We have a large frame structure consisting of tubing and plates. We originally just used the 'global bond' for contact, thinking this would give us the result/information we needed in displacement. When our model was built on the floor, there was a great difference in variance of our result and the actual displacement of the frame. We had another company come in take the same cad model, into another software (their own), run the same gravity load, and fixed conditions, and they were very close to the "real life" result. HELP? The only difference we can see in our study set up is the "bonding" or "contact" issue. We went back and made it all non-penetrating, and made "bonded -edge to surface contacts" to simulate the weld contacts. Ran the study, and it still seems to be seeing the whole thing as globally bonded.
ANY IDEAS? Has anyone else run into this? These are critical systems. (Our result is seeing .002 inches displacement, and in real life and in the other software, it is closer to .015 inches of displacement.)
Is there something we are missing in our setup for the contacts?