Numerical optimization of metal felt regenerators with axially varying matrix structures for Stirling engines

Abstract

The regenerator is the crucial component for the performance of any regenerative gas cycle. Due to the changing gas properties and flow conditions in the axial direction of a regenerator, a variation of its parameters in that direction is a promising approach to reduce its losses and thus increase the performance of the cycle. This contribution presents the first comprehensive analysis of the potential of metal felt regenerators with axially varying parameters operated in a Stirling engine. For this purpose, a numerical optimization of an existing regenerator in an experimental machine is conducted using a well-validated differential model. To isolate the positive effects of the regenerator optimization, the thermal efficiency of the engine is optimized under the constraint of a constant power density. In general, the optimized matrix is characterized by increased porosity and decreased fiber diameter at the hot end of the regenerator, and vice versa at the cold end. The thermal efficiency of the engine is increased to 27.54 % for the optimized matrix compared to 27.06 % for the matrix with axially constant parameters. The optimal parameters depend on the operating point. For instance, the difference between the parameters at the cold and hot end decreases at a reduced heater temperature. Nevertheless, a regenerator optimized at the design operating point yields efficiency enhancements throughout the entire operating range, even though these enhancements decrease with increasing deviation from the design point. These promising findings contribute to the further improvement of Stirling engines and suggest pursuing an experimental validation.