- Nektar-users - imperial.ac.uk

Nektar++ mailing list transition to new Nektar++ Forum
by Chris Cantwell 03 Jun '26

03 Jun '26

Dear Nektar++ users, We are in the process of transitioning this mailing list to a web-based forum (using Discourse). This will provide a more modern interface and additional features such as categories and tags which will enable posts to be more easily found and organised. Email notifications will continue to be available and can be configured to each user's preference. More information will also be shared during the workshop next week. The new forum is already live at https://forum.nektar.info. We encourage you to sign up and use it when you have your next Nektar++ question! While we will soon be disabling the mailing list for new posts, we will keep the archive available for reference for as long as we can. Kind regards, Chris -- Chris Cantwell Reader in Computational Engineering Director of Postgraduate Studies (PhD) Department of Aeronautics Imperial College London South Kensington Campus London SW7 2AZ Tel: +44 (0)20 759 45050 Email: c.cantwell(a)imperial.ac.uk www.imperial.ac.uk/people/c.cantwell

1 0

Important: Nektar++ workshop date change (8th-10th June 2026)
by Moxey, David 11 May '26

11 May '26

Hi all, A quick follow-up to my previous email on the Nektar++ workshop. We are trying to manage a clash between our workshop and another event that has been scheduled for the same week, but after we originally chose dates for the workshop. The least disruptive option was for us to adjust the dates by one day for the workshop, so the new dates are 8-10th June 2026. Apologies for the adjustment -- I hope this is sufficiently far out so that it doesn't affect anyone's travel plans, but if there is an issue please let me know! Many thanks, Dave -- Prof David Moxey, PhD MMath FHEA Professor of Computational Engineering & UKAEA/RAEng Research Chair david.moxey(a)kcl.ac.uk | web: davidmoxey.uk Department of Engineering, King's College London, Strand, London, WC2R 2LS.

1 2

Conversion of large .xml mesh file to HDF5 format.
by Nande, Ishan A. 07 May '26

07 May '26

CAUTION: This message came from outside Imperial. Do not click links or open attachments unless you recognise the sender and were expecting this email. Hello, I am trying to convert a large Nektar++ XML mesh file to HDF5 format using : NekMesh large_mesh.xml large_mesh.nekg The mesh size is of the order of (10^6) elements, and the conversion is failing. I suspect it may be due to a memory issue while reading/parsing the large XML file. I have also tried to run this conversion in parallel mode using srun --mpi=pmix(the HPC setup uses slurm), but there are a lot of warnings/errors printed on the screen. However, I see that in this parallel run there is a .nekg output and I am not sure whether it is correct. Has anyone encountered this before, or is there a recommended workflow for converting very large XML meshes to HDF5 without requiring extremely large memory? Any advice would be appreciated. Regards, Ishan PGR

1 0

Time information in ".chk" --> ".dat"
by Nande, Ishan A. 24 Apr '26

24 Apr '26

CAUTION: This message came from outside Imperial. Do not click links or open attachments unless you recognise the sender and were expecting this email. Hello, I switched to Nektar++ v5.9.0 some time ago and I am currently having an issue with .chk files when using the Linearized Euler Equations solver. I am using the Checkpoint filter to save .chk files at the desired output frequency. I then convert these .chk files to Tecplot .dat format using the standard FieldConvert workflow. The conversion itself appears to work correctly, and the field data can be opened in Tecplot. However, when I load the resulting .dat files into Tecplot, I do not see any time information associated with them. As a result, I am unable to create animations, because the .dat files are not treated as sequential time instances; instead, each newly loaded file appears to replace the previously loaded one. I also checked a single .dat file corresponding to a later physical time instance, but Tecplot still reports the solution time as 0 s, which is not correct. As a quick check, I opened the .dat file in Notepad++ and noticed that the file header does not appear to contain any time information. I am not sure whether this is the source of the problem, but it seems related. Could you please advise whether the checkpoint files should contain physical time information after conversion to Tecplot format, or whether an additional option is required in FieldConvert to preserve/export this metadata? Any help would be greatly appreciated. Regards, Ishan PGR

3 5

Writing .fld/.chk file from Python filter
by Nande, Ishan A. 06 Mar '26

06 Mar '26

******************* This email originates from outside Imperial. Do not click on links and attachments unless you recognise the sender. If you trust the sender, add them to your safe senders list https://spam.ic.ac.uk/SpamConsole/Senders.aspx to disable email stamping for this address. ******************* Hello, I am trying to set up a Python filter in V5.9.0 to perform some on-the-fly calculations on a 2D domain. I have been able to extract the coordinates, the solution variables' datasets and now, I want to write the calculated quantity in a .fld/.chk/HDF5 file. Can this be done ? Any info will be helpful. I have looked into the Demos/Python folder of the source code but that wasn't of much help. Regards, Ishan PGR

1 0

Fixed Flow Rate and Fixed Pressure Gradient Constraints
by sshokran＠uwo.ca 30 Jan '26

30 Jan '26

Dear all, I am studying 2D channel flow with spatially periodic wall modification, where a sinusoidal streamwise velocity profile is imposed at the lower wall. I first solve the problem using a fixed flow-rate constraint and compute the corresponding mean pressure gradient. I then use this pressure gradient in a separate simulation with a fixed pressure-gradient constraint to recover the flow rate. While this procedure works correctly for the baseline Poiseuille flow (no wall modification), it fails when the wall forcing is applied: the recovered flow rate does not match the prescribed one. I would appreciate any insight into possible causes of this inconsistency or guidance on how to implement this comparison correctly in Nektar++.

1 0

how to implement active flow control
by Ben 15 Dec '25

15 Dec '25

******************* This email originates from outside Imperial. Do not click on links and attachments unless you recognise the sender. If you trust the sender, add them to your safe senders list https://spam.ic.ac.uk/SpamConsole/Senders.aspx to disable email stamping for this address. ******************* Dear all, I am a novice in fluid simulation. Ask nektar++how to implement active flow control, such as injection and suction. Can you provide me with a classic case for my reference? Everyone can offer their opinions.Looking forward to your reply！

2 1

Nektar++ 5.9.0 Released.
by Chris Cantwell 15 Nov '25

15 Nov '25

Dear all, We are pleased to announce the release of Nektar++ v5.9.0. As always, if you encounter any problems please feel free to contact us on the mailing list nektar-users(a)imperial.ac.uk. You can also report bugs and issues on our issue tracker (https://gitlab.nektar.info/nektar/nektar/-/issues) This release includes a range of new features and improvements over the 5.8.0 release: **Library** - Fix BwdTrans and IProduct in Collection and MatrixFee for variable P (!2178) - Add vectorized min/max support to AVX (!2179) - Fix incorrect symmetry flux for interior penalty methods in DG (!1872) - Fix bug in ExpList (!2182) - Remove deprecated TimeIntegrationMethod in SolverInfo (!2061) - Some fix for PhysInterp1D (!2081) - Add 3D rotating mesh capability (!2184) - Use std::unordered_map instead of std::map for NekFactory (!2103) - Replace `std::shared_ptr` with `std::unique_ptr` inside SpatialDomains (! 2112, !2222) - Fix output messages clobbering error info in Lagrangian filter (!2200, !2202) - Unified IProductWRTDerivBaseOp align vector function for all 3D shapes (!2203) - Added a formula for the number of modes in a tetrahedron (!2205) - Added boundary condition and periodic boundary reader (!2221) - Remove duplicate IMEX Gear time-integration variant (!2225) - Fix bug in Neumann boundary condition for LinearAdvectionDiffusionReactionSolve (!2231) - Fix MatrixFree BwdTrans Tet kernel and update collection unit tests for non- homogeneous inputs (!2251) - Use MPI_Comm_split instead of MPI_Cart_create and MPI_Cart_sub (!2250) - Fix handling of PointGeom and PointExp for Collections and Fields (!2252) - Tidy-up GMRES verbose output (!1804) - Add a copy constructor method for boundary expansions and do not call partitioner in hdf5 input mesh if run in serial (!2241) - Remove duplicate StdPhysDeriv overload function (!2254) - Add EvaluateAtPoint interface in Equation (!2255) - Enable GenMatrix routine for SegExp (!2260) - Enforce const consitency in Comm class (!2261) - Add GetEquation to BoundaryCondition and ImposeNeumann/RobinConditions (!2257) - Adjustments to `NekLinSys` classes for future multigrid capability (!1811) - Avoid multiple data copy in NekFFTW (!2268) - Add FieldConvert modules to perform force decomposition for incompressible flows (!1531) - Deprecate use of IO_CheckSteps in favour of CheckpointFilter (!2197) **ADRSolver** - Add implicit UnsteadyAdvection for continuous Galerkin (!2231) - Fixed performance regression after removal of eVECTOR_WRAPPER in !1848 (!2239) **CardiacEPSolver** - Add script for converting CellML models into Nektar++ cell model code (!2185) **NekMesh** - Minor Bug-fix for industrial pipeline !2057 (!2135) - Scaling the mesh nodes (!2115) - Improved Isoparametric Splitting for O-type Prisms "bl" legacy, which kept as "blold" (!1757) - Read Gmsh V2.2 meshes without Physical Boundaries specified. Use CAD ID tag. (!2278) **CI and Packaging** - Disable CWIPI on Fedora (!2199) - Update Annulus test to run more efficiently (!1220) - Add testing and packaging for Debian Trixie (!2249) **FieldConvert** - Added a comprange feature to just select element touching a composite surface for more efficient boundary processing (!2241) Kind regards, Chris -- Chris Cantwell Reader in Computational Engineering Director of Postgraduate Studies (PhD) Department of Aeronautics Imperial College London South Kensington Campus London SW7 2AZ Tel: +44 (0)20 759 45050 Email: c.cantwell(a)imperial.ac.uk www.imperial.ac.uk/people/c.cantwell

1 0

Nektar++ workshop 9-11th June 2026 - save the date
by Moxey, David 13 Nov '25

13 Nov '25

Hi all, This is just a quick email with a date for your diary for the 2026 Nektar++ workshop, which will be held at King's College London between 9-11th June 2026. We will be in touch closer to the date with further details, but at present we plan to hold the event in a similar fashion as in previous years. Many thanks, Dave -- Prof David Moxey, PhD MMath FHEA Professor of Computational Engineering & UKAEA/RAEng Research Chair david.moxey(a)kcl.ac.uk | web: davidmoxey.uk Department of Engineering, King's College London, Strand, London, WC2R 2LS.

1 0

Pressure boundary conditions for velocity correction scheme
by Michałowski Krzysztof (STUD) 16 Oct '25

16 Oct '25

******************* This email originates from outside Imperial. Do not click on links and attachments unless you recognise the sender. If you trust the sender, add them to your safe senders list https://spam.ic.ac.uk/SpamConsole/Senders.aspx to disable email stamping for this address. ******************* Hello, I am a student at Warsaw University of Technology and I will be using Nektar++ for my masters thesis. I have a question about pressure boundary conditions for velocity correction scheme. To ask this question i had to write down some formulas that is why it was more conviniant to attache it to this message as a pdf document. Best Regards, Krzysztof Michałowski

5 6