IRstuff is right,
as you have an physical experiment, not an mathematical model, You will need to use experimental suited optimisation technique. The problem is with proper preparation of the experiment table. This is what he mean by DOE. For this You need to consider all your inputs and...
I would suggest using joints rather contact. But If you insist, try mark all your contact nodes as master (on the SE boundary), then Your SE will have them visible and contact (i.e. MPC) should be allowed.
JTMW,
Workbench is acting different than ANSYS GUI, If I get you right only way to read your model could be FE-Modeller - a FEA translation tool which is part of Workbench platform.
But really I dont see the reason why translate model to Workbench when it is made in ANSYS ?
Depending what You want to model but for click-clack,
I would suggest using rigid body module (You will find it in Accademic licences) and model Links as a joints. Contact can be adjusted easily for rigid bodies as frictionless.
