https://doi.org/10.1140/epjp/s13360-020-00597-0
Regular Article
A GMDH-type neural network model for predicting the effect of sister hole parameters on the film cooling effectiveness over a turbine blade
1
Mechanical Engineering Department, Faculty of Engineering, University of Guilan, Rasht, Iran
2
Faculty of Technology and Engineering, East of Guilan, University of Guilan, Rudsar, Iran
Received:
24
December
2019
Accepted:
7
July
2020
Published online:
15
July
2020
In this paper, a three-dimensional numerical analysis was employed to investigate the flow and thermal fields over a leading edge of AGTB-B1 high-pressure turbine blade. The effect of different parameters on the film cooling was studied. The computational methodology includes the use of a structured, non-uniform hexahedral grid consisting of the main flow channel, the coolant delivery tube and the feeding plenum. The SIMPLEC algorithm was implemented for pressure–velocity coupling. Computations were carried out for the following range of film cooling parameters: lateral position of sister holes 3, 3.5 and 4.75 mm; lateral angle of sister holes of − 8°, 0° and 8°; blowing ratio of 0.7, 1.1 and 1.5; and diameter of injection hole of 2, 2.5 and 3. Adiabatic effectiveness was used as a criterion to judge the performance of film cooling. The results show a good agreement compared with previous experimental and numerical data. Finally, the GMDH-type neural network was successfully employed for modeling and presenting a correlation for film cooling effectiveness as a function of effective parameters.
© Società Italiana di Fisica and Springer-Verlag GmbH Germany, part of Springer Nature 2020
