Ignition Delay and Heat-Release Rate for n-Heptane in a Direct-Injection Constant-Volume Combustion Chamber: Experiments and Computations

Abstract

The purpose of this study was to characterize combustion behavior for n-heptane using experimental measurements in a direct-injection constant-volume combustion chamber (CVCC) to validate chemical-kinetic mechanisms. This work is focused on compression-ignition (i.e., diesel) combustion, primarily because mechanisms for larger-chain diesel-relevant species are not well developed and require significant attention. The CVCC used in this work can be pressurized and heated to create engine-relevant conditions that enable study of autoignition behavior. In addition, the chamber is equipped with a high-pressure, common-rail diesel injector, making the study of autoignition and combustion in this system highly relevant to modern diesel engines. By varying injection pressure and duration, it is possible to control global equivalence ratio as well. Chamber pressure during injection and combustion is measured using a piezoelectric transducer, and can be subsequently used to infer heat-release rates. Experimental measurements for n-heptane mostly displayed expected trends. As initial chamber pressure increased, ignition delay decreased and peak pressure increased. As injection duration increased, ignition delay decreased due to faster ignition of richer mixtures, and peak pressure increased due to higher total heat release. The effect of temperature on ignition delay, however, was more complex and suggested some amount of NTC (negative temperature coefficient) behavior. For all conditions, heat-release rates indicated entirely premixed combustion with no hint of mixing-controlled combustion. Experimental data were compared with results from CHEMKIN-PRO simulations. The model simulated zero-dimensional combustion using a detailed n-heptane mechanism developed at Lawrence Livermore National Laboratory. These computations were used to infer local equivalence ratio information, based on equivalence ratio required in the model to match experimental ignition delay. For most test cases, the model required an equivalence ratio that was at least ∼2× richer than the global value. In addition, equivalence ratios in the model ranged a full order of magnitude, from ∼0.6 to 6, suggesting that local mixture equivalence ratios varied considerably as experimental conditions were varied. Results suggest that improved models that include details of spray physics are required in order to properly predict local equivalence ratios and resulting autoignition characteristics.