Help with stream tracer

I am having some problems calculating path trajectories. I am using the “Stream Tracer” filter with a 3D vector field and seeds along a line, and comparing it to a vizualisation of the field using “Surface LIC” in the same extracted subset plane. I thought the path trajectories would closely follow the LIC streaklines, but they do not. I am sure the path trajectories are incorrect - they seem to go into my solid boundaries (where the flow is zero), see pasted image - but I cannot figure out what I am doing wrong.

Any help appreciated.

I agree that the stream tracer output looks suspect. However, I’m not sure anyone can diagnose the problem by just looking at your image. Would it be possible for you to post some example data?

Hi, thanks for your help. I’ve uploaded the pvsm and one of the netcdf input files here. The other file is a little larger and I’ve posted it here: https://www.dropbox.com/s/6364g1qvpmnj0se/uvel.nc?dl=0.

depth.nc (549.9 KB) leeWave.pvsm (1.2 MB)

Hmm. This might be a bug in the stream tracer with rectilinear data (axis-aligned with uneven spacing).

Try running the data from uvel.nc through the Transform filter. Keep the default parameters. The data will not actually move, but it will be converted to a curvilinear mesh. That seems to work better with the stream tracer.

Thank you - adding the transform filter solved it.

I also tried the data on a uniform grid (i.e. I interpolated onto an evenly spaced grid before loading into ParaView), but that didn’t seem to fix it.

The figure is looking much better now (the simulation needs work, but the visualization is good) - thanks again for your help.

@utkarsh.ayachit If you get a chance, can you take a quick look to see why the stream tracer is giving different answers for structured data and rectilinear (or apparently image) data?