I think we're on the same page, then.

Friction causes a difference in system behavior, and that system behavior dictates the loads on restraints, but it's not fully accurate to state that the loads on restraints are necessarily exclusively to friction.

Friction will typically cause the centroid of thermal expansion to move in a system with sufficient degrees of freedom.

If you remove sufficient degrees of freedom, you will get the correct answer that LX-LY is exclusively friction loads.