High load limits of the controlled autoignition engine

Abstract

The controlled autoignition (CAI) engine is an engine concept that features very low soot and NOx emissions while achieving diesel-like efficiency. The CAI combustion process is characterized by a fast, volumetric burn of a premixed fuel and air mixture that is heavily diluted with either excess air or burned gas. One problem that limits the engine's introduction into the market is that the rate of combustion can be fast to the point that it is very loud and is destructive to the mechanical components of the engine. The nature of this problem is akin to spark-ignition knock and presents a high-load limit to the operating range. Misfire presents a second high-load limit. This work seeks to understand how various engine parameters affect the high-load knock limit and the high-load misfire limit of a CAI engine operating in the negative-valve-overlap mode. Valve timing, and therefore trapped residual gas fraction, turbocharging, intake air heating, and exhaust gas recirculation (EGR) are all explored for their effects on the high load limit. A single-cylinder research engine is used to assess each of these effects. 91 RON gasoline is used for all tests. The first part of the study explores the effects of boost, intake air temperature, and trapped residual fraction on the rate of pressure rise, which was used as the metric for knock, and on NIMEP. It was shown that if operation is constrained by maximum pressure rise rate (PRRmax,,), the maximum load condition always lies at the misfire limit. The dependencies of misfire on boost, intake air temperature, and residual fraction were also explored.

(cont.) The second part of the study examined how the use of EGR affected the high-load limit. Again, the high-load limit, constrained by maximum pressure rise rate, was maximized at the misfire limit. Varying boost and intake temperature did not materially change the high load limit because the misfire limit and the PRRmax changed simultaneously in such a way that the NIMEP at the new limit point did not change appreciably from the original value. Correlations were developed for the misfire limit, for the burn duration, and for combustion phasing.