Large Eddy simulations of premixed turbulent combustion

Download files
Access & Terms of Use
open access
Copyright: Chatakonda, Obulesu
Altmetric
Abstract
In the past two decades, Large-Eddy Simulation (LES) has become well established for modelling non-reacting flows. However, LES is still undergoing significant development for reacting flows. The most challenging topic in LES of combusting flows is modelling the reaction rate. To this end, several models have been proposed in the literature. Many of the models proposed so far are limited to corrugated flame (CF) regime. However, in many premixed combustion applications both CF and thin reaction zone (TRZ) regimes exist. There is, therefore, a need for accurate and robust physical modelling of the reaction rate across different regimes of premixed flames. In this thesis, a new flame wrinkling model is proposed based on fractal geometry concept requiring the modelling of an inner cut-off scale and fractal dimension. Damk¨ohler s limiting scaling laws are used to infer the cut-off and fractal dimension in limiting regimes. The assumptions made to obtain the fractal dimension in the proposed model are tested by analysing fractal characteristics of several thermochemical and thermo-nuclear direct numerical simulation (DNS) data sets featuring Damk¨ohler numbers. The fractal dimension is found to vary between 2.1 to 2.7 in the thermo-chemical hydrogen air turbulent premixed flames. The thermo-nuclear supernovae flames, which are in distributed regimes of combustion, yield fractal dimension about 2.7. The results for the maximum fractal dimensions are higher than previously reported in the literature and can be explained theoretically by a Reynolds number similarity argument which shows that the limiting value of the fractal dimension at low Damk¨ohler number is 8/3. To assess the performance of the new proposed model and also the models reported in the literature, a priori analysis of two DNS data sets are used. The sub-grid models are assessed using constant coefficient and dynamic coefficient versions using an approximate Germano identity. The effects of filter size on the performance of the models are also investigated. A priori assessments of the models show very good results for the models having power law form for all DNS data and filter sizes. In addition to this, a posteriori tests are conducted to assess these models in a slot Bunsen burner configuration in the TRZ regime. A combined progress variable and level set model has been used to model reaction rate. Chemistry is obtained from tabulated chemistry by solving strained and unstrained flamelets. The comparison of a priori and a posteriori results with the DNS data provides strong support for the validity of the proposed model.
Persistent link to this record
Link to Publisher Version
Link to Open Access Version
Additional Link
Author(s)
Chatakonda, Obulesu
Supervisor(s)
Hawkes, Evatt
Kook, Shawn
Creator(s)
Editor(s)
Translator(s)
Curator(s)
Designer(s)
Arranger(s)
Composer(s)
Recordist(s)
Conference Proceedings Editor(s)
Other Contributor(s)
Corporate/Industry Contributor(s)
Publication Year
2012
Resource Type
Thesis
Degree Type
PhD Doctorate
UNSW Faculty
Files
download whole.pdf 4.21 MB Adobe Portable Document Format
Related dataset(s)