From nek5000-users at lists.mcs.anl.gov Thu Feb 1 12:56:59 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 1 Feb 2018 18:56:59 +0000 Subject: [Nek5000-users] Clarification on PN-PN formulation with IFCHAR=T Message-ID: Dear all, I am doing DNS of an incompressible flow using the PN-PN formulation with IFCHAR=T. We chose the PN-PN formulation because we wanted to ensure continuous pressure field. We chose the characteristics based time-stepping for stability. More than a question, I would like to see if I am understanding the solution process correctly. I've read the "NEK_doc.pdf" file dated from July 2015 which contains information on the PN-PN formulation, and Dr. Fischer's notes on implementation of the OIFS/characteristics approach. I've also looked at the code. >From what I can tell, the main difference when using PN-PN with IFCHAR=T vs. IFCHAR=F is that instead of having the non-linear terms on the RHS (IFCHAR=F) (see Eq. 7.7 from NEK_doc.pdf), we have the \tilde{u}^n-q (q=1,...,k) terms from the material derivative which act as an explicit forcing term. Then, the pressure field is solved, decoupled from velocity, using Eq. 7.13 from NEK_doc.pdf. The computed pressure is then used to solve the Helmholtz system for the velocity field. Does this seem correct? Specially the part of the pressure solve being decoupled from velocity. From the code, it definitely seems that way. Also, if the flow is incompressible, that means Q_T in eq 7.13 is zero, but shouldn't grad(Q_T) be zero as well? Thanks for your attention, Juan Diego -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Fri Feb 2 10:50:08 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 2 Feb 2018 16:50:08 +0000 Subject: [Nek5000-users] (no subject) Message-ID: Hi all, Further my last request, since rm2 files contain the cross product running averages, how I can know the time-averaged , , and exactly? I know there is a way by calling auto_averager subroutine, but it doesn't have a good performance. To get the shear stress, what I did for the averaging is, 1. call averaging avg0f.0000*** files to get mean of u, v and w; 2. call averaging rm20f,0000** files to get mean of , , and (but I don't know exactly what they are since I dont know the order of results. 3. sqrt(mean-mean*mean) Do you think the process is correct? Kind regards, Jian -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Fri Feb 2 10:51:09 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 2 Feb 2018 16:51:09 +0000 Subject: [Nek5000-users] shear stress and auto_averager Message-ID: Hi all, Further my last request, since rm2 files contain the cross product running averages, how I can know the time-averaged , , and exactly? I know there is a way by calling auto_averager subroutine, but it doesn't have a good performance. To get the shear stress, what I did for the averaging is, 1. call averaging avg0f.0000*** files to get mean of u, v and w; 2. call averaging rm20f,0000** files to get mean of , , and (but I don't know exactly what they are since I dont know the order of results. 3. sqrt(mean-mean*mean) Do you think the process is correct? Kind regards, Jian -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Fri Feb 2 13:10:32 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sat, 3 Feb 2018 00:40:32 +0530 Subject: [Nek5000-users] Outflow Boundary Condition Clarification Message-ID: Im trying to simulate the formation of shear layer when a jet exits a nozzle and enters the flow domain. As the FIRST vortex leaves the outlet ( at which I have provided 'O' boundary condition using genbox) the solution diverges. I tried to implement the 'turbulent outflow' subroutine as suggested in the forum but now the vortex affects the upstream flow. But when I run the flow past circular cylinder example, the vortices leave the domain without any issue even though only 'O' boundary condition is used'. What is the difference between the two cases and what can i do to prevent the issue? -- 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 Fri Feb 2 13:27:50 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 2 Feb 2018 20:27:50 +0100 Subject: [Nek5000-users] Outflow Boundary Condition Clarification In-Reply-To: References: Message-ID: I think you should try the stabilised O condition, as described in a paper by Dong (2014). We have an implementation of that, we should probably provide it as a module in the repository. best regards, Philipp On 2018-02-02 20:10, nek5000-users at lists.mcs.anl.gov wrote: > Im trying to simulate the formation of shear layer when a jet exits a > nozzle and enters the flow domain. As the FIRST vortex leaves the outlet > ( at which I have provided 'O' boundary condition using genbox) the > solution diverges. I tried to implement the 'turbulent outflow' > subroutine as suggested in the forum but now the vortex affects the > upstream flow. But when I run the flow past circular cylinder example, > the vortices leave the domain without any issue even though only 'O' > boundary condition is used'. What is the difference between the two > cases and what can i do to prevent the issue? > > -- > Nitish Kovalam > Aerospace Engineering > Indian Institute of Space Science and Technology > > > _______________________________________________ > 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 Feb 2 14:19:54 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 2 Feb 2018 20:19:54 +0000 Subject: [Nek5000-users] Outflow Boundary Condition Clarification In-Reply-To: References: Message-ID: Attached is a .usr with turb_outflow that works for this case---there is no difficulty if you use the prescribed input values, as shown in the attached files. Here is a movie of the result - you can see that turb_outflow is turned on prior to the jet reaching the top of the domain. http://mcs.anl.gov/~obabko/jet_vm.mpeg This mesh is a bit under-resolved in the z-direction. The files for a new case, with more axial resolution, are posted here www.mcs.anl.gov/~fischer/jet2.tgz I'll be posting an example with a script on how to build this geometry to nekexamples soon. Paul ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Friday, February 2, 2018 1:10:32 PM To: nek5000-users at lists.mcs.anl.gov Subject: [Nek5000-users] Outflow Boundary Condition Clarification Im trying to simulate the formation of shear layer when a jet exits a nozzle and enters the flow domain. As the FIRST vortex leaves the outlet ( at which I have provided 'O' boundary condition using genbox) the solution diverges. I tried to implement the 'turbulent outflow' subroutine as suggested in the forum but now the vortex affects the upstream flow. But when I run the flow past circular cylinder example, the vortices leave the domain without any issue even though only 'O' boundary condition is used'. What is the difference between the two cases and what can i do to prevent the issue? -- Nitish Kovalam Aerospace Engineering Indian Institute of Space Science and Technology -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: jet2.usr Type: application/octet-stream Size: 18256 bytes Desc: jet2.usr URL: From nek5000-users at lists.mcs.anl.gov Fri Feb 2 22:33:35 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sat, 3 Feb 2018 10:03:35 +0530 Subject: [Nek5000-users] Nek5000-users Digest, Vol 108, Issue 2 In-Reply-To: References: Message-ID: Thank you Paul and Phillip for such a quick reply. I will run the jet case attached. So why was there no such problem when the first vortex from the wake of a circular cylinder was leaving the domain? On Sat, Feb 3, 2018 at 1:49 AM, 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. Outflow Boundary Condition Clarification > (nek5000-users at lists.mcs.anl.gov) > 2. Re: Outflow Boundary Condition Clarification > (nek5000-users at lists.mcs.anl.gov) > 3. Re: Outflow Boundary Condition Clarification > (nek5000-users at lists.mcs.anl.gov) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Sat, 3 Feb 2018 00:40:32 +0530 > From: nek5000-users at lists.mcs.anl.gov > To: nek5000-users at lists.mcs.anl.gov > Subject: [Nek5000-users] Outflow Boundary Condition Clarification > Message-ID: > > Content-Type: text/plain; charset="utf-8" > > Im trying to simulate the formation of shear layer when a jet exits a > nozzle and enters the flow domain. As the FIRST vortex leaves the outlet ( > at which I have provided 'O' boundary condition using genbox) the solution > diverges. I tried to implement the 'turbulent outflow' subroutine as > suggested in the forum but now the vortex affects the upstream flow. But > when I run the flow past circular cylinder example, the vortices leave the > domain without any issue even though only 'O' boundary condition is used'. > What is the difference between the two cases and what can i do to prevent > the issue? > > -- > Nitish Kovalam > Aerospace Engineering > Indian Institute of Space Science and Technology > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: attachments/20180203/87028828/attachment-0001.html> > > ------------------------------ > > Message: 2 > Date: Fri, 2 Feb 2018 20:27:50 +0100 > From: nek5000-users at lists.mcs.anl.gov > To: nek5000-users at lists.mcs.anl.gov > Subject: Re: [Nek5000-users] Outflow Boundary Condition Clarification > Message-ID: > > Content-Type: text/plain; charset=utf-8; format=flowed > > I think you should try the stabilised O condition, as described in a > paper by Dong (2014). We have an implementation of that, we should > probably provide it as a module in the repository. > > best regards, > Philipp > > On 2018-02-02 20:10, nek5000-users at lists.mcs.anl.gov wrote: > > Im trying to simulate the formation of shear layer when a jet exits a > > nozzle and enters the flow domain. As the FIRST vortex leaves the outlet > > ( at which I have provided 'O' boundary condition using genbox) the > > solution diverges. I tried to implement the 'turbulent outflow' > > subroutine as suggested in the forum but now the vortex affects the > > upstream flow. But when I run the flow past circular cylinder example, > > the vortices leave the domain without any issue even though only 'O' > > boundary condition is used'. What is the difference between the two > > cases and what can i do to prevent the issue? > > > > -- > > Nitish Kovalam > > Aerospace Engineering > > Indian Institute of Space Science and Technology > > > > > > _______________________________________________ > > Nek5000-users mailing list > > Nek5000-users at lists.mcs.anl.gov > > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > > > > > ------------------------------ > > Message: 3 > Date: Fri, 2 Feb 2018 20:19:54 +0000 > From: nek5000-users at lists.mcs.anl.gov > To: "nek5000-users at lists.mcs.anl.gov" > > Subject: Re: [Nek5000-users] Outflow Boundary Condition Clarification > Message-ID: > > Content-Type: text/plain; charset="us-ascii" > > > Attached is a .usr with turb_outflow that works for this case---there is > > no difficulty if you use the prescribed input values, as shown in the > attached > > files. > > > Here is a movie of the result - you can see that turb_outflow is turned on > prior to the jet reaching the top of the domain. > > > http://mcs.anl.gov/~obabko/jet_vm.mpeg > > > This mesh is a bit under-resolved in the z-direction. > > The files for a new case, with more axial resolution, are posted here > > > www.mcs.anl.gov/~fischer/jet2.tgz > > > > I'll be posting an example with a script on how to build this geometry to > nekexamples soon. > > > Paul > > > ________________________________ > From: Nek5000-users on behalf > of nek5000-users at lists.mcs.anl.gov > Sent: Friday, February 2, 2018 1:10:32 PM > To: nek5000-users at lists.mcs.anl.gov > Subject: [Nek5000-users] Outflow Boundary Condition Clarification > > Im trying to simulate the formation of shear layer when a jet exits a > nozzle and enters the flow domain. As the FIRST vortex leaves the outlet ( > at which I have provided 'O' boundary condition using genbox) the solution > diverges. I tried to implement the 'turbulent outflow' subroutine as > suggested in the forum but now the vortex affects the upstream flow. But > when I run the flow past circular cylinder example, the vortices leave the > domain without any issue even though only 'O' boundary condition is used'. > What is the difference between the two cases and what can i do to prevent > the issue? > > -- > Nitish Kovalam > Aerospace Engineering > Indian Institute of Space Science and Technology > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: attachments/20180202/7cbaf33d/attachment.html> > -------------- next part -------------- > A non-text attachment was scrubbed... > Name: jet2.usr > Type: application/octet-stream > Size: 18256 bytes > Desc: jet2.usr > URL: attachments/20180202/7cbaf33d/attachment.obj> > > ------------------------------ > > 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 108, Issue 2 > ********************************************* > -- 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 Fri Feb 2 22:52:05 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sat, 3 Feb 2018 04:52:05 +0000 Subject: [Nek5000-users] Nek5000-users Digest, Vol 108, Issue 2 In-Reply-To: References: , Message-ID: The turb_outflow boundary condition accelerates the flow at the exit (like a nozzle) so that all characteristics are outward pointing. At the outflow, the velocity becomes uout=u+Uout, where Uout is the additional velocity. This additional velocity is generated by setting the divergence to be positive in the last layer of elements (any element with an 'O ' BC). turb_outflow was originally written to vascular flows (i.e., internal flows) where there is nearly a 1-to-1 relationship between the mean inlet velocity and exit velocity. In that case Uout (and the consequent requirements for the amount of divergence required to generate Uout) could be determined automatically. In these cases, rq > 1 suffices to set the flux-ratio (qout/qin) = rq automatically. rq=2 ==> ratio of 2, which translates to a velocity ratio of 2 when the inlet and outlet areas are the same. If the parameter uin := 0, the code tries to estimate the velocity near the exit and then adjust divergence so that uout/ubar = rq. In some cases (such as a small jet in a big domain), the mean velocity (flux / area) is nowhere near what you want for a corrected velocity --- because you need to account for the high velocity at the jet. In this case, you prescribe an input value --- uin=1.0 is reasonable. If uin>0, the code sets ubar=uin to be your prescribed value and will set the divergence such that uout/ubar ~= rq. What is important is that (uout-uin) > u', where u' is the magnitude of your reversed flow. hth, Paul ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Friday, February 2, 2018 10:33:35 PM To: nek5000-users at lists.mcs.anl.gov Subject: Re: [Nek5000-users] Nek5000-users Digest, Vol 108, Issue 2 Thank you Paul and Phillip for such a quick reply. I will run the jet case attached. So why was there no such problem when the first vortex from the wake of a circular cylinder was leaving the domain? On Sat, Feb 3, 2018 at 1:49 AM, > 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. Outflow Boundary Condition Clarification (nek5000-users at lists.mcs.anl.gov) 2. Re: Outflow Boundary Condition Clarification (nek5000-users at lists.mcs.anl.gov) 3. Re: Outflow Boundary Condition Clarification (nek5000-users at lists.mcs.anl.gov) ---------------------------------------------------------------------- Message: 1 Date: Sat, 3 Feb 2018 00:40:32 +0530 From: nek5000-users at lists.mcs.anl.gov To: nek5000-users at lists.mcs.anl.gov Subject: [Nek5000-users] Outflow Boundary Condition Clarification Message-ID: > Content-Type: text/plain; charset="utf-8" Im trying to simulate the formation of shear layer when a jet exits a nozzle and enters the flow domain. As the FIRST vortex leaves the outlet ( at which I have provided 'O' boundary condition using genbox) the solution diverges. I tried to implement the 'turbulent outflow' subroutine as suggested in the forum but now the vortex affects the upstream flow. But when I run the flow past circular cylinder example, the vortices leave the domain without any issue even though only 'O' boundary condition is used'. What is the difference between the two cases and what can i do to prevent the issue? -- Nitish Kovalam Aerospace Engineering Indian Institute of Space Science and Technology -------------- next part -------------- An HTML attachment was scrubbed... URL: ------------------------------ Message: 2 Date: Fri, 2 Feb 2018 20:27:50 +0100 From: nek5000-users at lists.mcs.anl.gov To: nek5000-users at lists.mcs.anl.gov Subject: Re: [Nek5000-users] Outflow Boundary Condition Clarification Message-ID: > Content-Type: text/plain; charset=utf-8; format=flowed I think you should try the stabilised O condition, as described in a paper by Dong (2014). We have an implementation of that, we should probably provide it as a module in the repository. best regards, Philipp On 2018-02-02 20:10, nek5000-users at lists.mcs.anl.gov wrote: > Im trying to simulate the formation of shear layer when a jet exits a > nozzle and enters the flow domain. As the FIRST vortex leaves the outlet > ( at which I have provided 'O' boundary condition using genbox) the > solution diverges. I tried to implement the 'turbulent outflow' > subroutine as suggested in the forum but now the vortex affects the > upstream flow. But when I run the flow past circular cylinder example, > the vortices leave the domain without any issue even though only 'O' > boundary condition is used'. What is the difference between the two > cases and what can i do to prevent the issue? > > -- > Nitish Kovalam > Aerospace Engineering > Indian Institute of Space Science and Technology > > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > ------------------------------ Message: 3 Date: Fri, 2 Feb 2018 20:19:54 +0000 From: nek5000-users at lists.mcs.anl.gov To: "nek5000-users at lists.mcs.anl.gov" > Subject: Re: [Nek5000-users] Outflow Boundary Condition Clarification Message-ID: > Content-Type: text/plain; charset="us-ascii" Attached is a .usr with turb_outflow that works for this case---there is no difficulty if you use the prescribed input values, as shown in the attached files. Here is a movie of the result - you can see that turb_outflow is turned on prior to the jet reaching the top of the domain. http://mcs.anl.gov/~obabko/jet_vm.mpeg This mesh is a bit under-resolved in the z-direction. The files for a new case, with more axial resolution, are posted here www.mcs.anl.gov/~fischer/jet2.tgz I'll be posting an example with a script on how to build this geometry to nekexamples soon. Paul ________________________________ From: Nek5000-users > on behalf of nek5000-users at lists.mcs.anl.gov > Sent: Friday, February 2, 2018 1:10:32 PM To: nek5000-users at lists.mcs.anl.gov Subject: [Nek5000-users] Outflow Boundary Condition Clarification Im trying to simulate the formation of shear layer when a jet exits a nozzle and enters the flow domain. As the FIRST vortex leaves the outlet ( at which I have provided 'O' boundary condition using genbox) the solution diverges. I tried to implement the 'turbulent outflow' subroutine as suggested in the forum but now the vortex affects the upstream flow. But when I run the flow past circular cylinder example, the vortices leave the domain without any issue even though only 'O' boundary condition is used'. What is the difference between the two cases and what can i do to prevent the issue? -- Nitish Kovalam Aerospace Engineering Indian Institute of Space Science and Technology -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: jet2.usr Type: application/octet-stream Size: 18256 bytes Desc: jet2.usr URL: ------------------------------ 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 108, Issue 2 ********************************************* -- 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 Sat Feb 3 10:36:24 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sat, 03 Feb 2018 16:36:24 +0000 Subject: [Nek5000-users] Questions regarding avg_all Message-ID: Hi all, I understand that a call to avg_all computes the running averages of u, u^2 etc and then dumps the results after a fixed timestep as given by param68. What I do not understand is where in the domain does this computation take place? Does it average all the points in the domain? I understand how this would by useful in the turbChannel case where fully developed turbulence covers essentially the entire domain. However in cases where the flow develops turbulence within the domain and thus there are strong regions of turbulent anisotropy, it would be imperative to calculate these running averages along a specified plane or line. I did see some code defining these in the turbChannel example viz. common /plane/ uavg_pl(ly1*lely) $ , urms_pl(ly1*lely) $ , vrms_pl(ly1*lely) $ , wrms_pl(ly1*lely) $ , uvms_pl(ly1*lely) $ , yy(ly1*lely) $ , w1(ly1*lely),w2(ly1*lely) $ , ffx_avg, dragx_avg but I'm not exactly sure how this is defining a particular section of the geometry. In my particular case I'm trying to simulate a rectangular jet exhausting into a cuboidal domain. How can I set avg_all to compute the said average along a streamwise plane of the jet for example? Sincerely, -- *Amitvikram Dutta* Graduate Research Assistant Fluid Mechanics Research Lab Multi-Physics Interaction Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Sun Feb 4 04:16:54 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sun, 4 Feb 2018 11:16:54 +0100 Subject: [Nek5000-users] Questions regarding avg_all In-Reply-To: References: Message-ID: Dear Amitvikram, As far as I understand, the basic version of avg_all perform only a time average, i.e. the statistics are computed in 3D in every point of the domain. In case your flow has additional directions of homogeneity, then one can do an average over those directions as well; either by taking the output of avg_all and doing the average there, or direction during runtime (for the latter there is an implementation described here: http://www.ipd.anl.gov/anlpubs/2017/03/133610.pdf) Best regards, PHilipp On 2018-02-03 17:36, nek5000-users at lists.mcs.anl.gov wrote: > Hi all, > > I understand that a call to avg_all computes the running averages of u, > u^2 etc and then dumps the results after a fixed timestep as given by > param68. > > ?What I do not understand is where in the domain does this computation > take place? Does it average all the points in the domain? I understand > how this would by useful in the turbChannel case where fully developed > turbulence covers essentially the entire domain. However in cases where > the flow develops turbulence within the domain and thus there are strong > regions of turbulent anisotropy, it would be imperative to calculate > these running averages along a specified plane or line. I did see some > code defining these in the turbChannel example > viz. common /plane/? uavg_pl(ly1*lely) > ???? $???????????? ,? urms_pl(ly1*lely) > ???? $???????????? ,? vrms_pl(ly1*lely) > ???? $???????????? ,? wrms_pl(ly1*lely) > ???? $???????????? ,? uvms_pl(ly1*lely) > ???? $???????????? ,? yy(ly1*lely) > ???? $???????????? ,? w1(ly1*lely),w2(ly1*lely) > ???? $???????????? ,? ffx_avg, dragx_avg > > > but I'm not exactly sure how this is defining a particular section of > the geometry. In my particular case I'm trying to simulate a rectangular > jet exhausting into a cuboidal domain. How can I set avg_all to compute > the said average along a streamwise plane of the jet for example? > > Sincerely, > -- > > *Amitvikram Dutta* > > Graduate Research Assistant > > Fluid Mechanics Research Lab > > Multi-Physics Interaction 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 Sun Feb 4 14:08:08 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sun, 04 Feb 2018 20:08:08 +0000 Subject: [Nek5000-users] Questions regarding avg_all In-Reply-To: References: Message-ID: Hi Phillip, Thanks for the reply, that was very clear. However, I'm now facing a problem while trying to load the avvgblah.f000* files from my case. Currently I'm running visnek on these averaged files, generating a .nek5000 file and trying to read that using visit or paraview. This procedure results in an error. The reader complains that the first step of a nek file must contain a mesh which the files generated by avg_all do not seem to have. What is the procedure for reading this data then? Sincerely, On Sun, Feb 4, 2018 at 5:39 AM wrote: > Dear Amitvikram, > > As far as I understand, the basic version of avg_all perform only a time > average, i.e. the statistics are computed in 3D in every point of the > domain. In case your flow has additional directions of homogeneity, then > one can do an average over those directions as well; either by taking > the output of avg_all and doing the average there, or direction during > runtime (for the latter there is an implementation described here: > http://www.ipd.anl.gov/anlpubs/2017/03/133610.pdf) > > Best regards, > PHilipp > > On 2018-02-03 17:36, nek5000-users at lists.mcs.anl.gov wrote: > > Hi all, > > > > I understand that a call to avg_all computes the running averages of u, > > u^2 etc and then dumps the results after a fixed timestep as given by > > param68. > > > > What I do not understand is where in the domain does this computation > > take place? Does it average all the points in the domain? I understand > > how this would by useful in the turbChannel case where fully developed > > turbulence covers essentially the entire domain. However in cases where > > the flow develops turbulence within the domain and thus there are strong > > regions of turbulent anisotropy, it would be imperative to calculate > > these running averages along a specified plane or line. I did see some > > code defining these in the turbChannel example > > viz. common /plane/ uavg_pl(ly1*lely) > > $ , urms_pl(ly1*lely) > > $ , vrms_pl(ly1*lely) > > $ , wrms_pl(ly1*lely) > > $ , uvms_pl(ly1*lely) > > $ , yy(ly1*lely) > > $ , w1(ly1*lely),w2(ly1*lely) > > $ , ffx_avg, dragx_avg > > > > > > but I'm not exactly sure how this is defining a particular section of > > the geometry. In my particular case I'm trying to simulate a rectangular > > jet exhausting into a cuboidal domain. How can I set avg_all to compute > > the said average along a streamwise plane of the jet for example? > > > > Sincerely, > > -- > > > > *Amitvikram Dutta* > > > > Graduate Research Assistant > > > > Fluid Mechanics Research Lab > > > > Multi-Physics Interaction 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 Multi-Physics Interaction Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Sun Feb 4 14:27:36 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sun, 4 Feb 2018 20:27:36 +0000 Subject: [Nek5000-users] Questions regarding avg_all In-Reply-To: References: , Message-ID: Usually I just build a sequence of files with a different name such that the first one has the geometry. This is easily done with a unix script, e.g.: ln case0.f00001 view0.f00001 ln avgcase0.f00001 view0.f00002 ln avgcase0.f00002 view0.f00003 etc. Then use visit to look at case "view" Paul ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Sunday, February 4, 2018 2:08:08 PM To: nek5000-users at lists.mcs.anl.gov Subject: Re: [Nek5000-users] Questions regarding avg_all Hi Phillip, Thanks for the reply, that was very clear. However, I'm now facing a problem while trying to load the avvgblah.f000* files from my case. Currently I'm running visnek on these averaged files, generating a .nek5000 file and trying to read that using visit or paraview. This procedure results in an error. The reader complains that the first step of a nek file must contain a mesh which the files generated by avg_all do not seem to have. What is the procedure for reading this data then? Sincerely, On Sun, Feb 4, 2018 at 5:39 AM > wrote: Dear Amitvikram, As far as I understand, the basic version of avg_all perform only a time average, i.e. the statistics are computed in 3D in every point of the domain. In case your flow has additional directions of homogeneity, then one can do an average over those directions as well; either by taking the output of avg_all and doing the average there, or direction during runtime (for the latter there is an implementation described here: http://www.ipd.anl.gov/anlpubs/2017/03/133610.pdf) Best regards, PHilipp On 2018-02-03 17:36, nek5000-users at lists.mcs.anl.gov wrote: > Hi all, > > I understand that a call to avg_all computes the running averages of u, > u^2 etc and then dumps the results after a fixed timestep as given by > param68. > > What I do not understand is where in the domain does this computation > take place? Does it average all the points in the domain? I understand > how this would by useful in the turbChannel case where fully developed > turbulence covers essentially the entire domain. However in cases where > the flow develops turbulence within the domain and thus there are strong > regions of turbulent anisotropy, it would be imperative to calculate > these running averages along a specified plane or line. I did see some > code defining these in the turbChannel example > viz. common /plane/ uavg_pl(ly1*lely) > $ , urms_pl(ly1*lely) > $ , vrms_pl(ly1*lely) > $ , wrms_pl(ly1*lely) > $ , uvms_pl(ly1*lely) > $ , yy(ly1*lely) > $ , w1(ly1*lely),w2(ly1*lely) > $ , ffx_avg, dragx_avg > > > but I'm not exactly sure how this is defining a particular section of > the geometry. In my particular case I'm trying to simulate a rectangular > jet exhausting into a cuboidal domain. How can I set avg_all to compute > the said average along a streamwise plane of the jet for example? > > Sincerely, > -- > > *Amitvikram Dutta* > > Graduate Research Assistant > > Fluid Mechanics Research Lab > > Multi-Physics Interaction 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 Multi-Physics Interaction Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Feb 6 23:42:27 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 6 Feb 2018 23:42:27 -0600 Subject: [Nek5000-users] Getting planar average and outputting data Message-ID: Dear Neks, I am trying to average a box geometry along the xz-plane using the planar_average_s subroutine from turbChannel. But it is not outputting all the y values from my domain. I am using the average file from avg_all as a restart file and hope to output the planar averaged velocity and passive scalars along the y direction. Could someone help me fix my code? Below is what I have so far: call planar_average_s(yy,ym1,w1,w2) call planar_average_s(uavg_pl,vx,w1,w2) call planar_average_s(vavg_pl,vy,w1,w2) call planar_average_s(wavg_pl,vz,w1,w2) call planar_average_s(tavg_pl,t(:,1,1,1,1),w1,w2) call planar_average_s(ttavg_pl,t(:,1,1,1,2),w1,w2) call planar_average_s(tttavg_pl,t(:,1,1,1,3),w1,w2) if(nid.eq.0) then open(unit=57,file='mean_prof.dat') do i=1,ny1*nely/2 write(57,3) yy(i) & , uavg_pl(i) & , vavg_pl(i) & , wavg_pl(i) & , tavg_pl(i) & , ttavg_pl(i) & , tttavg_pl(i) enddo 3 format(7f17.9) close(57) endif call exitt Thanks for any help, Jon -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Feb 7 12:05:18 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 7 Feb 2018 19:05:18 +0100 (CET) Subject: [Nek5000-users] rotation of aero foil Message-ID: Hi Neks, Its a long time that i came up with something. So here we go, SInce I was doinf the FSI of aero foil, I have to deal with 2 degrees of freedom (its a 2D case) one in vertical and another in angular direction about the elastic axis of the profile. I could code the vertical motion of the foil but before starting the angular motion I have some doubts. Is there anyway to give the angular velocities directly to the mesh velocities? Apart from the translational velocity components Or Do you want me to covert the angluar velocities into translational components and give to mesh velocities? Do you have any example with a body which is rotating? Thanks in advance Sijo GEORGE PhD student Ecole centrale de Nantes -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Thu Feb 8 10:29:16 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 8 Feb 2018 17:29:16 +0100 (CET) Subject: [Nek5000-users] Rotating foil Message-ID: Hi Neks, Its a long time that i came up with something. So here we go, SInce I was doinf the FSI of aero foil, I have to deal with 2 degrees of freedom (its a 2D case) one in vertical and another in angular direction about the elastic axis of the profile. I could code the vertical motion of the foil but before starting the angular motion I have some doubts. Is there anyway to give the angular velocities directly to the mesh velocities? Apart from the translational velocity components Or Do you want me to covert the angluar velocities into translational components and give to mesh velocities? Do you have any example with a body which is rotating? Thanks in advance Sijo GEORGE PhD student Ecole centrale de Nantes -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Thu Feb 8 15:37:41 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 8 Feb 2018 21:37:41 +0000 Subject: [Nek5000-users] DNS of turbChannel Message-ID: Dear Nek users, I am new to NEK5000. I am aware of that two SGS models are available for the turbChannel. But if I am looking for a DNS, which line/subroutine shall I deactivate? Thanks a lot for the help. Regards, Emily -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Thu Feb 8 16:16:47 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Thu, 8 Feb 2018 22:16:47 +0000 Subject: [Nek5000-users] Getting planar average and outputting data Message-ID: Hi Jon, There is a set of new planar averaging routines that will merged into the repo soon.. these routines work for box type meshes (ex: meshes made using genbox since they have lexicographic element numbering). If you have a mesh that was made using genbox, you can use these routines from here: https://github.com/Nek5000/Nek5000/pull/411 Let me know if you have any questions about the usage of these routines. Ketan -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Thu Feb 8 20:00:50 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 9 Feb 2018 02:00:50 +0000 Subject: [Nek5000-users] DNS of turbChannel In-Reply-To: References: Message-ID: Dear Emily, Attached is a channel flow example with a 90-element mesh generated by genbox. It has 6 elements in the wall-normal direction, which for polynomial order N=7 (lx1=8 in the SIZE file) implies 42 points in the wall normal direction. I have run this case with lx1=12 (implying 6x11 = 66 points in the wall normal direction) as a reasonable LES for Re_tau=395. You can reach a DNS solution with increased resolution by increasing the polynomial order or by refining the mesh through changes to the box file and rerunning genbox and genmap. The .usr file should be fine in this case - it has no eddy-diffusivity. The .rea file does prescribe a filter here: param 103 = 0.05 and p101=1 implies 5% of the energy is removed from the Nth mode in the velocity field at each time step and 5% x (1/4) of the energy in the (N-1)th mode at each step. In the limit that you approach DNS, the amount of energy in these modes is in any case small, so you don't need to worry about turning them off, but you could do so if you wanted. hth, Paul ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Thursday, February 8, 2018 3:37:41 PM To: nek5000-users at lists.mcs.anl.gov Subject: [Nek5000-users] DNS of turbChannel Dear Nek users, I am new to NEK5000. I am aware of that two SGS models are available for the turbChannel. But if I am looking for a DNS, which line/subroutine shall I deactivate? Thanks a lot for the help. Regards, Emily -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: chan3d.tgz Type: application/x-gzip Size: 13456 bytes Desc: chan3d.tgz URL: From nek5000-users at lists.mcs.anl.gov Fri Feb 9 02:53:52 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 9 Feb 2018 09:53:52 +0100 Subject: [Nek5000-users] [PhD proposal] Instabilities and transition to turbulence of time-periodic flows Message-ID: Dear Nek's, Please find attached a PhD proposal on *instabilities and transition to turbulence of time-periodic flows* using Nek 5000. In a nutshell: building on previous work done at DynFluid (ENSAM, Paris, France), the PhD candidate will have to develop (among other things) a Newton-like solver in Nek5000 to compute unstable periodic orbits and to study their stability using Floquet stability theory. One of the possible flow configurations to be studied is that of the three-dimensional shear-driven cavity flow. The position starts early September/October 2018 In Paris, France. A three-year grant has been provided by the French Ministry of Higher-Education and Research to cover the wage of the PhD candidate. I would be very grateful if you could forward this PhD proposal to students of yours who would be interested. Best regards, Jean-Christophe Loiseau -- Jean-Christophe Loiseau Homepage -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: PhD_subject.pdf Type: application/pdf Size: 308998 bytes Desc: not available URL: From nek5000-users at lists.mcs.anl.gov Fri Feb 9 05:49:39 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 9 Feb 2018 12:49:39 +0100 Subject: [Nek5000-users] Computing aerodynamic coefficients Message-ID: Dear Neks: I have a file containing the lift and drag coefficient for a wing that was computed long back using NEK5000. The value of the aerodynamic coefficients look fine. Since computations were carried out for a very specific set of parameters we do not have any experimental or numerical result to check it against. The code that I am using presently for computing aerodynamic coefficients is slightly different from the one we used earlier. Can someone please take a look at the following snippet from earlier version of the code and check if it computes the right aerodynamic coefficients? Many thanks. Best wishes, NN. subroutine set_obj ! define objects for surface integrals c include 'SIZE' include 'TOTAL' integer e,f,eg nobj = 1 iobj = 0 do ii=nhis+1,nhis+nobj iobj = iobj+1 hcode(10,ii) = 'I' hcode( 1,ii) = 'F' hcode( 2,ii) = 'F' hcode( 3,ii) = 'F' lochis(1,ii) = iobj enddo nhis = nhis + nobj if (maxobj.lt.nobj) call exitti('increase maxobj in SIZE$',nobj) nxyz = nx1*ny1*nz1 nface = 2*ndim do e=1,nelv do f=1,nface if (cbc(f,e,1).eq.'W ') then iobj = 1 if (iobj.gt.0) then nmember(iobj) = nmember(iobj) + 1 mem = nmember(iobj) eg = lglel(e) object(iobj,mem,1) = eg object(iobj,mem,2) = f c write(6,1) iobj,mem,f,eg,e,nid,' OBJ' c 1 format(6i9,a4) endif endif enddo enddo c write(6,*) 'number',(nmember(k),k=1,4) c return end From nek5000-users at lists.mcs.anl.gov Fri Feb 9 06:59:47 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 9 Feb 2018 12:59:47 +0000 Subject: [Nek5000-users] Computing aerodynamic coefficients In-Reply-To: References: Message-ID: Your code looks fine to me. If you have not used Nek5000 in a while, I would recommend to look at the turbChannel example that contains the logic to compute the drag. Marco ?On 2/9/18, 6:49 AM, "Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov" wrote: object From nek5000-users at lists.mcs.anl.gov Fri Feb 9 07:48:34 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 9 Feb 2018 14:48:34 +0100 Subject: [Nek5000-users] Computing aerodynamic coefficients In-Reply-To: References: Message-ID: Thanks Marco. I will follow your recommendation. NN. On 02/09/2018 01:59 PM, nek5000-users at lists.mcs.anl.gov wrote: > Your code looks fine to me. If you have not used Nek5000 in a while, I would recommend to look at the turbChannel example that contains the logic to compute the drag. > > Marco > ?On 2/9/18, 6:49 AM, "Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov" wrote: > > object > > _______________________________________________ > 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 Feb 9 07:56:21 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 9 Feb 2018 14:56:21 +0100 (CET) Subject: [Nek5000-users] angular velocity to the mesh velocity Message-ID: Hi Neks, Its a long time that i came up with something. So here we go, SInce I was doinf the FSI of aero foil, I have to deal with 2 degrees of freedom (its a 2D case) one in vertical and another in angular direction about the elastic axis of the profile. I could code the vertical motion of the foil but before starting the angular motion I have some doubts. Is there anyway to give the angular velocities directly to the mesh velocities? Apart from the translational velocity components Or Do you want me to covert the angluar velocities into translational components and give to mesh velocities? Do you have any example with a body which is rotating? Thanks in advance Sijo GEORGE -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Fri Feb 9 08:03:26 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 9 Feb 2018 14:03:26 +0000 Subject: [Nek5000-users] angular velocity to the mesh velocity Message-ID: Sijo, The mv_cyl example is a 2-D moving body example. As for the mesh velocity, can you please be more specific? Marco From: Nek5000-users on behalf of "nek5000-users at lists.mcs.anl.gov" Reply-To: "nek5000-users at lists.mcs.anl.gov" Date: Friday, February 9, 2018 at 8:56 AM To: nek5000-users Subject: [Nek5000-users] angular velocity to the mesh velocity Hi Neks, Its a long time that i came up with something. So here we go, SInce I was doinf the FSI of aero foil, I have to deal with 2 degrees of freedom (its a 2D case) one in vertical and another in angular direction about the elastic axis of the profile. I could code the vertical motion of the foil but before starting the angular motion I have some doubts. Is there anyway to give the angular velocities directly to the mesh velocities? Apart from the translational velocity components Or Do you want me to covert the angluar velocities into translational components and give to mesh velocities? Do you have any example with a body which is rotating? Thanks in advance Sijo GEORGE -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Fri Feb 9 08:36:39 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 9 Feb 2018 15:36:39 +0100 (CET) Subject: [Nek5000-users] angular velocity to the mesh velocity In-Reply-To: References: Message-ID: Thank you for your replay Marco. I will explain from the begening clearly. My current task is to validate Fluid structre interaction on aero foil(NACA profile). The equations which i have to solve I have attached for aero foil. Before this case, I had succesfully done a case of vortex induced vibration of cylinder in NEK5000. And the code below shows a part of the forced oscillating cylinder. From this code I learned that the body motion prescribed through the mesh velocities but I also think that I can only give translational velocity components in NEK5000 to presribe a body motion (mesh as well). Since my aero foil has pitching (it rotates up and down about an axis at angle theta). Since it has to be a FSI problem my plan is by using " call torque_calc(scale,x0,.true.,.true.)" I will get the M (moment about the elastic axis (x0)) and it gives as an input to the Equation of angular motion (attached with this email). Then using Range Kutta method I get the angular velocity of the profile (r, theta,z co ordinates). Then I am planning to conver this angular velocity into linear velocity components (x, y, z direction) then impose into the wx,wy,wz components and I hope it leads to the movement of mesh thus the geometry. My question is could you think that its possible? Or is there any other simple way to give this angular velocity to my NACA profile about the elastic axis (x0)? And I feel difficulty to convert the angular velocity into linear velocty and gives its to each nodes which is on "mv" B.C . Could you please help me with this situation by an example if you have. Thanks in advance c Cylinder velocity components ucx = 0.0 ucy = amplitude * omega * cos(omega*time) ucz = 0.0 n = nx1*ny1*nz1*nelv do i=1,n ! Translational velocity wx(i,1,1,1) = basev(i)*ucx ! component. wy(i,1,1,1) = basev(i)*ucy wz(i,1,1,1) = basev(i)*ucz umeshx(i,1,1,1) = wx(i,1,1,1) umeshy(i,1,1,1) = wy(i,1,1,1) umeshz(i,1,1,1) = wz(i,1,1,1) enddo De: "nek5000-users" ?: "nek5000-users" Envoy?: Vendredi 9 F?vrier 2018 15:03:26 Objet: Re: [Nek5000-users] angular velocity to the mesh velocity Sijo, The mv_cyl example is a 2-D moving body example. As for the mesh velocity, can you please be more specific? Marco From: Nek5000-users on behalf of "nek5000-users at lists.mcs.anl.gov" Reply-To: "nek5000-users at lists.mcs.anl.gov" Date: Friday, February 9, 2018 at 8:56 AM To: nek5000-users Subject: [Nek5000-users] angular velocity to the mesh velocity Hi Neks, Its a long time that i came up with something. So here we go, SInce I was doinf the FSI of aero foil, I have to deal with 2 degrees of freedom (its a 2D case) one in vertical and another in angular direction about the elastic axis of the profile. I could code the vertical motion of the foil but before starting the angular motion I have some doubts. Is there anyway to give the angular velocities directly to the mesh velocities? Apart from the translational velocity components Or Do you want me to covert the angluar velocities into translational components and give to mesh velocities? Do you have any example with a body which is rotating? Thanks in advance 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: -------------- next part -------------- A non-text attachment was scrubbed... Name: sijo Type: image/png Size: 12534 bytes Desc: not available URL: From nek5000-users at lists.mcs.anl.gov Fri Feb 9 09:02:34 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 9 Feb 2018 15:02:34 +0000 Subject: [Nek5000-users] angular velocity to the mesh velocity In-Reply-To: References: , Message-ID: Sorry for the slow response but was on travel yesterday. For moving meshes with rigid-body motion of pitch+translation, say, you essentially need a smooth function that is 1 at the moving boundary and 0 at the far field boundaries. Call this function h(X,t)=h(x,y,z,t). Then your mesh velocity (W=[wx,wy,wz]), which is required by Nek's ALE formulation, is given by Translational Velocity(t)*h + Rotational Velocity*h Where h==1 (i.e., on the boundary) you will have the correct velocity. Where h==0 (on the far field), you will have no mesh velocity. h is in effect a blending function between these regions. My favorite way of generating h is to solve a Laplace equation with a variable diffusivity. A high diffusivity near the object ensures h=1 near the object and thus elements near the object move _with_ the object, which means that they do not get excessively deformed. I attach a .usr file from a case I ran several years ago -- I've not tested it recently so please take it as a guide and not a solution to your problem. Look for "Greens Function" to see how to compute h. Here is a paper describing the approach http://www.mcs.anl.gov/~fischer/pubhtml/ccfd.pdf Paul PS - Note - Nek requires _only_ the mesh velocity -- mesh position is updated by Nek. Note that you may, or may not, need to recompute the Green's function.... It is often a smooth function even after the mesh is deformed, so you can keep using it.... I suggest doing some experiments for your particular case. ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Friday, February 9, 2018 8:36:39 AM To: nek5000-users Subject: Re: [Nek5000-users] angular velocity to the mesh velocity Thank you for your replay Marco. I will explain from the begening clearly. My current task is to validate Fluid structre interaction on aero foil(NACA profile). The equations which i have to solve I have attached for aero foil. Before this case, I had succesfully done a case of vortex induced vibration of cylinder in NEK5000. And the code below shows a part of the forced oscillating cylinder. From this code I learned that the body motion prescribed through the mesh velocities but I also think that I can only give translational velocity components in NEK5000 to presribe a body motion (mesh as well). Since my aero foil has pitching (it rotates up and down about an axis at angle theta). Since it has to be a FSI problem my plan is by using " call torque_calc(scale,x0,.true.,.true.)" I will get the M (moment about the elastic axis (x0)) and it gives as an input to the Equation of angular motion (attached with this email). Then using Range Kutta method I get the angular velocity of the profile (r, theta,z co ordinates). Then I am planning to conver this angular velocity into linear velocity components (x, y, z direction) then impose into the wx,wy,wz components and I hope it leads to the movement of mesh thus the geometry. My question is could you think that its possible? Or is there any other simple way to give this angular velocity to my NACA profile about the elastic axis (x0)? And I feel difficulty to convert the angular velocity into linear velocty and gives its to each nodes which is on "mv" B.C . Could you please help me with this situation by an example if you have. Thanks in advance c Cylinder velocity components ucx = 0.0 ucy = amplitude * omega * cos(omega*time) ucz = 0.0 n = nx1*ny1*nz1*nelv do i=1,n ! Translational velocity wx(i,1,1,1) = basev(i)*ucx ! component. wy(i,1,1,1) = basev(i)*ucy wz(i,1,1,1) = basev(i)*ucz umeshx(i,1,1,1) = wx(i,1,1,1) umeshy(i,1,1,1) = wy(i,1,1,1) umeshz(i,1,1,1) = wz(i,1,1,1) enddo ________________________________ De: "nek5000-users" ?: "nek5000-users" Envoy?: Vendredi 9 F?vrier 2018 15:03:26 Objet: Re: [Nek5000-users] angular velocity to the mesh velocity Sijo, The mv_cyl example is a 2-D moving body example. As for the mesh velocity, can you please be more specific? Marco From: Nek5000-users on behalf of "nek5000-users at lists.mcs.anl.gov" Reply-To: "nek5000-users at lists.mcs.anl.gov" Date: Friday, February 9, 2018 at 8:56 AM To: nek5000-users Subject: [Nek5000-users] angular velocity to the mesh velocity Hi Neks, Its a long time that i came up with something. So here we go, SInce I was doinf the FSI of aero foil, I have to deal with 2 degrees of freedom (its a 2D case) one in vertical and another in angular direction about the elastic axis of the profile. I could code the vertical motion of the foil but before starting the angular motion I have some doubts. Is there anyway to give the angular velocities directly to the mesh velocities? Apart from the translational velocity components Or Do you want me to covert the angluar velocities into translational components and give to mesh velocities? Do you have any example with a body which is rotating? Thanks in advance 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: -------------- next part -------------- A non-text attachment was scrubbed... Name: pitch.usr Type: application/octet-stream Size: 6815 bytes Desc: pitch.usr URL: From nek5000-users at lists.mcs.anl.gov Fri Feb 9 09:15:15 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 9 Feb 2018 16:15:15 +0100 (CET) Subject: [Nek5000-users] angular velocity to the mesh velocity In-Reply-To: References: Message-ID: Thank you very much Paul for your detailed replay and I am very glad about it. But I am sorry if i wrongly said my question because the answer you said was not the one I was expecting. The thing which you have said I had already taken into account (the blending function ). In fact we already had a talk about this previously and I also had developed another blending function with another equation. my question was is apart from wx,wy,wz. Is there any way to give the angular velocity to the mesh (like wr,,wtheta,wz). Since my proile pitch about an axis I will have an angular velocity and I am bit confused how should I give this angular velocity as an input to the mesh velocity. Thanks with regards Sijo GEORGE De: "nek5000-users" ?: "nek5000-users" Envoy?: Vendredi 9 F?vrier 2018 16:02:34 Objet: Re: [Nek5000-users] angular velocity to the mesh velocity Sorry for the slow response but was on travel yesterday. For moving meshes with rigid-body motion of pitch+translation, say, you essentially need a smooth function that is 1 at the moving boundary and 0 at the far field boundaries. Call this function h(X,t)=h(x,y,z,t). Then your mesh velocity (W=[wx,wy,wz]), which is required by Nek's ALE formulation, is given by Translational Velocity(t)*h + Rotational Velocity*h Where h==1 (i.e., on the boundary) you will have the correct velocity. Where h==0 (on the far field), you will have no mesh velocity. h is in effect a blending function between these regions. My favorite way of generating h is to solve a Laplace equation with a variable diffusivity. A high diffusivity near the object ensures h=1 near the object and thus elements near the object move _with_ the object, which means that they do not get excessively deformed. I attach a .usr file from a case I ran several years ago -- I've not tested it recently so please take it as a guide and not a solution to your problem. Look for "Greens Function" to see how to compute h. Here is a paper describing the approach [ http://www.mcs.anl.gov/~fischer/pubhtml/ccfd.pdf | http://www.mcs.anl.gov/~fischer/pubhtml/ccfd.pdf ] Paul PS - Note - Nek requires _only_ the mesh velocity -- mesh position is updated by Nek. Note that you may, or may not, need to recompute the Green's function.... It is often a smooth function even after the mesh is deformed, so you can keep using it.... I suggest doing some experiments for your particular case. From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Friday, February 9, 2018 8:36:39 AM To: nek5000-users Subject: Re: [Nek5000-users] angular velocity to the mesh velocity Thank you for your replay Marco. I will explain from the begening clearly. My current task is to validate Fluid structre interaction on aero foil(NACA profile). The equations which i have to solve I have attached for aero foil. Before this case, I had succesfully done a case of vortex induced vibration of cylinder in NEK5000. And the code below shows a part of the forced oscillating cylinder. From this code I learned that the body motion prescribed through the mesh velocities but I also think that I can only give translational velocity components in NEK5000 to presribe a body motion (mesh as well). Since my aero foil has pitching (it rotates up and down about an axis at angle theta). Since it has to be a FSI problem my plan is by using " call torque_calc(scale,x0,.true.,.true.)" I will get the M (moment about the elastic axis (x0)) and it gives as an input to the Equation of angular motion (attached with this email). Then using Range Kutta method I get the angular velocity of the profile (r, theta,z co ordinates). Then I am planning to conver this angular velocity into linear velocity components (x, y, z direction) then impose into the wx,wy,wz components and I hope it leads to the movement of mesh thus the geometry. My question is could you think that its possible? Or is there any other simple way to give this angular velocity to my NACA profile about the elastic axis (x0)? And I feel difficulty to convert the angular velocity into linear velocty and gives its to each nodes which is on "mv" B.C . Could you please help me with this situation by an example if you have. Thanks in advance c Cylinder velocity components ucx = 0.0 ucy = amplitude * omega * cos(omega*time) ucz = 0.0 n = nx1*ny1*nz1*nelv do i=1,n ! Translational velocity wx(i,1,1,1) = basev(i)*ucx ! component. wy(i,1,1,1) = basev(i)*ucy wz(i,1,1,1) = basev(i)*ucz umeshx(i,1,1,1) = wx(i,1,1,1) umeshy(i,1,1,1) = wy(i,1,1,1) umeshz(i,1,1,1) = wz(i,1,1,1) enddo De: "nek5000-users" ?: "nek5000-users" Envoy?: Vendredi 9 F?vrier 2018 15:03:26 Objet: Re: [Nek5000-users] angular velocity to the mesh velocity Sijo, The mv_cyl example is a 2-D moving body example. As for the mesh velocity, can you please be more specific? Marco From: Nek5000-users on behalf of "nek5000-users at lists.mcs.anl.gov" Reply-To: "nek5000-users at lists.mcs.anl.gov" Date: Friday, February 9, 2018 at 8:56 AM To: nek5000-users Subject: [Nek5000-users] angular velocity to the mesh velocity Hi Neks, Its a long time that i came up with something. So here we go, SInce I was doinf the FSI of aero foil, I have to deal with 2 degrees of freedom (its a 2D case) one in vertical and another in angular direction about the elastic axis of the profile. I could code the vertical motion of the foil but before starting the angular motion I have some doubts. Is there anyway to give the angular velocities directly to the mesh velocities? Apart from the translational velocity components Or Do you want me to covert the angluar velocities into translational components and give to mesh velocities? Do you have any example with a body which is rotating? Thanks in advance 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 Fri Feb 9 09:28:48 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 9 Feb 2018 15:28:48 +0000 Subject: [Nek5000-users] angular velocity to the mesh velocity In-Reply-To: References: , Message-ID: Dear Sijo, (wx,wy,wz) = ( r cos theta, r sin theta , 0 ) * h(x,y,z) where r = sqrt(x*x+y*y), cos theta = x/r, sin theta = y/r Assuming that you are pitching about the origin. Paul ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Friday, February 9, 2018 9:15:15 AM To: nek5000-users Subject: Re: [Nek5000-users] angular velocity to the mesh velocity Thank you very much Paul for your detailed replay and I am very glad about it. But I am sorry if i wrongly said my question because the answer you said was not the one I was expecting. The thing which you have said I had already taken into account (the blending function ). In fact we already had a talk about this previously and I also had developed another blending function with another equation. my question was is apart from wx,wy,wz. Is there any way to give the angular velocity to the mesh (like wr,,wtheta,wz). Since my proile pitch about an axis I will have an angular velocity and I am bit confused how should I give this angular velocity as an input to the mesh velocity. Thanks with regards Sijo GEORGE ________________________________ De: "nek5000-users" ?: "nek5000-users" Envoy?: Vendredi 9 F?vrier 2018 16:02:34 Objet: Re: [Nek5000-users] angular velocity to the mesh velocity Sorry for the slow response but was on travel yesterday. For moving meshes with rigid-body motion of pitch+translation, say, you essentially need a smooth function that is 1 at the moving boundary and 0 at the far field boundaries. Call this function h(X,t)=h(x,y,z,t). Then your mesh velocity (W=[wx,wy,wz]), which is required by Nek's ALE formulation, is given by Translational Velocity(t)*h + Rotational Velocity*h Where h==1 (i.e., on the boundary) you will have the correct velocity. Where h==0 (on the far field), you will have no mesh velocity. h is in effect a blending function between these regions. My favorite way of generating h is to solve a Laplace equation with a variable diffusivity. A high diffusivity near the object ensures h=1 near the object and thus elements near the object move _with_ the object, which means that they do not get excessively deformed. I attach a .usr file from a case I ran several years ago -- I've not tested it recently so please take it as a guide and not a solution to your problem. Look for "Greens Function" to see how to compute h. Here is a paper describing the approach http://www.mcs.anl.gov/~fischer/pubhtml/ccfd.pdf Paul PS - Note - Nek requires _only_ the mesh velocity -- mesh position is updated by Nek. Note that you may, or may not, need to recompute the Green's function.... It is often a smooth function even after the mesh is deformed, so you can keep using it.... I suggest doing some experiments for your particular case. ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Friday, February 9, 2018 8:36:39 AM To: nek5000-users Subject: Re: [Nek5000-users] angular velocity to the mesh velocity Thank you for your replay Marco. I will explain from the begening clearly. My current task is to validate Fluid structre interaction on aero foil(NACA profile). The equations which i have to solve I have attached for aero foil. Before this case, I had succesfully done a case of vortex induced vibration of cylinder in NEK5000. And the code below shows a part of the forced oscillating cylinder. From this code I learned that the body motion prescribed through the mesh velocities but I also think that I can only give translational velocity components in NEK5000 to presribe a body motion (mesh as well). Since my aero foil has pitching (it rotates up and down about an axis at angle theta). Since it has to be a FSI problem my plan is by using " call torque_calc(scale,x0,.true.,.true.)" I will get the M (moment about the elastic axis (x0)) and it gives as an input to the Equation of angular motion (attached with this email). Then using Range Kutta method I get the angular velocity of the profile (r, theta,z co ordinates). Then I am planning to conver this angular velocity into linear velocity components (x, y, z direction) then impose into the wx,wy,wz components and I hope it leads to the movement of mesh thus the geometry. My question is could you think that its possible? Or is there any other simple way to give this angular velocity to my NACA profile about the elastic axis (x0)? And I feel difficulty to convert the angular velocity into linear velocty and gives its to each nodes which is on "mv" B.C . Could you please help me with this situation by an example if you have. Thanks in advance c Cylinder velocity components ucx = 0.0 ucy = amplitude * omega * cos(omega*time) ucz = 0.0 n = nx1*ny1*nz1*nelv do i=1,n ! Translational velocity wx(i,1,1,1) = basev(i)*ucx ! component. wy(i,1,1,1) = basev(i)*ucy wz(i,1,1,1) = basev(i)*ucz umeshx(i,1,1,1) = wx(i,1,1,1) umeshy(i,1,1,1) = wy(i,1,1,1) umeshz(i,1,1,1) = wz(i,1,1,1) enddo ________________________________ De: "nek5000-users" ?: "nek5000-users" Envoy?: Vendredi 9 F?vrier 2018 15:03:26 Objet: Re: [Nek5000-users] angular velocity to the mesh velocity Sijo, The mv_cyl example is a 2-D moving body example. As for the mesh velocity, can you please be more specific? Marco From: Nek5000-users on behalf of "nek5000-users at lists.mcs.anl.gov" Reply-To: "nek5000-users at lists.mcs.anl.gov" Date: Friday, February 9, 2018 at 8:56 AM To: nek5000-users Subject: [Nek5000-users] angular velocity to the mesh velocity Hi Neks, Its a long time that i came up with something. So here we go, SInce I was doinf the FSI of aero foil, I have to deal with 2 degrees of freedom (its a 2D case) one in vertical and another in angular direction about the elastic axis of the profile. I could code the vertical motion of the foil but before starting the angular motion I have some doubts. Is there anyway to give the angular velocities directly to the mesh velocities? Apart from the translational velocity components Or Do you want me to covert the angluar velocities into translational components and give to mesh velocities? Do you have any example with a body which is rotating? Thanks in advance 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 Fri Feb 9 09:29:16 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 9 Feb 2018 15:29:16 +0000 Subject: [Nek5000-users] angular velocity to the mesh velocity In-Reply-To: References: , Message-ID: PS - also, please have a look at the .usr file in the previous post. ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Friday, February 9, 2018 9:15:15 AM To: nek5000-users Subject: Re: [Nek5000-users] angular velocity to the mesh velocity Thank you very much Paul for your detailed replay and I am very glad about it. But I am sorry if i wrongly said my question because the answer you said was not the one I was expecting. The thing which you have said I had already taken into account (the blending function ). In fact we already had a talk about this previously and I also had developed another blending function with another equation. my question was is apart from wx,wy,wz. Is there any way to give the angular velocity to the mesh (like wr,,wtheta,wz). Since my proile pitch about an axis I will have an angular velocity and I am bit confused how should I give this angular velocity as an input to the mesh velocity. Thanks with regards Sijo GEORGE ________________________________ De: "nek5000-users" ?: "nek5000-users" Envoy?: Vendredi 9 F?vrier 2018 16:02:34 Objet: Re: [Nek5000-users] angular velocity to the mesh velocity Sorry for the slow response but was on travel yesterday. For moving meshes with rigid-body motion of pitch+translation, say, you essentially need a smooth function that is 1 at the moving boundary and 0 at the far field boundaries. Call this function h(X,t)=h(x,y,z,t). Then your mesh velocity (W=[wx,wy,wz]), which is required by Nek's ALE formulation, is given by Translational Velocity(t)*h + Rotational Velocity*h Where h==1 (i.e., on the boundary) you will have the correct velocity. Where h==0 (on the far field), you will have no mesh velocity. h is in effect a blending function between these regions. My favorite way of generating h is to solve a Laplace equation with a variable diffusivity. A high diffusivity near the object ensures h=1 near the object and thus elements near the object move _with_ the object, which means that they do not get excessively deformed. I attach a .usr file from a case I ran several years ago -- I've not tested it recently so please take it as a guide and not a solution to your problem. Look for "Greens Function" to see how to compute h. Here is a paper describing the approach http://www.mcs.anl.gov/~fischer/pubhtml/ccfd.pdf Paul PS - Note - Nek requires _only_ the mesh velocity -- mesh position is updated by Nek. Note that you may, or may not, need to recompute the Green's function.... It is often a smooth function even after the mesh is deformed, so you can keep using it.... I suggest doing some experiments for your particular case. ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Friday, February 9, 2018 8:36:39 AM To: nek5000-users Subject: Re: [Nek5000-users] angular velocity to the mesh velocity Thank you for your replay Marco. I will explain from the begening clearly. My current task is to validate Fluid structre interaction on aero foil(NACA profile). The equations which i have to solve I have attached for aero foil. Before this case, I had succesfully done a case of vortex induced vibration of cylinder in NEK5000. And the code below shows a part of the forced oscillating cylinder. From this code I learned that the body motion prescribed through the mesh velocities but I also think that I can only give translational velocity components in NEK5000 to presribe a body motion (mesh as well). Since my aero foil has pitching (it rotates up and down about an axis at angle theta). Since it has to be a FSI problem my plan is by using " call torque_calc(scale,x0,.true.,.true.)" I will get the M (moment about the elastic axis (x0)) and it gives as an input to the Equation of angular motion (attached with this email). Then using Range Kutta method I get the angular velocity of the profile (r, theta,z co ordinates). Then I am planning to conver this angular velocity into linear velocity components (x, y, z direction) then impose into the wx,wy,wz components and I hope it leads to the movement of mesh thus the geometry. My question is could you think that its possible? Or is there any other simple way to give this angular velocity to my NACA profile about the elastic axis (x0)? And I feel difficulty to convert the angular velocity into linear velocty and gives its to each nodes which is on "mv" B.C . Could you please help me with this situation by an example if you have. Thanks in advance c Cylinder velocity components ucx = 0.0 ucy = amplitude * omega * cos(omega*time) ucz = 0.0 n = nx1*ny1*nz1*nelv do i=1,n ! Translational velocity wx(i,1,1,1) = basev(i)*ucx ! component. wy(i,1,1,1) = basev(i)*ucy wz(i,1,1,1) = basev(i)*ucz umeshx(i,1,1,1) = wx(i,1,1,1) umeshy(i,1,1,1) = wy(i,1,1,1) umeshz(i,1,1,1) = wz(i,1,1,1) enddo ________________________________ De: "nek5000-users" ?: "nek5000-users" Envoy?: Vendredi 9 F?vrier 2018 15:03:26 Objet: Re: [Nek5000-users] angular velocity to the mesh velocity Sijo, The mv_cyl example is a 2-D moving body example. As for the mesh velocity, can you please be more specific? Marco From: Nek5000-users on behalf of "nek5000-users at lists.mcs.anl.gov" Reply-To: "nek5000-users at lists.mcs.anl.gov" Date: Friday, February 9, 2018 at 8:56 AM To: nek5000-users Subject: [Nek5000-users] angular velocity to the mesh velocity Hi Neks, Its a long time that i came up with something. So here we go, SInce I was doinf the FSI of aero foil, I have to deal with 2 degrees of freedom (its a 2D case) one in vertical and another in angular direction about the elastic axis of the profile. I could code the vertical motion of the foil but before starting the angular motion I have some doubts. Is there anyway to give the angular velocities directly to the mesh velocities? Apart from the translational velocity components Or Do you want me to covert the angluar velocities into translational components and give to mesh velocities? Do you have any example with a body which is rotating? Thanks in advance 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 Fri Feb 9 09:36:54 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 9 Feb 2018 16:36:54 +0100 (CET) Subject: [Nek5000-users] angular velocity to the mesh velocity In-Reply-To: References: Message-ID: Thanks Paul I give a try. Just a last question In order to find theta and r should i make a line from origin to each nodes in the whole mesh? Or is it possible to only the nodes which has "mv " boundary condition. Thanks regards Sijo GEORGE De: "nek5000-users" ?: "nek5000-users" Envoy?: Vendredi 9 F?vrier 2018 16:28:48 Objet: Re: [Nek5000-users] angular velocity to the mesh velocity Dear Sijo, (wx,wy,wz) = ( r cos theta, r sin theta , 0 ) * h(x,y,z) where r = sqrt(x*x+y*y), cos theta = x/r, sin theta = y/r Assuming that you are pitching about the origin. Paul From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Friday, February 9, 2018 9:15:15 AM To: nek5000-users Subject: Re: [Nek5000-users] angular velocity to the mesh velocity Thank you very much Paul for your detailed replay and I am very glad about it. But I am sorry if i wrongly said my question because the answer you said was not the one I was expecting. The thing which you have said I had already taken into account (the blending function ). In fact we already had a talk about this previously and I also had developed another blending function with another equation. my question was is apart from wx,wy,wz. Is there any way to give the angular velocity to the mesh (like wr,,wtheta,wz). Since my proile pitch about an axis I will have an angular velocity and I am bit confused how should I give this angular velocity as an input to the mesh velocity. Thanks with regards Sijo GEORGE De: "nek5000-users" ?: "nek5000-users" Envoy?: Vendredi 9 F?vrier 2018 16:02:34 Objet: Re: [Nek5000-users] angular velocity to the mesh velocity Sorry for the slow response but was on travel yesterday. For moving meshes with rigid-body motion of pitch+translation, say, you essentially need a smooth function that is 1 at the moving boundary and 0 at the far field boundaries. Call this function h(X,t)=h(x,y,z,t). Then your mesh velocity (W=[wx,wy,wz]), which is required by Nek's ALE formulation, is given by Translational Velocity(t)*h + Rotational Velocity*h Where h==1 (i.e., on the boundary) you will have the correct velocity. Where h==0 (on the far field), you will have no mesh velocity. h is in effect a blending function between these regions. My favorite way of generating h is to solve a Laplace equation with a variable diffusivity. A high diffusivity near the object ensures h=1 near the object and thus elements near the object move _with_ the object, which means that they do not get excessively deformed. I attach a .usr file from a case I ran several years ago -- I've not tested it recently so please take it as a guide and not a solution to your problem. Look for "Greens Function" to see how to compute h. Here is a paper describing the approach [ http://www.mcs.anl.gov/~fischer/pubhtml/ccfd.pdf | http://www.mcs.anl.gov/~fischer/pubhtml/ccfd.pdf ] Paul PS - Note - Nek requires _only_ the mesh velocity -- mesh position is updated by Nek. Note that you may, or may not, need to recompute the Green's function.... It is often a smooth function even after the mesh is deformed, so you can keep using it.... I suggest doing some experiments for your particular case. From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Friday, February 9, 2018 8:36:39 AM To: nek5000-users Subject: Re: [Nek5000-users] angular velocity to the mesh velocity Thank you for your replay Marco. I will explain from the begening clearly. My current task is to validate Fluid structre interaction on aero foil(NACA profile). The equations which i have to solve I have attached for aero foil. Before this case, I had succesfully done a case of vortex induced vibration of cylinder in NEK5000. And the code below shows a part of the forced oscillating cylinder. From this code I learned that the body motion prescribed through the mesh velocities but I also think that I can only give translational velocity components in NEK5000 to presribe a body motion (mesh as well). Since my aero foil has pitching (it rotates up and down about an axis at angle theta). Since it has to be a FSI problem my plan is by using " call torque_calc(scale,x0,.true.,.true.)" I will get the M (moment about the elastic axis (x0)) and it gives as an input to the Equation of angular motion (attached with this email). Then using Range Kutta method I get the angular velocity of the profile (r, theta,z co ordinates). Then I am planning to conver this angular velocity into linear velocity components (x, y, z direction) then impose into the wx,wy,wz components and I hope it leads to the movement of mesh thus the geometry. My question is could you think that its possible? Or is there any other simple way to give this angular velocity to my NACA profile about the elastic axis (x0)? And I feel difficulty to convert the angular velocity into linear velocty and gives its to each nodes which is on "mv" B.C . Could you please help me with this situation by an example if you have. Thanks in advance c Cylinder velocity components ucx = 0.0 ucy = amplitude * omega * cos(omega*time) ucz = 0.0 n = nx1*ny1*nz1*nelv do i=1,n ! Translational velocity wx(i,1,1,1) = basev(i)*ucx ! component. wy(i,1,1,1) = basev(i)*ucy wz(i,1,1,1) = basev(i)*ucz umeshx(i,1,1,1) = wx(i,1,1,1) umeshy(i,1,1,1) = wy(i,1,1,1) umeshz(i,1,1,1) = wz(i,1,1,1) enddo De: "nek5000-users" ?: "nek5000-users" Envoy?: Vendredi 9 F?vrier 2018 15:03:26 Objet: Re: [Nek5000-users] angular velocity to the mesh velocity Sijo, The mv_cyl example is a 2-D moving body example. As for the mesh velocity, can you please be more specific? Marco From: Nek5000-users on behalf of "nek5000-users at lists.mcs.anl.gov" Reply-To: "nek5000-users at lists.mcs.anl.gov" Date: Friday, February 9, 2018 at 8:56 AM To: nek5000-users Subject: [Nek5000-users] angular velocity to the mesh velocity Hi Neks, Its a long time that i came up with something. So here we go, SInce I was doinf the FSI of aero foil, I have to deal with 2 degrees of freedom (its a 2D case) one in vertical and another in angular direction about the elastic axis of the profile. I could code the vertical motion of the foil but before starting the angular motion I have some doubts. Is there anyway to give the angular velocities directly to the mesh velocities? Apart from the translational velocity components Or Do you want me to covert the angluar velocities into translational components and give to mesh velocities? Do you have any example with a body which is rotating? Thanks in advance 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 Fri Feb 9 10:26:43 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 9 Feb 2018 16:26:43 +0000 Subject: [Nek5000-users] angular velocity to the mesh velocity In-Reply-To: References: , Message-ID: Dear Sijo, Here is the code from the usr file that implements the rotation. The umesh values are used to set the velocity bcs in userbc. n = nx1*ny1*nz1*nelv x0 = 0.250 y0 = 0.000 amp = 2.*pi/180 aa = amp*omega*sin(omega*time) do i=1,n ! Translational velocity x = xm1(i,1,1,1) - x0 y = ym1(i,1,1,1) - y0 ucx = aa*y ucy = -aa*x ucz = 0.0 wx(i,1,1,1) = basev(i)*ucx ! component. wy(i,1,1,1) = basev(i)*ucy wz(i,1,1,1) = basev(i)*ucz umeshx(i,1,1,1) = wx(i,1,1,1) umeshy(i,1,1,1) = wy(i,1,1,1) umeshz(i,1,1,1) = wz(i,1,1,1) enddo ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Friday, February 9, 2018 9:36:54 AM To: nek5000-users Subject: Re: [Nek5000-users] angular velocity to the mesh velocity Thanks Paul I give a try. Just a last question In order to find theta and r should i make a line from origin to each nodes in the whole mesh? Or is it possible to only the nodes which has "mv " boundary condition. Thanks regards Sijo GEORGE ________________________________ De: "nek5000-users" ?: "nek5000-users" Envoy?: Vendredi 9 F?vrier 2018 16:28:48 Objet: Re: [Nek5000-users] angular velocity to the mesh velocity Dear Sijo, (wx,wy,wz) = ( r cos theta, r sin theta , 0 ) * h(x,y,z) where r = sqrt(x*x+y*y), cos theta = x/r, sin theta = y/r Assuming that you are pitching about the origin. Paul ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Friday, February 9, 2018 9:15:15 AM To: nek5000-users Subject: Re: [Nek5000-users] angular velocity to the mesh velocity Thank you very much Paul for your detailed replay and I am very glad about it. But I am sorry if i wrongly said my question because the answer you said was not the one I was expecting. The thing which you have said I had already taken into account (the blending function ). In fact we already had a talk about this previously and I also had developed another blending function with another equation. my question was is apart from wx,wy,wz. Is there any way to give the angular velocity to the mesh (like wr,,wtheta,wz). Since my proile pitch about an axis I will have an angular velocity and I am bit confused how should I give this angular velocity as an input to the mesh velocity. Thanks with regards Sijo GEORGE ________________________________ De: "nek5000-users" ?: "nek5000-users" Envoy?: Vendredi 9 F?vrier 2018 16:02:34 Objet: Re: [Nek5000-users] angular velocity to the mesh velocity Sorry for the slow response but was on travel yesterday. For moving meshes with rigid-body motion of pitch+translation, say, you essentially need a smooth function that is 1 at the moving boundary and 0 at the far field boundaries. Call this function h(X,t)=h(x,y,z,t). Then your mesh velocity (W=[wx,wy,wz]), which is required by Nek's ALE formulation, is given by Translational Velocity(t)*h + Rotational Velocity*h Where h==1 (i.e., on the boundary) you will have the correct velocity. Where h==0 (on the far field), you will have no mesh velocity. h is in effect a blending function between these regions. My favorite way of generating h is to solve a Laplace equation with a variable diffusivity. A high diffusivity near the object ensures h=1 near the object and thus elements near the object move _with_ the object, which means that they do not get excessively deformed. I attach a .usr file from a case I ran several years ago -- I've not tested it recently so please take it as a guide and not a solution to your problem. Look for "Greens Function" to see how to compute h. Here is a paper describing the approach http://www.mcs.anl.gov/~fischer/pubhtml/ccfd.pdf Paul PS - Note - Nek requires _only_ the mesh velocity -- mesh position is updated by Nek. Note that you may, or may not, need to recompute the Green's function.... It is often a smooth function even after the mesh is deformed, so you can keep using it.... I suggest doing some experiments for your particular case. ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Friday, February 9, 2018 8:36:39 AM To: nek5000-users Subject: Re: [Nek5000-users] angular velocity to the mesh velocity Thank you for your replay Marco. I will explain from the begening clearly. My current task is to validate Fluid structre interaction on aero foil(NACA profile). The equations which i have to solve I have attached for aero foil. Before this case, I had succesfully done a case of vortex induced vibration of cylinder in NEK5000. And the code below shows a part of the forced oscillating cylinder. From this code I learned that the body motion prescribed through the mesh velocities but I also think that I can only give translational velocity components in NEK5000 to presribe a body motion (mesh as well). Since my aero foil has pitching (it rotates up and down about an axis at angle theta). Since it has to be a FSI problem my plan is by using " call torque_calc(scale,x0,.true.,.true.)" I will get the M (moment about the elastic axis (x0)) and it gives as an input to the Equation of angular motion (attached with this email). Then using Range Kutta method I get the angular velocity of the profile (r, theta,z co ordinates). Then I am planning to conver this angular velocity into linear velocity components (x, y, z direction) then impose into the wx,wy,wz components and I hope it leads to the movement of mesh thus the geometry. My question is could you think that its possible? Or is there any other simple way to give this angular velocity to my NACA profile about the elastic axis (x0)? And I feel difficulty to convert the angular velocity into linear velocty and gives its to each nodes which is on "mv" B.C . Could you please help me with this situation by an example if you have. Thanks in advance c Cylinder velocity components ucx = 0.0 ucy = amplitude * omega * cos(omega*time) ucz = 0.0 n = nx1*ny1*nz1*nelv do i=1,n ! Translational velocity wx(i,1,1,1) = basev(i)*ucx ! component. wy(i,1,1,1) = basev(i)*ucy wz(i,1,1,1) = basev(i)*ucz umeshx(i,1,1,1) = wx(i,1,1,1) umeshy(i,1,1,1) = wy(i,1,1,1) umeshz(i,1,1,1) = wz(i,1,1,1) enddo ________________________________ De: "nek5000-users" ?: "nek5000-users" Envoy?: Vendredi 9 F?vrier 2018 15:03:26 Objet: Re: [Nek5000-users] angular velocity to the mesh velocity Sijo, The mv_cyl example is a 2-D moving body example. As for the mesh velocity, can you please be more specific? Marco From: Nek5000-users on behalf of "nek5000-users at lists.mcs.anl.gov" Reply-To: "nek5000-users at lists.mcs.anl.gov" Date: Friday, February 9, 2018 at 8:56 AM To: nek5000-users Subject: [Nek5000-users] angular velocity to the mesh velocity Hi Neks, Its a long time that i came up with something. So here we go, SInce I was doinf the FSI of aero foil, I have to deal with 2 degrees of freedom (its a 2D case) one in vertical and another in angular direction about the elastic axis of the profile. I could code the vertical motion of the foil but before starting the angular motion I have some doubts. Is there anyway to give the angular velocities directly to the mesh velocities? Apart from the translational velocity components Or Do you want me to covert the angluar velocities into translational components and give to mesh velocities? Do you have any example with a body which is rotating? Thanks in advance 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 Fri Feb 9 11:24:20 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 9 Feb 2018 18:24:20 +0100 (CET) Subject: [Nek5000-users] angular velocity to the mesh velocity In-Reply-To: References: Message-ID: Thanks paul. I am trying your method. And Could you please tell me what is the use of this subroutine? SUBROUTINE DLARTG( F, G, CS, SN, R ) Do you think that I can take any advantages of this function in my angular rotation case? If I could give the x and y components of the velocity as input to this function what would be the result? Sorry for asking again and again I am trying to understand your codes. Thanks again Sijo De: "nek5000-users" ?: "nek5000-users" Envoy?: Vendredi 9 F?vrier 2018 17:26:43 Objet: Re: [Nek5000-users] angular velocity to the mesh velocity Dear Sijo, Here is the code from the usr file that implements the rotation. The umesh values are used to set the velocity bcs in userbc. n = nx1*ny1*nz1*nelv x0 = 0.250 y0 = 0.000 amp = 2.*pi/180 aa = amp*omega*sin(omega*time) do i=1,n ! Translational velocity x = xm1(i,1,1,1) - x0 y = ym1(i,1,1,1) - y0 ucx = aa*y ucy = -aa*x ucz = 0.0 wx(i,1,1,1) = basev(i)*ucx ! component. wy(i,1,1,1) = basev(i)*ucy wz(i,1,1,1) = basev(i)*ucz umeshx(i,1,1,1) = wx(i,1,1,1) umeshy(i,1,1,1) = wy(i,1,1,1) umeshz(i,1,1,1) = wz(i,1,1,1) enddo From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Friday, February 9, 2018 9:36:54 AM To: nek5000-users Subject: Re: [Nek5000-users] angular velocity to the mesh velocity Thanks Paul I give a try. Just a last question In order to find theta and r should i make a line from origin to each nodes in the whole mesh? Or is it possible to only the nodes which has "mv " boundary condition. Thanks regards Sijo GEORGE De: "nek5000-users" ?: "nek5000-users" Envoy?: Vendredi 9 F?vrier 2018 16:28:48 Objet: Re: [Nek5000-users] angular velocity to the mesh velocity Dear Sijo, (wx,wy,wz) = ( r cos theta, r sin theta , 0 ) * h(x,y,z) where r = sqrt(x*x+y*y), cos theta = x/r, sin theta = y/r Assuming that you are pitching about the origin. Paul From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Friday, February 9, 2018 9:15:15 AM To: nek5000-users Subject: Re: [Nek5000-users] angular velocity to the mesh velocity Thank you very much Paul for your detailed replay and I am very glad about it. But I am sorry if i wrongly said my question because the answer you said was not the one I was expecting. The thing which you have said I had already taken into account (the blending function ). In fact we already had a talk about this previously and I also had developed another blending function with another equation. my question was is apart from wx,wy,wz. Is there any way to give the angular velocity to the mesh (like wr,,wtheta,wz). Since my proile pitch about an axis I will have an angular velocity and I am bit confused how should I give this angular velocity as an input to the mesh velocity. Thanks with regards Sijo GEORGE De: "nek5000-users" ?: "nek5000-users" Envoy?: Vendredi 9 F?vrier 2018 16:02:34 Objet: Re: [Nek5000-users] angular velocity to the mesh velocity Sorry for the slow response but was on travel yesterday. For moving meshes with rigid-body motion of pitch+translation, say, you essentially need a smooth function that is 1 at the moving boundary and 0 at the far field boundaries. Call this function h(X,t)=h(x,y,z,t). Then your mesh velocity (W=[wx,wy,wz]), which is required by Nek's ALE formulation, is given by Translational Velocity(t)*h + Rotational Velocity*h Where h==1 (i.e., on the boundary) you will have the correct velocity. Where h==0 (on the far field), you will have no mesh velocity. h is in effect a blending function between these regions. My favorite way of generating h is to solve a Laplace equation with a variable diffusivity. A high diffusivity near the object ensures h=1 near the object and thus elements near the object move _with_ the object, which means that they do not get excessively deformed. I attach a .usr file from a case I ran several years ago -- I've not tested it recently so please take it as a guide and not a solution to your problem. Look for "Greens Function" to see how to compute h. Here is a paper describing the approach [ http://www.mcs.anl.gov/~fischer/pubhtml/ccfd.pdf | http://www.mcs.anl.gov/~fischer/pubhtml/ccfd.pdf ] Paul PS - Note - Nek requires _only_ the mesh velocity -- mesh position is updated by Nek. Note that you may, or may not, need to recompute the Green's function.... It is often a smooth function even after the mesh is deformed, so you can keep using it.... I suggest doing some experiments for your particular case. From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Friday, February 9, 2018 8:36:39 AM To: nek5000-users Subject: Re: [Nek5000-users] angular velocity to the mesh velocity Thank you for your replay Marco. I will explain from the begening clearly. My current task is to validate Fluid structre interaction on aero foil(NACA profile). The equations which i have to solve I have attached for aero foil. Before this case, I had succesfully done a case of vortex induced vibration of cylinder in NEK5000. And the code below shows a part of the forced oscillating cylinder. From this code I learned that the body motion prescribed through the mesh velocities but I also think that I can only give translational velocity components in NEK5000 to presribe a body motion (mesh as well). Since my aero foil has pitching (it rotates up and down about an axis at angle theta). Since it has to be a FSI problem my plan is by using " call torque_calc(scale,x0,.true.,.true.)" I will get the M (moment about the elastic axis (x0)) and it gives as an input to the Equation of angular motion (attached with this email). Then using Range Kutta method I get the angular velocity of the profile (r, theta,z co ordinates). Then I am planning to conver this angular velocity into linear velocity components (x, y, z direction) then impose into the wx,wy,wz components and I hope it leads to the movement of mesh thus the geometry. My question is could you think that its possible? Or is there any other simple way to give this angular velocity to my NACA profile about the elastic axis (x0)? And I feel difficulty to convert the angular velocity into linear velocty and gives its to each nodes which is on "mv" B.C . Could you please help me with this situation by an example if you have. Thanks in advance c Cylinder velocity components ucx = 0.0 ucy = amplitude * omega * cos(omega*time) ucz = 0.0 n = nx1*ny1*nz1*nelv do i=1,n ! Translational velocity wx(i,1,1,1) = basev(i)*ucx ! component. wy(i,1,1,1) = basev(i)*ucy wz(i,1,1,1) = basev(i)*ucz umeshx(i,1,1,1) = wx(i,1,1,1) umeshy(i,1,1,1) = wy(i,1,1,1) umeshz(i,1,1,1) = wz(i,1,1,1) enddo De: "nek5000-users" ?: "nek5000-users" Envoy?: Vendredi 9 F?vrier 2018 15:03:26 Objet: Re: [Nek5000-users] angular velocity to the mesh velocity Sijo, The mv_cyl example is a 2-D moving body example. As for the mesh velocity, can you please be more specific? Marco From: Nek5000-users on behalf of "nek5000-users at lists.mcs.anl.gov" Reply-To: "nek5000-users at lists.mcs.anl.gov" Date: Friday, February 9, 2018 at 8:56 AM To: nek5000-users Subject: [Nek5000-users] angular velocity to the mesh velocity Hi Neks, Its a long time that i came up with something. So here we go, SInce I was doinf the FSI of aero foil, I have to deal with 2 degrees of freedom (its a 2D case) one in vertical and another in angular direction about the elastic axis of the profile. I could code the vertical motion of the foil but before starting the angular motion I have some doubts. Is there anyway to give the angular velocities directly to the mesh velocities? Apart from the translational velocity components Or Do you want me to covert the angluar velocities into translational components and give to mesh velocities? Do you have any example with a body which is rotating? Thanks in advance 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 -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Fri Feb 9 13:31:25 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 9 Feb 2018 19:31:25 +0000 Subject: [Nek5000-users] DNS of turbChannel Message-ID: Hello Paul, Thank you a lot for your explanation; it really helps. For now, I am using the most recent turChannel example of NEK5000 V.17. I successfully simulated a turbulent channel with a Retau of 180 using 1536 elements (I am aware of it's a lot of elements). The default settings were applied. I wonder is it in the .par file where you do the modification to convert it into DNS? Or somewhere else as well? Again, thank you for the help, Emily -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Fri Feb 9 14:12:18 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 9 Feb 2018 21:12:18 +0100 Subject: [Nek5000-users] DNS of turbChannel Message-ID: The turbChannel example is using a high pass filter relaxation term as SGS. For a DNS just - set lx1=8 and lxd=12 in SIZE - generate a finer mesh with more elements using genbox (if you change the mesh make sure to delete the old amg files) - set filtering=no in par -----Original message----- > From:nek5000-users at lists.mcs.anl.gov > Sent: Friday 9th February 2018 20:32 > To: nek5000-users at lists.mcs.anl.gov > Subject: Re: [Nek5000-users] DNS of turbChannel > > Hello Paul, > Thank you a lot for your explanation; it really helps. For now, I am using the most recent turChannel example of NEK5000 V.17. I successfully simulated a turbulent channel with a Retau of 180 using 1536 elements (I am aware of its a lot > of elements). The default settings were applied. I wonder is it in the .par file where you do the modification to convert it into DNS? Or somewhere else as well? > Again, thank you for the help, > Emily > _______________________________________________ > 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 Feb 10 03:52:19 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sat, 10 Feb 2018 10:52:19 +0100 Subject: [Nek5000-users] DNS of turbChannel In-Reply-To: References: Message-ID: does "filtering=no" turn off the filtering completely? As far as I understand, even for DNS* some low amounts of filtering is needed (1% or so). Or is that not true any longer? Philipp * Of course, the true definition of DNS might be a bit unclear; we typically say that a DNS resolves all relevant scales of motion, but what are these, really? On 2018-02-09 21:12, nek5000-users at lists.mcs.anl.gov wrote: > The turbChannel example is using a high pass filter relaxation term as SGS. For a DNS just > > - set lx1=8 and lxd=12 in SIZE > - generate a finer mesh with more elements using genbox (if you change the mesh make sure to delete the old amg files) > - set filtering=no in par > > > -----Original message----- >> From:nek5000-users at lists.mcs.anl.gov >> Sent: Friday 9th February 2018 20:32 >> To: nek5000-users at lists.mcs.anl.gov >> Subject: Re: [Nek5000-users] DNS of turbChannel >> >> Hello Paul, >> Thank you a lot for your explanation; it really helps. For now, I am using the most recent turChannel example of NEK5000 V.17. I successfully simulated a turbulent channel with a Retau of 180 using 1536 elements (I am aware of its a lot >> of elements). The default settings were applied. I wonder is it in the .par file where you do the modification to convert it into DNS? Or somewhere else as well? >> Again, thank you for the help, >> Emily >> _______________________________________________ >> 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 Feb 10 04:54:30 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sat, 10 Feb 2018 11:54:30 +0100 Subject: [Nek5000-users] DNS of turbChannel In-Reply-To: References: Message-ID: Yes it will turn it off. I agree this depends on the resolution. If you are really well resolved dealising is typicaly enough to make it stable otherwise you have to add filtering. -----Original message----- > From:nek5000-users at lists.mcs.anl.gov > Sent: Saturday 10th February 2018 10:52 > To: nek5000-users at lists.mcs.anl.gov > Subject: Re: [Nek5000-users] DNS of turbChannel > > does "filtering=no" turn off the filtering completely? As far as I > understand, even for DNS* some low amounts of filtering is needed (1% or > so). Or is that not true any longer? > > Philipp > > * Of course, the true definition of DNS might be a bit unclear; we > typically say that a DNS resolves all relevant scales of motion, but > what are these, really? > > On 2018-02-09 21:12, nek5000-users at lists.mcs.anl.gov wrote: > > The turbChannel example is using a high pass filter relaxation term as SGS. For a DNS just > > > > - set lx1=8 and lxd=12 in SIZE > > - generate a finer mesh with more elements using genbox (if you change the mesh make sure to delete the old amg files) > > - set filtering=no in par > > > > > > -----Original message----- > >> From:nek5000-users at lists.mcs.anl.gov > >> Sent: Friday 9th February 2018 20:32 > >> To: nek5000-users at lists.mcs.anl.gov > >> Subject: Re: [Nek5000-users] DNS of turbChannel > >> > >> Hello Paul, > >> Thank you a lot for your explanation; it really helps. For now, I am using the most recent turChannel example of NEK5000 V.17. I successfully simulated a turbulent channel with a Retau of 180 using 1536 elements (I am aware of its a lot > >> of elements). The default settings were applied. I wonder is it in the .par file where you do the modification to convert it into DNS? Or somewhere else as well? > >> Again, thank you for the help, > >> Emily > >> _______________________________________________ > >> 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 Feb 10 05:39:11 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sat, 10 Feb 2018 12:39:11 +0100 Subject: [Nek5000-users] DNS of turbChannel In-Reply-To: References: Message-ID: We typically have filtering on with low amplitude even for DNS (say 1%); in particular at high Re (whatever that means...). Dealiasing is always on for us. Philipp On 2018-02-10 11:54, nek5000-users at lists.mcs.anl.gov wrote: > Yes it will turn it off. > I agree this depends on the resolution. If you are really well resolved dealising is typicaly enough to make it stable otherwise you have to add filtering. > > -----Original message----- >> From:nek5000-users at lists.mcs.anl.gov >> Sent: Saturday 10th February 2018 10:52 >> To: nek5000-users at lists.mcs.anl.gov >> Subject: Re: [Nek5000-users] DNS of turbChannel >> >> does "filtering=no" turn off the filtering completely? As far as I >> understand, even for DNS* some low amounts of filtering is needed (1% or >> so). Or is that not true any longer? >> >> Philipp >> >> * Of course, the true definition of DNS might be a bit unclear; we >> typically say that a DNS resolves all relevant scales of motion, but >> what are these, really? >> >> On 2018-02-09 21:12, nek5000-users at lists.mcs.anl.gov wrote: >>> The turbChannel example is using a high pass filter relaxation term as SGS. For a DNS just >>> >>> - set lx1=8 and lxd=12 in SIZE >>> - generate a finer mesh with more elements using genbox (if you change the mesh make sure to delete the old amg files) >>> - set filtering=no in par >>> >>> >>> -----Original message----- >>>> From:nek5000-users at lists.mcs.anl.gov >>>> Sent: Friday 9th February 2018 20:32 >>>> To: nek5000-users at lists.mcs.anl.gov >>>> Subject: Re: [Nek5000-users] DNS of turbChannel >>>> >>>> Hello Paul, >>>> Thank you a lot for your explanation; it really helps. For now, I am using the most recent turChannel example of NEK5000 V.17. I successfully simulated a turbulent channel with a Retau of 180 using 1536 elements (I am aware of its a lot >>>> of elements). The default settings were applied. I wonder is it in the .par file where you do the modification to convert it into DNS? Or somewhere else as well? >>>> Again, thank you for the help, >>>> Emily >>>> _______________________________________________ >>>> 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 Feb 10 11:28:01 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sat, 10 Feb 2018 17:28:01 +0000 Subject: [Nek5000-users] turbChannel: Questions about targetCFL and dealiasing Message-ID: Dear Nek users, I have some fundamental questions in terms of CFL and dealiasing in the example of turbChannel. In the .par file, I noticed that the default setting of targetCFL is as high as 3.5, which really makes the code run very fast. But, typically, this CFL value is around 0.5 using implicit fractional time-step approach based on the FV method. Why CFL could be this high here while not compromising the stability? Is it because of the spectral element method? And Could you also please tell me when shall I turn on or off dealiasing? Regards, Emily -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Sun Feb 11 04:49:16 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sun, 11 Feb 2018 11:49:16 +0100 Subject: [Nek5000-users] turbChannel: Questions about targetCFL and dealiasing In-Reply-To: References: Message-ID: The code implements OIFS (operator integrator splitting factor) not an implicit fractional step method! A convection problem is integrated in substeps with RK4. For a targetCFL of 3.5 you need at least two substeps to be stable. You can increase the targetCFL even further depending on the desired accuracy. ?If this helps to lower time to solution depends on the cost of the pressure solve. If you want to run the solver as fast as possible turn off dealiasing. In all other cases turn it on (lxd has to have the right size e.g. 3/2 rule). On 10 Feb 2018, at 18:28, "nek5000-users at lists.mcs.anl.gov " > wrote: ?Dear Nek users, ? I have some fundamental questions in terms of CFL and dealiasing in the example of turbChannel. In the .par file, I noticed that the default setting of targetCFL is as high as 3.5, which really makes the code run very fast. But, typically, this CFL value is around 0.5 using implicit fractional time-step approach based on the FV method. Why CFL could be this high here while not compromising the stability? Is it because of the spectral element method? ? And Could you also please tell me when shall I turn on or off dealiasing? ? Regards, ? Emily _______________________________________________ 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 Sun Feb 11 05:07:36 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sun, 11 Feb 2018 12:07:36 +0100 Subject: [Nek5000-users] turbChannel: Questions about targetCFL and dealiasing In-Reply-To: References: Message-ID: if you use the characteristic scheme that is. In the other case (i.e. ifchar turned off), it is a classic IMEX scheme with target CFL of 0.5. I wonder, have there been any comparisons between the two schemes for standard cases like turbulent channel flow? On 2018-02-11 11:49, nek5000-users at lists.mcs.anl.gov wrote: > The code implements OIFS (operator integrator splitting factor) not an > implicit fractional step method! A convection problem is integrated in > substeps with RK4. For a targetCFL of 3.5 you need at least two substeps > to be stable. You can increase the targetCFL even further depending on > the desired accuracy. ?If this helps to lower time to solution depends > on the cost of the pressure solve. > > If you want to run the solver as fast as possible turn off dealiasing. > In all other cases turn it on (lxd has to have the right size e.g. 3/2 > rule). > > On 10 Feb 2018, at 18:28, "nek5000-users at lists.mcs.anl.gov > " > > wrote: >> >> Dear Nek users, >> >> I have some fundamental questions in terms of CFL and dealiasing in >> the example of turbChannel. In the .par file, I noticed that the >> default setting of targetCFL is as high as 3.5, which really makes the >> code run very fast. But, typically, this CFL value is around 0.5 using >> implicit fractional time-step approach based on the FV method. Why CFL >> could be this high here while not compromising the stability? Is it >> because of the spectral element method? >> >> And Could you also please tell me when shall I turn on or off dealiasing? >> >> Regards, >> >> Emily >> >> _______________________________________________ >> >> Nek5000-users mailing list >> >> Nek5000-users at lists.mcs.anl.gov >> >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> > > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > From nek5000-users at lists.mcs.anl.gov Sun Feb 11 16:57:55 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sun, 11 Feb 2018 22:57:55 +0000 Subject: [Nek5000-users] Transitioning from v16 to v17 Message-ID: Hi all, I'm thinking of transitioning from v16 to v17. I do however have a large number of previous simulations in v16. I'd like to be able to keep using them in v17. Additionally the cluster I have acess to has only v16 installed and the process to upgrade the version could take a significant amount of time. What is the back-compatibility level between v16 and v17? What factors must I keep in mind so that I can run my v16 versions on v17? Sincerely, -- *Amitvikram Dutta* Graduate Research Assistant Fluid Mechanics Research Lab Multi-Physics Interaction Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Feb 12 04:03:35 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 12 Feb 2018 11:03:35 +0100 Subject: [Nek5000-users] Transitioning from v16 to v17 Message-ID: Note, 16.0 was a development version of the code and not an official release. We do strongly recommend to update to v17.0. There is no installation process. Just untar the tarball and compile the code. In less than 2-3min you are done. The code is fully backward compatbile. For more details please check the release notes. -----Original message----- > From:nek5000-users at lists.mcs.anl.gov > Sent: Sunday 11th February 2018 23:58 > To: nek5000-users at lists.mcs.anl.gov > Subject: [Nek5000-users] Transitioning from v16 to v17 > > Hi all, > > Im thinking of transitioning from v16 to v17. I do however have a large number of previous simulations in v16. Id like to be able to keep using them in v17. Additionally the cluster I have acess to has only v16 installed and the process to upgrade the version could take a significant amount of time. > > What is the back-compatibility level between v16 and v17? What factors must I keep in mind so that I can run my v16 versions on v17? > > Sincerely, > -- > Amitvikram Dutta > Graduate Research Assistant > Fluid Mechanics Research Lab > Multi-Physics Interaction 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 Feb 12 12:08:18 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 12 Feb 2018 12:08:18 -0600 Subject: [Nek5000-users] Nek5000-users Digest, Vol 108, Issue 18 In-Reply-To: References: Message-ID: Hi Amitvikram, for all practical purposes there shouldn't be any backward compatibility issues between the two versions of Nek5000. It is recommended to use the latest, because there has been some improvements which has been found to reduce the solution time. Also, there is no installation involved with Nek5000, if you have v16 working, you can just download the v17 into your home folder and build/compile exactly the same way you built the previous version. I hope this helps. cheers Som Som Dutta Post Doctoral Research Associate Dept. of Computer Science University of Illinois at Urbana-Champaign On Mon, Feb 12, 2018 at 12:00 PM, 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. Transitioning from v16 to v17 (nek5000-users at lists.mcs.anl.gov) > 2. Re: Transitioning from v16 to v17 > (nek5000-users at lists.mcs.anl.gov) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Sun, 11 Feb 2018 22:57:55 +0000 > From: nek5000-users at lists.mcs.anl.gov > To: "nek5000-users at lists.mcs.anl.gov" > > Subject: [Nek5000-users] Transitioning from v16 to v17 > Message-ID: > > Content-Type: text/plain; charset="utf-8" > > Hi all, > > I'm thinking of transitioning from v16 to v17. I do however have a large > number of previous simulations in v16. I'd like to be able to keep using > them in v17. Additionally the cluster I have acess to has only v16 > installed and the process to upgrade the version could take a significant > amount of time. > > What is the back-compatibility level between v16 and v17? What factors must > I keep in mind so that I can run my v16 versions on v17? > > Sincerely, > -- > > *Amitvikram Dutta* > > Graduate Research Assistant > > Fluid Mechanics Research Lab > > Multi-Physics Interaction Lab > > University of Waterloo > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: attachments/20180211/b07455fe/attachment-0001.html> > > ------------------------------ > > Message: 2 > Date: Mon, 12 Feb 2018 11:03:35 +0100 > From: nek5000-users at lists.mcs.anl.gov > To: nek5000-users at lists.mcs.anl.gov > Subject: Re: [Nek5000-users] Transitioning from v16 to v17 > Message-ID: > > Content-Type: text/plain; charset=utf-8 > > Note, 16.0 was a development version of the code and not an official > release. We do strongly recommend to update to v17.0. There is no > installation process. Just untar the tarball and compile the code. In less > than 2-3min you are done. The code is fully backward compatbile. For more > details please check the release notes. > > -----Original message----- > > From:nek5000-users at lists.mcs.anl.gov > > Sent: Sunday 11th February 2018 23:58 > > To: nek5000-users at lists.mcs.anl.gov > > Subject: [Nek5000-users] Transitioning from v16 to v17 > > > > Hi all, > > > > Im thinking of transitioning from v16 to v17. I do however have a large > number of previous simulations in v16. Id like to be able to keep using > them in v17. Additionally the cluster I have acess to has only v16 > installed and the process to upgrade the version could take a significant > amount of time. > > > > What is the back-compatibility level between v16 and v17? What factors > must I keep in mind so that I can run my v16 versions on v17? > > > > Sincerely, > > -- > > Amitvikram Dutta > > Graduate Research Assistant > > Fluid Mechanics Research Lab > > Multi-Physics Interaction Lab > > University of Waterloo > > _______________________________________________ > > 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 108, Issue 18 > ********************************************** > -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Feb 12 18:46:27 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 13 Feb 2018 00:46:27 +0000 Subject: [Nek5000-users] Nek5000-users Digest, Vol 108, Issue 18 In-Reply-To: References: Message-ID: Thanks Som On Mon, Feb 12, 2018 at 6:13 PM wrote: > Hi Amitvikram, > > > for all practical purposes there shouldn't be any backward compatibility > issues between the > two versions of Nek5000. > > It is recommended to use the latest, because there has been some > improvements which > has been found to reduce the solution time. > > Also, there is no installation involved with Nek5000, if you have v16 > working, > you can just download the v17 into your home folder and build/compile > exactly the same way you built the previous version. > > I hope this helps. > > cheers > > Som > > Som Dutta > Post Doctoral Research Associate > Dept. of Computer Science > University of Illinois at Urbana-Champaign > > On Mon, Feb 12, 2018 at 12:00 PM, > 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. Transitioning from v16 to v17 (nek5000-users at lists.mcs.anl.gov) >> 2. Re: Transitioning from v16 to v17 >> (nek5000-users at lists.mcs.anl.gov) >> >> >> ---------------------------------------------------------------------- >> >> Message: 1 >> Date: Sun, 11 Feb 2018 22:57:55 +0000 >> From: nek5000-users at lists.mcs.anl.gov >> To: "nek5000-users at lists.mcs.anl.gov" >> >> Subject: [Nek5000-users] Transitioning from v16 to v17 >> Message-ID: >> >> Content-Type: text/plain; charset="utf-8" >> >> Hi all, >> >> I'm thinking of transitioning from v16 to v17. I do however have a large >> number of previous simulations in v16. I'd like to be able to keep using >> them in v17. Additionally the cluster I have acess to has only v16 >> installed and the process to upgrade the version could take a significant >> amount of time. >> >> What is the back-compatibility level between v16 and v17? What factors >> must >> I keep in mind so that I can run my v16 versions on v17? >> >> Sincerely, >> -- >> >> *Amitvikram Dutta* >> >> Graduate Research Assistant >> >> Fluid Mechanics Research Lab >> >> Multi-Physics Interaction Lab >> >> University of Waterloo >> -------------- next part -------------- >> An HTML attachment was scrubbed... >> URL: < >> http://lists.mcs.anl.gov/pipermail/nek5000-users/attachments/20180211/b07455fe/attachment-0001.html >> > >> >> ------------------------------ >> >> Message: 2 >> Date: Mon, 12 Feb 2018 11:03:35 +0100 >> From: nek5000-users at lists.mcs.anl.gov >> To: nek5000-users at lists.mcs.anl.gov >> Subject: Re: [Nek5000-users] Transitioning from v16 to v17 >> Message-ID: >> >> Content-Type: text/plain; charset=utf-8 >> >> Note, 16.0 was a development version of the code and not an official >> release. We do strongly recommend to update to v17.0. There is no >> installation process. Just untar the tarball and compile the code. In less >> than 2-3min you are done. The code is fully backward compatbile. For more >> details please check the release notes. >> >> -----Original message----- >> > From:nek5000-users at lists.mcs.anl.gov >> > Sent: Sunday 11th February 2018 23:58 >> > To: nek5000-users at lists.mcs.anl.gov >> > Subject: [Nek5000-users] Transitioning from v16 to v17 >> > >> > Hi all, >> > >> > Im thinking of transitioning from v16 to v17. I do however have a large >> number of previous simulations in v16. Id like to be able to keep using >> them in v17. Additionally the cluster I have acess to has only v16 >> installed and the process to upgrade the version could take a significant >> amount of time. >> > >> > What is the back-compatibility level between v16 and v17? What factors >> must I keep in mind so that I can run my v16 versions on v17? >> > >> > Sincerely, >> > -- >> > Amitvikram Dutta >> > Graduate Research Assistant >> > Fluid Mechanics Research Lab >> > Multi-Physics Interaction Lab >> > University of Waterloo >> > _______________________________________________ >> > 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 108, Issue 18 >> ********************************************** >> > > _______________________________________________ > 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 Multi-Physics Interaction Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Feb 13 05:30:35 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 13 Feb 2018 12:30:35 +0100 Subject: [Nek5000-users] Restart using second order scheme Message-ID: Dear Neks: I am using second order time time integration scheme for the computations. Since the second order scheme uses the flow field at (n) and (n-1) to compute the solution at (n+1) we need two files to do a full restart. If, however, we do not do a full restart and simply load the file using restart feature in .rea file, how is the second order scheme implemented for the first time step? Thank you. Best wishes, NN. From nek5000-users at lists.mcs.anl.gov Tue Feb 13 06:13:20 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 13 Feb 2018 13:13:20 +0100 Subject: [Nek5000-users] Restart using second order scheme In-Reply-To: References: Message-ID: Hi, if you do not do full restart, your first time step will be first order. Particularly important is also that if you do not have a good pressure initial condition, you might get large pressure errors, which might require you to use the coupled (Uzawa) algorithm (for PnPn-2). Philipp On 2018-02-13 12:30, nek5000-users at lists.mcs.anl.gov wrote: > Dear Neks: > > ? I am using second order time time integration scheme for the > computations. > Since the second order scheme uses the flow field at (n) and (n-1) to > compute > the solution at (n+1) we need two files to do a full restart. If, > however, we do > not do a full restart and simply load the file using restart feature in > .rea file, > how is the second order scheme implemented for the first time step? > > Thank you. > > Best wishes, > NN. > > _______________________________________________ > 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 Feb 13 06:16:16 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 13 Feb 2018 13:16:16 +0100 Subject: [Nek5000-users] Restart using second order scheme In-Reply-To: References: Message-ID: Thanks very much Philipp. NN. On 02/13/2018 01:13 PM, nek5000-users at lists.mcs.anl.gov wrote: > Hi, > if you do not do full restart, your first time step will be first > order. Particularly important is also that if you do not have a good > pressure initial condition, you might get large pressure errors, which > might require you to use the coupled (Uzawa) algorithm (for PnPn-2). > Philipp > > On 2018-02-13 12:30, nek5000-users at lists.mcs.anl.gov wrote: >> Dear Neks: >> >> I am using second order time time integration scheme for the >> computations. >> Since the second order scheme uses the flow field at (n) and (n-1) to >> compute >> the solution at (n+1) we need two files to do a full restart. If, >> however, we do >> not do a full restart and simply load the file using restart feature >> in .rea file, >> how is the second order scheme implemented for the first time step? >> >> Thank you. >> >> Best wishes, >> 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 > From nek5000-users at lists.mcs.anl.gov Tue Feb 13 09:53:15 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 13 Feb 2018 16:53:15 +0100 Subject: [Nek5000-users] v17.0 vs v17.0.0.beta2 regarding Stress Formulation and Pn/Pn-2 Message-ID: Dear All, I am using mixed 'W ' and 'SYM' boundary conditions on a boundary that is not aligned with x,y,z axes. (with IFSTRS=T) When using the previous v17.0.0.beta2 version, I could use these mixed boundary conditions with the Pn/Pn pressure formulation, giving a continuous pressure field. After upgrading to v17.0, I get a number of the following runtime errors when using Pn/Pn: --- Mixed B.C. on a side nonaligned with either the X,Y, or Z axis detected for Element 8 side 1 Requires PN/PN-2 STRESS FORMULATION --- Pn/Pn-2 runs, as expected. Granted that I have to use Pn/Pn-2 for mixed BC when not aligned with an axis, the problem is that the pressure is discontinuous, which is stated in the documentation for the Pn/Pn-2 formulation. This is evident when plotting the pressure in Visit. Is there a way to obtain a smooth (correct?) pressure field when using Pn/Pn-2? Or is there a way to use the Pn/Pn formulation with such boundary conditions? Thank you, Jean -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Feb 13 10:46:53 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 13 Feb 2018 16:46:53 +0000 Subject: [Nek5000-users] v17.0 vs v17.0.0.beta2 regarding Stress Formulation and Pn/Pn-2 In-Reply-To: References: Message-ID: Jean, I think what you want to do is to map the pressure on the velocity mesh. There is thread that explains how to do that: https://lists.mcs.anl.gov/mailman/htdig/nek5000-users/2010-July/000767.html Since the pressure is discontinuous I am not sure the pressure field will be smooth on the velocity mesh, though. Marco From: Nek5000-users on behalf of "nek5000-users at lists.mcs.anl.gov" Reply-To: "nek5000-users at lists.mcs.anl.gov" Date: Tuesday, February 13, 2018 at 10:54 AM To: nek5000-users Subject: [Nek5000-users] v17.0 vs v17.0.0.beta2 regarding Stress Formulation and Pn/Pn-2 Dear All, I am using mixed 'W ' and 'SYM' boundary conditions on a boundary that is not aligned with x,y,z axes. (with IFSTRS=T) When using the previous v17.0.0.beta2 version, I could use these mixed boundary conditions with the Pn/Pn pressure formulation, giving a continuous pressure field. After upgrading to v17.0, I get a number of the following runtime errors when using Pn/Pn: --- Mixed B.C. on a side nonaligned with either the X,Y, or Z axis detected for Element 8 side 1 Requires PN/PN-2 STRESS FORMULATION --- Pn/Pn-2 runs, as expected. Granted that I have to use Pn/Pn-2 for mixed BC when not aligned with an axis, the problem is that the pressure is discontinuous, which is stated in the documentation for the Pn/Pn-2 formulation. This is evident when plotting the pressure in Visit. Is there a way to obtain a smooth (correct?) pressure field when using Pn/Pn-2? Or is there a way to use the Pn/Pn formulation with such boundary conditions? Thank you, Jean -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Feb 14 05:00:04 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 14 Feb 2018 11:00:04 +0000 Subject: [Nek5000-users] subroutines of gradm1 and interp_v3 Message-ID: Hi Nek, May I know whether the following two subroutines can be used for different geometries without changing? 1. gradm1 ( http://www.mcs.anl.gov/research/projects/nek5/docs/html/navier5_8f_source.html ) 2. interp_v3 ( https://lists.mcs.anl.gov/pipermail/nek5000-users/2011-July/001409.html) Regards, Jian -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Feb 14 11:30:15 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 14 Feb 2018 17:30:15 +0000 Subject: [Nek5000-users] subroutines of gradm1 and interp_v3 Message-ID: Hi Jian, gradm1 is a routine in navier5.f that you should be able to use on any mesh to compute the gradient of any field. interp_v is an interpolation routine based on fgslib_findpts. In v17, a wrapper has been made available to use these routines. See intp_usr.f in the repo for these wrappers, and the hemi example in the Nek example repo for its usage. Ketan From nek5000-users at lists.mcs.anl.gov Wed Feb 14 14:02:26 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 14 Feb 2018 20:02:26 +0000 Subject: [Nek5000-users] turbChannel: Definition of CFL and write/print files Message-ID: Dear Nek users, I have a couple quick and preliminary questions. Could you please tell me where could I find the definition of CFL; more specifically, the file/subroutine where CFL is related to the time step, etc. The other question is that which file shall I go to add/modify lines to write out some field information in a file (like .dat files or so) or print on sreen? Thank you a lot for the help! Regards, Emily -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Feb 14 14:14:46 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 14 Feb 2018 20:14:46 +0000 Subject: [Nek5000-users] turbChannel: Definition of CFL and write/print files In-Reply-To: References: Message-ID: Hi Emily, You can write out field data at specific points within your domain by using the history points feature. Simply set up a file *casename*.his in the format specified in the user guide. Then in usrchk() in the .usr file include call hpts This will write the field data at the specified points at every timestep. Make sure that the max hpts parameter in your SIZE file is greater than the number of points you have specified in your .his file. The data will be written into the .his file as well. On Wed, Feb 14, 2018, 3:02 PM wrote: > Dear Nek users, > > > > I have a couple quick and preliminary questions. Could you please tell me > where could I find the definition of CFL; more specifically, the > file/subroutine where CFL is related to the time step, etc. The other > question is that which file shall I go to add/modify lines to write out > some field information in a file (like .dat files or so) or print on sreen? > > > > Thank you a lot for the help! > > > > Regards, > > > > Emily > _______________________________________________ > 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 Multi-Physics Interaction Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Thu Feb 15 18:29:16 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 16 Feb 2018 00:29:16 +0000 Subject: [Nek5000-users] turbChannel: Definition of CFL and write/print files Message-ID: Hello, Thank you a lot for the information. I will give it a shot. Would anyone tell me where could I find the file/subroutine containing the correlation between CFL and time step please.? Regards, Emily -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Fri Feb 16 09:00:09 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 16 Feb 2018 15:00:09 +0000 Subject: [Nek5000-users] turbChannel: Definition of CFL and write/print files In-Reply-To: References: Message-ID: Hi Emily, We use either compute_cfl in induct.f --- I think this is more accurate --- or a routine in setdt in subs1.f Paul ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Thursday, February 15, 2018 6:29:16 PM To: nek5000-users at lists.mcs.anl.gov Subject: Re: [Nek5000-users] turbChannel: Definition of CFL and write/print files Hello, Thank you a lot for the information. I will give it a shot. Would anyone tell me where could I find the file/subroutine containing the correlation between CFL and time step please.? Regards, Emily -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Feb 19 14:41:21 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 19 Feb 2018 20:41:21 +0000 Subject: [Nek5000-users] turbChannel: history points and instantaneous velocity field Message-ID: Dear Nek users, I am attempting to use History Points features in turbChannel example and unfortunately I run into an issue. In the subroutine userchck, I added: if(istep. eq. 1) then call hpts() !checkpoints endif where the file is name "check.his" , the format of which follows the user guide: 2 !number of checkpoints 1. -0.98 1. !x,y,z locations 2. -0.67 4. However, I received an error saying: Cannot open history file in subroutine hpts(). Has anyone run into this error before? Another question really bothering me is that I could not figure out how to write out the instantaneous velocity field (u(x,y,z),v(x,y,z),w(x,y,z)) at a specific time step (istep). Could anyone please help me on this as well? Regards, Emily -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Feb 19 14:52:39 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 19 Feb 2018 20:52:39 +0000 Subject: [Nek5000-users] turbChannel: history points and instantaneous velocity field In-Reply-To: References: Message-ID: Hi Emily, Did you call the hpts() subroutine without calling the istep loop? Sincerely, On Mon, Feb 19, 2018 at 3:41 PM wrote: > Dear Nek users, > > > > I am attempting to use History Points features in turbChannel example and > unfortunately I run into an issue. In the subroutine userchck, I added: > > > > if(istep. eq. 1) then > > call hpts() !checkpoints > > endif > > > > where the file is name ?check.his? , the format of which follows the user > guide: > > 2 !number of checkpoints > > 1. -0.98 1. !x,y,z locations > > 2. -0.67 4. > > > > However, I received an error saying: Cannot open history file in > subroutine hpts(). Has anyone run into this error before? > > > > Another question really bothering me is that I could not figure out how to > write out the instantaneous velocity field (u(x,y,z),v(x,y,z),w(x,y,z)) at > a specific time step (istep). Could anyone please help me on this as well? > > > > Regards, > > > > Emily > _______________________________________________ > 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 Multi-Physics Interaction Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Feb 20 08:20:23 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 20 Feb 2018 14:20:23 +0000 Subject: [Nek5000-users] Small random perturbations Message-ID: Hello everyone, I am just wondering about the validity of my way to initiate turbulence. My question is that instead of introducing random disturbances to the initial velocity, can I introduce them to the pressure gradient? Hint, I am using Nek5000 to simulate statistically non-stationary turbulent flow. Thank you in advance Ali Embry-Riddle Aeronautical University -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Feb 20 09:49:59 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 20 Feb 2018 16:49:59 +0100 Subject: [Nek5000-users] Small random perturbations In-Reply-To: References: Message-ID: what geometry are you interested in? Philipp On 2018-02-20 15:20, nek5000-users at lists.mcs.anl.gov wrote: > > Hello everyone, > > ? ?I am just wondering about the validity of my way to initiate > turbulence. My question is that instead of introducing random > disturbances to the initial velocity, can I introduce them to the > pressure gradient? Hint, I am using Nek5000 to simulate statistically > non-stationary turbulent flow. > > Thank you in advance > Ali > Embry-Riddle Aeronautical University > > > > > > > _______________________________________________ > 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 Feb 20 09:54:33 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 20 Feb 2018 15:54:33 +0000 Subject: [Nek5000-users] turbChannel: history points and instantaneous velocity field Message-ID: Hello, Thank you for the reply. The istep loop shall be fine. I tried some other lines in the "if..endif" loop here and they all worked. I also commented the if loop and I had the same error information. I really do not know why. Regards, Emily _______________________________________________________________________ Hi Emily, Did you call the hpts() subroutine without calling the istep loop? Sincerely, On Mon, Feb 19, 2018 at 3:41 PM > wrote: > Dear Nek users, > > > > I am attempting to use History Points features in turbChannel example and > unfortunately I run into an issue. In the subroutine userchck, I added: > > > > if(istep. eq. 1) then > > call hpts() !checkpoints > > endif > > > > where the file is name "check.his" , the format of which follows the user > guide: > > 2 !number of checkpoints > > 1. -0.98 1. !x,y,z locations > > 2. -0.67 4. > > > > However, I received an error saying: Cannot open history file in > subroutine hpts(). Has anyone run into this error before? > > > > Another question really bothering me is that I could not figure out how to > write out the instantaneous velocity field (u(x,y,z),v(x,y,z),w(x,y,z)) at > a specific time step (istep). Could anyone please help me on this as well? > > > > Regards, > > > > Emily -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Feb 20 11:04:37 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 20 Feb 2018 17:04:37 +0000 Subject: [Nek5000-users] turbChannel: history points and instantaneous velocity field In-Reply-To: References: Message-ID: Hi Emily, Could you send me your case off-list so that I could have a look? Sincerely, On Tue, Feb 20, 2018 at 10:55 AM wrote: > Hello, > > > > Thank you for the reply. The istep loop shall be fine. I tried some other lines in the ?if..endif? loop here and they all worked. > > I also commented the if loop and I had the same error information. I really do not know why. > > > > Regards, > > > > Emily > > > > _______________________________________________________________________ > > > > Hi Emily, > > > > Did you call the hpts() subroutine without calling the istep loop? > > > > Sincerely, > > > > On Mon, Feb 19, 2018 at 3:41 PM > wrote: > > > > >* Dear Nek users,* > > > > > > > > > > > >* I am attempting to use History Points features in turbChannel example and* > > >* unfortunately I run into an issue. In the subroutine userchck, I added:* > > > > > > > > > > > >* if(istep. eq. 1) then* > > > > > >* call hpts() !checkpoints* > > > > > >* endif* > > > > > > > > > > > >* where the file is name ?check.his? , the format of which follows the user* > > >* guide:* > > > > > >* 2 !number of checkpoints* > > > > > >* 1. -0.98 1. !x,y,z locations* > > > > > >* 2. -0.67 4.* > > > > > > > > > > > >* However, I received an error saying: Cannot open history file in* > > >* subroutine hpts(). Has anyone run into this error before?* > > > > > > > > > > > >* Another question really bothering me is that I could not figure out how to* > > >* write out the instantaneous velocity field (u(x,y,z),v(x,y,z),w(x,y,z)) at* > > >* a specific time step (istep). Could anyone please help me on this as well?* > > > > > > > > > > > >* Regards,* > > > > > > > > > > > >* Emily* > > > _______________________________________________ > 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 Multi-Physics Interaction Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Feb 20 12:13:27 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 20 Feb 2018 19:13:27 +0100 Subject: [Nek5000-users] turbChannel: history points and instantaneous velocity field In-Reply-To: References: Message-ID: Make sure you file is called .his On 20 Feb 2018, at 16:55, "nek5000-users at lists.mcs.anl.gov " > wrote: Hello, ? Thank you for the reply. The istep loop shall be fine. I tried some other lines in the ?if..endif? loop here and they all worked. I also commented the if loop and I had the same error information. I really do not know why. ? Regards, ? Emily ? _______________________________________________________________________ ? Hi Emily, ? Did you call the hpts() subroutine without calling the istep loop? ? Sincerely, ? On Mon, Feb 19, 2018 at 3:41 PM wrote: ? > Dear Nek users, >? >? >? > I am attempting to use History Points features in turbChannel example and > unfortunately I run into an issue. In the subroutine userchck, I added: >? >? >? > if(istep. eq. 1) then >? >?????????????? call hpts()?????????????????????? !checkpoints >? > endif >? >? >? > where the file is name ?check.his? , the format of which follows the user > guide: >? > 2???????????? !number of checkpoints >? > 1. -0.98 1.?? !x,y,z locations >? > 2. -0.67 4. >? >? >? > However, I received an error saying: Cannot open history file in > subroutine hpts(). Has anyone run into this error before? >? >? >? > Another question really bothering me is that I could not figure out how to > write out the instantaneous velocity field (u(x,y,z),v(x,y,z),w(x,y,z)) at > a specific time step (istep). Could anyone please help me on this as well? >? >? >? > Regards, >? >? >? > Emily ? _______________________________________________ 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 Feb 20 13:16:35 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 20 Feb 2018 19:16:35 +0000 Subject: [Nek5000-users] turbChannel: history points and instantaneous velocity field Message-ID: Hello, Could you please be more specific about the "case off-list"? I did not really get it. The file I used for the hpts() feature is named as "check.his". Regards, Emily ____________________________________________________ Make sure you file is called .his ____________________________________________________ Hi Emily, Could you send me your case off-list so that I could have a look? Sincerely, -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Feb 20 13:20:57 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 20 Feb 2018 20:20:57 +0100 Subject: [Nek5000-users] turbChannel: history points and instantaneous velocity field In-Reply-To: References: Message-ID: Is you case named check? If not it will fail.? On 20 Feb 2018, at 20:18, "nek5000-users at lists.mcs.anl.gov " > wrote: Hello, ? Could you please be more specific about the ?case off-list?? I did not really get it. ? The file I used for the hpts() feature is named as ?check.his?. ? Regards, ? Emily ? ? ____________________________________________________ ? Make sure you file is called .his ? ____________________________________________________ ? Hi Emily, ? Could you send me your case off-list so that I could have a look? ? Sincerely, ? _______________________________________________ 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 Feb 20 14:39:15 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 20 Feb 2018 20:39:15 +0000 Subject: [Nek5000-users] turbChannel: history points and instantaneous velocity field Message-ID: Hello, Yes, the file I used for hpts() has a name of "check.his". However an error occurs saying: Cannot open history file in subroutine hpts(). Thanks, Emily Is you case named check? If not it will fail. -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Feb 20 14:43:35 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 20 Feb 2018 20:43:35 +0000 Subject: [Nek5000-users] turbChannel: history points and instantaneous velocity field In-Reply-To: References: Message-ID: What is the name of your .usr and .rea file. Is it "check" too? On Tue, Feb 20, 2018 at 3:39 PM wrote: > Hello, > > > > Yes, the file I used for hpts() has a name of ?check.his?. However an > error occurs saying: Cannot open history file in subroutine hpts(). > > > > Thanks, > > > > Emily > > > > > > Is you case named check? If not it will fail. > > > _______________________________________________ > 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 Multi-Physics Interaction Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Feb 20 14:43:57 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 20 Feb 2018 21:43:57 +0100 Subject: [Nek5000-users] turbChannel: history points and instantaneous velocity field In-Reply-To: References: Message-ID: Emily, what your case name i.e. what the name of your rea (or par) file? On 20 Feb 2018, at 21:39, "nek5000-users at lists.mcs.anl.gov " > wrote: Hello, ? Yes, the file I used for hpts() has a name of ?check.his?. However an error occurs saying: Cannot open history file in subroutine hpts(). ? Thanks, ? Emily ? ? Is you case named check? If not it will fail.? ? _______________________________________________ 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 Feb 20 15:29:43 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 20 Feb 2018 21:29:43 +0000 Subject: [Nek5000-users] turbChannel: history points and instantaneous velocity field Message-ID: Hello, Thank you for the help. I was not aware of that the names had to be consistent with one another. Now it works after I changed the name. Thank you again. Emily Emily, what your case name i.e. what the name of your rea (or par) file? -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Feb 20 20:50:49 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 21 Feb 2018 09:50:49 +0700 Subject: [Nek5000-users] SYM bc Message-ID: Hi, Neks. I try to simulate the flow inside the channel between two parallel wall in 2D case. First run: box between two parallel walls which are orthogonal to y-axis with the vx being parabolic profile. Secong run: a half of the box where one wall are orthogonal with y-axis and 'SYM' bc in middle of the box, with the vx being parabolic profile. In the first case the vx profile is equal to exact solution with the amplitude in the middle of the box being equal to 1.5. So in the second case the vx profile also has the parobilic profile, but the amplitude on the 'SYM' boundary equals to 1.497 intead of 1.5. Is it some feature of bc ? And how I can fixed this ? -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Feb 20 22:40:36 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 21 Feb 2018 04:40:36 +0000 Subject: [Nek5000-users] Instantaneous velocity field Message-ID: Dear Nek users, Could anyone please teach me how to generate an instantaneous velocity field for the whole domain (let's say for just one time step and have it stored in a .dat file) for case of turbChannel. I have been tracking/reading many subroutines and files, but I could not find the location where I could add lines to generate the instantaneous velocity profile. Thank you. Tim -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Feb 21 04:55:40 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 21 Feb 2018 11:55:40 +0100 Subject: [Nek5000-users] Instantaneous velocity field In-Reply-To: References: Message-ID: The subroutine useric (in the usr file) is used to specify the initial condition for the whole domain. On 21 Feb 2018, at 05:41, "nek5000-users at lists.mcs.anl.gov " > wrote: Dear Nek users, ? Could anyone please teach me how to generate an instantaneous velocity field for the whole domain (let?s say for just one time step and have it stored in a .dat file) for case of turbChannel. I have been tracking/reading many subroutines and files, but I could not find the location where I could add lines to generate the instantaneous velocity profile. ? Thank you. ? Tim _______________________________________________ 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 Feb 21 05:35:40 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 21 Feb 2018 12:35:40 +0100 Subject: [Nek5000-users] Instantaneous velocity field In-Reply-To: References: Message-ID: do you mean how to write out an instantaneous field, or to generate an initial condition? On 2018-02-21 11:55, nek5000-users at lists.mcs.anl.gov wrote: > The subroutine useric (in the usr file) is used to specify the initial > condition for the whole domain. > > On 21 Feb 2018, at 05:41, "nek5000-users at lists.mcs.anl.gov > " > > wrote: > >> Dear Nek users, >> >> Could anyone please teach me how to generate an instantaneous velocity >> field for the whole domain (let?s say for just one time step and have >> it stored in a .dat file) for case of turbChannel. I have been >> tracking/reading many subroutines and files, but I could not find the >> location where I could add lines to generate the instantaneous >> velocity profile. >> >> Thank you. >> >> Tim >> >> _______________________________________________ >> >> Nek5000-users mailing list >> >> Nek5000-users at lists.mcs.anl.gov >> >> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users >> > > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > From nek5000-users at lists.mcs.anl.gov Wed Feb 21 09:22:25 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 21 Feb 2018 15:22:25 +0000 Subject: [Nek5000-users] Instantaneous velocity field Message-ID: Hello, Thank you for the reply. I meant the instantaneous velocity field for the domain (u(x,y,z),v(x,y,z),w(x,y,z)) at some certain time step (istep). Regards, Tim ________________________________ do you mean how to write out an instantaneous field, or to generate an initial condition? On 2018-02-21 11:55, nek5000-users at lists.mcs.anl.gov wrote: > The subroutine useric (in the usr file) is used to specify the initial > condition for the whole domain. > > On 21 Feb 2018, at 05:41, "nek5000-users at lists.mcs.anl.gov > >" > > >> wrote: > >> Dear Nek users, >> >> Could anyone please teach me how to generate an instantaneous velocity >> field for the whole domain (let's say for just one time step and have >> it stored in a .dat file) for case of turbChannel. I have been >> tracking/reading many subroutines and files, but I could not find the >> location where I could add lines to generate the instantaneous >> velocity profile. >> >> Thank you. >> >> Tim -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Feb 21 09:33:16 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 21 Feb 2018 16:33:16 +0100 Subject: [Nek5000-users] Instantaneous velocity field In-Reply-To: References: Message-ID: Doesn't outpost() (userchk subroutine) inside a conditional statement address your issue? Something like (if you have to flow information at istep=10): if (istep .eq. 10) then outpost(vx,vy,vz,pr,t,'vel') endif On 02/21/2018 04:22 PM, nek5000-users at lists.mcs.anl.gov wrote: > > Hello, > > Thank you for the reply. I meant the instantaneous velocity field for > the domain (u(x,y,z),v(x,y,z),w(x,y,z)) at some certain time step (istep). > > Regards, > > Tim > > ------------------------------------------------------------------------ > > do you mean how to write out an instantaneous field, or to generate an > initial condition? > On 2018-02-21 11:55, nek5000-users at lists.mcs.anl.gov > wrote: > >/The subroutine useric (in the usr file) is used to specify the initial / > >/condition for the whole domain./ > >// > >/On 21 Feb 2018, at 05:41, "nek5000-users at lists.mcs.anl.gov > / > >/ >" / > >/ / > >/ >> wrote:/ > >// > >>/Dear Nek users,/ > >>// > >>/Could anyone please teach me how to generate an instantaneous velocity / > >>/field for the whole domain (let?s say for just one time step and have / > >>/it stored in a .dat file) for case of turbChannel. I have been / > >>/tracking/reading many subroutines and files, but I could not find the / > >>/location where I could add lines to generate the instantaneous / > >>/velocity profile./ > >>// > >>/Thank you./ > >>// > >>/Tim/ > > > > _______________________________________________ > 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 Feb 21 09:34:46 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 21 Feb 2018 16:34:46 +0100 Subject: [Nek5000-users] Instantaneous velocity field In-Reply-To: <6ebc924d-3255-5673-f50c-f3013582599a@onera.fr> References: <6ebc924d-3255-5673-f50c-f3013582599a@onera.fr> Message-ID: I forgot to add 'call' before outpost. Please read the block as: if (istep .eq. 10) then call outpost(vx,vy,vz,pr,t,'vel') endif On 02/21/2018 04:33 PM, Navrose Navrose wrote: > > Doesn't outpost() (userchk subroutine) inside a conditional statement > address your issue? > Something like (if you have to flow information at istep=10): > > if (istep .eq. 10) then > outpost(vx,vy,vz,pr,t,'vel') > endif > > > On 02/21/2018 04:22 PM, nek5000-users at lists.mcs.anl.gov wrote: >> >> Hello, >> >> Thank you for the reply. I meant the instantaneous velocity field for >> the domain (u(x,y,z),v(x,y,z),w(x,y,z)) at some certain time step >> (istep). >> >> Regards, >> >> Tim >> >> ------------------------------------------------------------------------ >> >> do you mean how to write out an instantaneous field, or to generate an >> initial condition? >> On 2018-02-21 11:55, nek5000-users at lists.mcs.anl.gov >> wrote: >> >/The subroutine useric (in the usr file) is used to specify the initial / >> >/condition for the whole domain./ >> >// >> >/On 21 Feb 2018, at 05:41, "nek5000-users at lists.mcs.anl.gov >> / >> >/> >" / >> >/> / >> >/> >> wrote:/ >> >// >> >>/Dear Nek users,/ >> >>// >> >>/Could anyone please teach me how to generate an instantaneous velocity / >> >>/field for the whole domain (let?s say for just one time step and have / >> >>/it stored in a .dat file) for case of turbChannel. I have been / >> >>/tracking/reading many subroutines and files, but I could not find the / >> >>/location where I could add lines to generate the instantaneous / >> >>/velocity profile./ >> >>// >> >>/Thank you./ >> >>// >> >>/Tim/ >> >> >> >> _______________________________________________ >> 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 Feb 21 12:34:53 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 21 Feb 2018 18:34:53 +0000 Subject: [Nek5000-users] Instantaneous velocity field Message-ID: Hello, Thank you for the reply. Yes, outpost() in fact does yield the information I am looking for. But I hope I could generate the field information in .dat file similar to following pattern: open(unit=56,file='vel_fluc_prof.dat') write(56,'(A,1pe14.7)') '#time = ', time write(56,'(A)') & '# y y+ uu vv ww uv' Could you please teach me how to do this? Regards, Tim ________________________________ I forgot to add 'call' before outpost. Please read the block as: if (istep .eq. 10) then call outpost(vx,vy,vz,pr,t,'vel') endif On 02/21/2018 04:33 PM, Navrose Navrose wrote: > > Doesn't outpost() (userchk subroutine) inside a conditional statement > address your issue? > Something like (if you have to flow information at istep=10): > > if (istep .eq. 10) then > outpost(vx,vy,vz,pr,t,'vel') > endif > > > On 02/21/2018 04:22 PM, nek5000-users at lists.mcs.anl.gov wrote: >> >> Hello, >> >> Thank you for the reply. I meant the instantaneous velocity field for >> the domain (u(x,y,z),v(x,y,z),w(x,y,z)) at some certain time step >> (istep). >> >> Regards, >> >> Tim -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Feb 21 14:43:02 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 21 Feb 2018 21:43:02 +0100 Subject: [Nek5000-users] Instantaneous velocity field In-Reply-To: References: Message-ID: this is not really instantaneous data, but rather statistics. Then you would need to run some kind of averaging (in time, space etc.) and proper postprocessing afterwards, which strongly depends on the actual flow case. I am sure that there are people in this group that have done this type of analysis for a variety of flow cases/geometries. On 2018-02-21 19:34, nek5000-users at lists.mcs.anl.gov wrote: > Hello, > > Thank you for the reply. Yes, outpost() in fact does yield the > information I am looking for. ?But I hope I could generate the field > information in .dat file similar to following pattern: > > ??????? ?open(unit=56,file='vel_fluc_prof.dat') > > ???????? write(56,'(A,1pe14.7)') '#time = ', time > > ???????? write(56,'(A)') > > ???? &?? '#? y???? y+???? uu???? vv???? ww???? uv' > > Could you please teach me how to do this? > > Regards, > > Tim > > ------------------------------------------------------------------------ > > I forgot to add 'call' before outpost. Please read the block as: > > ? if (istep .eq. 10) then > > ?????? call outpost(vx,vy,vz,pr,t,'vel') > > ?? endif > > On 02/21/2018 04:33 PM, Navrose Navrose wrote: > >>// > >>/Doesn't outpost() (userchk subroutine) inside a conditional statement / > >>/address your issue?/ > >>/Something like (if you have to flow information at istep=10):/ > >>// > >>/? if (istep .eq. 10) then/ > >>/?????? outpost(vx,vy,vz,pr,t,'vel')/ > >>/?? endif/ > >>// > >>// > >>/On 02/21/2018 04:22 PM, nek5000-users at lists.mcs.anl.gov > wrote:/ > >>>// > >>>/Hello,/ > >>>// > >>>/Thank you for the reply. I meant the instantaneous velocity field for / > >>>/the domain (u(x,y,z),v(x,y,z),w(x,y,z)) at some certain time step / > >>>/(istep)./ > >>>// > >>>/Regards,/ > >>>// > >>>/Tim/ > > > > _______________________________________________ > 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 Feb 21 15:53:46 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 21 Feb 2018 21:53:46 +0000 Subject: [Nek5000-users] Instantaneous velocity field Message-ID: Hello, Thank you for the reply. I am aware of those lines are intended for time-averaging. I was wondering if anyone could teach me how to generate the instantaneous velocity field in a similar pattern like the following: open(unit=56,file='vel_fluc_prof.dat') write(56,'(A,1pe14.7)') '#time = ', time write(56,'(A)') & '# y y+ uu vv ww uv' Thanks a lot, Tim ________________________________ this is not really instantaneous data, but rather statistics. Then you would need to run some kind of averaging (in time, space etc.) and proper postprocessing afterwards, which strongly depends on the actual flow case. I am sure that there are people in this group that have done this type of analysis for a variety of flow cases/geometries. -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Feb 21 15:57:42 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 21 Feb 2018 21:57:42 +0000 Subject: [Nek5000-users] Instantaneous velocity field In-Reply-To: References: Message-ID: One of the ways to go about this would be to use the history points option. You would get your instantaneous velocity and pressure ields at certain timesteps written into a history file On Wed, Feb 21, 2018, 4:53 PM wrote: > Hello, > > > > Thank you for the reply. I am aware of those lines are intended for > time-averaging. I was wondering if anyone could teach me how to generate > the instantaneous velocity field > > in a similar pattern like the following: > > > > open(unit=56,file='vel_fluc_prof.dat') > > write(56,'(A,1pe14.7)') '#time = ', time > > write(56,'(A)') > > & '# y y+ uu vv ww uv' > > Thanks a lot, > > > > Tim > ------------------------------ > > > > this is not really instantaneous data, but rather statistics. Then you > > would need to run some kind of averaging (in time, space etc.) and > > proper postprocessing afterwards, which strongly depends on the actual > > flow case. I am sure that there are people in this group that have done > > this type of analysis for a variety of flow cases/geometries. > > > _______________________________________________ > 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 Multi-Physics Interaction Lab University of Waterloo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Fri Feb 23 11:29:43 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 23 Feb 2018 17:29:43 +0000 Subject: [Nek5000-users] Tripping mechanism Message-ID: Dear all, I've seen several papers using Nek5000 with a tripping mechanism based on a forcing term, explained in a paper by Schlatter & ?rl? (2012). However, they talk about using random harmonic signals h^{i}(z). My question is: What is the shape of such signals and how do you make them random? Thanks, Juan Diego -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Fri Feb 23 12:01:19 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 23 Feb 2018 18:01:19 +0000 Subject: [Nek5000-users] Help In-Reply-To: References: Message-ID: > On Feb 23, 2018, at 12:00 PM, "nek5000-users-request at lists.mcs.anl.gov" 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://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.mcs.anl.gov%2Fmailman%2Flistinfo%2Fnek5000-users&data=02%7C01%7Cvirgilio.a.gomez%40ttu.edu%7C86cccb39332f4a8f50c208d57ae7485d%7C178a51bf8b2049ffb65556245d5c173c%7C0%7C0%7C636550056137979126&sdata=kkTVB7y0TUaz9AOebv2GLx9Nk%2BCyS4G%2BjeT%2Bo4HW8dA%3D&reserved=0 > 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. Tripping mechanism (nek5000-users at lists.mcs.anl.gov) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Fri, 23 Feb 2018 17:29:43 +0000 > From: nek5000-users at lists.mcs.anl.gov > To: "nek5000-users at lists.mcs.anl.gov" > > Subject: [Nek5000-users] Tripping mechanism > Message-ID: > > Content-Type: text/plain; charset="iso-8859-1" > > Dear all, > > > I've seen several papers using Nek5000 with a tripping mechanism based on a forcing term, explained in a paper by Schlatter & ?rl? (2012). However, they talk about using random harmonic signals h^{i}(z). My question is: What is the shape of such signals and how do you make them random? > > > Thanks, > > > Juan Diego > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: > > ------------------------------ > > Subject: Digest Footer > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.mcs.anl.gov%2Fmailman%2Flistinfo%2Fnek5000-users&data=02%7C01%7Cvirgilio.a.gomez%40ttu.edu%7C86cccb39332f4a8f50c208d57ae7485d%7C178a51bf8b2049ffb65556245d5c173c%7C0%7C0%7C636550056137979126&sdata=kkTVB7y0TUaz9AOebv2GLx9Nk%2BCyS4G%2BjeT%2Bo4HW8dA%3D&reserved=0 > > > ------------------------------ > > End of Nek5000-users Digest, Vol 108, Issue 30 > ********************************************** From nek5000-users at lists.mcs.anl.gov Fri Feb 23 15:49:37 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Fri, 23 Feb 2018 22:49:37 +0100 Subject: [Nek5000-users] Tripping mechanism In-Reply-To: References: Message-ID: Hi Juan Diego, Indeed, we describe in our paper the trip forcing. The tripping has a temporal part (third-order Lagrange interpolants with specific time scale) and a spatial part (the h^i(z)). We typically use white noise for these h(z), low-pass filtered with a spectral cutoff filter at a wavenumber 2pi/z_s. So what one gets in h(z) is just random sines and cosines up to a spanwise scale z_s. The implementation of course needs to make sure i) that you have the same random numbers for the whole spanwise extent, and ii) that when restarting a simulation you continue the previous tripping in a smooth way as to not generate any peaks (discontinuities) at the restart. Please let me know if you have further questions, and I am happy to help more. Best regards, Philipp Schlatter On 2018-02-23 18:29, nek5000-users at lists.mcs.anl.gov wrote: > Dear all, > > > I've seen several papers using Nek5000 with a tripping mechanism based > on a forcing term, explained in a paper by Schlatter & ?rl? (2012). > However, they talk about using random harmonic signals h^{i}(z). My > question is: What is the shape of such signals and how do you make them > random? > > > Thanks, > > > Juan Diego > > > > _______________________________________________ > 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 Feb 24 01:53:01 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sat, 24 Feb 2018 15:53:01 +0800 Subject: [Nek5000-users] Stability utilities In-Reply-To: References: Message-ID: An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Sat Feb 24 13:20:35 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sat, 24 Feb 2018 19:20:35 +0000 Subject: [Nek5000-users] Blasius example Message-ID: Dear Nek users, I am new to the Nek5000 and have a few quick questions. I noticed that in the SIZE file of Blasius, the number of specified elements does not really match the actual number in the .rea file. I tried to re-generate the .rea file, but I do not have .box file or any other file I could use to modify the specifications in the Blasius folder. So could anyone tell me where could I find the corresponding .box file for this case? And there is a .sep file in the folder, is anyone aware of what is it for? Regards, Ming -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Sun Feb 25 09:06:33 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sun, 25 Feb 2018 15:06:33 +0000 Subject: [Nek5000-users] Further to the issue of ffx, ffy and ffz Message-ID: Hi, May I push the following question further? Is there anyone know what are the exact meaning for para(54) and (55) in rea file? Such as, what's the difference between: 0,0 and -3 1 for (54) and (55) respectively? what's meaning for sign (+) or (-)? Kind regards, Zhai Hi Philipp, Yes, they are shear velocity and bulk velocity. Can I say if the ffx, ffy or ffz equal to (u_t/u_b)^2, the flow is specified as a fixed pressure gradient in the channel, and parameter(54) and para(55) in rea file are fixed as 0. Otherwise, if the all ffx, ffy and ffz are 0, the flow is fixed at flow rate rather than pressure gradient? Cheers, Zhai -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Sun Feb 25 12:48:25 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sun, 25 Feb 2018 19:48:25 +0100 Subject: [Nek5000-users] Further to the issue of ffx, ffy and ffz In-Reply-To: References: Message-ID: Hi, The documentation says: P054: fixed flow rate dir: |p54|=1,2,3=x,y,z P055: vol.flow rate (p54>0) or Ubar (p54<0) So for instance if you want to run a pipe with bulk flow of unity, you set: p54: -3 (z) p55: 1 How this is done in the latest versions I do not know, as we do not use those for pipe production runs. Philipp On 2018-02-25 16:06, nek5000-users at lists.mcs.anl.gov wrote: > Hi, > > May I push the following question further? Is there anyone know what are > the exact meaning for para(54) and (55) in rea file? Such as, what's the > difference between: 0,0 and -3 1 for (54) and (55) respectively? what's > meaning for sign (+) or (-)? > > Kind regards, > > Zhai > > > Hi Philipp, > > Yes, they are shear velocity and bulk velocity. Can I say if the ffx, ffy > or ffz equal to (u_t/u_b)^2, the flow is specified as a fixed pressure > gradient in the channel, and parameter(54) and para(55) in rea file are > fixed as 0. Otherwise, if the all ffx, ffy and ffz are 0, the flow is fixed > at flow rate rather than pressure gradient? > > Cheers, > > Zhai > > > > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > From nek5000-users at lists.mcs.anl.gov Sun Feb 25 15:55:55 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sun, 25 Feb 2018 21:55:55 +0000 Subject: [Nek5000-users] turbChannel: Polynomial order Message-ID: Dear Nek users, I am aware of that the mesh resolution is related to the polynomial order. In the turbChannel, P-order is 8, which makes 8 grid for each element in the domain. I am wondering could I change (decrease or increase) this order just for one direction? Let's say in the spanwise direction (z), I only need P-order to be 6 while the other two directions to be 8. Could I do this? Thanks for your help! Regards, Emily -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Sun Feb 25 15:58:56 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Sun, 25 Feb 2018 22:58:56 +0100 Subject: [Nek5000-users] turbChannel: Polynomial order In-Reply-To: References: Message-ID: no, for the time being the order is the same in all directions. But of course you can put more elements in a certain directions, and refine in that way. Philipp On 2018-02-25 22:55, nek5000-users at lists.mcs.anl.gov wrote: > Dear Nek users, > > I am aware of that the mesh resolution is related to the polynomial > order. In the turbChannel, P-order is 8, which makes 8 grid for each > element in the domain. I am wondering could I change (decrease or > increase) this order just for one direction? Let?s say in the spanwise > direction (z), I only need P-order to be 6 while the other two > directions to be 8. Could I do this? > > Thanks for your help! > > Regards, > > Emily > > > > _______________________________________________ > Nek5000-users mailing list > Nek5000-users at lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users > From nek5000-users at lists.mcs.anl.gov Sun Feb 25 21:58:01 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 26 Feb 2018 10:58:01 +0700 Subject: [Nek5000-users] Dissipation rate Message-ID: Hi, Nek users. I am trying to calculate the dissaipation rate now. In my subroutine which computes the derivatives I use following procedure for smoothing: call col2 ( dui_dxj, bm1, ntot ) call dssum( dui_dxj, lx1, ly1, lz1 ) call col2 ( dui_dxj, binvm1, ntot ) and there are some unrealistic values on the surface between spectral elements. I set up also additional filtering of high modes of velocity with various transfer functions, but it did not help. Do you have any idea about the causes and how to fixed this ? Regards, Vatslav -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Feb 26 07:34:14 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 26 Feb 2018 14:34:14 +0100 (CET) Subject: [Nek5000-users] Mismatch of X,Y cordinates Message-ID: Hi Neks, I was trying to make sure the the moving wall condition has properly applied on .rea file by writing the x and y co -ordinates through the .usr file during the run time to a file called rectangle.dat (element number,face number,x,y respectively). But it looks bit different from what i have i given in .rea file. Esepacially if you check the output data file called rectangle.dat on line 5, the x and y co ordinates are 0.9375,0.9375 corresponding to the 133rd element and 2nd phase. But if we check the .rea file its not showing the right one. I have attached all the necessary files to run this case and could you please tell me what went wrong? Thanks in advance -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: rectangle.dat Type: application/octet-stream Size: 809 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: rectangle.usr Type: application/octet-stream Size: 11464 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: SESSION.NAME Type: application/octet-stream Size: 64 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: rectangle.rea Type: application/octet-stream Size: 131430 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: rectangle.map Type: application/octet-stream Size: 15456 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: SIZE Type: application/octet-stream Size: 1993 bytes Desc: not available URL: From nek5000-users at lists.mcs.anl.gov Mon Feb 26 07:38:31 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 26 Feb 2018 14:38:31 +0100 Subject: [Nek5000-users] Stability utilities In-Reply-To: References: Message-ID: An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Feb 26 09:59:47 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 26 Feb 2018 15:59:47 +0000 Subject: [Nek5000-users] Further to the issue of ffx, ffy and ffz In-Reply-To: References: , Message-ID: For a geometry with periodic boundary conditions in the flow direction (e.g., pipe flow with periodic bcs), p54 and p55 control the mass flow rate through the domain. p54 determines the flow direction to be controlled (which, right now, can only be x, y, or z; but this restriction is readily lifted if there was an application requiring such). So if |p54| = 1, 2, or 3, the flow rate will be fixed in the x, y, or z direction. If p54 < 0, then it means that p55 will be the mean velocity in the given direction. If p54 > 0, then it means that p55 will be the volumetric flow rate: Q= \int u dA. The flow rate is fixed according to the following algorithm. The Navier-Stokes equations are advanced in a semi-implicit way, using the following form: S(U,p) = NL(u) + f where NL includes all data that should be on the rhs of the equations and S is a linear Stokes problem to ensure div.U=0, and f is the forcing that corresponds to the mean pressure gradient. Because it is linear, the Stokes problem can be split into two parts: S(U',p') = NL(u) S(U0,p0) = 1 Then at every step, set U=U' + alpha U0 p=p' + alpha p0 such that \int U.nhat dA = desired flow rate, where nhat is the unit vector in the x,y, or z direction. hth, Paul ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Sunday, February 25, 2018 12:48:25 PM To: nek5000-users at lists.mcs.anl.gov Subject: Re: [Nek5000-users] Further to the issue of ffx, ffy and ffz Hi, The documentation says: P054: fixed flow rate dir: |p54|=1,2,3=x,y,z P055: vol.flow rate (p54>0) or Ubar (p54<0) So for instance if you want to run a pipe with bulk flow of unity, you set: p54: -3 (z) p55: 1 How this is done in the latest versions I do not know, as we do not use those for pipe production runs. Philipp On 2018-02-25 16:06, nek5000-users at lists.mcs.anl.gov wrote: > Hi, > > May I push the following question further? Is there anyone know what are > the exact meaning for para(54) and (55) in rea file? Such as, what's the > difference between: 0,0 and -3 1 for (54) and (55) respectively? what's > meaning for sign (+) or (-)? > > Kind regards, > > Zhai > > > Hi Philipp, > > Yes, they are shear velocity and bulk velocity. Can I say if the ffx, ffy > or ffz equal to (u_t/u_b)^2, the flow is specified as a fixed pressure > gradient in the channel, and parameter(54) and para(55) in rea file are > fixed as 0. Otherwise, if the all ffx, ffy and ffz are 0, the flow is fixed > at flow rate rather than pressure gradient? > > Cheers, > > Zhai > > > > > _______________________________________________ > 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 Feb 26 10:20:43 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 26 Feb 2018 16:20:43 +0000 Subject: [Nek5000-users] Tripping mechanism Message-ID: Dear Philipp, Thank you for your response. So if I understand correctly, first you generate a random signal as a function of Z and then use a low-pass filter to obtain h^i(z). Thinking about how to implement this in Nek5000, I guess one could impose random numbers at each node (collocation point) and then apply some kind of filter. Is this correct? Also, could you please clarify what you mean by: make sure that you have the same random numbers for the whole spanwise extent? Since h(z) is a function of z, I'm not sure I understand this statement. Thank you very much for you help. Juan Diego Colmenares PhD Candidate Department of Mechanical Engineering University of New Mexico -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Feb 26 11:15:05 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 26 Feb 2018 17:15:05 +0000 Subject: [Nek5000-users] Further to the issue of ffx, ffy and ffz In-Reply-To: References: , , Message-ID: If the question is how to use .par file with the fixed mass flow rate I would do it with one or two userparam there and then prescribing p54 & p55 in say userchk through the corresponding uparam as I recently tried for an avg_all call Aleks ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Monday, February 26, 2018 9:59:47 AM To: nek5000-users at lists.mcs.anl.gov Subject: Re: [Nek5000-users] Further to the issue of ffx, ffy and ffz For a geometry with periodic boundary conditions in the flow direction (e.g., pipe flow with periodic bcs), p54 and p55 control the mass flow rate through the domain. p54 determines the flow direction to be controlled (which, right now, can only be x, y, or z; but this restriction is readily lifted if there was an application requiring such). So if |p54| = 1, 2, or 3, the flow rate will be fixed in the x, y, or z direction. If p54 < 0, then it means that p55 will be the mean velocity in the given direction. If p54 > 0, then it means that p55 will be the volumetric flow rate: Q= \int u dA. The flow rate is fixed according to the following algorithm. The Navier-Stokes equations are advanced in a semi-implicit way, using the following form: S(U,p) = NL(u) + f where NL includes all data that should be on the rhs of the equations and S is a linear Stokes problem to ensure div.U=0, and f is the forcing that corresponds to the mean pressure gradient. Because it is linear, the Stokes problem can be split into two parts: S(U',p') = NL(u) S(U0,p0) = 1 Then at every step, set U=U' + alpha U0 p=p' + alpha p0 such that \int U.nhat dA = desired flow rate, where nhat is the unit vector in the x,y, or z direction. hth, Paul ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Sunday, February 25, 2018 12:48:25 PM To: nek5000-users at lists.mcs.anl.gov Subject: Re: [Nek5000-users] Further to the issue of ffx, ffy and ffz Hi, The documentation says: P054: fixed flow rate dir: |p54|=1,2,3=x,y,z P055: vol.flow rate (p54>0) or Ubar (p54<0) So for instance if you want to run a pipe with bulk flow of unity, you set: p54: -3 (z) p55: 1 How this is done in the latest versions I do not know, as we do not use those for pipe production runs. Philipp On 2018-02-25 16:06, nek5000-users at lists.mcs.anl.gov wrote: > Hi, > > May I push the following question further? Is there anyone know what are > the exact meaning for para(54) and (55) in rea file? Such as, what's the > difference between: 0,0 and -3 1 for (54) and (55) respectively? what's > meaning for sign (+) or (-)? > > Kind regards, > > Zhai > > > Hi Philipp, > > Yes, they are shear velocity and bulk velocity. Can I say if the ffx, ffy > or ffz equal to (u_t/u_b)^2, the flow is specified as a fixed pressure > gradient in the channel, and parameter(54) and para(55) in rea file are > fixed as 0. Otherwise, if the all ffx, ffy and ffz are 0, the flow is fixed > at flow rate rather than pressure gradient? > > Cheers, > > Zhai > > > > > _______________________________________________ > 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 Feb 26 11:59:19 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 26 Feb 2018 18:59:19 +0100 (CET) Subject: [Nek5000-users] Volume of an element Message-ID: Hi Neks, How to calculate volume of an element in a 3D mesh? If i write n = nx1*ny1*nz1*nelv do i=1,n v(i) = volume(1,1,1,i) enddo Can i store the volume of each element into a varibale called v? What happens if i use bm1(1,1,1,i)? Instead of volume(1,1,1,i) Thanks Sijo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Mon Feb 26 11:39:35 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 26 Feb 2018 18:39:35 +0100 Subject: [Nek5000-users] Tripping mechanism In-Reply-To: References: Message-ID: Hi Juan Diego, yes, first you create a random signal in Z with the appropriate spatial characteristics (in our case a low-pass filtered signal, but this could potentially be created in any other way). So the way you suggest should be fine, however, you just need to make sure that you signal does not depend on the mesh, i.e. give smaller scales close to element boundaries for instance. Regarding the second point. Essentially you just want to be able to restart your simulation. If you use the Lagrange-interpolants p as described in our paper, you would need to be able to re-create the same two h(z) you had when you stopped the simulation. That's what I meant. Best regards, Philipp On 2018-02-26 17:20, nek5000-users at lists.mcs.anl.gov wrote: > Dear Philipp, > > > Thank you for your response. So if I understand correctly, first you > generate a random signal as a function of Z and then use a low-pass > filter to obtain h^i(z). Thinking about how to implement this in > Nek5000, I guess one could impose random numbers at each node > (collocation point) and then apply some kind of filter. Is this correct? > > > Also, could you please clarify what you mean by: make sure that you have > the same random numbers for the whole spanwise extent? Since h(z) is a > function of z, I'm not sure I understand this statement. > > > Thank you very much for you help. > > > Juan Diego Colmenares > > PhD Candidate > > Department of Mechanical Engineering > > University of New Mexico > > > > > _______________________________________________ > 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 Feb 26 14:46:02 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Mon, 26 Feb 2018 20:46:02 +0000 Subject: [Nek5000-users] Volume of an element In-Reply-To: References: Message-ID: Dear Sijo, We typically refer to elements as the structures that have (N+1)^3 points. (Here, N+1 = lx1.) They are enumerated in nek5000 by the element number, e: integer e nxyz = nx1*ny1*nz1 do e=1,nelt volume = volel(e) volume = vlsum(bm1(1,1,1,e),nxyz) enddo would both yield the same value for the volume of element e. It appears to me that you are after something different, however. It seems like you want to associate a volume with each grid point. If that is the case, then the entries of the diagonal mass matrix, bm1, are indeed the quantities that you want. Each entry of bm1(i,1,1,1) for i=1 to n:=nxyz*nelt, corresponds to the integral of the underlying basis function. That is, if we think about a function phi_i(x) such that u(x) = sum_i=1^n u(i,1,1,1)*phi_i(x) then bm1(i,1,1,1) = \int phi_i(x) dx where the integral (\int) is taken over the entire computational domain. [ Note that we really don't have such a global phi_i(x) ... I use it simply for illustration, here.] hth Paul ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Monday, February 26, 2018 11:59:19 AM To: nek5000-users Subject: [Nek5000-users] Volume of an element Hi Neks, How to calculate volume of an element in a 3D mesh? If i write n = nx1*ny1*nz1*nelv do i=1,n v(i) = volume(1,1,1,i) enddo Can i store the volume of each element into a varibale called v? What happens if i use bm1(1,1,1,i)? Instead of volume(1,1,1,i) Thanks Sijo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Feb 27 04:45:30 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 27 Feb 2018 11:45:30 +0100 (CET) Subject: [Nek5000-users] Volume of an element In-Reply-To: References: Message-ID: Thanks Paul for your Kind replay. You were right. I want to get the volume associated with each grid point. And I tried the code below, but at each time when I run I get different answers for volume so I think my ocde is not rubust. Could you please verify my code below. nxyz = nx1*ny1*nz1 n = nx1*ny1*nz1*nelv vol = vlsum(bm1(1,1,1,1),nxyz) a=glmax(vol,1) b=glmin(vol,1) do e=1,n volume = vlsum(bm1(e,1,1,1),nxyz) open (unit=10,file="hcoeff.dat") write (10,7020) a,b,volume 7020 format(7f20.14) enddo Regards Sijo GEORGE De: "nek5000-users" ?: "nek5000-users" Envoy?: Lundi 26 F?vrier 2018 21:46:02 Objet: Re: [Nek5000-users] Volume of an element Dear Sijo, We typically refer to elements as the structures that have (N+1)^3 points. (Here, N+1 = lx1.) They are enumerated in nek5000 by the element number, e: integer e nxyz = nx1*ny1*nz1 do e=1,nelt volume = volel(e) volume = vlsum(bm1(1,1,1,e),nxyz) enddo would both yield the same value for the volume of element e. It appears to me that you are after something different, however. It seems like you want to associate a volume with each grid point. If that is the case, then the entries of the diagonal mass matrix, bm1, are indeed the quantities that you want. Each entry of bm1(i,1,1,1) for i=1 to n:=nxyz*nelt, corresponds to the integral of the underlying basis function. That is, if we think about a function phi_i(x) such that u(x) = sum_i=1^n u(i,1,1,1)*phi_i(x) then bm1(i,1,1,1) = \int phi_i(x) dx where the integral (\int) is taken over the entire computational domain. [ Note that we really don't have such a global phi_i(x) ... I use it simply for illustration, here.] hth Paul From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Monday, February 26, 2018 11:59:19 AM To: nek5000-users Subject: [Nek5000-users] Volume of an element Hi Neks, How to calculate volume of an element in a 3D mesh? If i write n = nx1*ny1*nz1*nelv do i=1,n v(i) = volume(1,1,1,i) enddo Can i store the volume of each element into a varibale called v? What happens if i use bm1(1,1,1,i)? Instead of volume(1,1,1,i) Thanks Sijo _______________________________________________ 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 Feb 27 05:12:12 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 27 Feb 2018 19:12:12 +0800 Subject: [Nek5000-users] Stability utilities In-Reply-To: References: Message-ID: Hi Guillaume, Thank you for clear my misunderstanding! I am not sure which exact the numerical scheme adopted in nek5000 to solve the eigenfunctions of direct and adjoint problems. I myself did not implement new things, just run a steady cylinder flow supplied in cyl_restart and used the setups of lin_chan2D to continue my simulation. I enclosed the parameter files here, please check where I was wrong. Thanks! Wei On Mon, Feb 26, 2018 at 9:38 PM, wrote: > Hi, > > If I understand the lin_channel2D example correctly, it is not using any > Krylov method but simply sets an initial perturbation and lets it evolve. > Are you using the Arnoldi algorithm? > > I can have a look if you upload your files somewhere. > > Guillaume > > > Le 24/02/2018 ? 08:53, nek5000-users at lists.mcs.anl.gov a ?crit : > > Dear Neks, > > Recently, I am trying to set up a linear stability analysis of flow past a > cylinder at very low Re, see Re=20, after I had a converged base flow, I > modified the example ?lin_chan*? and read the dns result as baseflow and > updated the *usr file, after testing groups of parameters in *.par, I > obtained different groups of growth rates, but none of them is closed to > the value documented by Luchini et tal, JFM or the value given by other > codes. Are there more details or rules to choose the parameters when > performing a linear stability, e.g. Krylov space dimension, number of > converged eigenvalues to be dumped? > > Thanks! > Wei > > > > _______________________________________________ > Nek5000-users mailing listNek5000-users at lists.mcs.anl.govhttps://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 > > -- Wei HE ?? School of Aeronautics Universidad Polit?cnica de Madrid http://www.cfm.upm.es/ -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: cyl_dir.par Type: application/octet-stream Size: 449 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: cyl_dir.usr Type: application/octet-stream Size: 4941 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: cyl_dir.re2 Type: application/octet-stream Size: 17636 bytes Desc: not available URL: From nek5000-users at lists.mcs.anl.gov Tue Feb 27 06:28:54 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 27 Feb 2018 12:28:54 +0000 Subject: [Nek5000-users] Volume of an element In-Reply-To: References: , Message-ID: Dear Sijo, The volume per grid point is given by the entries in bm1(i,j,k,e) For each element e, there are points [i,j,k] with index ranges i=1,...,nx1 j=1,...,ny1, k=1,...,nz1 You must reference elements by the 4th index (...,e). Fortunately, all the data is contiguous, so, if accessing _all_ points, you can use a loop like do i=1,n volume_of_point_i = bm1(i,1,1,1) enddo You certainly do not need to do any summing if you only want the point wise volume. Perhaps it'd be best if you indicate what exactly you are planning to do with the volume. Finally, I see your use of glmax and glmin. How many processors are you using? If more than one, are you intending that they are all writing to the file hcoeff.dat? hth, Paul ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Tuesday, February 27, 2018 4:45:30 AM To: nek5000-users Subject: Re: [Nek5000-users] Volume of an element Thanks Paul for your Kind replay. You were right. I want to get the volume associated with each grid point. And I tried the code below, but at each time when I run I get different answers for volume so I think my ocde is not rubust. Could you please verify my code below. nxyz = nx1*ny1*nz1 n = nx1*ny1*nz1*nelv vol = vlsum(bm1(1,1,1,1),nxyz) a=glmax(vol,1) b=glmin(vol,1) do e=1,n volume = vlsum(bm1(e,1,1,1),nxyz) open (unit=10,file="hcoeff.dat") write (10,7020) a,b,volume 7020 format(7f20.14) enddo Regards Sijo GEORGE ________________________________ De: "nek5000-users" ?: "nek5000-users" Envoy?: Lundi 26 F?vrier 2018 21:46:02 Objet: Re: [Nek5000-users] Volume of an element Dear Sijo, We typically refer to elements as the structures that have (N+1)^3 points. (Here, N+1 = lx1.) They are enumerated in nek5000 by the element number, e: integer e nxyz = nx1*ny1*nz1 do e=1,nelt volume = volel(e) volume = vlsum(bm1(1,1,1,e),nxyz) enddo would both yield the same value for the volume of element e. It appears to me that you are after something different, however. It seems like you want to associate a volume with each grid point. If that is the case, then the entries of the diagonal mass matrix, bm1, are indeed the quantities that you want. Each entry of bm1(i,1,1,1) for i=1 to n:=nxyz*nelt, corresponds to the integral of the underlying basis function. That is, if we think about a function phi_i(x) such that u(x) = sum_i=1^n u(i,1,1,1)*phi_i(x) then bm1(i,1,1,1) = \int phi_i(x) dx where the integral (\int) is taken over the entire computational domain. [ Note that we really don't have such a global phi_i(x) ... I use it simply for illustration, here.] hth Paul ________________________________ From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Monday, February 26, 2018 11:59:19 AM To: nek5000-users Subject: [Nek5000-users] Volume of an element Hi Neks, How to calculate volume of an element in a 3D mesh? If i write n = nx1*ny1*nz1*nelv do i=1,n v(i) = volume(1,1,1,i) enddo Can i store the volume of each element into a varibale called v? What happens if i use bm1(1,1,1,i)? Instead of volume(1,1,1,i) Thanks Sijo _______________________________________________ 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 Feb 27 06:39:46 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 27 Feb 2018 13:39:46 +0100 (CET) Subject: [Nek5000-users] Volume of an element In-Reply-To: References: Message-ID: De: "nek5000-users" ?: "nek5000-users" Envoy?: Mardi 27 F?vrier 2018 13:28:54 Objet: Re: [Nek5000-users] Volume of an element Dear Sijo, The volume per grid point is given by the entries in bm1(i,j,k,e) For each element e, there are points [i,j,k] with index ranges i=1,...,nx1 j=1,...,ny1, k=1,...,nz1 You must reference elements by the 4th index (...,e). Fortunately, all the data is contiguous, so, if accessing _all_ points, you can use a loop like do i=1,n volume_of_point_i = bm1(i,1,1,1) enddo You certainly do not need to do any summing if you only want the point wise volume. Perhaps it'd be best if you indicate what exactly you are planning to do with the volume. Finally, I see your use of glmax and glmin. How many processors are you using? If more than one, are you intending that they are all writing to the file hcoeff.dat? hth, Paul From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Tuesday, February 27, 2018 4:45:30 AM To: nek5000-users Subject: Re: [Nek5000-users] Volume of an element Thanks Paul for your Kind replay. You were right. I want to get the volume associated with each grid point. And I tried the code below, but at each time when I run I get different answers for volume so I think my ocde is not rubust. Could you please verify my code below. nxyz = nx1*ny1*nz1 n = nx1*ny1*nz1*nelv vol = vlsum(bm1(1,1,1,1),nxyz) a=glmax(vol,1) b=glmin(vol,1) do e=1,n volume = vlsum(bm1(e,1,1,1),nxyz) open (unit=10,file="hcoeff.dat") write (10,7020) a,b,volume 7020 format(7f20.14) enddo Regards Sijo GEORGE De: "nek5000-users" ?: "nek5000-users" Envoy?: Lundi 26 F?vrier 2018 21:46:02 Objet: Re: [Nek5000-users] Volume of an element Dear Sijo, We typically refer to elements as the structures that have (N+1)^3 points. (Here, N+1 = lx1.) They are enumerated in nek5000 by the element number, e: integer e nxyz = nx1*ny1*nz1 do e=1,nelt volume = volel(e) volume = vlsum(bm1(1,1,1,e),nxyz) enddo would both yield the same value for the volume of element e. It appears to me that you are after something different, however. It seems like you want to associate a volume with each grid point. If that is the case, then the entries of the diagonal mass matrix, bm1, are indeed the quantities that you want. Each entry of bm1(i,1,1,1) for i=1 to n:=nxyz*nelt, corresponds to the integral of the underlying basis function. That is, if we think about a function phi_i(x) such that u(x) = sum_i=1^n u(i,1,1,1)*phi_i(x) then bm1(i,1,1,1) = \int phi_i(x) dx where the integral (\int) is taken over the entire computational domain. [ Note that we really don't have such a global phi_i(x) ... I use it simply for illustration, here.] hth Paul From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Monday, February 26, 2018 11:59:19 AM To: nek5000-users Subject: [Nek5000-users] Volume of an element Hi Neks, How to calculate volume of an element in a 3D mesh? If i write n = nx1*ny1*nz1*nelv do i=1,n v(i) = volume(1,1,1,i) enddo Can i store the volume of each element into a varibale called v? What happens if i use bm1(1,1,1,i)? Instead of volume(1,1,1,i) Thanks Sijo _______________________________________________ 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 Feb 27 06:53:52 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 27 Feb 2018 13:53:52 +0100 (CET) Subject: [Nek5000-users] Volume of an element In-Reply-To: References: Message-ID: Sorry for an empty email before, it was a mistake. Here is my problem I was trying to make the blending coefficient(h) depending upon the volume of elements to have a mesh deoformation rigid near the walls (beasev = 1) and 0 at boundaries (beasev = 0). I had also seen a way to make blending coefficient from the papper that you had sent to me and it was there in an example called ocyl2.usr . But there is variable called delta which is an unknown scale to me and it depends up on the mesh. How ever, I have also seen another equation which deals with volumes of elements in mesh (you had mentioned in you papper also). I was successfully implemented that function on 2D cases by the code shown at the end. But when it becomes to 3D case, I realise that the volume of elements that i tried to capture is in a wrong way. So could you please help me to find the volume of element corresponding to all the grid points (I would like to have the element volume not the spectral elements volume for each grid points) for example if the spectral element order is 4. then it would have 25 small spectral elements in 1 big element right? What I would like to have what ever the grid point in that element I would like to have the element volume. that is if i tried like below,, n = nx1*ny1*nz1*nelv do e=1,n volume = bm1(e,1,1,1) enddo it will throw spectral elements volume right? and if i tried this, do e=1,nelt vol(e) = volel(e) enddo I will get element volume, but i would like to create smoothing varibale for all grid points. That is i need h(n) not h(nelt) (n and nelt are the array sizes) Code which works for 2D v = volume(1,1,1,1) a=glmax(v,1) b=glmin(v,1) do i=1,nelv denom = 1/a h1(i) = ((1-(b/a))/(denom*v)) enddo Regards Sijo De: "nek5000-users" ?: "nek5000-users" Envoy?: Mardi 27 F?vrier 2018 13:28:54 Objet: Re: [Nek5000-users] Volume of an element Dear Sijo, The volume per grid point is given by the entries in bm1(i,j,k,e) For each element e, there are points [i,j,k] with index ranges i=1,...,nx1 j=1,...,ny1, k=1,...,nz1 You must reference elements by the 4th index (...,e). Fortunately, all the data is contiguous, so, if accessing _all_ points, you can use a loop like do i=1,n volume_of_point_i = bm1(i,1,1,1) enddo You certainly do not need to do any summing if you only want the point wise volume. Perhaps it'd be best if you indicate what exactly you are planning to do with the volume. Finally, I see your use of glmax and glmin. How many processors are you using? If more than one, are you intending that they are all writing to the file hcoeff.dat? hth, Paul From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Tuesday, February 27, 2018 4:45:30 AM To: nek5000-users Subject: Re: [Nek5000-users] Volume of an element Thanks Paul for your Kind replay. You were right. I want to get the volume associated with each grid point. And I tried the code below, but at each time when I run I get different answers for volume so I think my ocde is not rubust. Could you please verify my code below. nxyz = nx1*ny1*nz1 n = nx1*ny1*nz1*nelv vol = vlsum(bm1(1,1,1,1),nxyz) a=glmax(vol,1) b=glmin(vol,1) do e=1,n volume = vlsum(bm1(e,1,1,1),nxyz) open (unit=10,file="hcoeff.dat") write (10,7020) a,b,volume 7020 format(7f20.14) enddo Regards Sijo GEORGE De: "nek5000-users" ?: "nek5000-users" Envoy?: Lundi 26 F?vrier 2018 21:46:02 Objet: Re: [Nek5000-users] Volume of an element Dear Sijo, We typically refer to elements as the structures that have (N+1)^3 points. (Here, N+1 = lx1.) They are enumerated in nek5000 by the element number, e: integer e nxyz = nx1*ny1*nz1 do e=1,nelt volume = volel(e) volume = vlsum(bm1(1,1,1,e),nxyz) enddo would both yield the same value for the volume of element e. It appears to me that you are after something different, however. It seems like you want to associate a volume with each grid point. If that is the case, then the entries of the diagonal mass matrix, bm1, are indeed the quantities that you want. Each entry of bm1(i,1,1,1) for i=1 to n:=nxyz*nelt, corresponds to the integral of the underlying basis function. That is, if we think about a function phi_i(x) such that u(x) = sum_i=1^n u(i,1,1,1)*phi_i(x) then bm1(i,1,1,1) = \int phi_i(x) dx where the integral (\int) is taken over the entire computational domain. [ Note that we really don't have such a global phi_i(x) ... I use it simply for illustration, here.] hth Paul From: Nek5000-users on behalf of nek5000-users at lists.mcs.anl.gov Sent: Monday, February 26, 2018 11:59:19 AM To: nek5000-users Subject: [Nek5000-users] Volume of an element Hi Neks, How to calculate volume of an element in a 3D mesh? If i write n = nx1*ny1*nz1*nelv do i=1,n v(i) = volume(1,1,1,i) enddo Can i store the volume of each element into a varibale called v? What happens if i use bm1(1,1,1,i)? Instead of volume(1,1,1,i) Thanks Sijo _______________________________________________ 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 Feb 27 09:13:53 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 27 Feb 2018 16:13:53 +0100 (CET) Subject: [Nek5000-users] blending coefficient in 3D Message-ID: Hi, In order to make a smooth blending coefficient for mesh motion, I made a code below, it works fine for a 2D case in my computer but it does not work as expected for a 3D case (I run 3D case in server with 1024 pocessors). Could you please tell me is there anything wrong? (The error is the volume that i am getting is some wrong values i think) and is it right way to make Dirichlet B.C on boundaries also calling hmholtz function to make a blending velocity gradient? Because I took that part from the NEK example which is 2D. So do i have to change anything to make working in 3D? subroutine my_base_meshv(basev,h1,h2,rhs,msk,tmp) include 'SIZE' include 'TOTAL' real basev(lx1,ly1,lz1,lelt),tmp(lx1,ly1,lz1,lelt) real h1(1),h2(1),rhs(1),msk(1) real m1,denom,volume(nx1*ny1*nz1*nelv) integer e,f n = nx1*ny1*nz1*nelv nxyz = nx1*ny1*nz1 do e=1,n volume(e) = vlsum(bm1(e,1,1,1),nxyz) enddo a=glmax(volume,n) b=glmin(volume,n) do i=1,n denom = 1/a h1(i) = ((1-(b/a))/(denom*volume(i))) enddo do i=1,n open (unit=10,file="h.dat") write (10,7020) h1(i),volume(i),a,b 7020 format(7f20.8) enddo m1 = -1. z0 = 0. nface = 2*ndim do e=1,nelv do f=1,nface c Set Dirichlet for mesh velocity on all boundaries if (cbc(f,e,1).ne.'E ') call facev(msk,e,f,z0,nx1,ny1,nz1) c Inhomogeneous Dirichlet on cylinder only if (cbc(f,e,1).eq.'mv ') then call fcaver(xavg,xm1,e,f) if (xavg.gt.-1.2) call facev(tmp,e,f,m1,nx1,ny1,nz1) endif enddo enddo call axhelm (rhs,tmp,h1,h2,1,1) tol = 1.e-4 imsh = 1 ifield = 1 call hmholtz('mshv',basev,rhs,h1,h2,msk,vmult,imsh,tol,200,1) call sub2(basev,tmp,n) c call outpost(basev,basev,basev,basev,basev,' ') c stop return end -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Feb 27 00:15:25 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 27 Feb 2018 15:15:25 +0900 Subject: [Nek5000-users] CPU, Memory and MPI Message-ID: Hello Neks,I want to know specific information about optimal usage of computing sources.I would like to show you my own environment for running a case.CPU: 3.3 GHz, 4 cores, 8 processors (Intel i7-6700HQ)RAM: 2133 MHz, 8 GB x 1 (DDR4-2133)Case: 1500 elements, N = 61.Based on Nek5000 document, total memory requirement is1500 (points) X 6^3 X 400 (words/point) X 8 (bytes/word) X (1 GB/1024^3 bytes) = 0.97 GBSo, is this calculation correct? so is 8 GB of memory enough?2.In the Nek5000 document, they also mention about "memory per core."Could you tell me the meaning of memory per core?I thought that, in my case, I assigned 2GB / core (If I use MPI with 4 cores. because I also have 8 GB of memory).If there are manual way to assign, how can I assign memory to cores?3.Actually, I also tried to use nekbmpi to use MPI process in my PC.But when I type like belownekbmpi 8(or 4)my CPU usage of 8 processors suddenly goes up but, there are no progress in simulation.(log file is remained as 0 bytes. when I check 'Task manager' of Windows 10,Nek5000 processes appear but they take just few MB of memory, but usage of CPU peaked up to 100%)I also tried to use nek5000 directly as below./nek5000 -np 8(or 4) > case.out &But I think, after typing this syntax, PC shows just moderate performance (same performance with ./nek5000 ).(log file also shows Number of processors: 1)I already installed MPI in my ubuntu environment in Windows 10.Also I compiled make file with mpif77 and mpicc.How can I use MPI in my PC?4.Finally,If you have idea to estimate time to run your case approximately(e.g. cycle per second), please give me.And I will also appreciate if you show how to calculate the appropriate number of processors to run a case.(no matter if it is based on your environment or my environment) Best Regards,Hyunduk. -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Tue Feb 27 10:08:39 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Tue, 27 Feb 2018 16:08:39 +0000 Subject: [Nek5000-users] Tripping mechanism In-Reply-To: References: Message-ID: Hi Philipp, Thank you for your response. It has been very helpful. Best, Juan Diego -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Feb 28 03:50:45 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 28 Feb 2018 10:50:45 +0100 (CET) Subject: [Nek5000-users] Base velocity problem Message-ID: Hi Neks, I am bit stuck on this occasion for past 2 days, the code below shows a part of the example called ocyl2.usr (from subroutine my_base_meshv(basev,h1,h2,rhs,msk,tmp) ). This code is to generate a base velocity (where basev=1 at wall of object and 0 at the boundary of the domain). I assume that I have a blending coefficient h1.(using Paul's method or volume of elements method). This subroutine working perfectly for the 2D cases. That is I have the base velocity ranges between 0 and 1. But when I tried it on an aero dynamic foil case(its a 3D case where just 1 element in Y direction), the value of base velocity does not even exceed 0. Almsot all values are like -0.0021564... Could you please tell me is there anything that I have to change in this subroutine to make it working. Tha value of m1 and z0 are just explicity giving the base velocity to the elements right? for example on the W or M.V conditions basev becomes 1 and on boundaries of the domain it becomes 0.0 m1 = -1. z0 = 0.0 nface = 2*ndim do e=1,nelv do f=1,nface c Set Dirichlet for mesh velocity on all boundaries if (cbc(f,e,1).ne.'E ') call facev(msk,e,f,z0,nx1,ny1,nz1) c Inhomogeneous Dirichlet on cylinder only if (cbc(f,e,1).eq.'mv ') then call fcaver(xavg,xm1,e,f) if (xavg.gt.-3.0) call facev(tmp,e,f,m1,nx1,ny1,nz1) endif enddo enddo call axhelm (rhs,tmp,h1,h2,1,1) tol = 1.e-4 imsh = 1 ifield = 1 call hmholtz('mshv',basev,rhs,h1,h2,msk,vmult,imsh,tol,200,1) call sub2(basev,tmp,n) c call outpost(basev,basev,basev,basev,basev,' ') c stop Regards Sijo -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Feb 28 02:36:56 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 28 Feb 2018 16:36:56 +0800 Subject: [Nek5000-users] periodic boundary conditions in mshconvert.py Message-ID: Hello Neks? I?m trying to use the mshconvert.py to get a periodic bc. I put in convert('test.msh',periodic_dx={(3,4):[0.0,3.0]},bcs={5:'O',6:'v'}). But there is always a mistake: Traceback (most recent call last): File "", line 1, in convert('test.msh',periodic_dx={(3,4):[0.0,3.0]},bcs={5:'O',6:'v'}) File "C:/Users/Administrator/Desktop/test/mshconvert.py", line 1477, in convert create_boundary_section(bcs, temperature, passive_scalars, mesh_format) File "C:/Users/Administrator/Desktop/test/mshconvert.py", line 668, in create_boundary_section periodic_cell_face_map[(c, local_face)] KeyError: (1, 2) The .msh file is the most simple one? (0 "GAMBIT to Fluent File") (0 "Dimension:") (2 2) (10 (0 1 8 1 2)) (10 (1 1 8 1 2)( 1.0000000000e+000 0.0000000000e+000 0.0000000000e+000 0.0000000000e+000 0.0000000000e+000 3.0000000000e+000 1.0000000000e+000 3.0000000000e+000 0.0000000000e+000 1.0000000000e+000 0.0000000000e+000 2.0000000000e+000 1.0000000000e+000 1.0000000000e+000 1.0000000000e+000 2.0000000000e+000 )) (0 "Faces:") (13(0 1 a 0)) (13(3 1 1 4 0)( 2 1 2 0 1 )) (13(4 2 2 e 0)( 2 3 4 0 3 )) (13(5 3 5 25 0)( 2 1 7 1 0 2 7 8 2 0 2 8 4 3 0 )) (13(6 6 8 3 0)( 2 2 5 0 1 2 5 6 0 2 2 6 3 0 3 )) (13(8 9 a 2 0)( 2 7 5 1 2 2 8 6 2 3 )) (0 "Cells:") (12 (0 1 3 0)) (12 (2 1 3 1 3)) (0 "Zones:") (45 (2 fluid fluid)()) (45 (3 inlet-vent p2)()) (45 (4 fan p1)()) (45 (5 axis out)()) (45 (6 wall in)()) (45 (8 interior default-interior)()) Thank you! Lily -------------- next part -------------- An HTML attachment was scrubbed... URL: From nek5000-users at lists.mcs.anl.gov Wed Feb 28 12:14:32 2018 From: nek5000-users at lists.mcs.anl.gov (nek5000-users at lists.mcs.anl.gov) Date: Wed, 28 Feb 2018 19:14:32 +0100 (CET) Subject: [Nek5000-users] Regarding base velocity (Mesh deformation) Message-ID: Hi Neks, I had sent you a detailed problem regarding base velocity (mesh motion and mesh deformation) in the morning unfortunately i have not get any replay yet . I think you guys are busy otherwise you guys are very resposive. I appreciate that. Even though I investigated my problem and I think i found where actually the problem is even though I still need some clarifications in that. The problem were in the boundary conditions since there are two type of B.Cs Homogenous (zero on all non- interior boundary faces ) and Inhomogenous( 1 at moving boundary faces). When i was trying on 2D there were no problem with B.Cs. But when i i tried my 3D mesh where a new B.C has come which is called SYMMETRY b.c. I imposed the SYM conditions on the horizontal walls of the fluid domain (not on the object) and also the exterior to the fluid domain on 3rd dimension.(3D means i put a thinkess as size of one element, which makes my foil case into 3D) As a result, for one element near the moving body, I have a m.v condition, SYM conditon on the exterior of the fluid domain in 3rd direction (lets say Z direction, so SYM on both sides) and E on the interior of the fluid domain. So the condition I gave is given below : m1 = -1. y0 = 0.0 nface = 2*ndim do e=1,nelv do f=1,nface c Set Dirichlet for mesh velocity on all boundaries if (cbc(f,e,1).ne.'E ') call facev(msk,e,f,y0,nx1,ny1,nz1) c Inhomogeneous Dirichlet on foil only if (cbc(f,e,1).eq.'mv '.or. cbc(f,e,1).eq.'SYM ') then if((ym1(1,1,f,e).lt.6.0).and.(ym1(1,1,f,e).gt.-6.0)) then !! where my domain +6 to -6 in Y direction and 0 to 0.1 in Z direction and -8 to 8 in X direction call facev(tmp,e,f,m1,nx1,ny1,nz1) endif endif enddo enddo Now I get a better basev (the values ranges between 0 and 1.2 before it was less than zero). But still I am getting a value which is slightly greater than 1.0 which is not good. So could you please tell me what should I do in this case, to have a nice smoothing velocity to all the mesh. Thanks with Regards Sijo GEORGE -------------- next part -------------- An HTML attachment was scrubbed... URL: