April 2018 - Nektar-users - imperial.ac.uk

NaN appears in using CompressibleFlowSolver
by 郭先文 07 May '18

07 May '18

Dear all, Recently I'm applying Nektar++ into calculating a 2D case of subsonic flow around a circular cylinder. The flow condition is under Re = 200000 and Ma = 0.75. And NaN in speed of sound keeps appearing within two specific element in the downstream and off-wall region of the cylinder, even after changing of grid size and time step. I wonder if the problem could be solved with the usage of artificial viscosity or smooth shock capture method (Non-smooth didn't work). But the user's guide doesn't cover enough information about those settings in detail. Where could I find some details and settings about those methods? Or is there any other solution to the problem? Thanks very much for your help. Regards, Xianwen. Xianwen Guo College of Engineering Peking University

2 1

ctest fails since Test 5 onwards
by #SHI ZIJI# 05 May '18

05 May '18

Dear all, I am having troubles compiling nektar 4.4.1 on Centos 7. I linked it with MKL, and was able to compile it, but it failed from test 5 till the last test (test 1 to 4 were successful though). My cmake command was ``` cmake\ -DCMAKE_CXX_FLAGS_RELEASE=" -I${MKLROOT}/include " \ -DNEKTAR_USE_MKL=ON \ -DMKL_LAPACK=${MKLROOT}/lib/intel64/libmkl_lapack95_lp64.a \ -DNEKTAR_USE_MPI=ON .. ``` I have tried linked with ThirdParty Boost following the FAQ page, but it does not help. I suspect the error was from linking with MKL? The deatailed error message was attached below. Start 4: Collections 4: Test command: /home/public/nektar/nektar++-4.4.1/build/library/UnitTests/Collections/CollectionsUnitTests "--detect_memory_leaks=0" 4: Test timeout computed to be: 9.99988e+06 4: Running 160 test cases... 4: 4: *** No errors detected 4/436 Test #4: Collections ....................................................................... Passed 0.37 sec test 5 Start 5: LibUtilities_NodalDemo_Tri_Deriv_P8 5: Test command: /home/public/nektar/nektar++-4.4.1/build/tests/Tester "/home/public/nektar/nektar++-4.4.1/library/Demos/LibUtilities/Tests/NodalDemo_Tri_Deriv_P8.tst" 5: Test timeout computed to be: 9.99988e+06 5: Error occurred running test: 5: Command: NodalDemo-g --order 8 --type 22 --deriv 1>output.out 2>output.err 5: === Output === 5: 5: === Errors === 5: sh: NodalDemo-g: command not found 5: 5/436 Test #5: LibUtilities_NodalDemo_Tri_Deriv_P8 ...............................................***Failed 0.02 sec Error occurred running test: Command: NodalDemo-g --order 8 --type 22 --deriv 1>output.out 2>output.err === Output === === Errors === sh: NodalDemo-g: command not found test 6 Start 6: LibUtilities_NodalDemo_Tri_Integral_P6 6: Test command: /home/public/nektar/nektar++-4.4.1/build/tests/Tester "/home/public/nektar/nektar++-4.4.1/library/Demos/LibUtilities/Tests/NodalDemo_Tri_Integral_P6.tst" 6: Test timeout computed to be: 9.99988e+06 6: Error occurred running test: 6: Command: NodalDemo-g --order 6 --type 23 --integral 1>output.out 2>output.err 6: === Output === 6: 6: === Errors === 6: sh: NodalDemo-g: command not found 6: 6/436 Test #6: LibUtilities_NodalDemo_Tri_Integral_P6 ............................................***Failed 0.02 sec Error occurred running test: Command: NodalDemo-g --order 6 --type 23 --integral 1>output.out 2>output.err === Output === === Errors === sh: NodalDemo-g: command not found Any help to resolve this issue or suggestions will be greatly appreciated. Thank you! Best Regards, Ziji

3 2

Nektar Arpack error
by Amitvikram Dutta 28 Apr '18

28 Apr '18

Hi all, I've installed nektar++ as directed and all the ctests pass. However when I try to run one of the example cases, I get the following error. Fatal : Level 0 assertion violation Value 'Arpack' is not valid for property 'DRIVER' I have installed Arpack separately on my machine but I'm not sure if I have to link it separately to nektar++ installation in some manner. Additionally I left all the cmake options on default as they were. I assumed this meant that Nektar++ had built the third party arpack installlation as directed. How should I proceed? Sincerely, -- *Amitvikram Dutta* Graduate Research Assistant Fluid Mechanics Research Lab Multi-Physics Interaction Lab University of Waterloo

2 1

Regarding FieldConvert utility
by Abhishek Kumar 24 Apr '18

24 Apr '18

Dear friends, I am trying to construct a three-dimensional field as an initial condition from a two-dimensional data. Here, the third direction is periodic (a Quasi-3D approach). Is there any way in Nektar++ to project a 2D field on specific number modes (or a number of z planes)? I tried --output-points-hom-z of FieldConvert utility, but this doesn't give the desired output. -- With regards Abhishek Kumar --------------------------------------------------------------------------------------- Abhishek Kumar Research Associate Applied Mathematics Research Centre Coventry University, Coventry CV1 5FB The United Kingdom Email # ac7600(a)coventry.ac.uk, abhishek.kir(a)gmail.com ---------------------------------------------------------------------------------------

2 2

Access to data published in 10.1017/jfm.2017.475
by Shervin Sammak 23 Apr '18

23 Apr '18

Hi, Can you please kindly share the data published in paper entitled "Direct numerical simulations of the flow around wings with spanwise waviness"? Regards, Shervin ----------------------------------------------------------- Shervin Sammak Research Assistant Professor Center for Research Computing University of Pittsburgh 4420 Bayard St Pittsburgh, PA 15213 E-mail: shervin.sammak(a)pitt.edu Website: www.crc.pitt.edu ~ You chase quality and quantity will chase you.

1 0

VTK and higher order elements
by Kurt Sansom 23 Apr '18

23 Apr '18

As of VTK 8.1 its looks like there is new functionality with "arbitrary-order Lagrange-interpolated cells for the following shapes: curve, tri, quad, tet, hex, wedge" https://blog.kitware.com/vtk-8-1-0/ Can anyone give some indication if this would be useful in nektar++ or be beneficial in post processing? Consider me ignorant here. Regards, ~Kurt -- Kurt Sansom

2 1

Re: [Nektar-users] Paraview single point with different values
by Kilian Lackhove 22 Apr '18

22 Apr '18

Hi Guillermo, this is neither a bug in nektar++ nor in paraview. You are using a discontinuous projection, so at the element edges, each field has two values defined by the two adjacent elements. So what you see in paraview is the actual soution. To avoid this effect, you could either increase the resolution in the respective region to reduce the discontinuity or apply a continuous projection in proprocessing, e.g. by using the "C0Projection" FieldConvert module. Cheers, Kilian Am Freitag, 20. April 2018, 20:14:27 CEST schrieb Guillermo: > Hi Kilian, > > Thank you. I attach the .vtu file. I also attach the mesh (.msh.xml) and > the conditions file (.xml). > To find the problem I mentioned I followed these steps: > Open the .vtu file and visualize the Mach number. Then, at the top left > corner you should see something like the image I have attached to this > email (Nozzle.png) (I am using the "Red to blue Rainbow" preset). You > can see that there is a point on the domain with a Mach number value > around 6.67e-01. Now use the "Hover Points on" tool from paraview and > set the cursor over that node. The pop up window should indicate that > the Mach number is 1.10772. > Using the spreadsheet view and searching for those Mach number values > you can see that its coordinates are the same. > > Please, let me know if there is anything else that I can do. > > Thank you. > > Kind regards, > Guillermo Suárez

1 0

FieldConvert binary VTU file
by Brinkerhoff, Joshua 22 Apr '18

22 Apr '18

Hello, I would like FieldConvert to output a binary .vtu file. I couldn't find any information on that in the user guide. Is there a convenient way of doing that? Thanks, and best regards, Joshua Dr. Joshua R. Brinkerhoff P.Eng. Assistant Professor Faculty of Applied Science | School of Engineering The University of British Columbia | Okanagan Campus EME 4217 3333 University Way | Kelowna British Columbia | V1V 1V7 Canada Phone 250 807 8238 joshua.brinkerhoff(a)ubc.ca<mailto:joshua.brinkerhoff@ubc.ca> UBC Okanagan CFD Laboratory http://okcfd.sites.olt.ubc.ca/ [UBC E-mail Signature]

2 1

Paraview single point with different values
by Guillermo 20 Apr '18

20 Apr '18

Dear all, I am visualizing a .vtu file with paraview and I noticed that one of the points from the mesh (visualize the data with the "Surface with edges" option") has different values for the same variable. When visualizing the variable it shows that, this specific variable has a value of 0.667, however, when using the "hover points over" tool, it shows that the variable has another different value, 1.107. Using spreadsheet view, I could locate the first value and its associate Point ID. Then I located the second value, the one from the hover tool, and it has a different Point ID. However, both coordinates are the same. I have contacted with paraview help and answered me that it must be due to some intended performance or an error from the solver, Nektar. My main concernis that maybe I am visualizing wrong results, it is showing the first value and it should be showing the second one instead. I use Point Data for that specific variable. Paraview version 5.0.1. Nektar version 4.4.1. Ubuntu OS. I do not attach the file because it is 9Mb size, but if you need, please let me know. Thank you very much for your help. Cheers, Guillermo Suarez

2 1

Field Convert Tecplot Data Format
by Rubini R. 20 Apr '18

20 Apr '18

Hello, I would like to import the output of Nektar++ to numpy array to do some post processing. I was thinking to convert the output to tecplot readable files and then import the results in numpy array format. The problem I'm facing is that the output generated by FieldConvert seems not to be in the classical DATAPACKING=POINT that makes the import straightforward. Do you know if there is any possibility to convert the output in that format or do you have any suggestion to solve my issue? Many Thanks and have a nice weekend! Regards, Riccardo

3 2