Difficulties in Computational Modelling of Piston/Liner Interaction

Abstract

From the very beginning of the piston dynamic behavior investigation, the engineers tended to use experimental methods and compare them with very simple calculations. However, the experimental tests did show some serious difficulties that needed to be managed in order to obtain reasonable results, e.g. separation of vibrations due to piston slap motion, signal destroyed by combustion process, etc. This is the reason why the test rigs were hardly able to provide detailed insights into the piston dynamics, since it could be lost in a plethora of interactive physical phenomena. Therefore, the focus was, gradually, aimed at the computational models, which were (and still are) much faster, but need to be compared with experimental results to ensure the adequacy of algorithms used. This paper discusses the difficulties and uncertainties occurring during the construction of such computational model.From the very beginning of the piston dynamic behavior investigation, the engineers tended to use experimental methods and compare them with very simple calculations. However, the experimental tests did show some serious difficulties that needed to be managed in order to obtain reasonable results, e.g. separation of vibrations due to piston slap motion, signal destroyed by combustion process, etc. This is the reason why the test rigs were hardly able to provide detailed insights into the piston dynamics, since it could be lost in a plethora of interactive physical phenomena. Therefore, the focus was, gradually, aimed at the computational models, which were (and still are) much faster, but need to be compared with experimental results to ensure the adequacy of algorithms used. This paper discusses the difficulties and uncertainties occurring during the construction of such computational model.