Reaction zones and their structure in MILD combustion
Combustion Science and Technology
Combustion Science and Technology Volume 186, Issue 8, 2014. DOI: 10.1080/00102202.2014.902814
MetadataShow full item record
Minamoto, Y., Swaminathan, N., Cant, R. S., & Leung, T. (2014). Reaction zones and their structure in MILD combustion. Combustion Science and Technology, 186 1075-1096. https://doi.org/10.1080/00102202.2014.902814
Three-dimensional direct numerical simulation (DNS) of turbulent combustion under moderate and intense low-oxygen dilution (MILD) conditions has been carried out inside a cuboid with inflow and outflow boundaries on the upstream and downstreamfaces respectively. The initial and inflowing mixture and turbulence fields are constructed carefully to be representative of MILD conditions involving partially mixed pockets of unburnt and burnt gases. The combustion kinetics is modelled using a skeletal mechanism for methane-air combustion, including non-unity Lewis numbers for species and temperature dependent transport properties. The DNS data is analysed to study theMILD reaction zone structure and its behaviour. The results show that the instantaneous reaction zones are convoluted and the degree of convolution increases with dilution and turbulence levels. Interactions of reaction zones occur frequently and are spread out in a large portion of the computational domain due to the mixture non-uniformity and high turbulence level. These interactions lead to local thickening of reaction zones yielding an appearance of distributed combustion despite the presence of local thin reaction zones. A canonical MILD flame element, called as MIFE, is proposed which represents the averaged mass fraction variation for major species reasonably well, although a fully representative canonical element needs to include the effect of reaction zone interactions and associated thickening effects on the mean reaction rate.
MILD combustion, flameless combustion, Direct numerical simulation (DNS), Reaction zones
YM acknowledges the financial support of Nippon Keidanren and Cambridge Overseas Trust. EPSRC support is acknowledged by NS. The support of Natural Sciences and Engineering Research Council of Canada is acknowledged by TL. This work made use of the facilities of HECToR, the UK’s national high-performance computing service, which is provided by UoE HPCx Ltd at the University of Edinburgh, Cray Inc and NAG Ltd, and funded by the Office of Science and Technology through EPSRCs High End Computing Programme.
External DOI: https://doi.org/10.1080/00102202.2014.902814
This record's URL: https://www.repository.cam.ac.uk/handle/1810/246755