I want to solve a coupled mechanoc-magnetostatic problem by using a nonlinear solver in a staggered scheme. My domain includes tow copper wires and a rectangular iron. All these are located inside vacuum. The rectangular iron is considered as a single supported beam. I solve the PDE of magnetostatic in whole domain, but I want to solve the mechanical problem for the iron beam. This idea works when I use the linear solver form the mechanical problem, but when I use the nonlinear solver, it does not work and any displacement solved for the beam. Does anybody know solve this problem?