Use of Drysteam as a material in Cfx for hp stage
 

I modeled a hp rotor blade by turbo grid and meshing was done turbo grid. That model was imported to cfx pre. First that model was run with the material as 'water ideal gas' and the results are coming 20% variation with practical results. After I changed the material as drysteam 'steam 5' to then if I see the results are something nonsense. Can anybody help to sortout this problem. What I feel is the mesh, boundary conditions and model are ok why because when the material is water ideal gas it giving appropriate results.Only the problem is with material.Any body who worked in this area. Please help me.