Creation of a parameterized model of a return channel flow path for CFD-researches

Abstract

An important element of the multistage centrifugal compressor flow path is a return channel, which has a complex configuration, its flow path is determined by a large number of parameters and sizes. Designing and optimizing stator elements is a long and laborious process. The use of modern computing systems can significantly reduce the time spent, make calculations of different variants for the return channel execution and choose the best. The aim of this work was to construct a parameterized model of stator elements. The process of constructing sketches in the DesingnModeler program of the ANSYS CFX software package with the output of variable values to the Workbench parameter table is considered in detail. It also describes the process of transferring data to a TurboGrid grid generator, the peculiarities of setting boundary conditions for a parameterized model in CFX-Pre, and the possibility of outputting data to a parameter table according to the calculation results from the CFD-Post program. The stator elements of the model stage 0028-058- 0373 were selected as the object under study. When constructing the flow path in the DesignModeler program, geometric parameters are determined that are subject to optimization. For simple optimization, it is enough to indicate the range of one parameter variation, in this case, the blade height at the outlet of the vane cascade, referred to the blade height at the inlet b6/b5. To assess the effectiveness of the vane cascade, the loss of total pressure was set Δp3−out*=p3*−pout*. The optimization results in the program are presented in tabular form, for clarity, a graph is built. Each calculated point can be opened in the CFD-Post, see the flow structure and calculate the necessary gas-dynamic characteristics. The use of modern computing systems makes it possible to carry out more complex optimization with several variable parameters and different performance criteria, which allows you to calculate a large number of variants and significantly saves time for setting a new task.