- Nektar-users - imperial.ac.uk

Nektar++ v4.4.0 Released
by Chris Cantwell 09 Mar '17

09 Mar '17

Dear all, We are pleased to announce the next release of Nektar++, v4.4.0. This release includes a wide range of new features including: - a new FieldUtils library which makes much of the FieldConvert functionality available to solvers on-the-fly, - support for HDF5 as an output format, - support for variable polynomial order in 3D CG discretisations, as well as many other improvements and bug fixes. Many of the solvers and utilities have also received major improvements. A full list of the changes is available in the CHANGELOG.md file distributed with the source code. As always, if you encounter any problems please feel free to contact us on the mailing list nektar-users(a)imperial.ac.uk. Cheers, Chris -- Chris Cantwell Imperial College London South Kensington Campus London SW7 2AZ Email: c.cantwell(a)imperial.ac.uk www.imperial.ac.uk/people/c.cantwell

1 0

extracting isosurfaces on runtime
by Asım Önder 08 Mar '17

08 Mar '17

Dear all, Is there any way to make isocontour module work in filter mode for derived quantities, e.g., Q field? I tried the following in config.xml file for a run using IncNavierStokesSolver solver: <FILTER TYPE="FieldConvert"> <PARAM NAME="OutputFile">Q_214.fld</PARAM> <PARAM NAME="OutputFrequency">10</PARAM> <PARAM NAME="Modules"> QCriterion </PARAM> </FILTER> <FILTER TYPE="FieldConvert"> <PARAM NAME="OutputFile">Q_214_iso2000.fld</PARAM> <PARAM NAME="OutputFrequency">10</PARAM> <PARAM NAME="Modules"> isocontour </PARAM> <PARAM NAME="fieldstr"> Q </PARAM> <PARAM NAME="fieldvalue"> 1000 </PARAM> </FILTER> The first filter works, and writes out the 3D Q field. However, second one does not work as Q is not defined as a field in the solver. Is there any possibility to address this other than extending the related solver to contain the calculation of the field? Thanks, Asim

2 2

Initializing a case with a specific boundary layer solution
by devabratasahoo＠campus.technion.ac.il 07 Mar '17

07 Mar '17

Hi All, I went through the CompressibleBL.cpp as well as the FalknerSkanBL.cpp routines provided in nektarinfo in order to initialize my case with a specific boundary layer solution. But, I am not clear how to implement it on the Nektar case file. Do I need to compile the BL code (CompressibleBL.cpp/FalknerSkanBL.cpp) separately and insert the output into the .xml file or is there a way to directly define the BL generator as a function of the initial condition section of the .xml file? I couldn't find any example using similar sort of case. Please suggest. Thanking you. Regards, Dev

1 0

Ununiform and unsteady source terms
by Song Ke 04 Mar '17

04 Mar '17

Hi Douglas: Thanks for your answer. I am testing a new case following your advise. Another problem. Sometimes the my source term model may be too complicate to be write in xml files as a formula or expression. I have to write c++ code for this model. Is the only solution to change the nektar source code? Regards, Song Ke

2 1

Ununiform and unsteady source terms
by Song Ke 28 Feb '17

28 Feb '17

Hi All, I am trying to define an ununiform and unsteady source term in CompressibleFlowSolver (nektar 4.3.5). The source term is independant on flow fields ,and can be defined as s(x,y,z, t) = s0(x,y,z)* s1(t) Here, "s0" can be obtain by a numerical model, and "s1" is a simple and discontinuous function. ( "t" is time.) Can anyone please give me some advise of how to set this source term to CompressibleFlowSolver ? Thanking you. Regards, S.Ke

2 1

Nektar++ Users Workshop 2017
by Chris Cantwell 22 Feb '17

22 Feb '17

Dear all, We are provisionally planning the Nektar++ Users Workshop for 2017 on the 14-16th June, to be held again at Imperial College London. As users of the code, we would like to encourage as many of you as possible to attend and share your work with us and other users. We hope to confirm the details in the next couple of weeks. In a similar format to last year, we plan to organise a programme of talks during the first half of the workshop, followed by interactive group sessions. Each group sessions will focus on addressing a challenge in the code, for example designing/implementing a desired new feature or capability, generalising an existing aspect of the code, or even just fixing a persistent bug! We hope both novice and seasoned users can contribute to these sessions. We would like to invite proposals for the group sessions from our users. Please reply directly to me with suggestions. Kind regards, Chris -- Chris Cantwell Imperial College London South Kensington Campus London SW7 2AZ Email: c.cantwell(a)imperial.ac.uk www.imperial.ac.uk/people/c.cantwell

1 0

Re: [Nektar-users] Source terms and mesh generation
by Selbach, Fabian 15 Feb '17

15 Feb '17

Dear users, Hi Kilian, sorry I have to ask another question and I hope you or somebody else can give me an answer. Actually I tried to use an analytical source term like this: <FUNCTION NAME="Source"> <E VAR="S" VALUE="0.5*exp((-0.34657*(((x-2)^2)+((y-0)^2))))*sin(2*3.14/30*t)" /> </FUNCTION> in the acoustic perturbation equation. I tried other kind of source terms too. Besides I am using a base flow as a result from an incompressible flow simulation. The APE reads them right, but it seems the solver does not read the source term. The results of field of p,u,v,rho are constant in time and do not change. Can somebody explain that? Best regards Fabian Von: Fabian Selbach [fabian.selbach(a)student.uni-siegen.de] Gesendet: Donnerstag, 9. Februar 2017 09:36 An: Kilian Lackhove Betreff: Re: [Nektar-users] Source terms and mesh generation Thanks Kilian, I thought so, but want to ask about it. Best regards Fabian Selbach > Am 09.02.2017 um 08:45 schrieb Kilian Lackhove <lackhove(a)ekt.tu-darmstadt.de>: > > Hi Fabian, > > there are a number of different source term formultations people use for the > APE, so our implementation is "source-term agnostic" in the way that you can > specify whatever RHS you want. > > I dont think we have a tool to compute derived fields (e.g. acoustic source > terms) from other fields (e.g. velocity, pressure, density ...) right now. So > you could either convert the fld files from the CompressibleFlowSolver to pts, > read those with python, matlab, etc., compute the source term, store it as pts > again and feed this into the APESolver. This is what other people do. The > alternative would be writing a FieldConvert module to take care of this. > > Cheers, > > Kilian > > Am Dienstag, 7. Februar 2017, 12:13:30 CET schrieb Selbach, Fabian: >> Dear users, >> >> can somebody send me an example of an APE-Simulation by a source from file >> of another simulation? Is it actually possible to do something like that in >> Nektar++? >> >> To take the base flow from a simulation is easy, but to get the acoustical >> source term is quit difficult and actually not possible, I think. Am I >> right? >> >> Best regards >> >> Fabian >> ________________________________ >> Von: Selbach, Fabian >> Gesendet: Freitag, 27. Januar 2017 16:14 >> An: nektar-users(a)imperial.ac.uk >> Betreff: Source terms and mesh generation >> >> Dear Users, >> >> which kind of mesh generator do you recommend to get a good mesh for an >> airfoil simulation? Actually I am using Gmsh. >> >> My next question deals with the CompressibleFlowSolver in Nektar++. Is there >> a possibility to get acoustical source terms during a simulation with its? >> >> Best regards >> >> Fabian > > -- > Kilian Lackhove, M.Sc. > Fachgebiet für Energie- und Kraftwerkstechnik > L1|08 114 > Technische Universität Darmstadt > Jovanka-Bontschits-Straße 2 > D-64287 Darmstadt > Germany > > Tel.: +49 6151 16 - 28915 > Fax: +49 6151 16 - 6555 > e-mail: lackhove(a)ekt.tu-darmstadt.de

2 3

Question about batch postprocessing
by Vidhi Zala 15 Feb '17

15 Feb '17

Hi all, I'm trying to post-process multiple files using FieldConvert and I run into the following issue (since slack channel is not very active, I posted it on mailing list): If we have multiple *.chk files and want to convert it to vtu/dat format to view in paraview, is there a way FieldConvert will process the files in bulk? I tried to write a script that does it in loop like this: for i in ../../../solvers/ADRSolver/Examples/UnAdvDiff_3DHom1D_*.chk; do ./FieldConvert-g ../../../solvers/ADRSolver/Examples/UnAdvDiff_3DHom1D.xml "${i%.chk}".vtu; done But it does not generate correct vtu files. Currently, I'm restricted to convert each chk to vtu file individually. I also read the module information of FieldConvert and was not able to figure out if there is a clever way to do this. Thanks in advance! Vidhi

3 2

Flow initialization with specific BL solution
by devabratasahoo＠campus.technion.ac.il 15 Feb '17

15 Feb '17

Hi Gianmarco, As per your suggestion, I went through the CompressibleBL.cpp as well as the FalknerSkanBL.cpp routines in order to initialize my case with a specific boundary layer solution. But, I am not clear how to implement it on the Nektar case file. Do I need to compile the BL code (CompressibleBL.cpp/FalknerSkanBL.cpp) separately and insert the output into the .xml file or is there a way to directly define the BL generator as a function of the initial condition section of the .xml file? I couldn't find any example using similar sort of case. Please suggest. Thanking you. Regards, Dev

1 0

Source terms and mesh generation
by Selbach, Fabian 09 Feb '17

09 Feb '17

Dear Users, which kind of mesh generator do you recommend to get a good mesh for an airfoil simulation? Actually I am using Gmsh. My next question deals with the CompressibleFlowSolver in Nektar++. Is there a possibility to get acoustical source terms during a simulation with its? Best regards Fabian

3 3