Research on load adaptability expansion of zero-dimensional heat release prediction for direct-injection diesel engines

Existing zero-dimensional heat release models cannot simultaneously give high predictive performance at both high and low loads. This paper presents a fully mechanism-based 0D heat release model with high level of model predictability over very broad operating range of direct-injection diesel engines. A transient momentum-based spray penetration and air-entrainment model is derived, and further combined with a non-steady evaporation model to predict combustible mixture preparation. The total fuel mass and corresponding average concentration for premixed combustion as well as the instantaneous amount of fuel available for diffusion combustion in both rich and lean spray region can be calculated based on ignition delay and mixture preparation. A modified mixing model considering combustion-induced turbulence is proposed to calculate the turbulent mixing intensity. The influence of total fuel mass for premixed combustion, evaporation limitation and combustion-induced turbulence on the prediction results of heat release rate are discussed, which confirms the requirement of load adaptability expansion on the physical depth of model. Moreover, extensive experimental validation of the proposed heat release model has been performed under different loads, engine speeds, injection timing and injection pressure using fixed model parameters. The maximum error of crank angle for 10%, 50% and 90% heat release are no more than 0.3 °CA, 0.7 °CA and 5.2 °CA respectively, which proves that the proposed heat release model has superior prediction performance.