Anybody have experience with NASTRAN's glue element approach of connecting surfaces with dissimilar meshes? They seem to be working out well in static stress models, but for dynamic models, I am having a lot of trouble getting 6 rigid body modes. I am working with shell elements, and some of...