Influence of Surface Roughness on the Performance of a Compressor Blade in a Linear Cascade: Experiment and Modeling

Abstract

Compressor blades experience significant surface degradation with service. Elevated levels of surface roughness reduce compressor efficiency and mass flow rate. This paper presents measurement and a new model of compressor blade performance degradation due to blade surface roughness. Performance tests have been conducted in a low-speed, linear cascade with roughened compressor blades. Equivalent sandgrain roughnesses of 12, 180, 300, 425, and 850 microns have been tested. These roughness values are representative of compressor blade roughnesses found in actual gas turbines in service. Flow angle, flow rate, and loss have been measured. For the tested roughnesses of 180, 300, 425, and 850 microns, the axial velocity ratio decreases by 0.1, 2.1, 2.5, and 5.4%, respectively. For the same cases, the deviation increases by 24, 38, 51, and 70%, respectively. Finally, the loss increases by 12, 44, 132, and 217%, respectively. Thus, among the three parameters, the loss responds most sensitively to changes in compressor blade roughness. Furthermore, a new mean-line model based on the assumption of 50% reaction stages has been developed to estimate the effects of roughness on the performance of a multi-stage compressor. The data from the cascade data are used as input to predict the performance of a single compressor stage. Subsequently, a stage-stacking method is used to enable prediction for a multi-stage compressor. According to the model, the pressure ratio, and mass flow rate are significantly influenced by the blade surface roughness.