contact penetrating the shell rigid body surface

[jisb]

Hi,

I am running abaqus/explicit, with deformable body confined within a rigid surface formed by shell elements. After running and viewed from CAE, I can see that the deformable material goes outside the shell surface during the dynamic event. But how could this happen? Initially the deformable body was entirely within the shell surface, shouldn't Abaqus contact formulation force the deformable body be on one side of the rigid surface all the time? If this is not something strange in abaqus, would this affect any modeling accuracy? Thanks very much!

Sean

 

[corus]

Check the normals on your surfaces and make sure that there is no initial overclosure of the surfaces.

corus

 

[jisb]

That's an important point, but -- Since in the initial configuration, the deformable body and the rigid body are very close to each other, when meshing, there was inevitably some amount of initial overclosure, which was subsequently adjusted by abaqus pre-processor. However, even so, there's still some unresolved initial overclosure stored as offsets. Do you think this is the problem? Can I replace the shell elements into solid elements (of course, need to replace the surface into a solid in geometry as well), to avoid this problem? Much appreciated!

Sean

 

[corus]

If you have some initial overclosure then one has passed through another before the job has started.
You may find it easier to use analytical rigid surfaces than using rigid shell surfaces, if it's possible. I believe there is some smoothing of the rigid shell surfaces but they look facetted to me and look as if they could cause problems when contacting another facetted deformable element.

corus