Learning Forum

[373352238]

I’m encountering an issue with the MAT_ADD_EROSION model in LS-DYNA and would appreciate your insights. Here’s the problem description:
Setup Material: Concrete
Erosion criterion: MXEPS = 0.8 (intended to delete elements when max principal strain reaches 0.8).
Observed Behavior
Premature Deletion:
During a restarted analysis, a large number of concrete elements are deleted immediately after initialization, even though the max principal strain is only on the order of 1e-3 (far below the 0.8 threshold).
Nearly the entire concrete part is erased prematurely.
Validation Test:
When I removed MAT_ADD_EROSION, the simulation ran normally, and the max principal strain remained very low (~1e-3) in the initial phase, confirming that the erosion criterion—not mechanics—caused the issue.
Question
Why does LS-DYNA delete these elements before reaching MXEPS = 0.8?
Could there be:
An unintended interaction with other erosion criteria?
A restart-related issue (e.g., strain initialization)?
A need for additional checks (e.g., time step stability)?

[igandiko]

Hello, which version of the solver are you running? I would try the latest version. Are you running SMP or MPP solver? Have you tried a simple restart first? Does the restart work without *mat_add_erosion? Can you run the analysis using dynain approach instead of relying on restarts? 

[herrylauu0]

 

 

The MXEPS criterion triggers element deletion when the effective plastic strain in an element exceeds a user-defined threshold. This is intended to simulate material failure, such as fracture or erosion, in dynamic simulations like impact or crash scenarios. Sprunked