Cécile Devaud

Modeling evaporation effects in conditional moment closure for spray autoignition

Simulations of an n-heptane spray autoigniting under conditions relevant to a diesel engine are performed using two-dimensional, first-order conditional moment closure (CMC) with full treatment of spray terms in the mixture fraction variance and CMC equations. The conditional evaporation term in the CMC equations is closed assuming interphase exchange to occur at the droplet saturation mixture fraction values only. Modeling of the unclosed terms in the mixture fraction variance equation is done accordingly. Comparison with experimental data for a range of ambient oxygen concentrations shows that the ignition delay is overpredicted. The trend of increasing ignition delay with decreasing oxygen concentration, however, is correctly captured. Good agreement is found between the computed and measured flame lift-off height for all conditions investigated. Analysis of source terms in the CMC temperature equation reveals that a convective–reactive balance sets in at the flame base, with spatial diffusion terms being important, but not as important as in lifted jet flames in cold air. Inclusion of droplet terms in the governing equations is found to affect the mixture fraction variance field in the region where evaporation is the strongest, and to slightly increase the ignition delay time due to the cooling associated with the evaporation. Both flame propagation and stabilization mechanisms, however, remain unaffected.

Assessment of the applicability of conditional moment closure to a lifted turbulent flame: first order model

This paper presents the results of a detailed study of Conditional Moment Closure (CMC) applied to a lifted turbulent flame. The objectives are to find out how first order, radially averaged CMC can represent a lifted flame and which mechanism of flame stabilization can be described by this modeling method. As a first stage of this study of CMC applied to ignition/extinction phenomena, turbulence and combustion calculations are decoupled, that is, the effect of turbulence upon combustion is included but the heat released due to combustion is not part of the turbulence calculations. A 10-step chemical mechanism is used to predict rates of reaction in hydrogen-air mixtures. Attention is focused on the lift-off region of the flame which is commonly considered as a cold flow. Comparison with published experimental data shows that the lift-off height is accurately predicted at 14 mm. The Favre averaged radial profiles of temperature and species mole fractions are also compared with the experimental values. The computational results agree well with the experimental points for lean mixtures but the temperatures are overpredicted in rich mixtures close to the centerline. Some of the current flame stabilization mechanisms are discussed in the context of the present results. Simple elliptic first-order CMC is shown to be able to reproduce some of these mechanisms. Modeling accurately the axial transport is a key factor to these simulations.

A new method of modeling the conditional scalar dissipation rate

A new method for calculating the conditional scalar dissipation rate 〈N|η〉 is derived from the probability density function (pdf) transport equation for the conserved scalar Z. Two different formulations are obtained. One is the result of direct integration of the pdf transport equation and the second is further developed assuming a two-parameter presumed form for the pdf. A linear model is used for the conditional velocity. The model is compared with a direct numerical simulation (DNS) of inhomogeneous turbulent mixing. The results are in very good agreement with the DNS and perform better than Girimaji’s model which is based on homogeneous flow properties. Further validation with some experimental data would be useful. The new method has also the potential of being easily implemented in a finite-volume computational fluid dynamics code.

RANS simulation of a turbulent premixed bluff body flame using conditional source-term estimation

In the present work, simulations are performed in order to study a turbulent premixed flame stabilised behind a bluff body burner. Conditional Source-term Estimation (CSE) is the combustion model adopted for this purpose and it is coupled with a trajectory gene- rated low-dimensional manifold method for chemistry reduction: conditional-averaged chemical source terms are closed by conditional-averaged scalars, which are obtained by inverting an integral equation. Two regularisation methods are implemented. The optimal regularisation parameter is determined and a sensitivity analysis is performed related to the effect of the value of the regularisation parameter. In the present study, only small differences in the conditional and unconditional averages are noticed for values of the regularisation parameter located in the optimal range determined by the L-curve. Two-equation k–ε and Shear Stress Transport (SST) turbulence models are used to solve the flow field in either the reactive or non-reactive case. Different values of the constant in the standard k–ε model and of γ2 in the SST k–ω model are tested. Mean axial velocity and progress variable profiles are compared with experimental data. It is found that SST k–ω gives better results for the velocity profile in both the reactive and non-reactive cases. A value of γ2 equal to 0.65 provides best predictions in the reactive case. The progress variable prediction exhibits similar profiles in the case of CSE and the flamelet combustion model approach. Very good agreement is achieved between the CSE predictions and experimental data.

Inverse analysis and regularisation in conditional source-term estimation modelling

Conditional Source-term Estimation (CSE) obtains the conditional species mass fractions by inverting a Fredholm integral equation of the first kind. In the present work, a Bayesian framework is used to compare two different regularisation methods: zeroth-order temporal Tikhonov regulatisation and first-order spatial Tikhonov regularisation. The objectives of the current study are: (i) to elucidate the ill-posedness of the inverse problem; (ii) to understand the origin of the perturbations in the data and quantify their magnitude; (iii) to quantify the uncertainty in the solution using different priors; and (iv) to determine the regularisation method best suited to this problem. A singular value decomposition shows that the current inverse problem is ill-posed. Perturbations to the data may be caused by the use of a discrete mixture fraction grid for calculating the mixture fraction PDF. The magnitude of the perturbations is estimated using a box filter and the uncertainty in the solution is determined based on the width of the credible intervals. The width of the credible intervals is significantly reduced with the inclusion of a smoothing prior and the recovered solution is in better agreement with the exact solution. The credible intervals for temporal and spatial smoothing are shown to be similar. Credible intervals for temporal smoothing depend on the solution from the previous time step and a smooth solution is not guaranteed. For spatial smoothing, the credible intervals are not dependent upon a previous solution and better predict characteristics for higher mixture fraction values. These characteristics make spatial smoothing a promising alternative method for recovering a solution from the CSE inversion process.

Conditional Moment Closure (CMC) applied to autoignition of high pressure methane jets in a shock tube

Autoignition of non-premixed methane–air mixtures is investigated using first-order Conditional Moment Closure (CMC). Turbulent velocity and mixing fields simulations are decoupled from the CMC calculations due to low temperature changes until ignition occurs. The CMC equations are cross-stream averaged and finite differences are applied to discretize the equations. A three-step fractional method is implemented to treat separately the stiff chemical source term. Two detailed chemical kinetics mechanisms are tested as well as two mixing models. The present results show good agreement with published experimental measurements for the magnitude of both ignition delay and kernel location. The slope of the predicted ignition delay is overpredicted and possible sources of discrepancy are identified. Both scalar dissipation rate models produce comparable results due to the turbulent flow homogeneity assumption. Further, ignition always occurs at low scalar dissipation rates, much lower than the flamelet critical value of ignition. Ignition is found to take place in lean mixtures for a value of mixture fraction around 0.02. The conditional species concentrations are in qualitative agreement with previous research. Homogeneous and inhomogeneous CMC calculations are also performed in order to investigate the role of physical transport in the present autoignition study. It is found that spatial transport is small at ignition time. Predicted ignition delays are shown to be sensitive to the chemical kinetics. Reasonable agreement with previous simulations is found. Improved formulations for the mixing model based on non-homogeneous turbulence are expected to have an impact.

Species and temperature predictions in a semi-industrial MILD furnace using a non-adiabatic conditional source-term estimation formulation

A Reynolds-Averaged Navier–Stokes (RANS) simulation of the semi-industrial International Flame Research Foundation (IFRF) furnace is performed using a non-adiabatic Conditional Source-term Estimation (CSE) formulation. This represents the first time that a CSE formulation, which accounts for the effect of radiation on the conditional reaction rates, has been applied to a large scale semi-industrial furnace. The objective of the current study is to assess the capabilities of CSE to accurately reproduce the velocity field, temperature, species concentration and nitrogen oxides (NOx) emission for the IFRF furnace. The flow field is solved using the standard k–ε turbulence model and detailed chemistry is included. NOx emissions are calculated using two different methods. Predicted velocity profiles are in good agreement with the experimental data. The predicted peak temperature occurs closer to the centreline, as compared to the experimental observations, suggesting that the mixing between the fuel jet and vitiated air jet may be overestimated. Good agreement between the species concentrations, including NOx, and the experimental data is observed near the burner exit. Farther downstream, the centreline oxygen concentration is found to be underpredicted. Predicted NOx concentrations are in good agreement with experimental data when calculated using the method of Peters and Weber. The current study indicates that RANS-CSE can accurately predict the main characteristics seen in a semi-industrial IFRF furnace.

Numerical Simulation of a Pool Fire and Large Object in a Cross-Wind

The present investigation is focused on assessing the capabilities of Large Eddy Simulations (LES) using simplified submodels for combustion and soot in a specific fire scenario. Fire development resulting from an aviation fuel spill close to a plane fuselage is considered. The computational domain and boundary conditions are defined according to the experimental configuration used in tests run by the Fire Research Group at the University of Waterloo. The present setup consists of a 2-m-diameter pool fueled with kerosene and located 1-m-upstream of a 2.7m-diameter culvert in a large enclosure. A cross-wind with a velocity of 13 m/s is imposed on the fire and culvert. The calculations are time-dependent and three-dimensional. Sensitivity to the grid refinement, size of the enclosure and wind profiles is first investigated. Comparison between measured temperatures and numerical results across the computational domain is made. Velocity profiles are also examined. Reasonable agreement with the experiments is found. In the light of the present results, directions for future work are also discussed. INTRODUCTION Accidental release of liquid fuels, either in industrial processes or in transportation systems, can pose a serious fire hazard. Once ignited, the flame spread will be rapid, smoke and toxic products will be released into the surroundings. The difficulties in understanding fire dynamics stem from the all correspondence to this author. 1 complex interactions between fluid motion, combustion, radiation and multiphase flow, which all take place in a turbulent buoyancy-driven environment. This results in a wide range of length scales, from the macroscopic to the molecular level. Fire growth is also strongly affected by external conditions such as wind and the presence of an enclosure for example. Systematic experimental investigations of spill fire scenarios are hindered by financial constraints, sensitivity to environmental conditions and detrimental effect on diagnostic equipment. Numerical simulations provide a promising tool to complement experimental studies and further develop understanding of fire safety and fire physics. At present, the wide range of length and time scales in large fires prohibits the use of three-dimensional direct numerical simulations where the governing equations are directly solved without any closure assumption. Consequently mathematical modelling of the major physical and chemical processes within the fire is still required and remains a challenging task. The present investigation is focused on Large Eddy Simulation (LES) applied to a pool fire in the vicinity of a large object in a crosswind. The experimental and computational configurations aim at reproducing the scenario of fire development from an aviation fuel spill close to a plane fuselage. In LES the large turbulent flow structures are fully resolved and only the dissipation scales require modelling. Flow in the plume region is associated with the large scale structures, and motion of the large eddies are expected to make the most significant contributions to the transport of heat, radiation, chemical species and soot [1]. Copyright c

Assessment of the capabilities of FireFOAM to model large-scale fires in a well-confined and mechanically ventilated multi-compartment structure

This article presents a large eddy simulation study of a pool fire in a well-confined and mechanically ventilated multi-room configuration. The capabilities of FireFOAM are assessed by comparing the numerical results to a well-documented set of experimental data available from Propagation d’un Incendie pour des Scénarios Multi-locaux Elémentaires. The eddy dissipation concept, finite volume discrete ordinate method, and one k-equation model are used for combustion, thermal radiation, and sub-grid scale closure, respectively. The main boundary conditions are imposed based on the experimental profiles. A detailed comparison is made with available experimental data. Good agreement between the large eddy simulation results and experimental values is achieved for temperatures, velocity, CO2 volume concentrations, and pressures for most compartments. There are some noticeable underpredictions of temperature in the outlet room. Overall, FireFOAM is shown to have good predictive capabilities for the present confined large-scale fire scenario.

Regenerative braking torque control of air hybrid engines with cam-based valvetrains

In this work, a novel air hybrid engine configuration is introduced in which cam-based valvetrain along with three-way and unidirectional valves make the implementation of different engine operational modes possible. In the proposed configuration, an electronic throttle system is used to manage the engine load in both conventional and braking modes. The necessity of engine torque control during regenerative braking is discussed and a lookup table/PI controller is applied to the engine model in GT-Power to control the engine torque at the regenerative mode. It is shown that utilising the proposed configuration, the regenerative braking mode can be simply implemented and the braking torque can be regulated by controlling the throttle angle.