Problem with ChanStability_Coupled.xml
Dear All, I'm a new user of Nektar++. I would like to carry out some stability investigations. I have two questions in regard to the usage of the software. *1*. I wanted to compare the eigenvectors calculated by the unsteady and steady Navier-Stokes solvers, but I encountered a problem. I tried to run nektar++-4.0.1/solvers/IncNavierStokesSolver/Tests/ChanStability_Coupled.xml, but I got the following error: /"// //IncNavierStokesSolver ChanStability_Coupled.xml// //// //Fatal : Level 0 assertion violation// //Base flow must be defined for linearised forms.// //"/ Then I defined the BaseFlow as: /<FUNCTION NAME="BaseFlow">// // <E VAR="u" VALUE="-y*y+1" />// // <E VAR="v" VALUE="0" />// //</FUNCTION>/ The simulation was carried out in this case, but the results are wrong: /"IncNavierStokesSolver ChanStability_Coupled_mod.xml// // //=======================================================================// // Solver Type: Coupled Linearised NS// //=======================================================================// // Arnoldi solver type : Arpack// // Arpack problem type : LI// // Single Fourier mode : false // // Beta set to Zero : false // // Shift (Real,Imag) : 0,0// // Krylov-space dimension : 16// // Number of vectors : 2// // Max iterations : 500// // Eigenvalue tolerance : 1e-06// //======================================================// //Initial Conditions:// // - Field u: from file ChanStability_Coupled.rst// // - Field v: from file ChanStability_Coupled.rst// //Writing: "ChanStability_Coupled_mod_0.chk"// //Matrix Setup Costs: 2.37365// //Multilevel condensation: 1.29072// // Inital vector : input file // //Iteration 16, output: 0, ido=99 // //Converged in 16 iterations// //Converged Eigenvalues: 3// // 0 0 0 -nan -nan// //Writing: "ChanStability_Coupled_mod_eig_0"// // 1 0 0 -nan -nan// //Writing: "ChanStability_Coupled_mod_eig_1"// // 2 0 0 -nan -nan// //Writing: "ChanStability_Coupled_mod_eig_2"// //L 2 error (variable u) : 2.589// //L inf error (variable u) : 1// //L 2 error (variable v) : 0// //L inf error (variable v) : 0// //"//// /Maybe I misunderstood something, could you provide help with this issue? *2*. Is it somehow possible to export the pressure field of the eigenvectors or should I calculate it by the Poisson equation? In this case could you send me an example how can I calculate the derivatives of a field variable. I saw the function FldAddScalGrad but I don't know how can I use. I really appreciate any help. Best regards, Péter <mailto:nektar-users@imperial.ac.uk>
Dear Péter Unfortunately, I am not entirely sure stability analysis with our coupled (steady) solver is working correctly, since all the regression tests related to it seem to have been disabled. This is probably something we need to investigate more thoroughly. Cheers, Chris On 07/07/15 10:03, Péter Tamás Nagy wrote:
Dear All,
I'm a new user of Nektar++. I would like to carry out some stability investigations. I have two questions in regard to the usage of the software. *1*. I wanted to compare the eigenvectors calculated by the unsteady and steady Navier-Stokes solvers, but I encountered a problem. I tried to run nektar++-4.0.1/solvers/IncNavierStokesSolver/Tests/ChanStability_Coupled.xml, but I got the following error: /"// //IncNavierStokesSolver ChanStability_Coupled.xml// //// //Fatal : Level 0 assertion violation// //Base flow must be defined for linearised forms.// //"/ Then I defined the BaseFlow as:
/<FUNCTION NAME="BaseFlow">// // <E VAR="u" VALUE="-y*y+1" />// // <E VAR="v" VALUE="0" />// //</FUNCTION>/
The simulation was carried out in this case, but the results are wrong:
/"IncNavierStokesSolver ChanStability_Coupled_mod.xml// // //=======================================================================// // Solver Type: Coupled Linearised NS// //=======================================================================// // Arnoldi solver type : Arpack// // Arpack problem type : LI// // Single Fourier mode : false // // Beta set to Zero : false // // Shift (Real,Imag) : 0,0// // Krylov-space dimension : 16// // Number of vectors : 2// // Max iterations : 500// // Eigenvalue tolerance : 1e-06// //======================================================// //Initial Conditions:// // - Field u: from file ChanStability_Coupled.rst// // - Field v: from file ChanStability_Coupled.rst// //Writing: "ChanStability_Coupled_mod_0.chk"// //Matrix Setup Costs: 2.37365// //Multilevel condensation: 1.29072// // Inital vector : input file // //Iteration 16, output: 0, ido=99 // //Converged in 16 iterations// //Converged Eigenvalues: 3// // 0 0 0 -nan -nan// //Writing: "ChanStability_Coupled_mod_eig_0"// // 1 0 0 -nan -nan// //Writing: "ChanStability_Coupled_mod_eig_1"// // 2 0 0 -nan -nan// //Writing: "ChanStability_Coupled_mod_eig_2"// //L 2 error (variable u) : 2.589// //L inf error (variable u) : 1// //L 2 error (variable v) : 0// //L inf error (variable v) : 0// //"////
/Maybe I misunderstood something, could you provide help with this issue?
*2*. Is it somehow possible to export the pressure field of the eigenvectors or should I calculate it by the Poisson equation? In this case could you send me an example how can I calculate the derivatives of a field variable. I saw the function FldAddScalGrad but I don't know how can I use.
I really appreciate any help.
Best regards, Péter
<mailto:nektar-users@imperial.ac.uk>
_______________________________________________ Nektar-users mailing list Nektar-users@imperial.ac.uk https://mailman.ic.ac.uk/mailman/listinfo/nektar-users
-- Chris Cantwell Imperial College London South Kensington Campus London SW7 2AZ Email: c.cantwell@imperial.ac.uk www.imperial.ac.uk/people/c.cantwell
Dear Peter, I have now fixed the ChanStability_Coupled.xml problem in a branch called fix/INSCoupledStabilitySolver It seems this regression test had been commented it. We then altered how the driver was iterating the coupled solver and so have added a specialised Forcing term which feed in the previous solution as the last iteration of the Coupled solver. I have not yet managed to get the 3D coupled solver running but hope to look at that in the near future before we can get this merged back into Master. Best regards, Spencer. On 7 Jul 2015, at 10:03, Péter Tamás Nagy <pnagy@hds.bme.hu<mailto:pnagy@hds.bme.hu>> wrote: Dear All, I'm a new user of Nektar++. I would like to carry out some stability investigations. I have two questions in regard to the usage of the software. 1. I wanted to compare the eigenvectors calculated by the unsteady and steady Navier-Stokes solvers, but I encountered a problem. I tried to run nektar++-4.0.1/solvers/IncNavierStokesSolver/Tests/ChanStability_Coupled.xml, but I got the following error: " IncNavierStokesSolver ChanStability_Coupled.xml Fatal : Level 0 assertion violation Base flow must be defined for linearised forms. " Then I defined the BaseFlow as: <FUNCTION NAME="BaseFlow"> <E VAR="u" VALUE="-y*y+1" /> <E VAR="v" VALUE="0" /> </FUNCTION> The simulation was carried out in this case, but the results are wrong: "IncNavierStokesSolver ChanStability_Coupled_mod.xml ======================================================================= Solver Type: Coupled Linearised NS ======================================================================= Arnoldi solver type : Arpack Arpack problem type : LI Single Fourier mode : false Beta set to Zero : false Shift (Real,Imag) : 0,0 Krylov-space dimension : 16 Number of vectors : 2 Max iterations : 500 Eigenvalue tolerance : 1e-06 ====================================================== Initial Conditions: - Field u: from file ChanStability_Coupled.rst - Field v: from file ChanStability_Coupled.rst Writing: "ChanStability_Coupled_mod_0.chk" Matrix Setup Costs: 2.37365 Multilevel condensation: 1.29072 Inital vector : input file Iteration 16, output: 0, ido=99 Converged in 16 iterations Converged Eigenvalues: 3 0 0 0 -nan -nan Writing: "ChanStability_Coupled_mod_eig_0" 1 0 0 -nan -nan Writing: "ChanStability_Coupled_mod_eig_1" 2 0 0 -nan -nan Writing: "ChanStability_Coupled_mod_eig_2" L 2 error (variable u) : 2.589 L inf error (variable u) : 1 L 2 error (variable v) : 0 L inf error (variable v) : 0 " Maybe I misunderstood something, could you provide help with this issue? 2. Is it somehow possible to export the pressure field of the eigenvectors or should I calculate it by the Poisson equation? In this case could you send me an example how can I calculate the derivatives of a field variable. I saw the function FldAddScalGrad but I don't know how can I use. I really appreciate any help. Best regards, Péter <mailto:nektar-users@imperial.ac.uk> _______________________________________________ Nektar-users mailing list Nektar-users@imperial.ac.uk<mailto:Nektar-users@imperial.ac.uk> https://mailman.ic.ac.uk/mailman/listinfo/nektar-users Spencer Sherwin McLaren Racing/Royal Academy of Engineering Research Chair, Professor of Computational Fluid Mechanics, Department of Aeronautics, Imperial College London South Kensington Campus London SW7 2AZ s.sherwin@imperial.ac.uk<mailto:s.sherwin@imperial.ac.uk> +44 (0) 20 759 45052
participants (3)
-
Chris Cantwell
-
Péter Tamás Nagy
-
Sherwin, Spencer J