I agree with caviac, sounds like the the solid has deformed too much. This could be caused by the element being a bad shape to begin with. It could also be caused by a time step that is too big. I'm guessing that dt=1e-6 not 1e+6 as written above. I think that ls-dyna calculates the time step based on the time it takes a longitudinal wave to travel accross the smallest element (so dt is based on E, density, poisson's and element geometry) and does not base it on stable time step size for a contact interface or the rate that a load is applied. So if you have big, soft elements and run them into a wall at high velocity they will invert. If this is a contact problem look at how far your part will travel in one time step and compare this to your element size. If it looks like dt is too big, reduce the time step scale factor. The problem could also be a result of dyna using 1 point integration, you can try modifying the hourglass stabilization method on the material card. Last idea is to write out your d3plot files at a high frequency so you can see what is happening.
Your description of the problem is brief so it's hard to say what's happening, but hopefully this helps.
Thanks for the replies.
I have got it working by changing the element formulation for the solid component from reduced Integration formulation(1) to Fully integrated formulation (2).
greetings
this problem usually incountered with soft materials such as foam *mat 57, change to *mat rigid type 20 or dig deep for proper contact, the maen reson i believe is could be due to adjacent materials of very diiferent stiffness.
hanibal