I'm trying to run a combo shell solid analysis and keep getting this error. The model is all shells except for the four pins which are solids. The structure has four legs with a clevis on the end which is fitted over a padeye and connected with the pin. I have a no penetration contact with the edge of the hole openings and the face of the pin that works correctly. When I run the analysis to check its behavior up to this point the clevis moves sideways down the length of the pin a couple of inches which is more than the throat on the clevis will allow. This is what I expected to happen and the no penetration contact appears to be acting correctly. I then added a no penetration contact on the two shell surfaces of the clevis ears versus the shell surface of the padeye to allow the ears to move laterally only the amount allowed by the gap in the clevis throat. I remeshed the model successfully. However when I try to run it, it errors out very quickly with "surface based contact normal not found". Did some searching, but I can't seem to find anything that points to a potential cause of this problem. I have run the model as a solid mesh also with the same type of contraints on the clevis and it appears to be working correctly. I want to get this shell model working also so I can get a better understanding of the pro's and con's between the two different model types for this structure. Any ideas?