Evaluation of a second order simulation for Sterling engine design and optimisation

This paper reports on the investigation of the simulation accuracy of a second order Stirling cycle simulation tool as developed by Urieli (2001) and improvements thereof against the known performance of the GPU-3 Stirling engine. The objective of this investigation is to establish a simulation tool to perform preliminary engine design and optimisation. The second order formulation under investigation simulates the engine based on the ideal adiabatic cycle, and parasitic losses are only accounted for afterwards. This approach differs from third order formulations that simulate the engine in a coupled manner incorporating non-idealities during cyclic simulation. While the second order approach is less accurate, it holds the advantage that the degradation of the ideal performance due to the various losses is more clearly defined and offers insight into improving engine performance. It is therefore particularly suitable for preliminary design of engines. Two methods to calculate the performance and efficiency of the data obtained from the ideal adiabatic cycle and the parasitic losses were applied, namely the method used by Urieli and a proposed alternative method. These two methods differ essentially in how the regenerator and pumping losses are accounted for. The overall accuracy of the simulations, especially using the proposed alternative method to calculate the different operational variables, proved to be satisfactory. Although significant inaccuracies occurred for some of the operational variables, the simulated trends in general followed the measurements and it is concluded that this second order Stirling cycle simulation tool using the proposed alternative method to calculate the different operational variables is suitable for preliminary engine design and optimisation.