I think I have discovered a problem with the FEA interface with NozzlePro. The global loads input for WRC107 should not be the same ones used in FEA analysis. According to Tony Paulin at PRG, NozzlePro applies the loads at the end of the nozzle. Therefore, in order to equal the forces and moments at the shell/nozzle interface the moments to be applied for FEA need to be reduced by the appropriate shear force * nozzle length. I think codecalc should preform this calculation prior to dumping the input to NozzlePro for analysis.