UC Berkeley

Reactive flow simulations involving turbulence-chemistry interaction are extremely challenging because of the strong non-linear coupling between chemistry and flow motions. Regardless the numerical method used, the chemical state is needed at each computational cell. Modeling of combustion processes with a large detailed chemistry requires tremendous computing resources making it impractical for routine engineering usage. It is becoming a common practice to use a reduced mechanism which contains only a subset of the species but still capable of capturing certain key combustion characteristics of interest such as ignition and major pollutant formation. The performances of the reduced mechanisms need to be assessed for the intended combustion regime but most test flames are of laminar nature or of zero dimensional.

With a one-dimensional representation, the Linear Eddy Model (LEM) and One-Dimensional Turbulence model (ODT) are intended for simulating turbulence-chemistry interactions among all physically relevant length-scales. The detailed mathematical formulation of LEM and ODT were presented and derived. Several numerical issues involved in the Monte-Carlo method used in this work are presented and explained. Additional details on the implementation in this work are given.

A standalone LEM model is used for simulating two premixed turbulent methane flames using two detailed mechanisms and six reduced mechanisms. The instantaneous consumption speed of the fuel is used for critical evaluation of the transient performances of these reduced mechanisms. Comparisons reveal that the computed flame speeds of reduced mechanisms can differ significantly in the transient turbulent flames.

ODT model is used to calculate a series of non-premixed turbulent jet flames (Sandia D,E,F flames) using two detailed reaction mechanisms and three reduced mechanisms. Overall, the ODT results with the two reduced mechanisms developed from non-premixed combustion are found in close agreement with those from the detailed mechanisms. The reduced chemistry developed from lean premixed flames gives results in poor agreement with the detailed chemistry. This is somewhat expected but it points out the importance of using the reduced chemistry that is developed only for the intended flames. These comparisons demonstrate the usefulness of LEM and ODT in critical assessment of reduced chemistry for turbulent flames.