question regarding post-processing for LAMMPS output

Hi all,

After running the LAMMPS to obtain the output, I can easily view it with OVITO. However building with USER-VTK doesn’t seem to work 100% (I can revisit it in about an hour or two).

Again, I don’t know if this question is right to be posed here, but I want to visualise the simulated flow through some form of either isosurface or contour - just to present a more ‘continuous’ solution instead (I know this contradicts the motivation of LAMMMPS… sorry to be the traitor…). Which framework is recommended? Is it vtk output -> Paraview -> some form of rendering there? I have heard of Blender but would it work as well?

Thanks,
Quang

Hi all,

After running the LAMMPS to obtain the output, I can easily view it with
OVITO. However building with USER-VTK doesn't seem to work 100% (I can
revisit it in about an hour or two).

vtk libraries come in different generations/versions requiring a
different set of libraries to link to.

Again, I don't know if this question is right to be posed here, but I want
to visualise the simulated flow through some form of either isosurface or
contour - just to present a more 'continuous' solution instead (I know this
contradicts the motivation of LAMMMPS... sorry to be the traitor...). Which
framework is recommended? Is it vtk output -> Paraview -> some form of
rendering there? I have heard of Blender but would it work as well?

this is very special requirement, and i am not sure, if that is
actually practical. what you are looking for is a good match for
continuum models, but at the scale of where LAMMPS is typically
operating with particle simulations, you will have very noisy
"density" data unless you do significant spatial and temporal
averaging. since the resulting gridded data, will typically require
much more storage as particle based data, i would recommend looking
into writing some post processing tool.

that said, VMD already has a tool/plugin included (volmap) that can
convert particle data into grids with different rules that can be
rendered as (static) isosurface or saved as (IBM data explorer format)
.dx files and it has a very efficient isosurface renderer (quicksurf)
for generating a simplified particle density isosurface on-the-fly.

if you want to go beyond that, you are definitely in the realm of
ParaView, VisIt and alike. personally, i am quite content with doing
particle based representations, they can be made to be very
informative and good looking visualization.

axel.