Numerical study of the performance parameters of the developed centrifugal compressor using a classical machine learning algorithm

In recent years, significant progress has been made in research related to turbomachine. Nowadays, due to the importance of optimization and finding the optimal model, the economic components and weight reduction of the structure are considered in each stage of compressors. Due to the use of these devices in air engines, regular and diesel engines, and use in petrorefineries, it can be counted among its importance. Due to the commercial nature of these data, research studies in this field are less clearly visible. The approach of the current research is to reduce the number of blades of a compressor and compensate this reduction by changing the blade angles with the help of a machine learning algorithm. In the next step, the flow analysis in the new geometry has been investigated with the Ansys CFX software, which is based on the finite element numerical method. In line with this research, the effect of changing the number of blades on functional parameters such as entropy and pressure ratio was evaluated. It was observed that it is possible to reduce the weight of the structure by reducing the number of blades, which reduces the pressure ratio by 5%. The k-ε sst model is used to model the flow turbulence. This model has been used in order to accurately observe the reverse flow in the downstream flow of the compressor.