Cooperative Optimization of Pre-swirl Nozzles and Receiver Holes in a Radial Pre-swirl System Using an ANN-PSO Approach

Abstract

Radial pre-swirl systems are widely applied in the aviation industry to supply cooling air to high-pressure turbine blades in aircraft engines. The efficiency of the film cooling can significantly decline when the air pressure is insufficient. This study explored the synergistic optimization of pre-swirl nozzles and receiver holes to improve the pressure ratio of a radial pre-swirl system. To attain this objective, we established a surrogate model using an artificial neural network and adopted the particle swarm optimization algorithm to pinpoint the optimized geometric parameters within the defined design scope. The results revealed that the optimal performance was achieved when the pre-swirl-nozzle tangential angle reached 40.4368°, the receiver-hole axial angle reached 2.0286°, and the tangential angle reached 30°. Additionally, multiple computational simulations were performed under diverse operational conditions to validate the efficacy of this optimization. The results revealed a significant enhancement in the pressure-boosting efficiency of the radial pre-swirl system, with negligible impact on temperature increment. The optimized model exhibited a 16.93% higher pressure ratio and 1.6% higher temperature ratio than the baseline model. This improvement can be attributed to enhancements in the flow field and reductions in local losses.