Loss Model Application To Through Flow Code For Low Pressure Turbine And Aerodynamic Optimization

The S2stream surface calculational method is widely applied to turbomachinary,while its accuracy depends on the loss model largely. The loss models used today areall empirical formula based on large number of cascade experiment data, whichindicates the limitation of their application. Meanwhile, the S2method takes animportant place in the profile aerodynamic optimization work, for its time-saving andhigh accuracy. So, it is significantly to investigate the application of different lossmodels to S2method.The five main control equations and one surface equation of S2method based onEuler was introduced firstly, and then the difference scheme was given. The generationand development of loss models was reviewed. The coefficienct assigned method, thethree A-M models, the K-O model, the M-B model, the A-A model, the BTИ model,the ЦИАМ model and the HIT model were described in detail correspondly, and theapplication of these loss models to a five-stage low presure turbine was shown in theaspect of gas flow angle, total pressure loss coefficient, sub-loss, reaction, etc. And thecomparison of these results with the three-dimensional viscous result was carried out.The loss of coefficient distribution method depends on the manual given entirly. TheA-M model and BTИ model computes larger than the CFD result. The loss distributionof K-O model is quite approprate, while the error of its value is larger. The M-B modeland A-A model have a better fitness. The ЦИАМ model was better used inshort-blade-stage than that in long-blade-stage. The loss of HIT model distributeduniform, and its value is credible.The profile optimization work based on M-B model, iSIGHT and AutoBlade wasconducted, and it took metal angle as variables, isentropic stage efficiency as targetfunction, and massflow as constrains. Numeca was applied to verify the optimizationalprofile result. There is a variation about0.08°~7.04°of each section geometric turningangle of before and after optimization. The efficient has an increase of1.45%of S2method, while massflow decrease0.63kg/s.