Design of an optimum combustion chamber across multiple speed/load points for a heavy-duty diesel engine: analytical design and experimental validation

Abstract

Analytical tools such as computational fluid dynamics (CFD) and micro-genetic optimization algorithm (GA) have been customized and applied at General Motors Research Laboratories (GMR) for designing a combustion chamber for GM heavy-duty diesel engines, capable of meeting present and future emission targets. In the combined analytical and experimental study described in this paper, the analytical design of the piston bowl shape spanned four key steady-state load points and the performance was validated by testing the analytical design in a single-cylinder engine (SCE). Computations were made to assess the fuel economy performance of the first analytical design. Based on the assessment, a second analytical design was performed with a different set of load points. Computations clearly revealed the progress of performance from the first to the second design. The two optimized piston bowl designs were tested in the SCE in order to validate the predicted progression of performance from the baseline bowl shape and the effect of the choice of the load points for analytical optimization. A multidimensional computer code (KIVA-3V) with a relatively simpler combustion model – the characteristic time-scale combustion (CTC) model – was used along with the GA optimization tool. Experimental results confirmed the predicted emission improvements and performance progression of the analytically designed piston bowls compared with the baseline. Detailed flow field analysis is presented for a selected load point to elucidate the physics behind the sensitivity of observed emission behaviour to variations in injector tip protrusion.