Flames Featuring Ignition - Extinction: Stochastic Modelling for the Prediction of Finite Rate Chemistry Effects
Author(s)
Gkagkas, Konstantinos
Type
Thesis or dissertation
Abstract
The current thesis presents a numerical study of steady and unsteady turbulent reacting
flows. The flow is calculated using Finite Volume based parabolic and elliptic flow solvers.
A transported probability density function (pdf) approach, closed at the joint–scalar level,
is used for the inclusion of the thermochemistry. A common characteristic of all the studied
cases are the strong finite rate chemistry effects which govern the flow. Two experimentally
well documented turbulent lifted flames were computed in order to explore the detailed
thermochemical flow structure and to reduce uncertainties associated with the chemical
kinetics. The effect of the applied detailed chemistry and its subsequent simplification on
the calculated thermochemical structure was also quantified. The two cases feature fuel jets
of methane or hydrogen issuing into a vitiated, high temperature coflow. Molecular mixing
is closed using the modified Curl’s mixing model and two algebraic closures are considered
for the closure of the mixing frequency. More complex flow patterns are considered through
the calculation of bluff body stabilised flames. These flames feature a recirculation region
and a neck zone of high strain rates, where significant levels of local extinction are found.
The transported pdf method captures the local extinction and can predict the pollutant
formation with high accuracy. The standard mixing frequency closure leads to over–
prediction of local extinction, while an algebraic extension leads to improved predictions.
When the flame is close to global extinction, strong instabilities occur, which lead to
questions regarding the use of a two–dimensional approach. For this reason, a three–
dimensional computational tool was developed and validated using both presumed and
transported pdf methods for the representation of the thermochemistry.
flows. The flow is calculated using Finite Volume based parabolic and elliptic flow solvers.
A transported probability density function (pdf) approach, closed at the joint–scalar level,
is used for the inclusion of the thermochemistry. A common characteristic of all the studied
cases are the strong finite rate chemistry effects which govern the flow. Two experimentally
well documented turbulent lifted flames were computed in order to explore the detailed
thermochemical flow structure and to reduce uncertainties associated with the chemical
kinetics. The effect of the applied detailed chemistry and its subsequent simplification on
the calculated thermochemical structure was also quantified. The two cases feature fuel jets
of methane or hydrogen issuing into a vitiated, high temperature coflow. Molecular mixing
is closed using the modified Curl’s mixing model and two algebraic closures are considered
for the closure of the mixing frequency. More complex flow patterns are considered through
the calculation of bluff body stabilised flames. These flames feature a recirculation region
and a neck zone of high strain rates, where significant levels of local extinction are found.
The transported pdf method captures the local extinction and can predict the pollutant
formation with high accuracy. The standard mixing frequency closure leads to over–
prediction of local extinction, while an algebraic extension leads to improved predictions.
When the flame is close to global extinction, strong instabilities occur, which lead to
questions regarding the use of a two–dimensional approach. For this reason, a three–
dimensional computational tool was developed and validated using both presumed and
transported pdf methods for the representation of the thermochemistry.
Date Issued
2008-09
Date Awarded
2009-04
Advisor
Lindstedt, Peter
Sponsor
BP Global Fuels
Creator
Gkagkas, Konstantinos
Publisher Department
Department of Mechanical Engineering
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)