A parametric optimization procedure for the suction system of reciprocating compressors

The design of the suction system of compressors is of fundamental importance for efficiency and reliability. This paper reports a method developed to optimize the suction system of a reciprocating compressor, by using the genetic algorithm NSGA-II. The isentropic and volumetric efficiencies are used as objective functions, while the bending fatigue stress is used as a constraint to meet valve reliability. A simulation model of the compression cycle was coupled to the optimization procedure, with correlations for flow and force effective areas in terms of geometric parameters of the suction valve. Valve dynamics was numerically solved via the finite element method. The proposed optimization procedure was applied to a reciprocating compressor adopted for household refrigeration, identifying suction system geometries more efficient than the original design.