- reverts commit d296732d - the scaling it correct
- there is a bug in Elmer that the heat source is not aware of the scaling
- write the scaling directly to the solver, not to the mesh itself. (make in principal no difference but we are closer to the solver)