Application of a multiple mapping conditioning mixing model to ECN Spray A

The engine combustion network (ECN) Spray A is modelled using the Reynolds-averaged Navier-Stokes-transported probability density function (RANS-TPDF) approach to validate the application of a new multiple mapping conditioning (MMC) mixing model to multiphase reactive flows. The composition TPDF equations are solved using a Lagrangian stochastic approach and the spray is modelled with a discrete particle approach. The model is first validated under non-reacting conditions (at 900 K) using experimental mixture-fraction data. Reactive simulations are then performed for three different ambient temperatures (800, 900, 1100 K) and oxygen concentrations (13, 15, 21%) at an ambient density of 22.8 kg/m(3). The MMC mixing model is compared with the interaction by exchange with the mean (IEM) mixing model. The ignition delay predictions are not sensitive to the mixing model and are predicted well by both the mixing models under all the tested ambient conditions. The IEM model overpredicts the flame lift-off length (FLOL) at high temperature and high oxygen conditions with a mixing constant C-phi = 2. The MMC model with C-phi = 2 and a target correlation coefficient r(t), = 0.935 between the mixture fraction and a reference variable used to condition mixing predicts good FLOL under all the conditions except 800 K. It is demonstrated that the lift-off length is controllable by changing the target correlation coefficient, while C-phi and therefore the mixing fields are held fixed. In comparison to the MMC model, the IEM model predicts a higher variance of temperature conditioned on mixture fraction near the flame base owing to its lacking the property of localness. The mixing distance between the notional TPDF particles in the composition space is also higher with the IEM model and it is demonstrated that by changing r(t), different levels of mixing locality can be achieved. (C) 2018 The Combustion Institute. Published by Elsevier Inc. All rights reserved.