National Center for Combustion Research and Development
Indian Institute Of Technology Madras & Indian Institute of Sciences, Bangalore
Modelling paradigms for MILD combustion
Speaker : N. SWAMINATHAN
Date : 11-07-2014
Venue: Aero / NCCRD Seminar Hall
N. SWAMINATHAN got his B.E. (Mech. 1987) from PSG Tech, M.E. (Aero, 1989) from IISc, PhD (Mech, 1994) from UofColorado, Boulder. Currently, he is a Professor of Mechanical Engineering at Cambridge University and his research focus is on reacting and non-reacting fluid flow phenomena and their modelling with an eye on practical applications. He was a consultant to Pratt & Whitney in association with Tata Consultancy Services and is a consultant to a number of gas turbine and IC engines OEMs worldwide. Swaminathan has published more than 120 papers in journals and conference proceedings and coedited a book on Turbulent Premixed Flames, published by Cambridge University Press in 2011. He is on the Advisory Committee for multi-disciplinary education on energy science programme at Tokyo Tech and serves on Editorial Board of many Open Access international journals on fluids, combustion, and energy topics. Abstract : Three-dimensional Direct Numerical Simulation (DNS) data of methane-air MILD combustion is analysed to study the behaviour of MILD reaction zones and to identify a suitable modelling paradigm for MILD combustion. The combustion kinetics in the DNS was modelled using a skeletal mechanism including non-unity Lewis number effects. The reaction zones under MILD conditions are highly convoluted and contorted resulting in their frequent interactions. This leads to combustion occurring over a large portion of the computational volume and giving an appearance of distributed combustion. Three paradigms, standard flamelets, mild flame elements (MIFEs) and PSR, along with a presumed PDF model are explored to estimate the mean and filtered reaction rate in MILD combustion. A beta function is used to estimate the presumed PDF shape. The variations of species mass fractions and reaction rate with temperature computed using these models are compared to the DNS results. The PSR-based model is found to be appropriate, since the conditional averages obtained from the DNS agree well with those obtained using the PSR model. The flamelets model with MIFEs gives only a qualitative agreement because it does not include the effects of reaction zone interactions.