Automatic Numerical Analyses and Optimization of Operating Maps Applied to a Radial Compressor

Abstract

Abstract This paper contributes to the field of radial compressor design by proposing an adaptive, automated workflow incorporating the analysis of the compressor performance for a multitude of operation points by means of the respective operating maps. Most state-of-the-art approaches do not consider that the operating map limits are not conserved while changing geometric parameters which constraints these analyses to a rather small design space. In contrast, the presented methodology considers the varying operating map limits in regards to the corresponding mass flow and with that expands the possible input parameter range. The presented workflow integrates different software solutions, starting with the automated generation of the compressor geometry based on a parametric CAD model. For each geometry a mesh is generated that is used for all subsequent CFD simulations which finally result in the operating map. For every speed line, the choke point is identified by an adaptive CFD computation (based on the principle of similarity). By using the calculated choke mass flow, supplementary CFD simulations obtain additional operating points on the current speed line by a stepwise reduction of the mass flow. However, the identification of the surge line is not within the scope of the presented approach. Therefore, the range covered by the map is determined by the mass flow at the maximum efficiency and the mass flow at the choke line. The developed framework is applied to optimize the operating map of a radial compressor. A successful optimization shows that the optimized design has an enlarged choke mass flow for lower compressor speed while the pressure ratio and polytropic efficiency are comparable. At the same time, this design has a comparable choke mass flow and efficiency for higher compressor speed, but an improved maximal pressure ratio. The obtained results from the optimization show that the methodology is applicable to a wide parameter range. By adaptively calculating the operating map limits, the approach is not restricted to a small design space.