TY - GEN A1 - Correa, Chrys UR - https://archiv.ub.uni-heidelberg.de/volltextserver/667/ KW - Combustion KW - Diesel Engines Y1 - 2000/// ID - heidok667 N2 - Three models were implemented, which are important for pollutant prediction in Diesel engines: ignition, chemistry and radiation. Ignition was tracked by means of a representative species (here CO), whose concentration remains small during the ignition period and which shows an increase at ignition. Its reaction rate was obtained from a detailed mechanism and combined with a presumed probability density function (pdf). The intrinsic low-dimensional manifold (ILDM) method was used as a chemistry model. It is an automatic reduction of a detailed chemical mechanism based on a local timescale analysis. It was also combined with a presumed pdf method. NOx and soot were predicted using a Zeldovich model and a phenomenological two-equation model, respectively. The radiative properties of the gases were described with a weighted sum of grey gases model (WSGGM). The radiative properties of soot were described by a grey model. The RTE was solved using the discrete ordinates method (DOM), which involves solving the RTE in discrete directions. The ignition and chemistry models were implemented in a standard CFD code, KIVA and used to simulate the combustion in a Caterpillar engine, for which experimental data were available. Ignition was observed to occur at the edge of the spray, in the lean region. Simulated pressure curves and mean NO concentrations were compared to experimental data and showed good agreement. Soot was strongly under-predicted due to the inability to identify the ILDM in the rich region. The DOM radiation model was tested in a furnace, and the wall fluxes were compared to analytical data. It was not used in the engine due to low quantities of soot predicted. Instead, an optically thin model was used in the engine and the radiative losses were seen to be negligible. AV - public TI - Combustion simulations in Diesel engines using reduced reaction mechanisms ER -