Efficient simulation and auto-calibration of soot particle processes in Diesel engines

Abstract

Parameters describing soot particle processes are generally derived from a limited number of experimental studies. These parameters then have to be carefully calibrated for different operating conditions in internal combustion engine applications. This paper presents an innovative calibration procedure for soot simulation in Diesel engines. A Diesel engine is simulated using the Stochastic Reactor Model engine code, which is implemented with the Moment Projection Method for handling the soot particle dynamics. The main advantage of the engine-soot model is its low computational cost. The model is then coupled with an advanced statistical toolkit, Model Development Suite, where the Hooke-Jeeves algorithm is adopted to calibrate seven soot model parameters automatically based on the measurement data. The ability of the integrated code for soot model calibration is evaluated by simulating the soot formation and oxidation processes in a heavy-duty Diesel engine which is operated under 18 different conditions. Results suggest that the integrated code is able to calibrate the soot model parameters effectively. A significant improvement in the match between the simulation results and experimental soot emission is obtained after calibration.