Thing went on and I found something to solve this question, and I just want to share with you my experience: PMSM (single phase) was too simple for my project, a better solution is FEM-Parametrized Rotary Actuator.
Whith this block, you can provide a flux mapping table .. phy = f(i;theta) (flux function of current and rotor position). I get the tables using FEMM and a matlab scrpit, but I think there could be some example tables.
Whith this solution .. I get the cogging torque, and for example, if I set a random initial position, the rotor, in the simulation, move to the minimum reluctance position without any current at the input.