%0 Journal Article %A Payandehdoost, M. %A Amanifard, Nima %A Naghashnejad, M. %A Deylami, Hamed Mohaddes %D 2014 %I Begell House %K film cooling, numerical analysis, heat transfer coefficient, turbine blade, GMDH-type neural network, turbulence %N 7 %P 643-657 %R 10.1615/HeatTransRes.2014007180 %T ROBUST MODEL FOR PREDICTING THE AVERAGE FILM COOLING HEAT TRANSFER COEFFICIENT OVER A TURBINE BLADE BASED ON THE FINITE VOLUME STUDY %U https://www.dl.begellhouse.com/journals/46784ef93dddff27,39f34fa02f25eb93,53ae7300634eed4a.html %V 45 %X In this paper, a 2D numerical approach was implemented to analyze the effect of parameters on the compressible turbulent film cooling performed by slot injection over a VKI rotor blade. In this connection, the flow and thermal fields were evaluated using the blowing ratio, total temperature of a coolant jet, injection angle, and the location of injection slots on the blade surface. The computational domain with a hybrid mesh system could provide the required foundations for using the realizable k−ε turbulence model as well as the SIMPLE algorithm. Finally, the group method of data handling (GMDH)-type neural networks which were optimized by the genetic algorithms have been successfully used to present separate polynomial relations for the area-weighted average film cooling heat transfer coefficient. The effective geometrical and flow parameters were separately involved on the pressure and suction sides of the film cooled blade. The achieved polynomials demonstrate the remarkable reliability of modeling in prediction of the film cooling heat transfer coefficient in terms of minimum training and prediction errors. %8 2014-07-09