• home
• current students
• graduate
• events

Presenter: Dr. Tarek Echekki - Sandia National Laboratories
Supervisor:

Date: Tue, April 3, 2001
Time: 10:00:00 - 00:00:00
Place: EOW 430

ABSTRACT

Abstract:

The process of auto-ignition is of fundamental interest to a number of practical applications that include Diesel and HCCI engine combustion and Gas Turbine combustion. This process presents many challenges for its modeling and simulation. It strongly depends upon the competition between mixing and chemistry at a very broad range of scales.

Preliminary results on the auto-ignition of hydrogen-air non-homogeneous mixtures using the Direct Numerical Simulation (DNS) approach and high-fidelity Linear-Eddy stochastic Model (LEM) are presented. The initial conditions are based on heated air and cold fuel mixtures. The present calculations illustrate a number of features of relevance to auto-ignition modeling for non-homogeneous mixtures. Auto-ignition occurs initially in spatially-homogeneous regions of very lean mixtures. The large differences in temperatures between the fuel and oxidizer streams, to a large extent, contributes to this mode of burning. The spatial homogeneity is consistent with the relatively low rates of dissipation.

Once auto-ignition is initiated in the few kernels that are both homogeneous and lean, burning shifts to more fuel-rich mixtures, and is most intense near stoichiometric conditions. Fronts that form from igniting kernels propagate, at least initially, as lean premixed flames. The simulations show that some of the salient features of the coupling between auto-ignition chemistry and mixing may adequately be captured by the LEM model. The model predictions are consistent with those of the multi-dimensional DNS.