Modelagem matemática e otimização de motores stirling com pistão livre

Abstract

Stirling engines, especially those with free piston configurations, have numerous advantages that make them promising for refrigeration applications, cryogenics, space systems applications, micro Combined Heat and Power (CHP) and naval applications such as those used in submarines. The free-piston Stirling engine configuration is currently the most profitable and promising configuration. Despite this, the technology still has limitations regarding its application because it has not yet been fully developed. To overcome these challenges and advance the development of Stirling engines, high-precision predictive mathematical methods are needed to model the functioning for design purposes. Difficulties in describing the operation of the engine limited the technology to simpler models with many simplifications of the real complex processes that take place in the engine. Among this extensive set of considerations needed to improve the Stirling engine performance, highlights the required to perform a combined analysis of the engine thermal operating process and the pistons operating dynamics. Resulting, therefore, in high prediction accuracy of the design parameters necessary for the good performance of the engine. The main contribution of this work was the adoption of an approach that allows the simultaneous modeling of the work process of free-piston Stirling engines and the dynamics of piston operation applied to second-order Stirling engine modeling and the modeling of operating conditions.in quasi-teady approach. The optimization of ten geometric parameters through the application of genetic algorithms optimization and multi-objective particle swarm algorithm was applied to the numerical model of quasi-steady conditions developed. The prediction model resulted in an error in power of only 4.1% and 0.7% of efficiency. The optimized geometry produced 1419.0 W of power and efficiency of 39.7Ho when using the genetic algorithm and power of 1227,8.0 W and thermal efficiency of 39,99a using the MOPSO algorithm.