Description of problem: ========== I have a model with superelements rotated 90 degrees in another coordinate system via the SETRAN command. These superelements are connected to non-superelements via contact elements. (The contact elements are not part of the superelement definition.) After I expand the results, the displacements look reversed (x displacement should be z displacement). Resolution: ========== When transforming superelements with the SETRAN command, the default is to rotate its nodal coordinate systems (NOROT argument of SETRAN command). If the superelement has nodes shared with other elements, these nodes should also be rotated in the same coordinate system – otherwise, the contact elements are using one coordinate system whereas the superelement is using another. This explains why the displacements seem swapped if the rotated coordinate system is 90 degrees from global Cartesian.