The Effects of Different Property Models in a Computational Fluid Dynamics Simulation of a Reciprocating Compressor

Computational fluid dynamics was applied to model a simple reciprocating compressor using R134a (1,1,1,2-tetrafluoroethane) as the working fluid. The sensitivity of the compressor model to various property models was quantitatively assessed by calculating the work required to carry out several compression cycles. The ideal gas equation, a virial equation using only the second virial coefficient, and the Peng-Robinson equation were compared to a referencequality Helmholtz energy equation of state. Significant errors, up to 12% in the density of the outflowing gas, can result from the use of the ideal gas model. The Peng-Robinson equation resulted in density errors up to 6.3%. The virial equation gave values closest to those calculated using the Helmholtz energy equation of state, with errors in density up to 4.7%. The results also show that an increase in accuracy in work and mass flow calculations achieved by using the Helmholtz energy equation of state is obtainable without an impractical increase in computation time