A comparison of different generalised modelling approaches for a scroll refrigerant compressor

A comparison was made between the errors for the simulated and test performance of a scroll compressor based on three generalised modelling approaches. Four types of refrigerants, namely R22, R134a, R407C and R410A, were analysed, each using a number of compressors which cover the low to medium capacity range. It was found that the use of the constant-polytropic-coefficient approach (CPC) yielded much higher errors in the simulated isentropic efficiencies than those based on the constant-polytropic-efficiency approach (CPE) and the Mallen–Saville approach (MS). The CPE was the best among the three approaches considered, although it was computationally the least efficient. The average errors in the simulated isentropic efficiencies ranged from 8.3 to 10.1% for the four types of refrigerants considered while the corresponding errors in the simulated refrigerant mass flow rates and compressor power inputs varied from 1.1 to 1.8% and 7.7 to 10%. Meanwhile, the simplest CPC should be dismissed on account of the large resultant error and the yielding of incorrect trends in the simulated isentropic efficiencies.

► Compressor performances were simulated based on various generalised models.
► Comparisons were made with test data from the manufacturer.
► The constant-polytropic-efficiency approach (CPE) was the best.
► The constant-polytropic-coefficient approach (CPC) yielded the largest error.
► The CPC should be dismissed.