Hi,
The questions look quite general. Here is what I think.
In terms of some essential properties that you may want to extract from a simulation, well, it depends on your needs. As you noted, the normal and shear stress can be important, and then maybe the strain rate, etc. People may have different ways of deriving them in the case of granular material.
As for processing and visualizing the data, in my case I always write my own scripts to do that, because the way I need them to be massaged tends to be specific. Granted, ParaView can visualize both point data and volumetric data, but I tend to finish all the post-processing work before shipping the data to ParaView, instead of attempting to combine them in ParaView, or any other specialized visualization tool, for that matter.
Writing scripts for post-processing is simple, you can use Matlab or Python. For example, if you have force data as a point cloud, then you can
use scipy to subsample them at the locations of your granular particles and associate the result with each particle, then output these data-infused particles to a CSV file, then visualize it in ParaView. That is the easiest way for me.
We can get more creative. An example in DEME is the Indentation demo, where in the simulation script, we directly calculate the particle strain based on the difference of particle locations between 2 frames, and ask the solver to stream that information into the output file directly (I was too lazy to bother post-processing there). It might be more difficult to pull off, but what I want to say is that what you do to the data obtained from the simulation is up to you, and it is easier and more explicit to use scripts to do that.
Thank you,
Ruochun