Mean-line model development for off-design performance prediction of transonic axial compressor of an industrial gas turbine based on computational fluid dynamics database

Abstract

In this work, a mean-line model with the row-by-row stacking scheme is developed to simulate the off-design operation of a multi-stage transonic axial compressor of an industrial gas turbine. A database of computational fluid dynamics (CFD) simulation results validated with the field data is generated at different rotating speeds and pressure ratios by using CFD simulation tools and available detail geometry of the studied compressor. The compressor characteristic parameters including inlet/exit blockage factor, deviation angle parameter, rotor temperature and pressure coefficients, and the pressure loss of the stator for all stages, plus variable inlet guide vane, are extracted from CFD simulation results. Based on analytical/empirical relations in the literature, a generic formulation for the characteristic parameters is defined as a function of airflow parameters such as inlet Mach number, inlet flow angle, and velocity ratio. Since the accuracy of this type of model depends on the formulated characteristic parameters, and due to the error propagation potential of the sequential stacking solutions, a multi-objective optimization program is developed to specify the coefficients and exponent values of the generic formulations. The functionality of characteristic curves versus airflow parameters for different compressor rows is comprehensively investigated. The comparison between the results of the mean-line model and CFD simulations shows the high accuracy of the model. The developed model can be used in applications such as condition monitoring of in-service gas turbines, studying the effects of rescheduling the variable stator vanes, and investigating the wet compression process.