November 2016 - Nek5000-users - mailman3.cels.anl.gov

Re: [Nek5000-users] Error Compiling
by nek5000-users＠lists.mcs.anl.gov 20 Nov '16

20 Nov '16

Hi, I think it might be helpful to familiarize yourself with the Nek’s user guide first: http://nek5000.github.io/NekDoc/Nek_users.pdf Specifically chapter 2 which talks about setting up your path. Using Visit might be a better option if you are just starting with Nek since there is a lot more documentation for it. Let’s say you have result files for your case (let’s say a1) - a10.f0000* or a10.fld0* formats. To visualize using Visit, type "visnek a1” which will generate a file a1.nek5000. Now you are ready to visualize.. open Visit (I have visit 2.9.1), and click “Open” and navigate to the folder where you ran your calculation. Once you are there double click on a1.nek5000. It should open fine and you can then use the options (open - pseudocolor - velocity) to visualize velocity results etc… If you really want to use postx, type in postx to get started. This will pop up a black window. Depending on your display settings, the text at the bottom of the window might not be visible. If you don’t see any text, you will have to use the terminal to read the text (the text is duplicated there). I usually move the black window to the right side of the screen and terminal to the left side of the screen. Once you have done that, click on the black window and answer the question that you read on terminal window “Enter Session Name”. Type in a1 and hit enter. The postx window should populate with a plot of the mesh and a bunch of options on the right side which are self descriptive. Thanks, Ketan

1 0

Error Compiling
by nek5000-users＠lists.mcs.anl.gov 18 Nov '16

18 Nov '16

Hi all, I'm ver new at nek5000. I'm trying to compiling a first example case (eddy) and I get a error. After ./makenek eddy_uv I get: ./makenek: line 92: /home/mateus/nek5000/core/makenek.inc: No such file or directory make: makefile: No such file or directory make: *** No rule to make target 'makefile'. Stop. Can someone help me? -- Mateus Peixoto Avanci Engenheiro Mecânico Aluno do Programa de Pós-Graduação em Engenharia Mecânica Universidade Federal Fluminense *http://lattes.cnpq.br/1773896725783894 <http://lattes.cnpq.br/1773896725783894>*

1 10

Autosphere's radius
by nek5000-users＠lists.mcs.anl.gov 17 Nov '16

17 Nov '16

Hi, Neks! I am trying to create an own mesh, based on expansion example. When I build a circle in pretex, it is always has radius = 0.5. Can I change it in the pretex directly? (Now I could vary only numbers of elements in inner and outer shells). Thanks in advance, Jacob.

1 1

Linear Systems and Solvers in Nek5000
by nek5000-users＠lists.mcs.anl.gov 14 Nov '16

14 Nov '16

Hello Nek users, This could be a simple question to some, but any help would be great here. I am working on a project that is aiming to suggest appropriate solvers for the system Ax = b, using Machine Learning, given the matrix A. I know that there are many matrix datasets online, but for specific reasons, we want to extract matrices from Nek5000 (compute features from that) and note the solvers it uses for those linear systems. The Navier-Stokes equation is non-linear due to the advection term, but in certain cases, the governing equations reduce to being linear. I am trying to understand what those cases are and the solvers that Nek uses to handle those linear systems. Based on my reading from various sources, it looks like Nek handles the following linear system cases. Any comments on this would be very helpful: 1) *System:* Poisson equation for Pressure (incompressible case?) *Solvers:* Nek supports GMRES and PCG with Parameter P042 (from http://nek5000.github.io/NekDoc/Nek_usersch7.html#x9-750007.1) 2) https://www.mcs.anl.gov/~fischer/Nek5000/nek_tutorial_1.pdf - Page 44 What is the Helmholtz solve for velocity here? Is this the only "solver" for velocity? 3) http://nek5000.github.io/NekDoc/Nek_usersch1.html#x3-50001.1 *System*: Energy equation behaves as a transport equation for temperature right? *Solvers*: Helmholz, CVODE, CVODE with user-supplied Jacobian. 4) All Stokes flow cases with constant viscosity? Does this seem correct? If you need any other information, please let me know. Thanking You, K.N.Ramachandran (Ram) Ph: 814-441-4279

1 0

How to know if a solution if converged?
by nek5000-users＠lists.mcs.anl.gov 14 Nov '16

14 Nov '16

Dear Nek users, I have a numerical solution for a turbulent flow and would like to know if my solution is well converged. My current method consists in monitoring the average velocity. I would like to know if there is a more advanced method/technique I could use in order to determine whether or not my solution is converged. Thanks, Marco

1 0

Boundary condition on velocity
by nek5000-users＠lists.mcs.anl.gov 11 Nov '16

11 Nov '16

Hi, neks. I am trying to simulate a 3d pipe, fluid flows along z. And I need an inflow boundary condition on velocity on forming of cylinder. How I can set a radial velocity in userbc (Vr = sqrt( vx**2+vy**2 ) ) ? I need to set Vr and dVr/dr. Thank you so mach Vatslav

1 1

答复: Nek5000-users Digest, Vol 93, Issue 6
by nek5000-users＠lists.mcs.anl.gov 09 Nov '16

09 Nov '16

Hello, Nicolas I read the documentation again carefully and very happy that my problem has been solved. Thank you very much for advice. I would like to share the curve definition parts in my mesh converting commands for a ring below: Curves={4:{'type':'C','depth':4,'circle_center':(0,0), 'radius':1},3:{'type':'C','depth':4,'circle_center':(0,0),'radius':0.5}} For 4 and 3 are outer and inner circle. ---------------------------------------------------------------- Bolun Xu University of Science and Technology of China Hefei, Anhui, China -----邮件原件----- 发件人: nek5000-users-bounces(a)lists.mcs.anl.gov [mailto:[email protected]] 代表 nek5000-users-request(a)lists.mcs.anl.gov 发送时间: 2016年11月9日 23:27 收件人: nek5000-users(a)lists.mcs.anl.gov 主题: Nek5000-users Digest, Vol 93, Issue 6 Send Nek5000-users mailing list submissions to nek5000-users(a)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(a)lists.mcs.anl.gov You can reach the person managing the list at nek5000-users-owner(a)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. Re: A problem in Prenek (nek5000-users(a)lists.mcs.anl.gov) 2. 2D circular ring mesh converting by mshconvert.py (nek5000-users(a)lists.mcs.anl.gov) 3. Re: 2D circular ring mesh converting by mshconvert.py (nek5000-users(a)lists.mcs.anl.gov) 4. Re: Read Nek output in Matlab (nek5000-users(a)lists.mcs.anl.gov) 5. Name of the variable in userf and userq (nek5000-users(a)lists.mcs.anl.gov) ---------------------------------------------------------------------- Message: 1 Date: Tue, 8 Nov 2016 21:32:25 +0000 From: nek5000-users(a)lists.mcs.anl.gov To: "nek5000-users(a)lists.mcs.anl.gov" <nek5000-users(a)lists.mcs.anl.gov> Subject: Re: [Nek5000-users] A problem in Prenek Message-ID: <mailman.5887.1478640760.3602.nek5000-users(a)lists.mcs.anl.gov> Content-Type: text/plain; charset="utf-8" Hi Saikat, I think this is what is expected. If we look at postpro.f in nek?s source code (subroutine gen_rea_bc), there are 4 different formats in which the boundary conditions will be written depending on the element count. In your case, since the element count is greater than 1000 but less than 100,000, we expect the format to be: first 4 character spaces for boundary condition, next 5 characters for element number, 1 character for edge/face number depending on 2d/3d and so on. As a result, your boundary condition line shows as: E 30724 2731.00 2.00000 0.00000 0.00000 0.00000 Thanks, Ketan

1 0

Name of the variable in userf and userq
by nek5000-users＠lists.mcs.anl.gov 09 Nov '16

09 Nov '16

Hi Neks, I am trying to set volume forcing in my equations through userf and userq. However it is not for the baseflow but for perturbations. Does anyone know the name of the perturbations variables which are equivalent to ffx, ffy, ffz, and qvol? So I can write something like : ffxp = ... ffyp = ... ffzp = ... qvolp = ... Thank you for your time, Best wishes, Arnold

1 0

PostDoc Position using NEK5000
by nek5000-users＠lists.mcs.anl.gov 09 Nov '16

09 Nov '16

Dear Nek5000 users, We have two immediate postdoc positions in our group in CORIA lab. (CNRS), Rouen, France. One of these position is on the use of Nek5000 software for converging and diverging pipe flows. We are looking for a talented young researcher who has a good knowledge of Nek5000 to join to our group for Jan./Feb. 2017. Please feel free to send your CV along with your references to me. More details about the position can be found in: http://www.cfd-online.com/Jobs/showjob.php?record_id=13809 Best, Mostafa -- Kind Regards, ****************************************** Mostafa SAFDARI SHADLOO Associate Professor, CNRS-University & INSA of Rouen 76800 St Etienne du Rouvray, France Office: (+33) 02 32 95 97 76 Emailto: msshadloo(a)coria.fr <msshadloo(a)ec-nantes.fr>, mostafa(a)sabanciuniv.edu ******************************************

1 0

Read Nek output in Matlab
by nek5000-users＠lists.mcs.anl.gov 09 Nov '16

09 Nov '16

Hi Neks, I would like to make some postprocessing with the Nek output. Can I read that files in Matlab (or C)? Just to make it clear: If my .rea file is called whatever.rea and the output is whatever0.f00**** I would like to read whatever0.f00**** and make some postprocessing. It will be also very helpful to learn how to extrapolate the mesh so that I can plot combinations of the solutions for instance. Thank you so much, Alessandro -- Dr Alessandro Alla Florida State University Department of Scienfitic Computing 400 Dirac Science Library Tallahassee FL 32306-4120 Room: 462 Phone: +1 8506441010 E-Mail: aalla(a)fsu.edu Skype: alessandro.alla www.alessandroalla.com

1 2