Diesel engines are being more commonly used for light automotive applications, due to their higher efficiency. However, pollutant emissions can be higher than their gasoline counterparts, being difficult to reduce and control because reducing one pollutant increases another. One way to reduce emissions is by using multiple injection strategies. However, understanding multiple injections is no easy task, so far done by trial and error and experience.
Therefore, a numerical 1D model is to be adapted to simulate multiple injection situations in a diesel engine. In a previous paper by the authors, an existing model was adapted with a thermal dilatation model to consider both radial and axial dilatations in the diesel spray. The base model used is that of Ma et al (based on the Eulerian model of Musculus and Kattke for inert diesel jets). One limitation found on this model was the difference in the heat release and pressure rise curves with respect to the experimental data, resulting in an erroneous estimation of the combustion chamber pressure, which will be crucial to the modeling of multiple injections.
To improve this calculation, a mapping of the mixture fraction and reactants fraction is proposed, following the work of Tap and Veynante for diffusion flames, which in turn is based on the Burke-Schumann approach of infinitely fast chemistry for the combustion modeling. A gradual transition from a mixture regime to a steady combustion is modeled by the average reaction rate calculated using Chemkin. A radial distribution of fractions is made by using the mixture fraction distribution instead of the fuel fraction distribution (such as described in Desantes et al.), and then applying the previous mapping. The resulting plots of heat release and pressure rise have improved precision, leading to a better estimation of the combustion chamber pressure, critical to a good modeling of the multiple injection phenomena.