Large Eddy Simulation of Swirl Flames

Document Sample
Large Eddy Simulation of Swirl Flames Powered By Docstoc
					                           Application of Large Eddy Simulation technique to
                           Premixed and Non-premixed Combustion Problems

         Professor W. Malalasekera, School of Mechanical and Manufacturing Engineering,
                                    Loughborough University

Advancement in the design and operation of combustion devices used in automobile, air
transport and power generation industries is very important for the reduction of emissions
contributing to global warming. Such advances depend upon making combustion equipment
operate at higher efficiencies so that more power can be extracted for the same amount of fuel
burnt. This will, in the long run, reduce emissions or maintain at the present levels while meeting
the present and future demand for power and energy. To this end Computational Fluid Dynamics
(CFD) has become a vital tool in the design process and more and more industries are now
using CFD to explore flow behaviour of various designs and simulate temperature, heat transfer
and emissions in combustion equipment before prototypes are built for testing. Such CFD studies
have various benefits – design cycle can be shortened, new ideas can be tested without prior
experimentation which can be very costly and incremental changes can be made to the design to
achieve a desired effect. CFD models for combustion simulations are, however, far from perfect
to use in such studies. There are many issues that make combustion modelling one of the most
difficult areas in CFD applications. Complexities such as turbulence/chemistry interactions,
chemical kinetics, coupling of flow turbulence and temperature to density, heat transfer and
radiation effects make the CFD modelling of combustion very challenging.

Combustion modelling success entirely depend on accurate modelling of turbulence, flow and
heat transfer including radiation. Considerable developments in the recent past has been
directed towards improving RANS based turbulence models yet there are practical situations
where even the most advanced RANS based models have not predicted very well. More recently
Large Eddy Simulation technique has become more popular for complex type applications. In this
lecture Professor Malalasekera will present LES based validation efforts of combustion modelling
which his group has conducted. There is a great need to validate combustion models in isolation
with good quality data to understand important features of combustion models and to gain
confidence in their application to real life combustion situations. In this presentation Professor
Malalasekera explains experience and validation of LES based combustion models for non-
premixed, pre-mixed and partially premixed combustion situations which are relevant to practical
combustion applications. The basics of the presentation covers

•        Current status of the LES based combustion models.
•        Advantages of LES based combustion models over RANS based models
•        Applications to bluff-body, swirl stabilised and lifted flames
•        Detailed validation of LES combustion models
•        A discussion on future directions and applications.

Modelling of swirl flames, application to vented explosions and modelling of a laboratory lifted
flame are discussed in this presentation. Experimental data from high quality experiments are
used to compare the LES results and in general it is shown that LES is a promising technique
and capable of producing good quality results in a variety combustion situations.


                               RESEARCH SYNOPSYS AND CAREER RECORD
    Professor W. Malalasekera, School of Mechanical and Manufacturing Engineering, Loughborough University
Weeratunge Malalasekera is a Professor in the Wolfson School of Mechanical and Manufacturing Engineering at
Loughborough University who has worked in the development of numerical techniques for CFD, radiative heat
transfer and combustion for over 20 years. He gained his Ph.D. at Imperial College in 1988 where he worked with
Professor Fred Lockwood in the development of CFD based combustion models and radiation calculation
techniques. Prior to joining Loughborough as a lecturer in 1989, he was a postdoctoral researcher at Imperial
College, London.
          Early work of Dr. Malalasekera has been mainly on the development of CFD-based combustion prediction
procedures for practical combustors and modelling of compartment fires. Development and implementation of pdf-
based turbulent combustion models, soot models, NO x calculation algorithms and accurate radiation calculation
techniques into CFD procedures were undertaken in this early work.
          In a number of studies Prof. Malalasekera has contributed to I.C. engine combustion research. Combustion
and radiation algorithms developed in the context of general combustion situations have been successfully
implemented in I.C. combustion prediction procedures. Inclusion of advanced sub-models for auto ignition and soot
formation and combustion have been shown to greatly enhance predictions in engines. Prof. Malalasekera’s other
published research work in I.C. engine combustion includes, full cycle simulation of radiative heat transfer in
complex i.c. engine geometries and experimental validation, investigations into engine auto ignition of end gas,
effects of near-wall turbulence and heat transfer on combustion. These publications have been used by many other
international researchers in their work.
          Prof. Malalasekera’s group has made a considerable contribution to the development of numerical radiative
heat transfer calculation techniques. In these studies a suite of radiation calculation techniques for body-fitted grid
systems using the Discrete Transfer method, Monte Carlo method, Discrete Ordinates method and the YIX method
have been successfully developed. The impact of this work on the international radiation community was
demonstrated by the adoption of several cases developed by the group by other researchers as benchmarks for the
comparison and validation of methods. Numerically based fundamental work includes investigations into the error
analysis and convergence behaviour of various techniques, and development of advanced grid adaptation
techniques for the discrete transfer method. Experimental studies to measure thermal radiation in complex geometry
configurations and comparison with numerical methods have also been undertaken.
          The group has also been actively engaged in fundamental combustion modelling research. Development of
laminar flamelet based combustion models has been a major thrust. Inclusion of radiation in laminar flamelet models
and special procedures to improve NOx predictions has been addressed in these studies. More recent work has
used large eddy simulation (LES) techniques for combustion. Some of the current work includes the LES modelling
of the Sydney swirl burner series, LES modelling of vented explosions, development of a progress variable approach
for non-premixed LES, development of coupled radiation/laminar flamelet models for LES and NO x modelling with
detailed chemistry. These publications have resulted increased international recognition for Loughborough work at
the TNF workshops and the establishment of useful links such as the collaboration with Sydney university
combustion group.
          Development and application of CFD techniques for the modelling of sonic jets, numerical and experimental
investigation of dust explosion phenomena in inter-connected vessels, and experimental and numerical investigation
of heat transfer in condensing boilers are some of the other work undertaken by Prof. Malalasekera in the past.
Recently completed research activities also include combined CFD/finite element calculations of engine heat
transfer and structural interactions, development of porous media flow and heat transfer models and their application
to fibre bonding, modelling of flow and heat transfer in natural environments and investigations into energy storage
and distribution strategies.
          Since joining Loughborough in 1989, Dr Malalasekera has provided the inspiration and leadership to
develop CFD related combustion and heat transfer research and today the group is well recognised nationally and
internationally within academic circles. This is evident from the Ph.D. completions, range of quality publications and
presentations of their work in international conferences.
          Prof. Malalasekera has made various contributions to the international workshop on Turbulent Non-
Premixed combustion (TNF) and delivered keynote lectures in CFD and numerical heat transfer at international
conferences and conducted invited lectures and short courses in CFD at various overseas organisations.
Lougborough group work closely with many industrial companies. Combustion, automotive and thermo-fluid related
activities are industrially supported by Shell, Jaguar, Ford, Lotus, Perkins/Caterpillar, Ricardo, 3M Health-Care and
many other partners. The group has excellent collaborative links with many UK and international universities and
work closely with industry.
          Prof. Malalasekera contributes to all levels in undergraduate and M.Sc. courses. In collaboration with a
colleague, Prof Malalasekera was a pioneer in introducing CFD at undergraduate level back in 1990. At that time
CFD was very much in the arena of Ph.D. researchers. Prof. Malalasekera developed accessible and
comprehensive course material and later published a text book with a colleague entitled “An Introduction to
Computational Fluid Dynamics: The Finite Volume Method”. First published by Longman Higher Education in 1995,
this book has become a widely used popular CFD course text at many universities world-wide. An enhanced second
edition of this book covering advanced topics and recent developments in CFD has been published in February
2007, by Pearson Higher Education. A list of his publications could be obtained                                     via

                                                                                                              Paper   2/2