Simulations of diesel–methanol dual-fuel engine combustion with large eddy simulation and Reynolds-averaged Navier–Stokes model

Abstract

A comparison of large eddy simulation and Reynolds-averaged Navier–Stokes combustion models was performed using simulations of a diesel–methanol dual-fuel engine. The two models share the assumption that the reaction rate for each species is determined by the chemical kinetic processes as well as the relative magnitude of mixing and reaction effects, which are characterized by a kinetic timescale and a turbulence timescale. The main difference in the two models lies in their formulations of the turbulence timescale. The turbulence timescale in the Reynolds-averaged Navier–Stokes model is proportional to the eddy turnover time, while the turbulence timescale in the large eddy simulation model is the time needed for the mixture to reach perfectly mixed conditions based on the scalar dissipation rate. The models were tested using data from experiments on a modified heavy-duty diesel engine. Test cases using low methanol ratios, high methanol ratios and different initial temperatures were examined. In general, the ignition delay increases as the methanol ratio increases. In the models, this was found to occur due to the chemical interaction of the methanol and the diesel fuel. In addition, methanol auto-ignition occurs when the initial temperature is over 425 K. The large eddy simulation model and Reynolds-averaged Navier–Stokes model give similar predictions for the low methanol ratio cases. The large eddy simulation model shows improved capability to predict the high methanol ratio cases and methanol auto-ignition cases, represented by good agreement with experimental results.