Postprocessing PHAST output - ParaView?

[Leo]

Hello,

I appreciate that PHAST writes a hdf file that can be read by modelviewer. Using this tool, we can create some insightful figures. Also, there are a few options to also output machine readable text files that can be further used to visualize e. g. to map the modeled groundwater level.

Trying to find a way to visualize streamlines from a velocity vector field (vel.xyz), I came across ParaView, an open source program used widely in CFD simulations. I managed to load the output file from phast and generate stream lines.

However, I was thinking, why not use this tool to visualize all of PHAST's output. ParaView has the capability to read hdf files but it seems that the structure of PHAST's hdf output file is not compatible. So far, I don't understand what makes a hdf file a hdf file and if it is possible to easily convert this to a readable format. The advantage of Paraview is, that it offers much more features to display and analyze the simulation results than modelviewer.

Does anyone know of a workaround to get the hdf file loaded in Paraview OR is it possible to adjust PHAST's output file(s) so it could be read more easily?

Thanks

[dlparkhurst]

We have looked at ParaView in the past, but did not make an interface. We might do a little checking to see how hard it is to make a file that ParaView can read.

There is a program phasthdf.exe distributed with the batch version of PHAST that might be useful to you. Normally, it is installed in C:\Program Files\USGS\phast-3.7.15749\bin. The program allows you to extract data to a .csv file. You can select any combination of x, y, z, time, scalars, and vectors.

The program needs a Java Runtime Environment. If you get a warning message, it is out of date, but the following corrected warning includes the information to download the Corretto RTE.

"Sorry, phasthdf could not locate a suitable Java Runtime Environment.\nPlease download and install Amazon Corretto 8 x64 JDK from https://aws.amazon.com/corretto."

[Leo]

Thanks for your reply. I use the hdf exporter sometimes and find it a useful tool for further analysis. However, I tend to use the provided xyz files that allow easy automation of postprocessing tasks (which the hdf exporter and the hdf file do not).

[Leo]

I came across another possibility, that might work: providing a xdmf file which "explains" how the data in the *.h5 file is structured. So far I couldn't manage to successfully write such file nor fully understand the structure of the hdf file which is written by PHAST.

I would be happy about some advice if anybody has prior experience with hdf/xdmf or vtk files used to visualize PHAST (or other) simulation results.

[dlparkhurst]

Xdmf might be a good way to go. We'll look into it.