Multi-objective optimization of discrete film hole arrangement on a high pressure turbine end-wall with conjugate heat transfer simulations

This paper discussed a method of combining a full automatic multi-objective optimization and conjugate heat transfer calculation to obtain optimal cooling layouts on a transonic high pressure guide vane under a realistic turbine working condition. The improvement in cooling design from the optimized models was analyzed in detail, along with a discussion of sensitivities of two objective functions to five design variables. The full automatic method comprises the process of geometry creation, mesh generation, numerical solution and post data analysis. The vane is solid and the end-wall is arranged in a linear cascade. On the end-wall, film holes are all cylindrical and classified in five regions, with region A near the leading edge of the vane, region B near the suction side, regions C and D near the pressure side, and region E for the rest. Five design variables are three pitch-to-hole ratios in regions B, D, E and two compound angles of film holes in regions A and D. Two selected objective functions are area averaged overall cooling effectiveness of the end-wall and aerodynamic losses in a cross-plane at x/C_ax = 1.06 just downstream of the outlet of the cascade. For the optimization process, the multi-objective genetic algorithm based on the Non-dominated Sorted Genetic Algorithm-II was applied. The Latin hypercube sampling method was used to choose 21 experimental design points in the design space, which are also the sources for constructing the surrogate models with the Kriging model. The results demonstrate that the method using full automatic optimization and conjugate heat transfer calculation has achieved an increase of 8.7%–9.5% in area-averaged overall cooling effectiveness and a reduction of about 4.8%–6.1% in aerodynamic losses. The highest increase in cooling effectiveness exists in the region near the pressure side with a mild increase in the middle of the passage. The largest heat flux reduction exists in the regions near the pressure side and the crown of the suction side. The change of compound angle in region A near the leading edge has a negligible influence on overall cooling effectiveness but a high impact on aerodynamic losses. It's advisable to maintain the compound angle and pitch-to-diameter ratio at low values in region D near the pressure side to obtain high cooling performance.