Stipendium Xiangyi Chen

Development of advanced flow simulation models and investigation of flow mechanisms in an axial compressor based on data mining approaches

As the flow interaction is naturally complicated in turbomachinery, prediction and analysis of the internal flow has always been a critical issue. One effective method in predicting the flow field is called numerical simulation using Computational Fluid Dynamics (CFD) technique, and in industrial application, the most popular CFD method should be solving Reynolds-averaged Navier-Stokes (RANS) equations. However, as this method introduces some artificial models, it has a limited resolution in capturing the flow structures and presented under-performing results according to the previous research. Therefore, some high-resolution simulation methods like scale adaptive simulation (SAS) will be taken into consideration in this research, and these advanced flow simulation models will be further developed based on the specific flow condition.

Due to the inherent unsteadiness and nonlinearity of the flow in compressor, researchers have not fully understood the flow mechanisms and the correlations in flow data even though the numerical simulation offers an insight into the flow field. In this research, some data mining approaches like clustering, classification and dimensional reduction will be applied in analyzing the flow data, aiming at providing a better understanding of the flow structure in compressor and offering a design guideline to improve the performance of turbomachines. Based on this, the philosophy behind variation of blade tip clearance and flow control approach like casing treatment will be specifically studied