There were a couple of problems here originally.
1. One analysis was a modal, the contact is ignored hence unexpected results.
2. The results scaling of the static contact study was a little enthusuastic and the components appeared to pass through each other.
