Negative volume and complex speed of sound termination 1

[Ed93]

Hello everyone!

I have a problem to simulate an impact onto soil ground (Material Model 5 Soild and Foam) with LS-Dyna, because I am getting negative volumes which lead to complex speed of sound in the material.

My model consists of two layer of soil (Automatic_Surface_To_Surface) and an round impactor. Does anyone have any advice how to solve this problem?

What I tried so far:

- Reduce the timestep scale factor to 0.67 --> no change
- Improve the hourglas formulation (I used type 2,3 and 5) --> same problem, but later error termination.


I am glad for any help and advices!


Greetings
Ed93

 

[Dave442]

If you are happy that your issue is not caused by incorrect units, material properties, constraints or loads:

- Identify the region where elements are turning inside out -> can you improve the mesh in that region to prevent the issue?

- If not, then you could look into 'adaptive re-meshing' to try to avoid this.

 

[btrueblood]

Are you hand meshing? Negative volume (negative Jacobian) implies a nodal numbering scheme that did not follow the assumed element formulation, i.e. used left-handed vs. right-handed numbering.

 

[Dave442]

OP mentioned using LS-Dyna so I assumed the negative volume errors are due to excessive mesh distortion.

This would also make sense for impact studies on soft foam materials.

 

[btrueblood]

Ah, good point.

 

[Ed93]

Hey,

thanks for your answers!

Normaly negative volumes in foams are caused by massive compression. In my case the compression was not that big, so I have been looking for another reason and solved it by enlarging the mesh of the second soil. I think the problem was the contact between the two soils. The nodes where directly above each other and maybe some errors accured cause I used a nodes to surface contact. Now everything works fine, because I have now a little offset between the nodes.

Thanks again for your help and fast replies!