lz1,lelt) > > ntot = nx1*ny1*nz1*nelv > > call opgrad (w1,w2,w3,Z) > call opdssum(w1,w2,w3) > call opcolv (w1,w2,w3,binvm1) > > call magsqr (chi,w1,w2,w3,ntot) > call col2 (chi,D,ntot) > > return > end > c----------------------------------------------------------------------- > > > Le 2017-11-08 10:55, Samuel Ahizi a ?crit?: >> Dear Neks, >> >> I would like to compute the turbulent dissipation rate, in order to >> calculate the Kolmogorov microscales. Does a subroutine already exist >> or should I compute it from userchk ? >> >> Thank you very much for your answers, >> Samuel From nek5000-users at lists.mcs.anl.gov Wed Nov 15 09:08:12 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 15 Nov 2017 16:08:12 +0100 (CET) Subject: [Nek5000-users] Problem with drag coefficient of ext_cyl solved Message-ID: Hi Ketan, Sorry for late replay. Infact, I have found the mistake. It was from my side. I called call object two times in usr check routine to calculate drag coefficient. so the results got double and tripled each time. anyway thanks for the replay. Now I am working on Forced motion of cylinder by changing some parameters in ocyl example. Thanks with regards Sijo George -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Nov 15 12:41:33 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 15 Nov 2017 18:41:33 +0000 Subject: [Nek5000-users] gen_rea for cylinder Message-ID: Hi, I am trying to generate a rea file with curved side data for a pipe (radius of 1 and length of 13) by calling gen_rea_full(2) in usrdat2. When I open newer.out, the curved side data block is empty even though my geometry has faces with curvature. I know I can generate the same geometry with genbox but this is for testing purpose. My questions: - is this the correct call for this function? - if it is, are there any limitations for this function? Thanks, Marco From nek5000-users at lists.mcs.anl.gov Wed Nov 15 13:14:37 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 15 Nov 2017 19:14:37 +0000 Subject: [Nek5000-users] gen_rea for cylinder In-Reply-To: References: Message-ID: Marco, do you mean you called gen_rea(2)? Aleks ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Wednesday, November 15, 2017 12:41:33 PM To: Nek5000-users at lists.mcs.anl.gov Subject: [Nek5000-users] gen_rea for cylinder Hi, I am trying to generate a rea file with curved side data for a pipe (radius of 1 and length of 13) by calling gen_rea_full(2) in usrdat2. When I open newer.out, the curved side data block is empty even though my geometry has faces with curvature. I know I can generate the same geometry with genbox but this is for testing purpose. My questions: - is this the correct call for this function? - if it is, are there any limitations for this function? Thanks, Marco _______________________________________________ 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 Wed Nov 15 13:21:17 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 15 Nov 2017 19:21:17 +0000 Subject: [Nek5000-users] gen_rea for cylinder In-Reply-To: References: Message-ID: I tried with both. As far as I understand, gen_rea(2) only writes the mesh and boundary information. gen_rea_full writes a full rea file. Does the mesh have to be imported in a exodus format with hex27 elements for this routine to work? On Nov 15, 2017, at 2:14 PM, nek5000-users at lists.mcs.anl.gov > wrote: Marco, do you mean you called gen_rea(2)? Aleks ________________________________ From: Nek5000-users > on behalf of nek5000-users at lists.mcs.anl.gov > Sent: Wednesday, November 15, 2017 12:41:33 PM To: Nek5000-users at lists.mcs.anl.gov Subject: [Nek5000-users] gen_rea for cylinder Hi, I am trying to generate a rea file with curved side data for a pipe (radius of 1 and length of 13) by calling gen_rea_full(2) in usrdat2. When I open newer.out, the curved side data block is empty even though my geometry has faces with curvature. I know I can generate the same geometry with genbox but this is for testing purpose. My questions: - is this the correct call for this function? - if it is, are there any limitations for this function? Thanks, Marco _______________________________________________ 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 Wed Nov 15 13:35:14 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 15 Nov 2017 19:35:14 +0000 Subject: [Nek5000-users] gen_rea for cylinder In-Reply-To: References: Message-ID: I answered my own question: the curved side data option of the gen_rea function only works when the mesh is saved in the exodus format with hex27 type element. Marco On Nov 15, 2017, at 2:21 PM, nek5000-users at lists.mcs.anl.gov wrote: I tried with both. As far as I understand, gen_rea(2) only writes the mesh and boundary information. gen_rea_full writes a full rea file. Does the mesh have to be imported in a exodus format with hex27 elements for this routine to work? On Nov 15, 2017, at 2:14 PM, nek5000-users at lists.mcs.anl.gov > wrote: Marco, do you mean you called gen_rea(2)? Aleks ________________________________ From: Nek5000-users > on behalf of nek5000-users at lists.mcs.anl.gov > Sent: Wednesday, November 15, 2017 12:41:33 PM To: Nek5000-users at lists.mcs.anl.gov Subject: [Nek5000-users] gen_rea for cylinder Hi, I am trying to generate a rea file with curved side data for a pipe (radius of 1 and length of 13) by calling gen_rea_full(2) in usrdat2. When I open newer.out, the curved side data block is empty even though my geometry has faces with curvature. I know I can generate the same geometry with genbox but this is for testing purpose. My questions: - is this the correct call for this function? - if it is, are there any limitations for this function? Thanks, Marco _______________________________________________ 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 Wed Nov 15 18:39:51 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 15 Nov 2017 21:39:51 -0300 Subject: [Nek5000-users] Undefined references when compiling Message-ID: Hi all, I've never had any problem when compiling (in my personal computer or in the cluster server where I run simulations). Now I'm trying to use the latest version of Nek, and it worked in my computer but when I try to compile a case in the cluster I get errors. Attached is the full compiler.out file, but the final part is: [...] /user/r/robinson/Nek5000/3rd_party/gslib/src/libgs.a(amg.o): In function `dread_mpi.isra.22': amg.c:(.text+0x53b): undefined reference to `ompi_mpi_double' /user/r/robinson/Nek5000/3rd_party/gslib/src/libgs.a(amg.o): In function `dopen_mpi.constprop.27': amg.c:(.text+0x5eb): undefined reference to `ompi_mpi_info_null' /user/r/robinson/Nek5000/3rd_party/gslib/src/libgs.a(amg.o): In function `gslib_crs_amg_setup': amg.c:(.text+0x18d2): undefined reference to `ompi_mpi_unsigned_char' amg.c:(.text+0x1cc1): undefined reference to `ompi_mpi_unsigned_char' amg.c:(.text+0x1cde): undefined reference to `ompi_mpi_unsigned_char' amg.c:(.text+0x1cfb): undefined reference to `ompi_mpi_unsigned_char' amg.c:(.text+0x1dcb): undefined reference to `ompi_mpi_unsigned_char' /user/r/robinson/Nek5000/3rd_party/gslib/src/libgs.a(amg.o):amg.c:(.text+0x1df1): more undefined references to `ompi_mpi_unsigned_char' follow collect2: ld returned 1 exit status make: *** [nek5000] Error 1 I notice that there are a lot of undefined references related to MPI. One differences is that in my computer I have openmpi but the cluster has mpich. Perhaps it is related with that, although when I use previous versions of Nek I have no problem in the cluster. I can notice too that all the errors come from Nek5000/3rd_party/gslib/src. I went into that folder and realized that maybe I should install gslib libraries, and that means asking for the cluster administrator to do it. What do you think I should do? Thanks in advance, JP. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: compiler.out Type: application/octet-stream Size: 32716 bytes Desc: not available URL: From nek5000-users at lists.mcs.anl.gov Wed Nov 15 19:14:45 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 16 Nov 2017 01:14:45 +0000 Subject: [Nek5000-users] Undefined references when compiling In-Reply-To: References: Message-ID: Hi JP, Perhaps makenek clean first ? Paul ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Wednesday, November 15, 2017 6:39:51 PM To: nek5000-users at lists.mcs.anl.gov Subject: [Nek5000-users] Undefined references when compiling Hi all, I've never had any problem when compiling (in my personal computer or in the cluster server where I run simulations). Now I'm trying to use the latest version of Nek, and it worked in my computer but when I try to compile a case in the cluster I get errors. Attached is the full compiler.out file, but the final part is: [...] /user/r/robinson/Nek5000/3rd_party/gslib/src/libgs.a(amg.o): In function `dread_mpi.isra.22': amg.c:(.text+0x53b): undefined reference to `ompi_mpi_double' /user/r/robinson/Nek5000/3rd_party/gslib/src/libgs.a(amg.o): In function `dopen_mpi.constprop.27': amg.c:(.text+0x5eb): undefined reference to `ompi_mpi_info_null' /user/r/robinson/Nek5000/3rd_party/gslib/src/libgs.a(amg.o): In function `gslib_crs_amg_setup': amg.c:(.text+0x18d2): undefined reference to `ompi_mpi_unsigned_char' amg.c:(.text+0x1cc1): undefined reference to `ompi_mpi_unsigned_char' amg.c:(.text+0x1cde): undefined reference to `ompi_mpi_unsigned_char' amg.c:(.text+0x1cfb): undefined reference to `ompi_mpi_unsigned_char' amg.c:(.text+0x1dcb): undefined reference to `ompi_mpi_unsigned_char' /user/r/robinson/Nek5000/3rd_party/gslib/src/libgs.a(amg.o):amg.c:(.text+0x1df1): more undefined references to `ompi_mpi_unsigned_char' follow collect2: ld returned 1 exit status make: *** [nek5000] Error 1 I notice that there are a lot of undefined references related to MPI. One differences is that in my computer I have openmpi but the cluster has mpich. Perhaps it is related with that, although when I use previous versions of Nek I have no problem in the cluster. I can notice too that all the errors come from Nek5000/3rd_party/gslib/src. I went into that folder and realized that maybe I should install gslib libraries, and that means asking for the cluster administrator to do it. What do you think I should do? Thanks in advance, JP. -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Nov 15 19:20:03 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 15 Nov 2017 22:20:03 -0300 Subject: [Nek5000-users] Undefined references when compiling Message-ID: Hi all again, I realized that there was a change in Nek5000/3rd_party/gslib a few hours ago, and now I have no problem. Thnaks. JP. -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Nov 15 19:34:39 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 16 Nov 2017 02:34:39 +0100 Subject: [Nek5000-users] Undefined references when compiling In-Reply-To: References: Message-ID: Unless you have a very good reason not to use the release tarball I don't recommend the latest source on GitHub. Cheers, Stefan -----Original message----- > From:nek5000-users at lists.mcs.anl.gov > Sent: Thursday 16th November 2017 2:21 > To: nek5000-users at lists.mcs.anl.gov > Subject: Re: [Nek5000-users] Undefined references when compiling > > Hi all again, > > I realized that there was a change in?Nek5000/3rd_party/gslib a few hours ago, and now I have no problem. > > Thnaks. > > JP. > _______________________________________________ > 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 Nov 16 14:19:27 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 16 Nov 2017 21:19:27 +0100 (CET) Subject: [Nek5000-users] Mesh motion(mesh deformation parameter) Message-ID: Hello Nek, I was trying to validate moving cylinder test cases (ocyl2.usr more specifically). I could validate it properly. Even though I have some doubts regarding the mesh motion. Mesh deformation more precisely. I have seen a function to send the deformation to far field apart from the near wall region (Laplace equation with h1 as blending coefficient).So in order to find h1 in the equation there is a parameter called delta (which is 2 in this case). I have also read a paper regarding this equation. In that paper its clearly states that it is the average thickness of the element which is close to the object. So my question is, is there anyway to calculate it automatically or easily to calculate the thickness of that particular elements from any input files? because if I change the mesh (suppose if I refine the mesh size everything will change right?) how can I find the new average thickness of the elements which is close to the object? or how can I measure the area or volume of the elements which is close to the wall. so could you please help me with this? Thanks with regards Sijo George -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Nov 20 06:46:12 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 20 Nov 2017 12:46:12 +0000 Subject: [Nek5000-users] libxsmm within Nek5000 Message-ID: Hi, Do you have any performance results of the libxsmm comparing with stand-alone "mxm" ? In the master branch of Nek5000, the libxsmm function "libxsmm_dgemm" is only called in file " mxm_wrapper.f" ... #ifdef XSMM call libxsmm_dgemm('N','N',n1,n3,n2,1.0,a,n1,b,n2,0.0,c,n1) #endif ... I could not find any optimization works such as dispatch and streaming updates in the kernels that were used in e.g. NekBox. What is the status about the implementation of libxsmm or other SIMD techniques within Nek5000 right now? Thanks. /Jing -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Nov 20 06:59:01 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 20 Nov 2017 13:59:01 +0100 Subject: [Nek5000-users] libxsmm within Nek5000 Message-ID: At the moment there are no tuned "high-level "operators like axhelm(). Also my experience is that libxsmm is faster for lx1>14 which is rarely used for production runs. Cheers, Stefan -----Original message----- > From:nek5000-users at lists.mcs.anl.gov > Sent: Monday 20th November 2017 13:49 > To: nek5000-users at lists.mcs.anl.gov > Subject: [Nek5000-users] libxsmm within Nek5000 > > Hi, > > Do you have any performance results of the libxsmm comparing with stand-alone "mxm" ? > > In the master branch of Nek5000, the libxsmm function "libxsmm_dgemm" is only called in file " > mxm_wrapper.f" > ... > #ifdef XSMM > ???????? call libxsmm_dgemm('N','N',n1,n3,n2,1.0,a,n1,b,n2,0.0,c,n1) > #endif > ... > > I could not find any optimization works such as dispatch and streaming updates in the kernels that were used in e.g. NekBox. > > What is the status about the implementation of libxsmm or?other?SIMD techniques within Nek5000 right now? > > Thanks. /Jing > _______________________________________________ > 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 Nov 20 09:06:48 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 20 Nov 2017 16:06:48 +0100 Subject: [Nek5000-users] libxsmm within Nek5000 In-Reply-To: References: Message-ID: Let me add some more comments about this: NekBox implements three tuned high-levels operators: - opgrad - interpolation - axhelm Note, in a typical production the time spend axhelm dominates (25-45% of the total runtime) where the other two are around 5% in total. From what I recall they managed to get a speedup of 1.5x for axhelm (lx1=8) compared to Nek5000 default implementation. This suggests that the overall speedup for lx1=8 is about 10-20% at most. Having this in mind I didn't explore it further because it doesn't buy you much. Cheers, Stefan -----Original message----- > From:nek5000-users at lists.mcs.anl.gov > Sent: Monday 20th November 2017 14:00 > To: nek5000-users at lists.mcs.anl.gov > Subject: Re: [Nek5000-users] libxsmm within Nek5000 > > At the moment there are no tuned "high-level "operators like axhelm(). Also my experience is that libxsmm is faster for lx1>14 which is rarely used for production runs. > > Cheers, > Stefan > > -----Original message----- > > From:nek5000-users at lists.mcs.anl.gov > > Sent: Monday 20th November 2017 13:49 > > To: nek5000-users at lists.mcs.anl.gov > > Subject: [Nek5000-users] libxsmm within Nek5000 > > > > Hi, > > > > Do you have any performance results of the libxsmm comparing with stand-alone "mxm" ? > > > > In the master branch of Nek5000, the libxsmm function "libxsmm_dgemm" is only called in file " > > mxm_wrapper.f" > > ... > > #ifdef XSMM > > ???????? call libxsmm_dgemm('N','N',n1,n3,n2,1.0,a,n1,b,n2,0.0,c,n1) > > #endif > > ... > > > > I could not find any optimization works such as dispatch and streaming updates in the kernels that were used in e.g. NekBox. > > > > What is the status about the implementation of libxsmm or?other?SIMD techniques within Nek5000 right now? > > > > Thanks. /Jing > > _______________________________________________ > > 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 Nov 20 17:20:39 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 20 Nov 2017 23:20:39 +0000 Subject: [Nek5000-users] Plane Jet - Geometry Message-ID: Hi all, I'm trying to use do a LES study of a plane jet exhausting into a domain in nek5000. I'd like to know which geometry in the examples folder is the closest to my needs. I suspect its the turbchannel one, but I'm not quite certain as to how to set the boundaries of the jet body as solid, within the domain. Sincerely, -- *Amitvikram Dutta* Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Nov 20 19:10:50 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 21 Nov 2017 01:10:50 +0000 Subject: [Nek5000-users] Error while compiling exo2nek Message-ID: Hi all, I keep getting an error while compiling exo2nek. Warning: Type mismatch in argument ?a? at (1); passed REAL(8) to CHARACTER(1) gcc -mcmodel=medium -c -DUNDERSCORE ../../core/byte.c gfortran -I/home/adutta/lib/seacas/include -mcmodel=medium -c -fdefault-real-8 ../../core/speclib.f gfortran -I/home/adutta/lib/seacas/include -mcmodel=medium -c -fdefault-real-8 mxm.f gfortran -I/home/adutta/lib/seacas/include -mcmodel=medium -fdefault-real-8 -o /home/adutta/Nek5000/bin/exo2nek exo2nek.o byte.o speclib.o mxm.o -L/home/adutta/lib/netcdf/lib -L/home/adutta/lib/seacas/lib -lnetcdf -lexodus -lexoIIv2for32 make[1]: Leaving directory '/home/adutta/Nek5000/tools/exo2nek' It comes up as a warning but exo2nek is probably not being built correctly as whenever I try to run it I get a *seg fault error,* I have made all the changes as directed by the README file and installed seacas and netcdf. The maketools file has been suitably modified as well. bin_nek_tools=`cd ../bin && pwd` # specify your compilers here F77="gfortran -I$HOME/lib/seacas/include" CC="gcc" # linking flags #USR_LFLAGS="-L$HOME/lib" USR_LFLAGS="-L$HOME/lib/netcdf/lib -L$HOME/lib/seacas/lib" Has anyone else run into this issue? Sincerely, -- *Amitvikram Dutta* Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Nov 21 01:07:10 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 21 Nov 2017 07:07:10 +0000 Subject: [Nek5000-users] Plane Jet - Geometry In-Reply-To: References: Message-ID: Hi Amitvikram, I'm guessing you can build this case with genbox, which allows you to assemble several boxes together, each having varying resolution. The boxes need to match in a conforming way, but this is not difficult. You presumably could build a 2D mesh first, then extrude with n2to3. Or you could build the 3D mesh directly with genbox. In the span wise direction you can have periodicity, symmetry, or walls. You can either use synthetic turbulence at an upstream inlet or you can use a recycling BC that effectively gives you periodic channel flow to generate incoming fully-developed turbulence. Once generated, you probably want to apply a mesh smoother before starting the simulations. We've found that this can improve accuracy and reduce time to solution. If you want assistance, feel free to email me off-list. hth, Paul ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Monday, November 20, 2017 5:20:39 PM To: nek5000-users at lists.mcs.anl.gov Subject: [Nek5000-users] Plane Jet - Geometry Hi all, I'm trying to use do a LES study of a plane jet exhausting into a domain in nek5000. I'd like to know which geometry in the examples folder is the closest to my needs. I suspect its the turbchannel one, but I'm not quite certain as to how to set the boundaries of the jet body as solid, within the domain. Sincerely, -- Amitvikram Dutta Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Nov 21 02:16:57 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 21 Nov 2017 09:16:57 +0100 Subject: [Nek5000-users] Error while compiling exo2nek In-Reply-To: References: Message-ID: Please update to the latest version on GitHub and try again (just run ./maketools exo2nek) -----Original message----- > From:nek5000-users at lists.mcs.anl.gov > Sent: Tuesday 21st November 2017 2:12 > To: nek5000-users at lists.mcs.anl.gov > Subject: [Nek5000-users] Error while compiling exo2nek > > Hi all, > > I keep getting an error while compiling exo2nek. > > Warning: Type mismatch in argument ?a? at (1); passed REAL(8) to CHARACTER(1) > gcc -mcmodel=medium? -c -DUNDERSCORE ../../core/byte.c > gfortran -I/home/adutta/lib/seacas/include -mcmodel=medium -c -fdefault-real-8 ../../core/speclib.f > gfortran -I/home/adutta/lib/seacas/include -mcmodel=medium -c -fdefault-real-8 mxm.f > gfortran -I/home/adutta/lib/seacas/include -mcmodel=medium -fdefault-real-8 -o /home/adutta/Nek5000/bin/exo2nek exo2nek.o byte.o speclib.o mxm.o? -L/home/adutta/lib/netcdf/lib -L/home/adutta/lib/seacas/lib -lnetcdf -lexodus -lexoIIv2for32 > make[1]: Leaving directory /home/adutta/Nek5000/tools/exo2nek > > It comes up as a warning but exo2nek is probably not being built correctly as whenever I try to run it I get a seg fault error, > > I have made all the changes as directed by the README file and installed seacas and netcdf. The maketools file has been suitably modified as well. > > bin_nek_tools=`cd ../bin && pwd` > > # specify your compilers here > F77="gfortran -I$HOME/lib/seacas/include" > CC="gcc" > > # linking flags > #USR_LFLAGS="-L$HOME/lib" > USR_LFLAGS="-L$HOME/lib/netcdf/lib -L$HOME/lib/seacas/lib" > > Has anyone else run into this issue? > > Sincerely, > > > > > -- > Amitvikram Dutta > Graduate Research Assistant > Fluid Mechanics Research Lab > University of Waterloo > _______________________________________________ > 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 Nov 20 21:39:38 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 21 Nov 2017 03:39:38 +0000 Subject: [Nek5000-users] Error while compiling exo2nek In-Reply-To: References: Message-ID: Can you try the exo2nek script on the attached exo mesh file and see if it works. It works on my machine. It has 2496 elements and 3 sidesets. Marco -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: cylinder.exo Type: application/octet-stream Size: 980572 bytes Desc: cylinder.exo URL: -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Nov 21 08:37:24 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 21 Nov 2017 14:37:24 +0000 Subject: [Nek5000-users] Error while compiling exo2nek In-Reply-To: References: Message-ID: Hi Marco, I can't get exo2nek to compile properly at all. After downloading the latest version I get the following error messages when I run sudo ./maketools exo2nek Make exo2nek... ---------------------- make[1]: Entering directory '/home/adutta/Nek5000/tools/exo2nek' make[1]: Warning: File 'makefile' has modification time 17453 s in the future env: ?./install?: Permission denied makefile:19: recipe for target 'lib' failed make[1]: *** [lib] Error 126 make[1]: Leaving directory '/home/adutta/Nek5000/tools/exo2nek' ERROR: exo2nek failed to compile! makefile:4: recipe for target 'all' failed make: *** [all] Error 1 On Tue, Nov 21, 2017 at 9:33 AM wrote: > Can you try the exo2nek script on the attached exo mesh file and see if it > works. It works on my machine. It has 2496 elements and 3 sidesets. > > Marco > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > -- *Amitvikram Dutta* Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Nov 21 08:45:10 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 21 Nov 2017 15:45:10 +0100 Subject: [Nek5000-users] Error while compiling exo2nek In-Reply-To: References: Message-ID: > env: ?./install?: Permission denied Looks like you cannot run the install. Can you check the permissions. -----Original message----- > From:nek5000-users at lists.mcs.anl.gov > Sent: Tuesday 21st November 2017 15:38 > To: nek5000-users at lists.mcs.anl.gov > Subject: Re: [Nek5000-users] Error while compiling exo2nek > > Hi Marco, > > I cant get exo2nek to compile properly at all. After downloading the latest version I get the following error messages when I run sudo ./maketools exo2nek > > Make exo2nek... > ---------------------- > make[1]: Entering directory /home/adutta/Nek5000/tools/exo2nek > make[1]: Warning: File makefile has modification time 17453 s in the future > env: ?./install?: Permission denied > makefile:19: recipe for target lib failed > make[1]: *** [lib] Error 126 > make[1]: Leaving directory /home/adutta/Nek5000/tools/exo2nek > > ERROR: exo2nek failed to compile! > makefile:4: recipe for target all failed > make: *** [all] Error 1 > > > > On Tue, Nov 21, 2017 at 9:33 AM > wrote: > Can you try the exo2nek script on the attached exo mesh file and see if it works. It works on my machine. It has 2496 elements and 3 sidesets. > > Marco > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > -- > Amitvikram Dutta > Graduate Research Assistant > Fluid Mechanics Research Lab > University of Waterloo > _______________________________________________ > 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 Nov 21 09:08:52 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 21 Nov 2017 15:08:52 +0000 Subject: [Nek5000-users] Error while compiling exo2nek In-Reply-To: References: Message-ID: Changed the permission of the install file and it works fine now. Also the mesh supplied by Marco has been converted sucessfully and a .re2 file has been generated. Thanks all! On Tue, Nov 21, 2017 at 9:45 AM wrote: > > > env: ?./install?: Permission denied > > Looks like you cannot run the install. Can you check the permissions. > > -----Original message----- > > From:nek5000-users at lists.mcs.anl.gov > > Sent: Tuesday 21st November 2017 15:38 > > To: nek5000-users at lists.mcs.anl.gov > > Subject: Re: [Nek5000-users] Error while compiling exo2nek > > > > Hi Marco, > > > > I cant get exo2nek to compile properly at all. After downloading the > latest version I get the following error messages when I run sudo > ./maketools exo2nek > > > > Make exo2nek... > > ---------------------- > > make[1]: Entering directory /home/adutta/Nek5000/tools/exo2nek > > make[1]: Warning: File makefile has modification time 17453 s in the > future > > env: ?./install?: Permission denied > > makefile:19: recipe for target lib failed > > make[1]: *** [lib] Error 126 > > make[1]: Leaving directory /home/adutta/Nek5000/tools/exo2nek > > > > ERROR: exo2nek failed to compile! > > makefile:4: recipe for target all failed > > make: *** [all] Error 1 > > > > > > > > On Tue, Nov 21, 2017 at 9:33 AM > wrote: > > Can you try the exo2nek script on the attached exo mesh file and see if > it works. It works on my machine. It has 2496 elements and 3 sidesets. > > > > Marco > > > > _______________________________________________ > > Nek5000-users mailing list > > Nek5000-users at lists.mcs.anl.gov > > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users < > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users> > > -- > > Amitvikram Dutta > > Graduate Research Assistant > > Fluid Mechanics Research Lab > > University of Waterloo > > _______________________________________________ > > 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 > -- *Amitvikram Dutta* Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Nov 21 09:26:51 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 21 Nov 2017 16:26:51 +0100 Subject: [Nek5000-users] Error while compiling exo2nek In-Reply-To: References: Message-ID: I guess the mistake was to run ./maketools exo2nek with sudo. -----Original message----- > From:nek5000-users at lists.mcs.anl.gov > Sent: Tuesday 21st November 2017 16:13 > To: nek5000-users at lists.mcs.anl.gov > Subject: Re: [Nek5000-users] Error while compiling exo2nek > > Changed the permission of the install file and it works fine now. > > Also the mesh supplied by Marco has been converted sucessfully and a .re2 file has been generated. > > Thanks all! > > On Tue, Nov 21, 2017 at 9:45 AM > wrote: > > > env: ?./install?: Permission denied > > Looks like you cannot run the install. Can you check the permissions. > > -----Original message----- > > From:nek5000-users at lists.mcs.anl.gov > > > Sent: Tuesday 21st November 2017 15:38 > > To: nek5000-users at lists.mcs.anl.gov > > Subject: Re: [Nek5000-users] Error while compiling exo2nek > > > > Hi Marco, > > > > I cant get exo2nek to compile properly at all. After downloading the latest version I get the following error messages when I run sudo ./maketools exo2nek > > > > Make exo2nek... > > ---------------------- > > make[1]: Entering directory /home/adutta/Nek5000/tools/exo2nek > > make[1]: Warning: File makefile has modification time 17453 s in the future > > env: ?./install?: Permission denied > > makefile:19: recipe for target lib failed > > make[1]: *** [lib] Error 126 > > make[1]: Leaving directory /home/adutta/Nek5000/tools/exo2nek > > > > ERROR: exo2nek failed to compile! > > makefile:4: recipe for target all failed > > make: *** [all] Error 1 > > > > > > > > On Tue, Nov 21, 2017 at 9:33 AM >> wrote: > > Can you try the exo2nek script on the attached exo mesh file and see if it works. It works on my machine. It has 2496 elements and 3 sidesets. > > > > Marco > > > > _______________________________________________ > > Nek5000-users mailing list > > Nek5000-users at lists.mcs.anl.gov > > > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > > > -- > > Amitvikram Dutta > > Graduate Research Assistant > > Fluid Mechanics Research Lab > > University of Waterloo > > _______________________________________________ > > 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 > -- > Amitvikram Dutta > Graduate Research Assistant > Fluid Mechanics Research Lab > University of Waterloo > _______________________________________________ > 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 Nov 21 09:40:48 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 21 Nov 2017 15:40:48 +0000 Subject: [Nek5000-users] Error while compiling exo2nek In-Reply-To: References: Message-ID: No, I ran it without sudo as well, and cchanged to sudo when I got the permission error. On an additional issue, I'm generating my meshes in the exodusII format by using PoinWise and then exporting the exodusII file. exo2nek fails with the following error message when I try to use it on these files. EXODUS: ERROR: Attempting to open the netcdf-4 file: 'pw_exo1.exo' . Either the netcdf library does not support netcdf-4 or there is a filesystem or some other issue ERROR: cannot open file pw_exo1.exo On checking the properties of the .exo file I found that its an HDF document and not a netcdf file. Since Marco's file converts without problems I'm sure the issue lies in how pointwise is exporting the exodusII files. Is exo2nek not a viable alternative for exodusII files then? On Tue, Nov 21, 2017 at 10:33 AM wrote: > I guess the mistake was to run ./maketools exo2nek with sudo. > > -----Original message----- > > From:nek5000-users at lists.mcs.anl.gov > > Sent: Tuesday 21st November 2017 16:13 > > To: nek5000-users at lists.mcs.anl.gov > > Subject: Re: [Nek5000-users] Error while compiling exo2nek > > > > Changed the permission of the install file and it works fine now. > > > > Also the mesh supplied by Marco has been converted sucessfully and a > .re2 file has been generated. > > > > Thanks all! > > > > On Tue, Nov 21, 2017 at 9:45 AM > wrote: > > > > > env: ?./install?: Permission denied > > > > Looks like you cannot run the install. Can you check the permissions. > > > > -----Original message----- > > > From:nek5000-users at lists.mcs.anl.gov From%3Anek5000-users at lists.mcs.anl.gov> > > > > Sent: Tuesday 21st November 2017 15:38 > > > To: nek5000-users at lists.mcs.anl.gov nek5000-users at lists.mcs.anl.gov> > > > Subject: Re: [Nek5000-users] Error while compiling exo2nek > > > > > > Hi Marco, > > > > > > I cant get exo2nek to compile properly at all. After downloading the > latest version I get the following error messages when I run sudo > ./maketools exo2nek > > > > > > Make exo2nek... > > > ---------------------- > > > make[1]: Entering directory /home/adutta/Nek5000/tools/exo2nek > > > make[1]: Warning: File makefile has modification time 17453 s in the > future > > > env: ?./install?: Permission denied > > > makefile:19: recipe for target lib failed > > > make[1]: *** [lib] Error 126 > > > make[1]: Leaving directory /home/adutta/Nek5000/tools/exo2nek > > > > > > ERROR: exo2nek failed to compile! > > > makefile:4: recipe for target all failed > > > make: *** [all] Error 1 > > > > > > > > > > > > On Tue, Nov 21, 2017 at 9:33 AM nek5000-users at lists.mcs.anl.gov >> > wrote: > > > Can you try the exo2nek script on the attached exo mesh file and see > if it works. It works on my machine. It has 2496 elements and 3 sidesets. > > > > > > Marco > > > > > > _______________________________________________ > > > Nek5000-users mailing list > > > Nek5000-users at lists.mcs.anl.gov Nek5000-users at lists.mcs.anl.gov> > > > > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users < > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users> < > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users < > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users>> > > > -- > > > Amitvikram Dutta > > > Graduate Research Assistant > > > Fluid Mechanics Research Lab > > > University of Waterloo > > > _______________________________________________ > > > Nek5000-users mailing list > > > Nek5000-users at lists.mcs.anl.gov Nek5000-users at lists.mcs.anl.gov> > > > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users < > 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 < > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users> > > -- > > Amitvikram Dutta > > Graduate Research Assistant > > Fluid Mechanics Research Lab > > University of Waterloo > > _______________________________________________ > > 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 > -- *Amitvikram Dutta* Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Nov 21 09:49:46 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 21 Nov 2017 16:49:46 +0100 Subject: [Nek5000-users] Error while compiling exo2nek In-Reply-To: References: Message-ID: Ok so how did you fix the permission issue? Is there something we need to change on our end? -----Original message----- > From:nek5000-users at lists.mcs.anl.gov > Sent: Tuesday 21st November 2017 16:44 > To: nek5000-users at lists.mcs.anl.gov > Subject: Re: [Nek5000-users] Error while compiling exo2nek > > No, I ran it without sudo as well, and cchanged to sudo when I got the permission error. > > On an additional issue, Im generating my meshes in the exodusII format by using PoinWise and then exporting the exodusII file. exo2nek fails with the following error message when I try to use it on these files. > > EXODUS: ERROR: Attempting to open the netcdf-4 file: > ??? pw_exo1.exo > ??? . Either the netcdf library does not support netcdf-4 or there is a filesystem or some other issue > ERROR: cannot open file pw_exo1.exo???????????????????? > > On checking the properties of the .exo file I found that its an HDF document and not a netcdf file. > Since Marcos file converts without problems Im sure the issue lies in how pointwise is exporting the exodusII files. > > Is exo2nek not a viable alternative for exodusII files then? > > On Tue, Nov 21, 2017 at 10:33 AM > wrote: > I guess the mistake was to run ./maketools exo2nek with sudo. > > -----Original message----- > > From:nek5000-users at lists.mcs.anl.gov > > > Sent: Tuesday 21st November 2017 16:13 > > To: nek5000-users at lists.mcs.anl.gov > > Subject: Re: [Nek5000-users] Error while compiling exo2nek > > > > Changed the permission of the install file and it works fine now. > > > > Also the mesh supplied by Marco has been converted sucessfully and a .re2 file has been generated. > > > > Thanks all! > > > > On Tue, Nov 21, 2017 at 9:45 AM >> wrote: > > > > > env: ?./install?: Permission denied > > > > Looks like you cannot run the install. Can you check the permissions. > > > > -----Original message----- > > > From:nek5000-users at lists.mcs.anl.gov > >> > > > Sent: Tuesday 21st November 2017 15:38 > > > To: nek5000-users at lists.mcs.anl.gov > > > > Subject: Re: [Nek5000-users] Error while compiling exo2nek > > > > > > Hi Marco, > > > > > > I cant get exo2nek to compile properly at all. After downloading the latest version I get the following error messages when I run sudo ./maketools exo2nek > > > > > > Make exo2nek... > > > ---------------------- > > > make[1]: Entering directory /home/adutta/Nek5000/tools/exo2nek > > > make[1]: Warning: File makefile has modification time 17453 s in the future > > > env: ?./install?: Permission denied > > > makefile:19: recipe for target lib failed > > > make[1]: *** [lib] Error 126 > > > make[1]: Leaving directory /home/adutta/Nek5000/tools/exo2nek > > > > > > ERROR: exo2nek failed to compile! > > > makefile:4: recipe for target all failed > > > make: *** [all] Error 1 > > > > > > > > > > > > On Tue, Nov 21, 2017 at 9:33 AM > >>> wrote: > > > Can you try the exo2nek script on the attached exo mesh file and see if it works. It works on my machine. It has 2496 elements and 3 sidesets. > > > > > > Marco > > > > > > _______________________________________________ > > > Nek5000-users mailing list > > > Nek5000-users at lists.mcs.anl.gov > >> > > > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > >> > > > -- > > > Amitvikram Dutta > > > Graduate Research Assistant > > > Fluid Mechanics Research Lab > > > University of Waterloo > > > _______________________________________________ > > > 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 > > > -- > > Amitvikram Dutta > > Graduate Research Assistant > > Fluid Mechanics Research Lab > > University of Waterloo > > _______________________________________________ > > 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 > -- > Amitvikram Dutta > Graduate Research Assistant > Fluid Mechanics Research Lab > University of Waterloo > _______________________________________________ > 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 Nov 21 09:52:42 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 21 Nov 2017 15:52:42 +0000 Subject: [Nek5000-users] Error while compiling exo2nek In-Reply-To: References: Message-ID: I fixed the permission issue by opening up the permissions menu on the install file and clicking "Allow running as a program" On Tue, Nov 21, 2017 at 10:49 AM wrote: > Ok so how did you fix the permission issue? Is there something we need to > change on our end? > > -----Original message----- > > From:nek5000-users at lists.mcs.anl.gov > > Sent: Tuesday 21st November 2017 16:44 > > To: nek5000-users at lists.mcs.anl.gov > > Subject: Re: [Nek5000-users] Error while compiling exo2nek > > > > No, I ran it without sudo as well, and cchanged to sudo when I got the > permission error. > > > > On an additional issue, Im generating my meshes in the exodusII format > by using PoinWise and then exporting the exodusII file. exo2nek fails with > the following error message when I try to use it on these files. > > > > EXODUS: ERROR: Attempting to open the netcdf-4 file: > > pw_exo1.exo > > . Either the netcdf library does not support netcdf-4 or there is a > filesystem or some other issue > > ERROR: cannot open file pw_exo1.exo > > > > On checking the properties of the .exo file I found that its an HDF > document and not a netcdf file. > > Since Marcos file converts without problems Im sure the issue lies in > how pointwise is exporting the exodusII files. > > > > Is exo2nek not a viable alternative for exodusII files then? > > > > On Tue, Nov 21, 2017 at 10:33 AM > wrote: > > I guess the mistake was to run ./maketools exo2nek with sudo. > > > > -----Original message----- > > > From:nek5000-users at lists.mcs.anl.gov From%3Anek5000-users at lists.mcs.anl.gov> > > > > Sent: Tuesday 21st November 2017 16:13 > > > To: nek5000-users at lists.mcs.anl.gov nek5000-users at lists.mcs.anl.gov> > > > Subject: Re: [Nek5000-users] Error while compiling exo2nek > > > > > > Changed the permission of the install file and it works fine now. > > > > > > Also the mesh supplied by Marco has been converted sucessfully and a > .re2 file has been generated. > > > > > > Thanks all! > > > > > > On Tue, Nov 21, 2017 at 9:45 AM nek5000-users at lists.mcs.anl.gov >> > wrote: > > > > > > > env: ?./install?: Permission denied > > > > > > Looks like you cannot run the install. Can you check the permissions. > > > > > > -----Original message----- > > > > From:nek5000-users at lists.mcs.anl.gov From%3Anek5000-users at lists.mcs.anl.gov> From%3Anek5000-users at lists.mcs.anl.gov From%253Anek5000-users at lists.mcs.anl.gov>> < > nek5000-users at lists.mcs.anl.gov > nek5000-users at lists.mcs.anl.gov>>> > > > > Sent: Tuesday 21st November 2017 15:38 > > > > To: nek5000-users at lists.mcs.anl.gov nek5000-users at lists.mcs.anl.gov> > > > > > Subject: Re: [Nek5000-users] Error while compiling exo2nek > > > > > > > > Hi Marco, > > > > > > > > I cant get exo2nek to compile properly at all. After downloading the > latest version I get the following error messages when I run sudo > ./maketools exo2nek > > > > > > > > Make exo2nek... > > > > ---------------------- > > > > make[1]: Entering directory /home/adutta/Nek5000/tools/exo2nek > > > > make[1]: Warning: File makefile has modification time 17453 s in the > future > > > > env: ?./install?: Permission denied > > > > makefile:19: recipe for target lib failed > > > > make[1]: *** [lib] Error 126 > > > > make[1]: Leaving directory /home/adutta/Nek5000/tools/exo2nek > > > > > > > > ERROR: exo2nek failed to compile! > > > > makefile:4: recipe for target all failed > > > > make: *** [all] Error 1 > > > > > > > > > > > > > > > > On Tue, Nov 21, 2017 at 9:33 AM nek5000-users at lists.mcs.anl.gov > > nek5000-users at lists.mcs.anl.gov> >>> wrote: > > > > Can you try the exo2nek script on the attached exo mesh file and see > if it works. It works on my machine. It has 2496 elements and 3 sidesets. > > > > > > > > Marco > > > > > > > > _______________________________________________ > > > > Nek5000-users mailing list > > > > Nek5000-users at lists.mcs.anl.gov Nek5000-users at lists.mcs.anl.gov> > Nek5000-users at lists.mcs.anl.gov > Nek5000-users at lists.mcs.anl.gov>>> > > > > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users < > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users> < > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users < > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users>> < > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users < > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users> < > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users < > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users>>> > > > > -- > > > > Amitvikram Dutta > > > > Graduate Research Assistant > > > > Fluid Mechanics Research Lab > > > > University of Waterloo > > > > _______________________________________________ > > > > Nek5000-users mailing list > > > > Nek5000-users at lists.mcs.anl.gov Nek5000-users at lists.mcs.anl.gov> > > > > > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users < > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users> < > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users < > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users>> > > > _______________________________________________ > > > Nek5000-users mailing list > > > Nek5000-users at lists.mcs.anl.gov Nek5000-users at lists.mcs.anl.gov> > > > > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users < > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users> < > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users < > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users>> > > > -- > > > Amitvikram Dutta > > > Graduate Research Assistant > > > Fluid Mechanics Research Lab > > > University of Waterloo > > > _______________________________________________ > > > Nek5000-users mailing list > > > Nek5000-users at lists.mcs.anl.gov Nek5000-users at lists.mcs.anl.gov> > > > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users < > 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 < > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users> > > -- > > Amitvikram Dutta > > Graduate Research Assistant > > Fluid Mechanics Research Lab > > University of Waterloo > > _______________________________________________ > > 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 > -- *Amitvikram Dutta* Graduate Research Assistant Fluid Mechanics Research Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Nov 21 21:25:10 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 22 Nov 2017 06:25:10 +0300 Subject: [Nek5000-users] =?utf-8?q?=D0=A1ollocation_of_i=2Cj=2Ck_in_each_s?= =?utf-8?q?pectral_element?= Message-ID: Hi, Neks! I am working with a complex geometry witch was built in gambit. And I want to change positions of some points in spectral elements. I've read in documentation that i,j,k,e in xm1(i,j,k,e), for example, are changed from 1 to nx1,ny1,nz1 and nelv respectively. But during my tests it's seemed that in different elements x,y,z axis not always correspond to the global coordinate system. Am I right? And how could ?I find out the?principle of collocation of i,j,k in each spectral element? ? ? ? ? ? ? ? Here there is a picture to show what I mean: https://drive.google.com/file/d/13mpVldmdHOIDZG3M3AsqZvwPffilS0q_/view?usp=sharing ? Best regards, Vlad -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Nov 22 03:31:11 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 22 Nov 2017 10:31:11 +0100 (CET) Subject: [Nek5000-users] previous value of drag coefficient Message-ID: Hi Neks, Since I am trying to implement Adams bashforth scheme on my FSI problem (for equation of motion), I would like to pass the lift coefficients of the current timestep and the just previous time step to the function where I implemented the Adams Bashforth module. But its seems a bit difficult to fetch the just previous value of dragy value since it calls a new value at every time step. I simply tried a swapping method. But it gets zero value. Could you please tell me whether is there any way to fetch the just previous value of a variable (from an array or something?) Thanks regards Sijo GEORGE -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Nov 22 04:05:45 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 22 Nov 2017 11:05:45 +0100 Subject: [Nek5000-users] previous value of drag coefficient In-Reply-To: References: Message-ID: Hi, I suggest to use lag arrays as done for the BDF scheme, that is to have an array of size #multisteps and at each new time step to copy the old RHS (as you do AB) one step backwards. This is then somewhat consistent with Nek. Just be careful during the startup. Hth, Philipp On November 22, 2017 10:31:11 AM GMT+01:00, nek5000-users at lists.mcs.anl.gov wrote: >Hi Neks, > >Since I am trying to implement Adams bashforth scheme on my FSI problem >(for equation of motion), I would like to pass the lift coefficients of >the current timestep and the just previous time step to the function >where I implemented the Adams Bashforth module. But its seems a bit >difficult to fetch the just previous value of dragy value since it >calls a new value at every time step. I simply tried a swapping method. >But it gets zero value. Could you please tell me whether is there any >way to fetch the just previous value of a variable (from an array or >something?) > > >Thanks regards > >Sijo GEORGE -- Sent from my Android device with K-9 Mail. Please excuse my brevity. From nek5000-users at lists.mcs.anl.gov Wed Nov 22 04:18:04 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 22 Nov 2017 11:18:04 +0100 (CET) Subject: [Nek5000-users] previous value of drag coefficient In-Reply-To: References: Message-ID: Thank you Philip. I am sorry I just started to use NEK5000 and in Fortran too. I could not find an example of the concept of lag array. I know its a silly question. But could you please help me how to initialize such an array. Thanks again Sijo GEORGE ----- Mail original ----- De: "nek5000-users" ?: "nek5000-users" Envoy?: Mercredi 22 Novembre 2017 11:05:45 Objet: Re: [Nek5000-users] previous value of drag coefficient Hi, I suggest to use lag arrays as done for the BDF scheme, that is to have an array of size #multisteps and at each new time step to copy the old RHS (as you do AB) one step backwards. This is then somewhat consistent with Nek. Just be careful during the startup. Hth, Philipp On November 22, 2017 10:31:11 AM GMT+01:00, nek5000-users at lists.mcs.anl.gov wrote: >Hi Neks, > >Since I am trying to implement Adams bashforth scheme on my FSI problem >(for equation of motion), I would like to pass the lift coefficients of >the current timestep and the just previous time step to the function >where I implemented the Adams Bashforth module. But its seems a bit >difficult to fetch the just previous value of dragy value since it >calls a new value at every time step. I simply tried a swapping method. >But it gets zero value. Could you please tell me whether is there any >way to fetch the just previous value of a variable (from an array or >something?) > > >Thanks regards > >Sijo GEORGE -- Sent from my Android device with K-9 Mail. Please excuse my brevity. _______________________________________________ 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 Nov 22 04:20:25 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 22 Nov 2017 11:20:25 +0100 (CET) Subject: [Nek5000-users] Moving Mesh Message-ID: Hello Nek, I was trying to validate moving cylinder test cases (ocyl2.usr more specifically). I could validate it properly. Even though I have some doubts regarding the mesh motion. Mesh deformation more precisely. I have seen a function to send the deformation to far field apart from the near wall region (Laplace equation with h1 as blending coefficient).So in order to find h1 in the equation there is a parameter called delta (which is 2 in this case). I have also read a paper regarding this equation. In that paper its clearly states that it is the average thickness of the element which is close to the object. So my question is, is there anyway to calculate it automatically or easily to calculate the thickness of that particular elements from any input files? because if I change the mesh (suppose if I refine the mesh size everything will change right?) how can I find the new average thickness of the elements which is close to the object? or how can I measure the area or volume of the elements which is close to the wall. so could you please help me with this? Thanks with regards Sijo George -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Nov 22 04:23:14 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 22 Nov 2017 11:23:14 +0100 Subject: [Nek5000-users] previous value of drag coefficient In-Reply-To: References: Message-ID: just declare an array "lift(nabstep)", and at each step you push the content up by one index: do i=nabstep-1,-1,1 lift(i+1) = lift(i) end do lift(1) = current_lift something like that. philipp On 2017-11-22 11:18, nek5000-users at lists.mcs.anl.gov wrote: > Thank you Philip. I am sorry I just started to use NEK5000 and in Fortran too. I could not find an example of the concept of lag array. I know its a silly question. But could you please help me how to initialize such an array. > > Thanks again > Sijo GEORGE > > ----- Mail original ----- > De: "nek5000-users" > ?: "nek5000-users" > Envoy?: Mercredi 22 Novembre 2017 11:05:45 > Objet: Re: [Nek5000-users] previous value of drag coefficient > > Hi, > I suggest to use lag arrays as done for the BDF scheme, that is to have an array of size #multisteps and at each new time step to copy the old RHS (as you do AB) one step backwards. > This is then somewhat consistent with Nek. Just be careful during the startup. > Hth, Philipp > > On November 22, 2017 10:31:11 AM GMT+01:00, nek5000-users at lists.mcs.anl.gov wrote: >> Hi Neks, >> >> Since I am trying to implement Adams bashforth scheme on my FSI problem >> (for equation of motion), I would like to pass the lift coefficients of >> the current timestep and the just previous time step to the function >> where I implemented the Adams Bashforth module. But its seems a bit >> difficult to fetch the just previous value of dragy value since it >> calls a new value at every time step. I simply tried a swapping method. >> But it gets zero value. Could you please tell me whether is there any >> way to fetch the just previous value of a variable (from an array or >> something?) >> >> >> Thanks regards >> >> Sijo GEORGE > From nek5000-users at lists.mcs.anl.gov Wed Nov 22 07:25:51 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 22 Nov 2017 14:25:51 +0100 (CET) Subject: [Nek5000-users] previous value of drag coefficient In-Reply-To: References: Message-ID: Hi, I tried but still the same problem, here is the code what i tried, only 100 time steps I have given in .rea file. The output looks like this: for the first step it gives 0.5 as expected but from next time step onwards it gives zero. I think after each time step it deletes all its memory and store again. If suppose I give: write (6,7020) y(istep) instead of write (6,7020) y(istep-1) it will print all the time steps as it is stored. So is there anyway to get just previous value from an array at each time step? remember that i need lift coefficient instead of istep. istep is just a dummy parameter to check. Thanks Regards Sijo GEORGE c----------------------------------------------------------------------- subroutine userchk include 'SIZE' include 'TOTAL' include 'RESTART' real x0(3),y(100) save x0 data x0 /3*0/ parameter (lt=lx1*ly1*lz1*lelv) common /scrns/ vort(lt,3), w1(lt), w2(lt) n = nx1*ny1*nz1*nelv call comp_vort3(vort , w1, w2, vx, vy, vz) call copy (T,vort,n) ! Vorticity --> T ifto = .true. ! Dump vorticity as T if (istep.eq.0) call set_obj ! define objects for surface integrals scale = 2. ! Cd = F/(.5 rho U^2 ) = 2*F c if (mod(istep,10).eq.0) call torque_calc(scale,x0,.true.,.false.) y(0)=0.5 y(istep) = istep write (6,7020) y(istep-1) 7020 format(7f20.8) ifusermv = .true. if (ifusermv) call my_meshv ! Compute our own mesh velocity return end c----------------------------------------------------------------------- ----- Mail original ----- De: "nek5000-users" ?: "nek5000-users" Envoy?: Mercredi 22 Novembre 2017 11:23:14 Objet: Re: [Nek5000-users] previous value of drag coefficient just declare an array "lift(nabstep)", and at each step you push the content up by one index: do i=nabstep-1,-1,1 lift(i+1) = lift(i) end do lift(1) = current_lift something like that. philipp On 2017-11-22 11:18, nek5000-users at lists.mcs.anl.gov wrote: > Thank you Philip. I am sorry I just started to use NEK5000 and in Fortran too. I could not find an example of the concept of lag array. I know its a silly question. But could you please help me how to initialize such an array. > > Thanks again > Sijo GEORGE > > ----- Mail original ----- > De: "nek5000-users" > ?: "nek5000-users" > Envoy?: Mercredi 22 Novembre 2017 11:05:45 > Objet: Re: [Nek5000-users] previous value of drag coefficient > > Hi, > I suggest to use lag arrays as done for the BDF scheme, that is to have an array of size #multisteps and at each new time step to copy the old RHS (as you do AB) one step backwards. > This is then somewhat consistent with Nek. Just be careful during the startup. > Hth, Philipp > > On November 22, 2017 10:31:11 AM GMT+01:00, nek5000-users at lists.mcs.anl.gov wrote: >> Hi Neks, >> >> Since I am trying to implement Adams bashforth scheme on my FSI problem >> (for equation of motion), I would like to pass the lift coefficients of >> the current timestep and the just previous time step to the function >> where I implemented the Adams Bashforth module. But its seems a bit >> difficult to fetch the just previous value of dragy value since it >> calls a new value at every time step. I simply tried a swapping method. >> But it gets zero value. Could you please tell me whether is there any >> way to fetch the just previous value of a variable (from an array or >> something?) >> >> >> Thanks regards >> >> Sijo GEORGE > _______________________________________________ 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 Nov 22 09:49:47 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 22 Nov 2017 16:49:47 +0100 (CET) Subject: [Nek5000-users] Storing dragy coefficient Message-ID: Hi, I was working on storing dragy coefficient on an array. But its not getting permanently stored in the varibale after the next time step. That means I could not access the previous stored value from the array. Is there any way to store and access dragy of different time steps? Thanks with regards Sijo GEORGE -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Nov 22 09:58:15 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 22 Nov 2017 15:58:15 +0000 Subject: [Nek5000-users] Storing dragy coefficient In-Reply-To: References: Message-ID: if I correctly understand your problem, you will need to look into common block or save data that are Fortran capabilities. Marco On Nov 22, 2017, at 10:49 AM, nek5000-users at lists.mcs.anl.gov wrote: Hi, I was working on storing dragy coefficient on an array. But its not getting permanently stored in the varibale after the next time step. That means I could not access the previous stored value from the array. Is there any way to store and access dragy of different time steps? Thanks with regards Sijo GEORGE _______________________________________________ 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 Wed Nov 22 10:10:13 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 22 Nov 2017 17:10:13 +0100 (CET) Subject: [Nek5000-users] Storing dragy coefficient In-Reply-To: References: Message-ID: Thank you marco. So you mean if my problem is for 100 time steps and if i declare a variable like below real dum(100) save dum and if i write dum(istep)= dragy(1) # I have one object I could store all the dragy coefficient into this array permenently? Because I just want to use the previous values that means (dum(istep-1)) in some other functions. Thanks with regards Sijo GEORGE De: "nek5000-users" ?: "nek5000-users" Envoy?: Mercredi 22 Novembre 2017 16:58:15 Objet: Re: [Nek5000-users] Storing dragy coefficient if I correctly understand your problem, you will need to look into common block or save data that are Fortran capabilities. Marco On Nov 22, 2017, at 10:49 AM, [ mailto:nek5000-users at lists.mcs.anl.gov | nek5000-users at lists.mcs.anl.gov ] wrote: Hi, I was working on storing dragy coefficient on an array. But its not getting permanently stored in the varibale after the next time step. That means I could not access the previous stored value from the array. Is there any way to store and access dragy of different time steps? Thanks with regards Sijo GEORGE _______________________________________________ Nek5000-users mailing list [ mailto:Nek5000-users at lists.mcs.anl.gov | 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 Wed Nov 22 10:18:24 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 22 Nov 2017 16:18:24 +0000 Subject: [Nek5000-users] Storing dragy coefficient In-Reply-To: References: Message-ID: I think it should do the trick. If you want to use the values in a different subroutine, you must use common block. If the values are used in the same subroutine as they are stored, then save data should be sufficient. Marco On Nov 22, 2017, at 11:10 AM, nek5000-users at lists.mcs.anl.gov wrote: Thank you marco. So you mean if my problem is for 100 time steps and if i declare a variable like below real dum(100) save dum and if i write dum(istep)= dragy(1) # I have one object I could store all the dragy coefficient into this array permenently? Because I just want to use the previous values that means (dum(istep-1)) in some other functions. Thanks with regards Sijo GEORGE ________________________________ De: "nek5000-users" > ?: "nek5000-users" > Envoy?: Mercredi 22 Novembre 2017 16:58:15 Objet: Re: [Nek5000-users] Storing dragy coefficient if I correctly understand your problem, you will need to look into common block or save data that are Fortran capabilities. Marco On Nov 22, 2017, at 10:49 AM, nek5000-users at lists.mcs.anl.gov wrote: Hi, I was working on storing dragy coefficient on an array. But its not getting permanently stored in the varibale after the next time step. That means I could not access the previous stored value from the array. Is there any way to store and access dragy of different time steps? Thanks with regards Sijo GEORGE _______________________________________________ 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 Wed Nov 22 10:22:30 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 22 Nov 2017 17:22:30 +0100 (CET) Subject: [Nek5000-users] Storing dragy coefficient In-Reply-To: References: Message-ID: Thanks Marco I give a try, I am trying to implement FSI with adams bashforth method. RK4 scheme is validated. Now I am doing the stability study (comparing adams and RK4). So thanks and please do keep in touch. Sijo De: "nek5000-users" ?: "nek5000-users" Envoy?: Mercredi 22 Novembre 2017 17:18:24 Objet: Re: [Nek5000-users] Storing dragy coefficient I think it should do the trick. If you want to use the values in a different subroutine, you must use common block. If the values are used in the same subroutine as they are stored, then save data should be sufficient. Marco On Nov 22, 2017, at 11:10 AM, [ mailto:nek5000-users at lists.mcs.anl.gov | nek5000-users at lists.mcs.anl.gov ] wrote: Thank you marco. So you mean if my problem is for 100 time steps and if i declare a variable like below real dum(100) save dum and if i write dum(istep)= dragy(1) # I have one object I could store all the dragy coefficient into this array permenently? Because I just want to use the previous values that means (dum(istep-1)) in some other functions. Thanks with regards Sijo GEORGE De: "nek5000-users" < [ mailto:nek5000-users at lists.mcs.anl.gov | nek5000-users at lists.mcs.anl.gov ] > ?: "nek5000-users" < [ mailto:nek5000-users at lists.mcs.anl.gov | nek5000-users at lists.mcs.anl.gov ] > Envoy?: Mercredi 22 Novembre 2017 16:58:15 Objet: Re: [Nek5000-users] Storing dragy coefficient if I correctly understand your problem, you will need to look into common block or save data that are Fortran capabilities. Marco BQ_BEGIN On Nov 22, 2017, at 10:49 AM, [ mailto:nek5000-users at lists.mcs.anl.gov | nek5000-users at lists.mcs.anl.gov ] wrote: Hi, I was working on storing dragy coefficient on an array. But its not getting permanently stored in the varibale after the next time step. That means I could not access the previous stored value from the array. Is there any way to store and access dragy of different time steps? Thanks with regards Sijo GEORGE _______________________________________________ Nek5000-users mailing list [ mailto:Nek5000-users at lists.mcs.anl.gov | Nek5000-users at lists.mcs.anl.gov ] [ https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users | https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users ] _______________________________________________ Nek5000-users mailing list [ mailto:Nek5000-users at lists.mcs.anl.gov | Nek5000-users at lists.mcs.anl.gov ] https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users _______________________________________________ Nek5000-users mailing list [ mailto:Nek5000-users at lists.mcs.anl.gov | Nek5000-users at lists.mcs.anl.gov ] https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users BQ_END _______________________________________________ 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 Nov 23 04:02:23 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 23 Nov 2017 17:02:23 +0700 Subject: [Nek5000-users] read the *f000* files Message-ID: Hi, Neks. I want to read the *f000* files not in visit, but for post-proccesing in matlab, for example. Is it possible to get some part of data from theese files and save it in *.dat ? Regards, Vatslav. -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Thu Nov 23 04:05:31 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 23 Nov 2017 11:05:31 +0100 Subject: [Nek5000-users] read the *f000* files In-Reply-To: References: Message-ID: Hi, you can check the python library by Jacopo and Nicolo: https://github.com/jcanton/pymech There are also matlab routines around. Perhpas Adam can send you a link to these. Philipp On 2017-11-23 11:02, nek5000-users at lists.mcs.anl.gov wrote: > Hi, Neks. > > I want to read the *f000* files not in visit, but for post-proccesing in > matlab, for example. > Is it possible to get some part of data from theese files and save it in > *.dat ? > > Regards, Vatslav. > > > _______________________________________________ > 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 Nov 23 04:09:13 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 23 Nov 2017 11:09:13 +0100 Subject: [Nek5000-users] read the *f000* files In-Reply-To: References: Message-ID: actually, the Matlab reader is here: https://github.com/nfabbiane/nekmatlab PHilipp On 2017-11-23 11:02, nek5000-users at lists.mcs.anl.gov wrote: > Hi, Neks. > > I want to read the *f000* files not in visit, but for post-proccesing in > matlab, for example. > Is it possible to get some part of data from theese files and save it in > *.dat ? > > Regards, Vatslav. > > > _______________________________________________ > 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 Nov 23 04:13:30 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 23 Nov 2017 17:13:30 +0700 Subject: [Nek5000-users] read the *f000* files Message-ID: Thanks a lot, Philipp. I will try it. -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Thu Nov 23 11:05:02 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 23 Nov 2017 17:05:02 +0000 Subject: [Nek5000-users] Specifying the Froude Number Message-ID: Hi All, I'm trying to work out how to specify the Froude number in my simulation of a buoyant plume. I can specify Reynolds and Peclet in the .rea file but there doesn't seem to be anywhere analogous for Fr. I thought it might be in the .usr function in the subroutine userf, mine currently looks like this: subroutine userf (ix,iy,iz,iel) include 'SIZE' include 'TSTEP' c include 'TOTAL' include 'NEKUSE' integer e,f,eg ffx = -0.06976*temp ffy = 0.9976*temp ffz = 0.0 return end Does anyone have any idea (if this is the correct place, and) how I can specify Fr here. I also want to ensure gravity is in the correct direction (namely, towards the floor), and how does what I have at the moment do this? My simulation outputs do have gravity in the correct direction, so it must be correct. Thank you in advance. Best, Daniel -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Thu Nov 23 11:49:17 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 23 Nov 2017 18:49:17 +0100 Subject: [Nek5000-users] Specifying the Froude Number In-Reply-To: References: Message-ID: I guess the Froude number and the way to exactly specify it depends on the scaling of your equations. Your numbers indicate that your gravity is in the x/y plane, 4 degrees off the x axis? But in principle, this is the correct place, you just need to add the correct factor in front of "temp". I suggest also to use a cos/sin for specifying the direction (if I understand your code). Philipp On 2017-11-23 18:05, nek5000-users at lists.mcs.anl.gov wrote: > Hi All, > > > I'm trying to work out how to specify the Froude number in my simulation > of a buoyant plume.? I can specify Reynolds and Peclet in the .rea file > but there doesn't seem to be anywhere analogous for Fr.? I thought it > might be in the .usr function in the subroutine userf, mine currently > looks like this: > > > ????? subroutine userf? (ix,iy,iz,iel) > ????? include 'SIZE' > ????? include 'TSTEP' > c???? include 'TOTAL' > ????? include 'NEKUSE' > > ????? integer e,f,eg > > ????? ffx = -0.06976*temp > ????? ffy = 0.9976*temp > ????? ffz = 0.0 > > ????? return > ????? end > > > Does anyone have any idea (if this is the correct place, and) how I can > specify Fr here.? I also want to ensure gravity is in the correct > direction (namely, towards the floor), and how does what I have at the > moment do this?? My simulation outputs do have gravity in the correct > direction, so it must be correct. > > > Thank you in advance. > > > Best, > > > Daniel > > > > _______________________________________________ > 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 Nov 23 22:07:14 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 24 Nov 2017 01:07:14 -0300 Subject: [Nek5000-users] Energy source term Message-ID: Hi all, If I need to add an oscillating source term in the energy equation for a small region in a 2D domain would it be fine to do it this way, with these variables: ? c ============================= subroutine userq (ix,iy,iz,ieg) include 'SIZE' include 'TOTAL' include 'NEKUSE' real pii,frec,ampli pii=4.0*atan(1.0) frec=0.005 ampli=1.0 if (x .ge. 0 .and. x .le. 0.02 .and. $ y .ge. 0 .and. y .le. 0.02) then qvol = ampli*sin(2.0*pii*frec*time) else qvol = 0.0 end if source = 0.0 return end c ================================== Thanks in advance, JP. -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Fri Nov 24 03:20:49 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 24 Nov 2017 10:20:49 +0100 Subject: [Nek5000-users] Energy source term In-Reply-To: References: Message-ID: Hi, The lines appear fine, however I would not use a force that goes from zero to the final value within on grid point; this might lead to oscillations in the solution. Perhaps a smoothing might be a better solution. Philipp On 2017-11-24 05:07, nek5000-users at lists.mcs.anl.gov wrote: > Hi all, > > If I need to add an oscillating source term in the energy equation for a > small region in a 2D domain would it be fine to do it this way, with > these variables: ? > > c ============================= > ? ? ? subroutine userq? (ix,iy,iz,ieg) > ? ? ? include 'SIZE' > ? ? ? include 'TOTAL' > ? ? ? include 'NEKUSE' > > > ? ? ? real pii,frec,ampli > > ? ? ? pii=4.0*atan(1.0) > ? ? ? frec=0.005 > ? ? ? ampli=1.0 > > > ? ? ? if (x .ge. 0 .and. x .le. 0.02 .and. > ? ? ?$ y .ge. 0 .and. y .le. 0.02) then > > ? ? ? ? ? qvol = ampli*sin(2.0*pii*frec*time) > > ? ? ? else > > ? ? ? ? ? qvol = 0.0 > > ? ? ? end if > > > ? ? ? source = 0.0 > > ? ? ? return > ? ? ? end > c ================================== > > > Thanks in advance, > > JP. > > > _______________________________________________ > 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 Nov 24 08:41:44 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 24 Nov 2017 14:41:44 +0000 Subject: [Nek5000-users] Energy source term In-Reply-To: References: , Message-ID: Yes -- you will want a smooth forcing in space - unless the jump occurs at an element boundary. If the forcing has a jump discontinuity across element boundaries it's usually easier to set a flag in (say) usrdat2() that identifies which elements are to have the source term and which are not. Then add something like the following to userq: common /myflag/ qflag(lx1,ly1,lz1,lelt) integer qflag,e qvol=0 e=gllel(ieg) if (qflag(ix,iy,iz,e).eq.1) qvol=ampli*sin(2.0*pi*frec*time) ! pi defined by nek etc. In usrdat2(), use something like common /myflag/ qflag(lx1,ly1,lz1,lelt) integer qflag,e integer e n = lx1*ly1*lz1*nelt nxyz=lx1*ly1*lz1 call izero(qflag,n) do e=1,nelt xtest = xm1(2,2,1,e) ytest = ym1(2,2,1,e) if ..... xtest and ytest in your bounds .... call ione(qtest(1,1,1,e),nxyz) enddo hth, Paul ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Friday, November 24, 2017 3:20 AM To: nek5000-users at lists.mcs.anl.gov Subject: Re: [Nek5000-users] Energy source term Hi, The lines appear fine, however I would not use a force that goes from zero to the final value within on grid point; this might lead to oscillations in the solution. Perhaps a smoothing might be a better solution. Philipp On 2017-11-24 05:07, nek5000-users at lists.mcs.anl.gov wrote: > Hi all, > > If I need to add an oscillating source term in the energy equation for a > small region in a 2D domain would it be fine to do it this way, with > these variables: ? > > c ============================= > subroutine userq (ix,iy,iz,ieg) > include 'SIZE' > include 'TOTAL' > include 'NEKUSE' > > > real pii,frec,ampli > > pii=4.0*atan(1.0) > frec=0.005 > ampli=1.0 > > > if (x .ge. 0 .and. x .le. 0.02 .and. > $ y .ge. 0 .and. y .le. 0.02) then > > qvol = ampli*sin(2.0*pii*frec*time) > > else > > qvol = 0.0 > > end if > > > source = 0.0 > > return > end > c ================================== > > > Thanks in advance, > > JP. > > > _______________________________________________ > 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 Nov 24 10:18:31 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 24 Nov 2017 17:18:31 +0100 (CET) Subject: [Nek5000-users] Storing dragy coefficient In-Reply-To: References: Message-ID: Hi, It worked. And now I just have one more question. I store the drag coefficient in each time step lets say real dragycoeff(NSTEP) save dragycoeff dragycoeff(istep) = dragy(1) The reason why I store these values : I would like to use them for adams bashforth scheme. (Since I am working on FSI problem and to solve equation of motion) so i may need the previous for 4 values maximum to compute the displacement and velocity values for the next time step. So my question is, If I have a very large time steps (lets say 500000) will it be efficient? I am also thingking another way like dynamic memory allocation. Since I just need the previous 4 values at each time step, I dont have to store all the drag coefficient for all the time steps. So is there any way in NEK5000 to store the just 4 previous values and update these values at every time step. So that i can decalare a varibale of just siwe of 4 (lets say : real dragycoeff(4) ). Or Can I make any function for that? Thanks in advance SIjo GEORGE De: "nek5000-users" ?: "nek5000-users" Envoy?: Mercredi 22 Novembre 2017 17:22:30 Objet: Re: [Nek5000-users] Storing dragy coefficient Thanks Marco I give a try, I am trying to implement FSI with adams bashforth method. RK4 scheme is validated. Now I am doing the stability study (comparing adams and RK4). So thanks and please do keep in touch. Sijo De: "nek5000-users" ?: "nek5000-users" Envoy?: Mercredi 22 Novembre 2017 17:18:24 Objet: Re: [Nek5000-users] Storing dragy coefficient I think it should do the trick. If you want to use the values in a different subroutine, you must use common block. If the values are used in the same subroutine as they are stored, then save data should be sufficient. Marco On Nov 22, 2017, at 11:10 AM, [ mailto:nek5000-users at lists.mcs.anl.gov | nek5000-users at lists.mcs.anl.gov ] wrote: Thank you marco. So you mean if my problem is for 100 time steps and if i declare a variable like below real dum(100) save dum and if i write dum(istep)= dragy(1) # I have one object I could store all the dragy coefficient into this array permenently? Because I just want to use the previous values that means (dum(istep-1)) in some other functions. Thanks with regards Sijo GEORGE De: "nek5000-users" < [ mailto:nek5000-users at lists.mcs.anl.gov | nek5000-users at lists.mcs.anl.gov ] > ?: "nek5000-users" < [ mailto:nek5000-users at lists.mcs.anl.gov | nek5000-users at lists.mcs.anl.gov ] > Envoy?: Mercredi 22 Novembre 2017 16:58:15 Objet: Re: [Nek5000-users] Storing dragy coefficient if I correctly understand your problem, you will need to look into common block or save data that are Fortran capabilities. Marco BQ_BEGIN On Nov 22, 2017, at 10:49 AM, [ mailto:nek5000-users at lists.mcs.anl.gov | nek5000-users at lists.mcs.anl.gov ] wrote: Hi, I was working on storing dragy coefficient on an array. But its not getting permanently stored in the varibale after the next time step. That means I could not access the previous stored value from the array. Is there any way to store and access dragy of different time steps? Thanks with regards Sijo GEORGE _______________________________________________ Nek5000-users mailing list [ mailto:Nek5000-users at lists.mcs.anl.gov | Nek5000-users at lists.mcs.anl.gov ] [ https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users | https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users ] _______________________________________________ Nek5000-users mailing list [ mailto:Nek5000-users at lists.mcs.anl.gov | Nek5000-users at lists.mcs.anl.gov ] https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users _______________________________________________ Nek5000-users mailing list [ mailto:Nek5000-users at lists.mcs.anl.gov | Nek5000-users at lists.mcs.anl.gov ] https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users BQ_END _______________________________________________ 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 Nov 24 12:20:15 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 24 Nov 2017 19:20:15 +0100 Subject: [Nek5000-users] Storing dragy coefficient In-Reply-To: References: Message-ID: Hi If I understand well, you should just do it as I wrote earlier: just declare an array "lift(nabstep)", and at each step you push the content up by one index: real, parameter :: nabstep=4 real cd(nabstep) save cd .... do i=nabstep-1,-1,1 cd(i+1) = cd(i) end do cd(1) = new_cd or something like that. philipp On 2017-11-24 17:18, nek5000-users at lists.mcs.anl.gov wrote: > Hi, > > It worked. And now I just have one more question. I store the drag > coefficient in each time step > lets say > > real dragycoeff(NSTEP) > save dragycoeff > > dragycoeff(istep) = dragy(1) > > The reason why I store these values : I would like to use them for adams > bashforth scheme. (Since I am working on FSI problem and to solve > equation of motion) so i may need the previous for 4 values maximum to > compute the displacement and velocity values for the next time step. > > So my question is, If I have a very large time steps (lets say 500000) > will it be efficient? I am also thingking another way like dynamic > memory allocation. Since I just need the previous 4 values at each time > step, I dont have to store all the drag coefficient for all the time > steps. So is there any way in NEK5000 to store the just 4 previous > values and update these values at every time step. So that i can > decalare a varibale of just siwe of 4 (lets say : real dragycoeff(4) ). > Or Can I make any function for that? > > Thanks in advance > > SIjo GEORGE > > ------------------------------------------------------------------------ > *De: *"nek5000-users" > *?: *"nek5000-users" > *Envoy?: *Mercredi 22 Novembre 2017 17:22:30 > *Objet: *Re: [Nek5000-users] Storing dragy coefficient > > Thanks Marco I give a try, I am trying to implement FSI with adams > bashforth method. RK4 scheme is validated. Now I am doing the stability > study (comparing adams and RK4). So thanks and please do keep in touch. > > Sijo > ------------------------------------------------------------------------ > *De: *"nek5000-users" > *?: *"nek5000-users" > *Envoy?: *Mercredi 22 Novembre 2017 17:18:24 > *Objet: *Re: [Nek5000-users] Storing dragy coefficient > > I think it should do the trick. > > If you want to use the values in a different subroutine, you must use > common block. If the values are used in the same subroutine as they are > stored, then save data should be sufficient. > > Marco > > > On Nov 22, 2017, at 11:10 AM, nek5000-users at lists.mcs.anl.gov > wrote: > > Thank you marco. > > So you mean if my problem is for 100 time steps and if i declare a > variable like below > > ????? real dum(100) > ????? save dum > > ?and if i write > > ???? dum(istep)= dragy(1)???????????????????? # I have one object > > I could store all the dragy coefficient into this array permenently? > > Because I just want to use the previous values that means > (dum(istep-1)) in some other functions. > > > Thanks with regards > > > Sijo GEORGE > > > ------------------------------------------------------------------------ > *De: *"nek5000-users" > > *?: *"nek5000-users" > > *Envoy?: *Mercredi 22 Novembre 2017 16:58:15 > *Objet: *Re: [Nek5000-users] Storing dragy coefficient > > if I correctly understand your problem, you will need to look into > common block or save data that are Fortran capabilities. > > Marco > > On Nov 22, 2017, at 10:49 AM, nek5000-users at lists.mcs.anl.gov > wrote: > > Hi, > > I was working on storing dragy coefficient on an array. But its > not getting permanently stored in the varibale after the next > time step. That means I could not access the previous stored > value from the array. Is there any way to store and access dragy > of different time steps? > > Thanks > > with regards > > Sijo GEORGE > _______________________________________________ > 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 > > > _______________________________________________ > 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 Nov 25 04:04:02 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sat, 25 Nov 2017 10:04:02 +0000 Subject: [Nek5000-users] link between local and global direction Message-ID: Hi, Neks! What is the principle of creation of link between local ( i, j, k ) in some spectral element and the global coordinate system ( x, y, z ) in sophisticated geometry ? And how to find out how does these links changes when I move or rotate the elements ? Regards, Vatslav. -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Sat Nov 25 22:17:58 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sun, 26 Nov 2017 07:17:58 +0300 Subject: [Nek5000-users] =?utf-8?q?mesh_deformation?= Message-ID: Hi, Neks! I am interesting in how could I deform only the bottom layer of points in a spectal element? In usrdat2 i,j,k,?indices?correspond to different axes of the global mesh. For example, xm1(1,1,1,e) could be a?coordinate of different vertices for different elements. Best regards, Mark? -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Sat Nov 25 23:11:52 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sun, 26 Nov 2017 05:11:52 +0000 Subject: [Nek5000-users] link between local and global direction In-Reply-To: References: Message-ID: Hi Vatslav, There is some discussion of mesh deformation here: http://nek5000.github.io/NekDoc/geometry.html Look for mesh modification. As long as your transformation is a relatively smooth function of the input geometry, i.e., X' = f(X), where X=(x,y,z) is the input geometry and X'=(x',y',z') is the output geometry, then things are likely to be ok. A call to fix_geom just before the "return" statement in usrdat2 is often a good idea. That is: subroutine usrdat2 your code here (e.g., following the url above) call fix_geom return end Note that the comments about outputting the geometry in the URL above are somewhat dated. Nek5000 always writes the geometry to the first 0.f00001 file, so you don't have to explicitly request to have the geometry written with the other fields (velocity, pressure, temperature...) hth, Paul ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Saturday, November 25, 2017 4:04:02 AM To: nek5000-users at lists.mcs.anl.gov Subject: [Nek5000-users] link between local and global direction Hi, Neks! What is the principle of creation of link between local ( i, j, k ) in some spectral element and the global coordinate system ( x, y, z ) in sophisticated geometry ? And how to find out how does these links changes when I move or rotate the elements ? Regards, Vatslav. -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Sun Nov 26 08:44:45 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sun, 26 Nov 2017 14:44:45 +0000 Subject: [Nek5000-users] mesh deformation In-Reply-To: References: Message-ID: Hi Mark, One example of non-affine (hex8's vertex) coordinate transformation is in usrdat() of https://github.com/Nek5000/NekExamples/blob/master/vortex/r1854a.usr Aleks ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Saturday, November 25, 2017 10:17:58 PM To: nek5000-users Subject: [Nek5000-users] mesh deformation Hi, Neks! I am interesting in how could I deform only the bottom layer of points in a spectal element? In usrdat2 i,j,k, indices correspond to different axes of the global mesh. For example, xm1(1,1,1,e) could be a coordinate of different vertices for different elements. Best regards, Mark -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Nov 27 07:52:52 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 27 Nov 2017 14:52:52 +0100 (CET) Subject: [Nek5000-users] mesh deformation Message-ID: Hello Nek, I was trying to validate moving cylinder test cases (ocyl2.usr more specifically). I could validate it properly. Even though I have some doubts regarding the mesh motion. Mesh deformation more precisely. I have seen a function to send the deformation to far field apart from the near wall region (Laplace equation with h1 as blending coefficient).So in order to find h1 in the equation there is a parameter called delta (which is 2 in this case). I have also read a paper regarding this equation. In that paper its clearly states that it is the average thickness of the element which is close to the object. So my question is, is there anyway to calculate it automatically or easily to calculate the thickness of that particular elements from any input files? because if I change the mesh (suppose if I refine the mesh size everything will change right?) how can I find the new average thickness of the elements which is close to the object? or how can I measure the area or volume of the elements which is close to the wall. so could you please help me with this? Thanks with regards Sijo George -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Nov 27 08:47:35 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 27 Nov 2017 14:47:35 +0000 Subject: [Nek5000-users] mesh deformation In-Reply-To: References: Message-ID: Dear Sijo, In the ocyl2 case, the delta is based on a dimension that is related to the model. (Here, delta=2*D, where D=1 is the diameter of the cylinder.) Such a choice would be reasonable for all flow past a cylinder cases for that particular geometry, independent of the mesh resolution. The boundary-layer thickness criterion is important when there are multiple objects that are potentially touching. Keep in mind that any of these choices are moderated by the fact that we are simply computing a blending function that is the solution to the (modified) Laplace equation. It will have more or less the correct shape --- even the unmodified Laplace solution is not a terrible blending function. The modification is designed simply to help preserve mesh sizes near boundaries. I paste below a snippet of code that computes the average thickness of elements near the moving wall. Paul integer e,f nxz = nx1*nz1 nxyz = nx1*ny1*nz1 n = nxyz*nelv srfbl = 0. ! Surface area of elements in b.l. volbl = 0. ! Volume of elements in boundary layer do e=1,nelv do f=1,nface if (cbc(f,e,1).eq.'mv ') then srfbl = srfbl + vlsum(area(1,1,f,e),nxz ) volbl = volbl + vlsum(bm1 (1,1,1,e),nxyz) endif enddo enddo srfbl = glsum(srfbl,1) ! Sum over all processors volbl = glsum(volbl,1) delta = volbl / srfbl ! Avg thickness of b.l. elements call rone (h1,n) call rzero(h2,n) call cheap_dist(d,1,'mv ') deltap = 2*delta ! Protected b.l. thickness do i=1,n arg = -(d(i)/deltap)**2 h1(i) = h1(i) + 9*exp(arg) enddo ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Monday, November 27, 2017 7:52 AM To: nek5000-users Subject: [Nek5000-users] mesh deformation Hello Nek, I was trying to validate moving cylinder test cases (ocyl2.usr more specifically). I could validate it properly. Even though I have some doubts regarding the mesh motion. Mesh deformation more precisely. I have seen a function to send the deformation to far field apart from the near wall region (Laplace equation with h1 as blending coefficient).So in order to find h1 in the equation there is a parameter called delta (which is 2 in this case). I have also read a paper regarding this equation. In that paper its clearly states that it is the average thickness of the element which is close to the object. So my question is, is there anyway to calculate it automatically or easily to calculate the thickness of that particular elements from any input files? because if I change the mesh (suppose if I refine the mesh size everything will change right?) how can I find the new average thickness of the elements which is close to the object? or how can I measure the area or volume of the elements which is close to the wall. so could you please help me with this? Thanks with regards Sijo George -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Nov 27 09:17:26 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 27 Nov 2017 16:17:26 +0100 (CET) Subject: [Nek5000-users] mesh deformation In-Reply-To: References: Message-ID: Thanks Paul thank you very much. I could understand the things what you mentioend about the laplace equation since I have read your publication reagrding moving boundary and ALE formulation. I am gonna try it on my FSI problem with 2D single cylinder. Since its been some time I had also developed a code which pushes the most of the deformation to the outer region: Here it is: v = volume(1,1,1,1) a=glmax(v,1) b=glmin(v,1) L = 1 do i=1,n vdiv = b/a num = 1-vidv denom = 1/a h1(i) = h1(i) + ((1- (b/a) **L ) / (( denom * v ) **L)) enddo which also works same as like you have created (I have only verified visually). But I have some questions regarding this: Before my doubts I have to be clear on one thing. Here I will be talking about two different elements. First one is Elements (lets say I have 1348 elements in my domain) and Second is Spectral elements (6*6*1348=48528) where nx1 = 6. 1) Since the loop runs from 1 to n, the blending coefficient also makes h1 for all speactral elements right? 2) In that case, my blending coefficient is uniform across a particular Element of the domain? more specifically lets say my element number is 1A and in 1A there are 36 spectral elements of different size. As per my code above, is my blending coefficient same for all the spectral elements which is inside 1A element?or is it different?(because I consider the volume of each element as "v") My idea was to give different h1 coefficient for each Elements and give constant h1 to spectral elements in each elements. More precisily lets sat there are 2 elements called 1A and 1B. 1A has a h1 = 2.0 so all the spectral elements in the 1A has h1 of 2.0 and 1 B has h1 = 10.0 so all the spectral elements in the 1B has h1 of 10.0 I know its a quite big email. Even though could you please correct me if I am wrong. Because gradually I am going to run a very complex problem. So I Have to be sure in these cases before starting it. Thanks with regards Sijo George De: "nek5000-users" ?: "nek5000-users" Envoy?: Lundi 27 Novembre 2017 15:47:35 Objet: Re: [Nek5000-users] mesh deformation Dear Sijo, In the ocyl2 case, the delta is based on a dimension that is related to the model. (Here, delta=2*D, where D=1 is the diameter of the cylinder.) Such a choice would be reasonable for all flow past a cylinder cases for that particular geometry, independent of the mesh resolution. The boundary-layer thickness criterion is important when there are multiple objects that are potentially touching. Keep in mind that any of these choices are moderated by the fact that we are simply computing a blending function that is the solution to the (modified) Laplace equation. It will have more or less the correct shape --- even the unmodified Laplace solution is not a terrible blending function. The modification is designed simply to help preserve mesh sizes near boundaries. I paste below a snippet of code that computes the average thickness of elements near the moving wall. Paul integer e,f nxz = nx1*nz1 nxyz = nx1*ny1*nz1 n = nxyz*nelv srfbl = 0. ! Surface area of elements in b.l. volbl = 0. ! Volume of elements in boundary layer do e=1,nelv do f=1,nface if (cbc(f,e,1).eq.'mv ') then srfbl = srfbl + vlsum(area(1,1,f,e),nxz ) volbl = volbl + vlsum(bm1 (1,1,1,e),nxyz) endif enddo enddo srfbl = glsum(srfbl,1) ! Sum over all processors volbl = glsum(volbl,1) delta = volbl / srfbl ! Avg thickness of b.l. elements call rone (h1,n) call rzero(h2,n) call cheap_dist(d,1,'mv ') deltap = 2*delta ! Protected b.l. thickness do i=1,n arg = -(d(i)/deltap)**2 h1(i) = h1(i) + 9*exp(arg) enddo From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Monday, November 27, 2017 7:52 AM To: nek5000-users Subject: [Nek5000-users] mesh deformation Hello Nek, I was trying to validate moving cylinder test cases (ocyl2.usr more specifically). I could validate it properly. Even though I have some doubts regarding the mesh motion. Mesh deformation more precisely. I have seen a function to send the deformation to far field apart from the near wall region (Laplace equation with h1 as blending coefficient).So in order to find h1 in the equation there is a parameter called delta (which is 2 in this case). I have also read a paper regarding this equation. In that paper its clearly states that it is the average thickness of the element which is close to the object. So my question is, is there anyway to calculate it automatically or easily to calculate the thickness of that particular elements from any input files? because if I change the mesh (suppose if I refine the mesh size everything will change right?) how can I find the new average thickness of the elements which is close to the object? or how can I measure the area or volume of the elements which is close to the wall. so could you please help me with this? Thanks with regards Sijo George _______________________________________________ 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 Nov 27 09:35:03 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 27 Nov 2017 15:35:03 +0000 Subject: [Nek5000-users] mesh deformation In-Reply-To: References: , Message-ID: Dear Sijo, There's no reason to have h1 uniform in each element, so I did not bother to do that. In the SEM we usually view fields as a continuum and rarely concern with element-to-element variations as those would then depend on the mesh itself. There is no hard and fast rule here, save that you probably want the mesh deformation to be smooth within each element. hth, Paul ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Monday, November 27, 2017 9:17:26 AM To: nek5000-users Subject: Re: [Nek5000-users] mesh deformation Thanks Paul thank you very much. I could understand the things what you mentioend about the laplace equation since I have read your publication reagrding moving boundary and ALE formulation. I am gonna try it on my FSI problem with 2D single cylinder. Since its been some time I had also developed a code which pushes the most of the deformation to the outer region: Here it is: v = volume(1,1,1,1) a=glmax(v,1) b=glmin(v,1) L = 1 do i=1,n vdiv = b/a num = 1-vidv denom = 1/a h1(i) = h1(i) + ((1- (b/a) **L ) / (( denom * v ) **L)) enddo which also works same as like you have created (I have only verified visually). But I have some questions regarding this: Before my doubts I have to be clear on one thing. Here I will be talking about two different elements. First one is Elements (lets say I have 1348 elements in my domain) and Second is Spectral elements (6*6*1348=48528) where nx1 = 6. 1) Since the loop runs from 1 to n, the blending coefficient also makes h1 for all speactral elements right? 2) In that case, my blending coefficient is uniform across a particular Element of the domain? more specifically lets say my element number is 1A and in 1A there are 36 spectral elements of different size. As per my code above, is my blending coefficient same for all the spectral elements which is inside 1A element?or is it different?(because I consider the volume of each element as "v") My idea was to give different h1 coefficient for each Elements and give constant h1 to spectral elements in each elements. More precisily lets sat there are 2 elements called 1A and 1B. 1A has a h1 = 2.0 so all the spectral elements in the 1A has h1 of 2.0 and 1 B has h1 = 10.0 so all the spectral elements in the 1B has h1 of 10.0 I know its a quite big email. Even though could you please correct me if I am wrong. Because gradually I am going to run a very complex problem. So I Have to be sure in these cases before starting it. Thanks with regards Sijo George ________________________________ De: "nek5000-users" ?: "nek5000-users" Envoy?: Lundi 27 Novembre 2017 15:47:35 Objet: Re: [Nek5000-users] mesh deformation Dear Sijo, In the ocyl2 case, the delta is based on a dimension that is related to the model. (Here, delta=2*D, where D=1 is the diameter of the cylinder.) Such a choice would be reasonable for all flow past a cylinder cases for that particular geometry, independent of the mesh resolution. The boundary-layer thickness criterion is important when there are multiple objects that are potentially touching. Keep in mind that any of these choices are moderated by the fact that we are simply computing a blending function that is the solution to the (modified) Laplace equation. It will have more or less the correct shape --- even the unmodified Laplace solution is not a terrible blending function. The modification is designed simply to help preserve mesh sizes near boundaries. I paste below a snippet of code that computes the average thickness of elements near the moving wall. Paul integer e,f nxz = nx1*nz1 nxyz = nx1*ny1*nz1 n = nxyz*nelv srfbl = 0. ! Surface area of elements in b.l. volbl = 0. ! Volume of elements in boundary layer do e=1,nelv do f=1,nface if (cbc(f,e,1).eq.'mv ') then srfbl = srfbl + vlsum(area(1,1,f,e),nxz ) volbl = volbl + vlsum(bm1 (1,1,1,e),nxyz) endif enddo enddo srfbl = glsum(srfbl,1) ! Sum over all processors volbl = glsum(volbl,1) delta = volbl / srfbl ! Avg thickness of b.l. elements call rone (h1,n) call rzero(h2,n) call cheap_dist(d,1,'mv ') deltap = 2*delta ! Protected b.l. thickness do i=1,n arg = -(d(i)/deltap)**2 h1(i) = h1(i) + 9*exp(arg) enddo ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Monday, November 27, 2017 7:52 AM To: nek5000-users Subject: [Nek5000-users] mesh deformation Hello Nek, I was trying to validate moving cylinder test cases (ocyl2.usr more specifically). I could validate it properly. Even though I have some doubts regarding the mesh motion. Mesh deformation more precisely. I have seen a function to send the deformation to far field apart from the near wall region (Laplace equation with h1 as blending coefficient).So in order to find h1 in the equation there is a parameter called delta (which is 2 in this case). I have also read a paper regarding this equation. In that paper its clearly states that it is the average thickness of the element which is close to the object. So my question is, is there anyway to calculate it automatically or easily to calculate the thickness of that particular elements from any input files? because if I change the mesh (suppose if I refine the mesh size everything will change right?) how can I find the new average thickness of the elements which is close to the object? or how can I measure the area or volume of the elements which is close to the wall. so could you please help me with this? Thanks with regards Sijo George _______________________________________________ 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 Nov 27 09:54:55 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 27 Nov 2017 16:54:55 +0100 (CET) Subject: [Nek5000-users] mesh deformation In-Reply-To: References: Message-ID: Thanks you. But I am sorry I was not able to follow your answer. But could you please tell me that, if i write like this n = nx1*ny1*nz1*nelv do i=1,n v = volume(i,1,1,1) x = xm1(i,1,1,1,) enddo the volume which gives corresponding the elements which has the position of x. Regars Sijo GEORGE De: "nek5000-users" ?: "nek5000-users" Envoy?: Lundi 27 Novembre 2017 16:35:03 Objet: Re: [Nek5000-users] mesh deformation Dear Sijo, There's no reason to have h1 uniform in each element, so I did not bother to do that. In the SEM we usually view fields as a continuum and rarely concern with element-to-element variations as those would then depend on the mesh itself. There is no hard and fast rule here, save that you probably want the mesh deformation to be smooth within each element. hth, Paul From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Monday, November 27, 2017 9:17:26 AM To: nek5000-users Subject: Re: [Nek5000-users] mesh deformation Thanks Paul thank you very much. I could understand the things what you mentioend about the laplace equation since I have read your publication reagrding moving boundary and ALE formulation. I am gonna try it on my FSI problem with 2D single cylinder. Since its been some time I had also developed a code which pushes the most of the deformation to the outer region: Here it is: v = volume(1,1,1,1) a=glmax(v,1) b=glmin(v,1) L = 1 do i=1,n vdiv = b/a num = 1-vidv denom = 1/a h1(i) = h1(i) + ((1- (b/a) **L ) / (( denom * v ) **L)) enddo which also works same as like you have created (I have only verified visually). But I have some questions regarding this: Before my doubts I have to be clear on one thing. Here I will be talking about two different elements. First one is Elements (lets say I have 1348 elements in my domain) and Second is Spectral elements (6*6*1348=48528) where nx1 = 6. 1) Since the loop runs from 1 to n, the blending coefficient also makes h1 for all speactral elements right? 2) In that case, my blending coefficient is uniform across a particular Element of the domain? more specifically lets say my element number is 1A and in 1A there are 36 spectral elements of different size. As per my code above, is my blending coefficient same for all the spectral elements which is inside 1A element?or is it different?(because I consider the volume of each element as "v") My idea was to give different h1 coefficient for each Elements and give constant h1 to spectral elements in each elements. More precisily lets sat there are 2 elements called 1A and 1B. 1A has a h1 = 2.0 so all the spectral elements in the 1A has h1 of 2.0 and 1 B has h1 = 10.0 so all the spectral elements in the 1B has h1 of 10.0 I know its a quite big email. Even though could you please correct me if I am wrong. Because gradually I am going to run a very complex problem. So I Have to be sure in these cases before starting it. Thanks with regards Sijo George De: "nek5000-users" ?: "nek5000-users" Envoy?: Lundi 27 Novembre 2017 15:47:35 Objet: Re: [Nek5000-users] mesh deformation Dear Sijo, In the ocyl2 case, the delta is based on a dimension that is related to the model. (Here, delta=2*D, where D=1 is the diameter of the cylinder.) Such a choice would be reasonable for all flow past a cylinder cases for that particular geometry, independent of the mesh resolution. The boundary-layer thickness criterion is important when there are multiple objects that are potentially touching. Keep in mind that any of these choices are moderated by the fact that we are simply computing a blending function that is the solution to the (modified) Laplace equation. It will have more or less the correct shape --- even the unmodified Laplace solution is not a terrible blending function. The modification is designed simply to help preserve mesh sizes near boundaries. I paste below a snippet of code that computes the average thickness of elements near the moving wall. Paul integer e,f nxz = nx1*nz1 nxyz = nx1*ny1*nz1 n = nxyz*nelv srfbl = 0. ! Surface area of elements in b.l. volbl = 0. ! Volume of elements in boundary layer do e=1,nelv do f=1,nface if (cbc(f,e,1).eq.'mv ') then srfbl = srfbl + vlsum(area(1,1,f,e),nxz ) volbl = volbl + vlsum(bm1 (1,1,1,e),nxyz) endif enddo enddo srfbl = glsum(srfbl,1) ! Sum over all processors volbl = glsum(volbl,1) delta = volbl / srfbl ! Avg thickness of b.l. elements call rone (h1,n) call rzero(h2,n) call cheap_dist(d,1,'mv ') deltap = 2*delta ! Protected b.l. thickness do i=1,n arg = -(d(i)/deltap)**2 h1(i) = h1(i) + 9*exp(arg) enddo From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Monday, November 27, 2017 7:52 AM To: nek5000-users Subject: [Nek5000-users] mesh deformation Hello Nek, I was trying to validate moving cylinder test cases (ocyl2.usr more specifically). I could validate it properly. Even though I have some doubts regarding the mesh motion. Mesh deformation more precisely. I have seen a function to send the deformation to far field apart from the near wall region (Laplace equation with h1 as blending coefficient).So in order to find h1 in the equation there is a parameter called delta (which is 2 in this case). I have also read a paper regarding this equation. In that paper its clearly states that it is the average thickness of the element which is close to the object. So my question is, is there anyway to calculate it automatically or easily to calculate the thickness of that particular elements from any input files? because if I change the mesh (suppose if I refine the mesh size everything will change right?) how can I find the new average thickness of the elements which is close to the object? or how can I measure the area or volume of the elements which is close to the wall. so could you please help me with this? Thanks with regards Sijo George _______________________________________________ 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 Nov 27 10:08:46 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 27 Nov 2017 17:08:46 +0100 (CET) Subject: [Nek5000-users] mesh deformation In-Reply-To: References: Message-ID: Sorry again and my last question is, Finally I want to use a Naca profile, So the deltap = 2* delta would be still valid ? Or I have to use anyother scaling information? regards Sijo GEORGE De: "nek5000-users" ?: "nek5000-users" Envoy?: Lundi 27 Novembre 2017 16:54:55 Objet: Re: [Nek5000-users] mesh deformation Thanks you. But I am sorry I was not able to follow your answer. But could you please tell me that, if i write like this n = nx1*ny1*nz1*nelv do i=1,n v = volume(i,1,1,1) x = xm1(i,1,1,1,) enddo the volume which gives corresponding the elements which has the position of x. Regars Sijo GEORGE De: "nek5000-users" ?: "nek5000-users" Envoy?: Lundi 27 Novembre 2017 16:35:03 Objet: Re: [Nek5000-users] mesh deformation Dear Sijo, There's no reason to have h1 uniform in each element, so I did not bother to do that. In the SEM we usually view fields as a continuum and rarely concern with element-to-element variations as those would then depend on the mesh itself. There is no hard and fast rule here, save that you probably want the mesh deformation to be smooth within each element. hth, Paul From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Monday, November 27, 2017 9:17:26 AM To: nek5000-users Subject: Re: [Nek5000-users] mesh deformation Thanks Paul thank you very much. I could understand the things what you mentioend about the laplace equation since I have read your publication reagrding moving boundary and ALE formulation. I am gonna try it on my FSI problem with 2D single cylinder. Since its been some time I had also developed a code which pushes the most of the deformation to the outer region: Here it is: v = volume(1,1,1,1) a=glmax(v,1) b=glmin(v,1) L = 1 do i=1,n vdiv = b/a num = 1-vidv denom = 1/a h1(i) = h1(i) + ((1- (b/a) **L ) / (( denom * v ) **L)) enddo which also works same as like you have created (I have only verified visually). But I have some questions regarding this: Before my doubts I have to be clear on one thing. Here I will be talking about two different elements. First one is Elements (lets say I have 1348 elements in my domain) and Second is Spectral elements (6*6*1348=48528) where nx1 = 6. 1) Since the loop runs from 1 to n, the blending coefficient also makes h1 for all speactral elements right? 2) In that case, my blending coefficient is uniform across a particular Element of the domain? more specifically lets say my element number is 1A and in 1A there are 36 spectral elements of different size. As per my code above, is my blending coefficient same for all the spectral elements which is inside 1A element?or is it different?(because I consider the volume of each element as "v") My idea was to give different h1 coefficient for each Elements and give constant h1 to spectral elements in each elements. More precisily lets sat there are 2 elements called 1A and 1B. 1A has a h1 = 2.0 so all the spectral elements in the 1A has h1 of 2.0 and 1 B has h1 = 10.0 so all the spectral elements in the 1B has h1 of 10.0 I know its a quite big email. Even though could you please correct me if I am wrong. Because gradually I am going to run a very complex problem. So I Have to be sure in these cases before starting it. Thanks with regards Sijo George De: "nek5000-users" ?: "nek5000-users" Envoy?: Lundi 27 Novembre 2017 15:47:35 Objet: Re: [Nek5000-users] mesh deformation Dear Sijo, In the ocyl2 case, the delta is based on a dimension that is related to the model. (Here, delta=2*D, where D=1 is the diameter of the cylinder.) Such a choice would be reasonable for all flow past a cylinder cases for that particular geometry, independent of the mesh resolution. The boundary-layer thickness criterion is important when there are multiple objects that are potentially touching. Keep in mind that any of these choices are moderated by the fact that we are simply computing a blending function that is the solution to the (modified) Laplace equation. It will have more or less the correct shape --- even the unmodified Laplace solution is not a terrible blending function. The modification is designed simply to help preserve mesh sizes near boundaries. I paste below a snippet of code that computes the average thickness of elements near the moving wall. Paul integer e,f nxz = nx1*nz1 nxyz = nx1*ny1*nz1 n = nxyz*nelv srfbl = 0. ! Surface area of elements in b.l. volbl = 0. ! Volume of elements in boundary layer do e=1,nelv do f=1,nface if (cbc(f,e,1).eq.'mv ') then srfbl = srfbl + vlsum(area(1,1,f,e),nxz ) volbl = volbl + vlsum(bm1 (1,1,1,e),nxyz) endif enddo enddo srfbl = glsum(srfbl,1) ! Sum over all processors volbl = glsum(volbl,1) delta = volbl / srfbl ! Avg thickness of b.l. elements call rone (h1,n) call rzero(h2,n) call cheap_dist(d,1,'mv ') deltap = 2*delta ! Protected b.l. thickness do i=1,n arg = -(d(i)/deltap)**2 h1(i) = h1(i) + 9*exp(arg) enddo From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Monday, November 27, 2017 7:52 AM To: nek5000-users Subject: [Nek5000-users] mesh deformation Hello Nek, I was trying to validate moving cylinder test cases (ocyl2.usr more specifically). I could validate it properly. Even though I have some doubts regarding the mesh motion. Mesh deformation more precisely. I have seen a function to send the deformation to far field apart from the near wall region (Laplace equation with h1 as blending coefficient).So in order to find h1 in the equation there is a parameter called delta (which is 2 in this case). I have also read a paper regarding this equation. In that paper its clearly states that it is the average thickness of the element which is close to the object. So my question is, is there anyway to calculate it automatically or easily to calculate the thickness of that particular elements from any input files? because if I change the mesh (suppose if I refine the mesh size everything will change right?) how can I find the new average thickness of the elements which is close to the object? or how can I measure the area or volume of the elements which is close to the wall. so could you please help me with this? Thanks with regards Sijo George _______________________________________________ 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 Mon Nov 27 12:11:24 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 27 Nov 2017 13:11:24 -0500 Subject: [Nek5000-users] Conjugate Problem Message-ID: Hello Neks, I am working on a conjugate problem with Nek5000. Is there any reference that discusses how nek5000 solves the conjugate heat transfer problem? Thanks. Mu Xu -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Nov 27 14:55:02 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 27 Nov 2017 20:55:02 +0000 Subject: [Nek5000-users] Conjugate Problem In-Reply-To: References: Message-ID: Dear Mu, The best way to think of CHT is as if solving an advection-diffusion equation with u==0 in the solid. In Nek5000, the fluid elements are numbered first (both globally, and on each processor) so that an fluid/pressure updates are performed on elements e=1,...,nelv; while thermal updates are performed on elements e=1,...,nelt > nelv. Your mesh must be build such that all solid elements are numbered after all fluid elements. hth, Paul ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Monday, November 27, 2017 12:11:24 PM To: nek5000-users at lists.mcs.anl.gov Subject: [Nek5000-users] Conjugate Problem Hello Neks, I am working on a conjugate problem with Nek5000. Is there any reference that discusses how nek5000 solves the conjugate heat transfer problem? Thanks. Mu Xu -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Nov 27 18:59:56 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 27 Nov 2017 18:59:56 -0600 Subject: [Nek5000-users] Simulating two isothermal fluids with different densities Message-ID: Hi Nek users, I am simulating two miscible fluids with different density and viscosity. There is no temperature difference. I am using the Boussinesq approximation to account for buoyancy within userf. And to calculate density, I am transporting concentration (kmol/m3) using the "temperature" scalar field. Attached is an example box case (4 walls with no slip and zero flux BCs). For ICs, the left half of the box is one concentration and the right half is another. Fluid then moves due to buoyancy. Density and viscosity are calculated as a function of concentration. These equations work well with larger mass diffusivity (D), specified in p08. However when I decrease this value, the maximum and minimum concentrations exceed the initial values. If I want to consider a real fluid, D needs to be extremely small (eg for water-ethyl, D=1e-9 m2/s). In the attached example, the concentration scalar diverges. Can somebody help me understand why concentration diverges with low D values and how to fix it? If I am approaching this problem incorrectly, please let me know. Thank you, Jonathan -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: boussinesq_box_example.zip Type: application/zip Size: 164562 bytes Desc: not available URL: From nek5000-users at lists.mcs.anl.gov Tue Nov 28 11:27:29 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 28 Nov 2017 18:27:29 +0100 Subject: [Nek5000-users] All elements deformed :: regenerate geometry data Message-ID: Hi, I have prepared a 3D mesh in GAMBIT and used the matlab converter to make the .rea and .map file. When I use these files for computation, things appear to go on nicely and the flow field that I get is not different from what I expect. However, I notice the following lines in the output file which says that the geometry has been regenerated. What does this mean? Would it comprise the quality of the mesh near the wall? Reading checkpoint data byte swap: F 6.543210 -2.9312772E+35 FILE: restart.f00001 0 2.7300E+02 done :: Read checkpoint data avg data-throughput = 137.8MBps io-nodes = 1736 xyz min -3.5000 0.0000 -10.000 uvwpt min -1.5151 -1.7105 -1.6631 -0.87366 0.0000 PS min 0.0000 0.99000E+22 xyz max 60.000 3.1200 10.000 uvwpt max 2.2376 0.91592 1.7789 0.47959 0.0000 PS max 0.0000 -0.99000E+22 Restart: recompute geom. factors. regenerate geometry data 1 NOTE: All elements deformed , param(59) ^=0 done :: regenerate geometry data 1 done :: set initial conditions call userchk Many thanks, NN. From nek5000-users at lists.mcs.anl.gov Tue Nov 28 13:20:05 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 28 Nov 2017 14:20:05 -0500 Subject: [Nek5000-users] Conjugate Problem In-Reply-To: References: Message-ID: Hello, Paul Thank you. This is helpful. Is there any document that discuss the detail about the algorithm of conjugate problem in Nek5000. We can read and reference this document. Mu Xu On Mon, Nov 27, 2017 at 3:55 PM, wrote: > > Dear Mu, > > > The best way to think of CHT is as if solving an advection-diffusion > equation with u==0 in the solid. > > > In Nek5000, the fluid elements are numbered first (both globally, and on > each processor) so that an fluid/pressure updates are performed on elements > e=1,...,nelv; while thermal updates are performed on elements e=1,...,nelt > > nelv. > > > Your mesh must be build such that all solid elements are numbered after > all fluid elements. > > > hth, > > > Paul > > > ------------------------------ > *From:* Nek5000-users on behalf > of nek5000-users at lists.mcs.anl.gov > *Sent:* Monday, November 27, 2017 12:11:24 PM > *To:* nek5000-users at lists.mcs.anl.gov > *Subject:* [Nek5000-users] Conjugate Problem > > Hello Neks, > > I am working on a conjugate problem with Nek5000. Is there any reference > that discusses how > > nek5000 solves the conjugate heat transfer problem? > > Thanks. > > Mu Xu > > _______________________________________________ > 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 Wed Nov 29 03:40:20 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 29 Nov 2017 10:40:20 +0100 Subject: [Nek5000-users] Conjugate Problem In-Reply-To: References: Message-ID: Hi, There is example in NekExample repository called conj_ht. You can take a look at it. Regards Adam On 28/11/17 20:20, nek5000-users at lists.mcs.anl.gov wrote: > Hello, Paul > > Thank you. This is helpful. Is there any document that discuss the > detail about the algorithm of conjugate problem in Nek5000. We can > read and reference this document. > > Mu Xu > > On Mon, Nov 27, 2017 at 3:55 PM, > wrote: > > > Dear?Mu, > > > The best way to think of ?CHT is as if?solving an > advection-diffusion equation with u==0 in the solid. > > > In Nek5000, the fluid elements are numbered first (both globally, > and on each processor) so that an fluid/pressure updates are > performed on elements e=1,...,nelv; while thermal updates are > performed on elements e=1,...,nelt > nelv. > > > Your mesh must be build such that all solid elements are numbered > after all fluid elements. > > > hth, > > > Paul > > ------------------------------------------------------------------------ > *From:* Nek5000-users > on behalf of > nek5000-users at lists.mcs.anl.gov > > > > *Sent:* Monday, November 27, 2017 12:11:24 PM > *To:* nek5000-users at lists.mcs.anl.gov > > *Subject:* [Nek5000-users] Conjugate Problem > Hello Neks, > > I am working on a conjugate problem with Nek5000. Is there any > reference that discusses how > > nek5000 solves the conjugate heat transfer problem? > > Thanks. > > Mu Xu > > _______________________________________________ > 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 Wed Nov 29 15:03:09 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 29 Nov 2017 21:03:09 +0000 Subject: [Nek5000-users] avg_all averaging Message-ID: Hi nek, May I know how I can use the files generated by calling avg_all subroutine in userchk to get the averaged velocities? Bascially, I want to have some plot some lines. Kind regards, ZJ -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Nov 28 13:02:53 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 29 Nov 2017 00:32:53 +0530 Subject: [Nek5000-users] Nek5000 Low Mach Jet Flow Message-ID: Hello Sir, Need to solve Low Mach number Helium (or Hydrogen ) jet opening to atmospheric air. Need to solve the species diffusion equation for this. Is it possible to solve using Nek5000? If yes,how? Thank you in advance -- Nitish Kovalam Aerospace Engineering Indian Institute of Space Science and Technology -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Nov 29 12:14:31 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 29 Nov 2017 18:14:31 +0000 Subject: [Nek5000-users] Failing to converge when running in parallel Message-ID: When running with mpi, I am currently experiencing an error with helmholtz which causes the solution to blow up quickly. The issue seems to disappear when running on lower mpi ranks. Is there a known issue which may cause this, or a work around I can employ? I haven't seen this before for similar mesh sizes. This case is a small mesh, but I plan on upscaling quite a bit. See attached for the run log and user file. Thanks, Jefferson Davis -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: run.log Type: text/x-log Size: 60107 bytes Desc: run.log URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: turbChannel.usr Type: application/octet-stream Size: 38173 bytes Desc: turbChannel.usr URL: From nek5000-users at lists.mcs.anl.gov Wed Nov 29 21:26:10 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 30 Nov 2017 06:26:10 +0300 Subject: [Nek5000-users] =?utf-8?q?tools_in_new_version?= Message-ID: Hi, Neks! I am trying to create a complex geometry and use nektools. I use nekmerge2 script to merge two parts of my mesh, and then I want to use reatore2 and genmap. In previous version the same script worked find, but now (~/Nek5000/bin/) reatore2 gives a note:? Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DENORMAL and after that genmap gives a error: ABORT: SELF-CHK 1 2 4 0 Try to tighten the mesh tolerance! I have tested this script with old tools (on another computer) and it's still works fine. Do you have any ideas why does this problem occur? Best regards, Mark -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Nov 29 23:01:23 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 30 Nov 2017 05:01:23 +0000 Subject: [Nek5000-users] avg_all averaging Message-ID: Hi ZJ, A quick search of the archives with keyword ?avg_all? will give you answer to related questions: https://lists.mcs.anl.gov/mailman/mmsearch/nek5000-users?config=nek5000-users.htsearch&restrict=&exclude=&method=and&format=short&sort=score&words=avg_all avg_all is a collection of time-averaged information in the avg, rms and rm2 files. The avg files has time-averaged u,v and w; rms has time-averaged u^2, v^2 and w^2; and rm2 has uv, vw, and uw. The avg_all routines is located in navier5.f in case you want to get more information. Typically the first step in post processing is to use the routine ?auto_averager? to combine all the files output from avg_all. autoaverager takes as input the name of a file (example: fname1.list) which has a list of all the files that you want to combine (example: the contents of fname1.list will be: avgcasename0.f00001 avgcasename0.f00002 ? Note- there should be only 1 filename per line in the fname1.list file). You can use the auto_averager routine to combine all your avgcasename0.f00* files and output them to a single file or do more post-processing with it. The quantities available in avg, rms and rm2 files are sufficient to get the actual urms, vrms velocities etc? Additionally, once you time-average the data using auto_averager, you can use space averaging routines such as z_avg etc? to average quantities in homogeneous direction of your domain. It would also help if you specify what exactly are you trying to get from these avg_all files i.e. are you looking for rms velocities or something else? Here is a snippet of my code that I have used in the past to post-process these files: call auto_averager(fname1) call copy(uk,vx) call copy(vk,vy) call copy(wk,vz) call col2(uk,uk,lt) call col2(vk,vk,lt) call col2(wk,wk,lt) call auto_averager(fname2) call sub2(vx,uk,lt) call sub2(vy,vk,lt) call sub2(vz,wk,lt) call z_avg(vxa,vx,gs_avg_hndl,nelxy,ifld) call z_avg(vya,vy,gs_avg_hndl,nelxy,ifld) call z_avg(vza,vz,gs_avg_hndl,nelxy,ifld) call outpost(vxa,vya,vza,vxa,vya,' ') Here fname1 has the list of all avgcasename0.f00* files and fname2 has the list of all the rmscasename0.f00* files. Ketan -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Thu Nov 30 04:11:43 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 30 Nov 2017 11:11:43 +0100 Subject: [Nek5000-users] tools in new version In-Reply-To: References: Message-ID: Can you try again with the latest version on GitHub (just submitted a potential fix). -----Original message----- > From:nek5000-users at lists.mcs.anl.gov > Sent: Thursday 30th November 2017 4:27 > To: nek5000-users at lists.mcs.anl.gov > Subject: [Nek5000-users] tools in new version > > Hi, Neks! > > I am trying to create a complex geometry and use nektools. I use nekmerge2 script to merge two parts of my mesh, and then I want to use reatore2 and genmap. In previous version the same script worked find, but now (~/Nek5000/bin/) reatore2 gives a note:? > > Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DENORMAL > > and after that genmap gives a error: > > ABORT: SELF-CHK 1 2 4 0 > Try to tighten the mesh tolerance! > > I have tested this script with old tools (on another computer) and it's still works fine. Do you have any ideas why does this problem occur? > > Best regards, > Mark > > > > _______________________________________________ > 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 Nov 30 17:05:30 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 30 Nov 2017 23:05:30 +0000 Subject: [Nek5000-users] Nek5000-users Digest, Vol 105, Issue 32 In-Reply-To: References: Message-ID: Dear Ketan, Many thanks for your reply. Basically I want to get the averaged velocity, rms velocities and shear stress where I got two questions. 1. Once the subroutines were writen in usr file, Is calling avg_all still needed? In other words, the post-processing is done after longtime running, and usr file is changed instead of calling avg_all, isn't it? 2. After trying the subroutines you provided, there isnt no files generated in the directory. Can you give me more infomation about the averaging process. Best wishes, ZJ On 30 November 2017 at 18:00, wrote: > Send Nek5000-users mailing list submissions to > nek5000-users at lists.mcs.anl.gov > > To subscribe or unsubscribe via the World Wide Web, visit > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > or, via email, send a message with subject or body 'help' to > nek5000-users-request at lists.mcs.anl.gov > > You can reach the person managing the list at > nek5000-users-owner at lists.mcs.anl.gov > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Nek5000-users digest..." > > > Today's Topics: > > 1. tools in new version (nek5000-users at lists.mcs.anl.gov) > 2. Re: avg_all averaging (nek5000-users at lists.mcs.anl.gov) > 3. Re: tools in new version (nek5000-users at lists.mcs.anl.gov) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Thu, 30 Nov 2017 06:26:10 +0300 > From: nek5000-users at lists.mcs.anl.gov > To: nek5000-users at lists.mcs.anl.gov > Subject: [Nek5000-users] tools in new version > Message-ID: > > Content-Type: text/plain; charset="utf-8" > > Hi, Neks! > > I am trying to create a complex geometry and use nektools. I use nekmerge2 > script to merge two parts of my mesh, and then I want to use reatore2 and > genmap. In previous version the same script worked find, but now > (~/Nek5000/bin/) reatore2 gives a note:? > > Note: The following floating-point exceptions are signalling: > IEEE_INVALID_FLAG IEEE_DENORMAL > > and after that genmap gives a error: > > ABORT: SELF-CHK 1 2 4 0 > Try to tighten the mesh tolerance! > > I have tested this script with old tools (on another computer) and it's > still works fine. Do you have any ideas why does this problem occur? > > Best regards, > Mark > > > > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: attachments/20171130/1450ceba/attachment-0001.html> > > ------------------------------ > > Message: 2 > Date: Thu, 30 Nov 2017 05:01:23 +0000 > From: nek5000-users at lists.mcs.anl.gov > To: "nek5000-users at lists.mcs.anl.gov" > > Subject: Re: [Nek5000-users] avg_all averaging > Message-ID: > > Content-Type: text/plain; charset="utf-8" > > Hi ZJ, > > A quick search of the archives with keyword ?avg_all? will give you answer > to related questions: > https://lists.mcs.anl.gov/mailman/mmsearch/nek5000- > users?config=nek5000-users.htsearch&restrict=&exclude=& > method=and&format=short&sort=score&words=avg_all > > avg_all is a collection of time-averaged information in the avg, rms and > rm2 files. The avg files has time-averaged u,v and w; rms has time-averaged > u^2, v^2 and w^2; and rm2 has uv, vw, and uw. The avg_all routines is > located in navier5.f in case you want to get more information. > > Typically the first step in post processing is to use the routine > ?auto_averager? to combine all the files output from avg_all. autoaverager > takes as input the name of a file (example: fname1.list) which has a list > of all the files that you want to combine (example: the contents of > fname1.list will be: avgcasename0.f00001 avgcasename0.f00002 ? Note- there > should be only 1 filename per line in the fname1.list file). You can use > the auto_averager routine to combine all your avgcasename0.f00* files and > output them to a single file or do more post-processing with it. The > quantities available in avg, rms and rm2 files are sufficient to get the > actual urms, vrms velocities etc? Additionally, once you time-average the > data using auto_averager, you can use space averaging routines such as > z_avg etc? to average quantities in homogeneous direction of your domain. > > It would also help if you specify what exactly are you trying to get from > these avg_all files i.e. are you looking for rms velocities or something > else? > > Here is a snippet of my code that I have used in the past to post-process > these files: > call auto_averager(fname1) > call copy(uk,vx) > call copy(vk,vy) > call copy(wk,vz) > call col2(uk,uk,lt) > call col2(vk,vk,lt) > call col2(wk,wk,lt) > call auto_averager(fname2) > call sub2(vx,uk,lt) > call sub2(vy,vk,lt) > call sub2(vz,wk,lt) > call z_avg(vxa,vx,gs_avg_hndl,nelxy,ifld) > call z_avg(vya,vy,gs_avg_hndl,nelxy,ifld) > call z_avg(vza,vz,gs_avg_hndl,nelxy,ifld) > call outpost(vxa,vya,vza,vxa,vya,' ') > > Here fname1 has the list of all avgcasename0.f00* files and fname2 has the > list of all the rmscasename0.f00* files. > > > Ketan > > > > > > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: attachments/20171130/b19f2925/attachment-0001.html> > > ------------------------------ > > Message: 3 > Date: Thu, 30 Nov 2017 11:11:43 +0100 > From: nek5000-users at lists.mcs.anl.gov > To: nek5000-users at lists.mcs.anl.gov > Subject: Re: [Nek5000-users] tools in new version > Message-ID: > > Content-Type: text/plain; charset=utf-8 > > Can you try again with the latest version on GitHub (just submitted a > potential fix). > > -----Original message----- > > From:nek5000-users at lists.mcs.anl.gov > > Sent: Thursday 30th November 2017 4:27 > > To: nek5000-users at lists.mcs.anl.gov > > Subject: [Nek5000-users] tools in new version > > > > Hi, Neks! > > > > I am trying to create a complex geometry and use nektools. I use > nekmerge2 script to merge two parts of my mesh, and then I want to use > reatore2 and genmap. In previous version the same script worked find, but > now (~/Nek5000/bin/) reatore2 gives a note:? > > > > Note: The following floating-point exceptions are signalling: > IEEE_INVALID_FLAG IEEE_DENORMAL > > > > and after that genmap gives a error: > > > > ABORT: SELF-CHK 1 2 4 0 > > Try to tighten the mesh tolerance! > > > > I have tested this script with old tools (on another computer) and it's > still works fine. Do you have any ideas why does this problem occur? > > > > Best regards, > > Mark > > > > > > > > _______________________________________________ > > Nek5000-users mailing list > > Nek5000-users at lists.mcs.anl.gov > > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > > > ------------------------------ > > Subject: Digest Footer > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > > > ------------------------------ > > End of Nek5000-users Digest, Vol 105, Issue 32 > ********************************************** > -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Nov 1 01:22:22 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 01 Nov 2017 06:22:22 -0000 Subject: [Nek5000-users] =?utf-8?q?Axis_in_spectral_elements?= Message-ID: Hi, Neks! I am working with a complex geometry and want to change positions of some points in spectral elements. I've read in documentation that i,j,k,e in xm1(i,j,k,e), for example, are changed from 1 to nx1,ny1,nz1 and nelv respectively. But during my tests it's seemed that in different elements x,y,z axis not always correspond to the global coordinate system. Am I right? And how could ?I change only the lowest slice of an element, for example? Best regards, Elizabeth -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Nov 15 11:06:54 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 15 Nov 2017 17:06:54 -0000 Subject: [Nek5000-users] (no subject) Message-ID: Hi, Neks! I am working with a complex geometry witch was built in gambit. I want to know a parameter "iside" for each of faces ( outflow, inflow and all walls) on zero step. If I ask in userbc ( in *.usr file) write "iside", I receive only iside=1,?that matches inflow. It means that faces with type "wall" don't have "iside" value,am i right?? And how?each surface is assigned value "iside"? Best regards, Elizabeth -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Nov 15 11:29:21 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 15 Nov 2017 17:29:21 -0000 Subject: [Nek5000-users] =?utf-8?q?Question_about_iside_value?= Message-ID: Hi, Neks! I am working with a complex geometry witch was built in gambit. I want to know a parameter "iside" for each of faces ( outflow, inflow and all walls) on zero step. If I ask in userbc ( in *.usr file) write "iside", I receive only iside=1,?that matches inflow. It means that faces with type "wall" don't have "iside" value,am i right?? And how?each surface is assigned value "iside"? Best regards, Elizabeth -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Nov 21 21:12:26 2017 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 22 Nov 2017 03:12:26 -0000 Subject: [Nek5000-users] =?utf-8?q?=D0=A1ollocation_of_i=2Cj=2Ck_in_each_s?= =?utf-8?q?pectral_element?= Message-ID: Hi, Neks! I am working with a complex geometry witch was built in gambit. And I want to change positions of some points in spectral elements. I've read in documentation that i,j,k,e in xm1(i,j,k,e), for example, are changed from 1 to nx1,ny1,nz1 and nelv respectively. But during my tests it's seemed that in different elements x,y,z axis not always correspond to the global coordinate system. Am I right? And how could ?I find out the?principle of collocation of i,j,k in each spectral element? ? ? ? ? ? ? ? Best regards, Elizabeth -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: SHeqBH5bWnY.jpg Type: image/jpeg Size: 31884 bytes Desc: not available URL: