Waiting for a simulation is a good way to become interested in CPU settings. Before changing them, though, look at the mesh and the amount of output being written.
An unexpectedly tiny element or unnecessary output can explain some of the waiting. Keep a baseline whose result you understand so a faster trial has something useful to compare against.
Find the small elements
Display Characteristic length in the relevant states. Did a fine region come from the initial mesh, outline refinement or adaptive settings?
Some detail may be essential. Some may be an unintended effect of another choice. Locate it before reducing mesh quality across the whole model.
The explicit time-step post explains why small elements matter. For a quicker exploratory setup, also read feasibility versus validation.
Look at what is going to disk
Review saved states and contours in Control. Keep what you need to investigate the result, but reconsider output with no purpose in this trial.
Restart files also take space. Use the restart guide before removing them so you understand which recovery route you are giving up.
Then look at CPU Control
Speed-up and mass scaling are different controls. Their interactions deserve attention rather than treating every faster option as equivalent.
For rate-dependent materials, tool speed needs particular care. The documented velocity-correction example uses 0.1 for a numerical speed of 10 m/s and an intended effective speed of 1 m/s. That explains the relationship; it is not a value to copy into every model.
Compare runtime and result together
Record the host, core count and changed settings. Compare equivalent result states using the same legend and relevant shape or force measurements.
There isn’t one best core count for every model. Measure it on your setup. If you publish the comparison, show both the time and what happened to the result. “It ran faster” becomes much more useful when the reader can also see what changed—or stayed sufficiently close for the study’s purpose.
Previous: Measure Draw-in and Sections in PAM-STAMP 2025
Next: Set Up a Tailored Blank in PAM-STAMP 2025
[…] Previous: Reduce PAM-STAMP 2025 Runtime with Controlled Changes […]