Technical Report 8, c4e-Preprint Series, Cambridge

Numerical Analysis of a Natural Gas Fuelled HCCI Engine with Exhaust Gas Recirculation Using a Stochastic Reactor Model

Authors: Amit Bhave, Michael Balthasar, Markus Kraft*, and Fabian Mauss

Reference: Technical Report 8, c4e-Preprint Series, Cambridge, 2002

Associated Themes:
  Theme icon Theme icon


Abstract

Combustion and emissions formation in a Volvo TD 100 series diesel engine running in a homogeneous charge compression ignition (HCCI) mode and fuelled with natural gas is simulated and compared to experimental measurements for both without and with exhaust gas recirculation (EGR). A new stochastic approach is introduced to model the convective heat transfer, which accounts for fluctuations and fluid-wall interaction effects. This model is included in the Partially stirred plug flow reactor (PaSPFR), a stochastic reactor model (SRM), to study the effect of EGR on pressure, auto-ignition timing and emissions of CO and unburned hydrocarbons (HCs). The model accounts for temperature inhomogeneities and includes a detailed chemical mechanism to simulate the chemical reactions within the combustion chamber. Turbulent mixing is described by the interaction by exchange with the mean (IEM) model. A Monte Carlo method with second order time splitting technique is employed to obtain the numerical solution. The model is validated by comparing the simulated in-cylinder pressure history and emissions, with the measurements taken from the ref.: Christensen, M.et. al., SAE paper 982454. Excellent agreement is obtained between the peak pressure, ignition timing and CO and HC emissions predicted by the model and those obtained from the measurements for the non-EGR, 38% EGR and 47% EGR cases. A comparison between the pressure profiles for the cases studied, reveals that the ignition timing and the peak pressure are dependent on the EGR. With EGR, the peak pressure reduces and the auto-ignition is delayed. The trend observed in the experimentally measured emission values with EGR is also predicted correctly by the model.

Material from this preprint has been published in The International Journal of Engine Research.

Download

PDF (330.5 KB)