Cold Start on Low Compression Ratio Diesel Engine: Experimental and 3D RANS Computation Investigations

Diesel engines' future emission standards require greatly reducing the pollutant emissions directly at the exit of the engine before post-processing. The reduction of NOx “production” with very low levels of Particulate Matter (PM), HC and CO must be performed without damaging the drivability, and keeping fuel consumption and combustion noise under control. The reduction of the Compression Ratio (CR) is one of the most promising ways to achieve this challenge. The counterpart of reducing Diesel engine CR is the decrease in start ability. As a consequence, a stringent limitation of reducing Diesel CR is cold start requirements. Indeed, reduction of ambient temperature leads to penalties in fuel vaporization and auto-ignition capabilities, even more at very low temperature (–20°C and below). This paper presents the work performed on a HSDI Common-Rail Diesel 4-cylinder engine. Three kinds of investigations were used: experiments performed at very low temperature (down to –25°C); incylinder imaging (videoscope) and three-dimensional Computational Fluid Dynamics (CFD) computations in cold start conditions. The combustion chamber was adapted in order to reach a low compression ratio (CR 13.7:1) by modifying the piston bowl shape. First, experimental results obtained with a low CR engine are compared with those obtained with the original conventional CR engine. Then, a complete recalibration of injection settings (fuel quantity and timing, injection pressure, etc.) was carried out. It allows one to significantly reduce start delay with a low CR engine and the reference start delay with the conventional CR becomes reachable. In addition, effects of combustion chamber design such as spray position according to the glow plug were studied and show a great potential regarding behavior in cold conditions in particular, in order to reduce start delay again in such conditions. In order to complete the engine tests, CFD calculations were performed during the starting operation in ambient cold conditions (–20°C). The obtained results are in good agreement with optical observations and in-cylinder pressure measurements. Correlations between experiments and calculations give consistent explanations concerning the different phenomena occurring during cold start.