CFD Online Discussion Forums - Sliding mesh vs MRF in axial turbine simulation

Hello folk, I'm a student and I'm actually dealing with a simulation for estimating the torque in an axial turbine. The turbine could have different blade shapes, and I would evaluate the torque for each of them. The turbine has 20 mm diameter and 8 blades, while the medium is water. The mesh has been generated by Gambit, using the turbo plugin also. In addition to the volumes surrounding the blade I have 2 volumes more: one at the inlet and one at the outlet of the turbine, which have got a considerable lenght. According to the volumes related to the turbine, the inlet and outlet are modelled for 1/8 of the diameter. I created interfaces between the inlet-blade_volume and the blade_volume-outlet. In Fluent I tried to perform analisys with sliding mesh and MRF, leaving the boundaries, solver and other parameter the same for both simulation. The simulations converges in both case, but when I get the torque from turbo report panel, I obtain values which are considerable different. A relative error between the calculated torque in MRF and the calculated torque in sliding mesh is more than 55%. Is it normal? Which is you best suggestion for this kind on simulation? Thanks Vito