Consortium On Computational Combustion For Engineering Applications (COCCFEA)

K. H. Luo and C. J. Lawn (Queen Mary, London)

R. S. Cant (Cambridge)

D. R. Emerson and X. J. Gu (Daresbury Laboratory)

Background

Turbulent combustion has been a major challenge in engineering science in the past century and will require significant research and investment well into the 21st century. The underlying difficulty has always been the coupling of the chemical reaction with turbulence, which itself is generally regarded as the most difficult problem in classical physics. A major development in the 20th century was the application of computational methods to the simulation of turbulent combustion process, made possible by the quantum leap in supercomputer technologies in the latter part of the last century. The appearance and continuing development of computational combustion, especially direct numerical simulation (DNS) and large eddy simulation (LES), as a powerful tool complementary to experiment, has considerably advanced our fundamental knowledge and accelerated the engineering applications of turbulent combustion. Both DNS and LES can, to different degrees, resolve the structure of turbulent flames as well as provide mean and higher order statistics. This yields greatly enhanced insight into the mechanisms behind phenomena and more accurate statistics than available from traditional methods including experiment, which are invaluable for turbulent combustion modelling. The limitations of DNS and LES are not due to any drawback in the methods themselves but due to practical limits on computer memory and speed, and cost considerations. The fundamental requirements of these methods are that they should resolve time and length scales, which is not possible or necessary by the more conventional Reynolds Averaged Navier-Stokes (RANS) approach. The difficulties are that:

1. both turbulence and chemical reactions have a wide range of length and time scales to be resolved;
2. scales associated with chemical reactions are usually smaller than those of turbulence e.g. the flame thickness compared to the Kolmogorov length scale;
3. coupling of turbulence and combustion introduces new length scales;
4. the relations between the scales are often problem dependent i.e. flames in internal combustion engines and those in gas turbine combustors need different treatment.

The implication is that combustion modelling, which in essence involves representing the effects of scales by mean quantities, or approximating higher-order fluctuations with lower order statistics, will continue to play an important and vital role in combustion research, especially in engineering applications.

As a joint effort to tackle the afore-mentioned problems in turbulent combustion, a Consortium On Computational Combustion For Engineering Applications (COCCFEA), which was a subset of the CCP12 Steering Group, was initially set up in 1994 under the chairmanship of Professor K. N. C. Bray FRS (Cambridge). Its membership included researchers with a spectrum of expertise in the field ranging from DNS to combustion modelling with RANS. EPSRC supported the project with a grant (GR/K4160) to enable meetings of the consortium and with funds for a PDRA for two years under the High Performance Computing Initiative (HPCI). The work was highly rated and laid the foundation for a further grant (GR/L06843) for a 12 month extension. A further three year proposal, led by Professor D. Bradley (Leeds) and Professor W. P. Jones (Imperial College), was supported by EPSRC under grant (GR/M19918/01) and finished in September 2001 with a highly successful 2-day workshop held in Chester on September 11-12. At each stage, the consortium has successfully expanded its membership and a new proposal, led by Dr. K. H. Luo, Professor C. J. Lawn (Queen Mary, London) and Dr. R. S. Cant (Cambridge), has been successful (GR/R66197/01) and started on January 1st, 2002. The consortium membership consists of 8 UK institutions and 15 permanent staff members, with many other PhD researchers and post-doctoral staff:

• Drs. R. S. Cant and N. Mastorakis (Cambridge)

• Professor G. M. Makhviladze (Central Lancashire)

• Professor J. B. Moss and Dr. P. A. Rubini (Cranfield)

• Drs. D. R. Emerson and X. J. Gu (Daresbury Laboratory)

• Professors W. P. Jones and R. P. Lindstedt and Dr. A. Kronenberg (Imperial College)

• Professors D. Bradley FRS and P. Gaskell (University of Leeds)

• Professors C. J. Lawn and K. H. Luo (Queen Mary, University of London)

• Dr. R. Prosser (UMIST)

The consortium is supported by CCP12 staff at Daresbury Laboratory, as indicated above. The proposed work, which will be funded through individual grant proposals, will involve modelling flame instabilities in turbulent flames, DNS and LES of premixed, nonpremixed and partially premixed turbulent combustion, numerical methods for DNS and LES, burning in turbulent premixtures, explosions, and conditional moment closure (CMC) modelling.

The figures show the temperature field from a DNS simulation of Methane-Air flame
(courtesy of Prof. K. H. Luo, Queen Mary, University of London)

Contact Details

For further information on the combustion consortium please contact:

Professor David Emerson

Science and Technology Facilities Council
Daresbury Laboratory
Daresbury Science and Innovation Campus
Warrington WA4 4AD
Cheshire
United Kingdom