Do I need to have a unique inp file for each load case? I tried multiple *STEP cards, however I think the results from previous steps get carried across because subsequent steps do not converge this way.
You can use:
*STEP, PERTURBATION *STATIC
for this. It will work as a series of independent steps (equivalent to
*LOAD CASE in Abaqus).
What does PERTURBATION do to a series of static steps? The manual says it doesn’t make sense in that case.
If they are linear (no NLGEOM or nonlinear material/etc.) then I don’t think it makes sense for previous step’s results to be carried through anyway since they’d be obliterated by the next step so you shouldn’t need any special parameters.
Another way to do independent steps is to put each one in a separate file with *INCLUDE to use a common model definition.
Multiple static perturbation steps are used in Abaqus as an alternative to load case feature (not available in CalculiX). Assuming that they work in the same way in CalculiX, they could be useful, at least in terms of analysis time and ensuring linearity.