Christopher J. Rutland

Modeling high-speed viscous liquid sheet atomization

Abstract A linear stability analysis is presented for a liquid sheet that includes the effects of the surrounding gas, surface tension and the liquid viscosity on the wave growth process. An inviscid dispersion relation is used to identify the transition from a long wavelength regime to a short wavelength regime, analogous to the first and second wind induced breakup regimes of cylindrical liquid jets. This transition, which is found to occur at a gas Weber number of 27/16, is used to simplify the viscous dispersion relation for use in multi-dimensional simulations of sheet breakup. The resulting dispersion relation is used to predict the maximum unstable growth rate and wave length, the sheet breakup length and the resulting drop size for pressure-swirl atomizers. The predicted drop size is used as a boundary condition in a multi-dimensional spray model. The results show that the model is able to accurately predict liquid spray penetration, local Sauter mean diameter and overall spray shape.

Development and testing of diesel engine CFD models

The development and validation of Computational Fluid Dynamic (CFD) models for diesel engine combustion and emissions is described. The complexity of diesel combustion requires simulations with many complex, interacting submodels in order to be successful. The review focuses on the current status of work at the University of Wisconsin Engine Research Center. The research program, which has been ongoing for over five years, has now reached the point where significant predictive capability is in place. A modified version of the KIVA code is used for the computations, with improved submodels for liquid breakup, drop distortion and drag, spray-wall impingement with rebounding, sliding and breaking-up drops, wall heat transfer with unsteadiness and compressibility, multistep kinetics ignition and laminar-turbulent characteristic time combustion models, Zeldovich NOx formation, and soot formation with Nagle-Strickland-Constable oxidation. The code also considers piston-cylinder-liner crevice flows and allows computations of the intake flow process in the realistic engine geometry with two moving intake valves. A multicomponent fuel vaporization model and a flamelet combustion model have also been implemented. Significant progress has been made using a modified RNG k-e turbulence model. This turbulence model is capable of predicting the large-scale structures that are produced by the squish flows and generated by the spray. These flow structures have an important impact on the prediction of NOx formation since it is very sensitive to the local temperatures in the combustion chamber. Model validation experiments have been performed using a single-cylinder version of a heavy duty truck engine that features state-of-the-art high-pressure electronic fuel injection and emissions instrumentation. In addition to cylinder pressure, heat release, and emissions measurements, combustion visualization experiments have been performed using an endoscope system that takes the place of one of the exhaust valves. In-cylinder gas velocity (PIV) and gas temperature measurements have also been made in the motored engine using optical techniques. Modifications to the engine geometry for optical access were minimal, thus ensuring that the results represent the actual engine. Experiments have also been conducted to study the effect of injection characteristics, including injection pressure and rate, nozzle inlet condition and multiple injections on engine performance and emissions. The results show that multiple pulsed injections can be used to significantly reduce both soot and NOx simultaneously in the engine. In addition, when combined with exhaust gas recirculation to further lower NOx, pulsed injections are found to be still very effective at reducing soot. The intake flow CFD modeling results show that the details of the intake flow process influence the engine performance. Comparisons with the measured engine cylinder pressure, heat release, soot and NOx emission data, and the combustion visualization flame images show that the CFD model results are generally in good agreement with the experiments. In particular, the model is able to correctly predict the soot—NOx trade-off trend as a function of injection timing. However, further work is needed to improve the accuracy of predictions of combustion with late injection, and to assess the effect of intake flows on emissions.

Direct simulations of premixed turbulent flames with nonunity Lewis numbers

Abstract A principal effect of turbulence on premixed flames in the flamelet regime is to wrinkle the flame fronts. For nonunity Lewis numbers, Le ≠ 1, the local flame structure is altered in curved regions. This effect is examined using direct numerical simulations of three-dimensional isotropic turbulence with constant density, single-step Arrhenius kinetics chemistry. Simulations of Lewis numbers 0.8, 1.0, and 1.2 are compared. At the local level, curvature effects dominated changes to the flame structure while strain effects were insignificant. A strong Lewis-number-dependent correlation was found between surface curvature and the local flame speed. The correlation was positive for Le 1. At the global level, strain-related effects were more significant than curvature effects. The turbulent flame speed changed significantly with Lewis number, increasing as Le decreased. This was found to be due to strain effects that have a nonzero mean over the flame surface, rather than to curvature effects that have a nearly zero mean. The mean product temperature was also found to vary with Lewis number, being higher for Le > 1 and lower for Le

Modeling thermally thick pyrolysis of wood

Abstract A general model of the pyrolysis of a wood slab is presented and validated with a set of heat release data. The model is applied to particle half-thicknesses from 5 μm to 5 cm , temperatures from 800 to 2000 K , and moisture contents from 0% to 30%. Internal temperatures, pyrolysis rates and yields of tar, hydrocarbons and char are presented. Four pyrolysis regimes are identified, depending on external temperature and particle size: thermally thin—kinetically limited, thermally thin—heat transfer limited, thermally thick, and thermal wave regimes.

Large-eddy simulations for internal combustion engines – a review

A review of using large-eddy simulation (LES) in computational fluid dynamic studies of internal combustion engines is presented. Background material on turbulence modelling, LES approaches, specifically for engines, and the expectations of LES results are discussed. The major modelling approaches for turbulence, combustion, scalars, and liquid sprays are discussed. In each of these areas, a taxonomy is presented for the various types of models appropriate for engines. Advantages, disadvantages, and examples of use in the literature are described for the various types of models. Several recent examples of engine studies using LES are discussed. Recommendations and future prospects are included.

Multi-dimensional modeling of thin liquid films and spray-wall interactions resulting from impinging sprays

Abstract The focus of this work is to formulate and validate a multi-dimensional, fuel film model to help account for the fuel distribution during combustion in internal combustion engines. Spray-wall interaction and spray-film interaction are also incorporated into the model. The fuel film model simulates thin fuel film flow on solid surfaces of arbitrary configuration. This is achieved by solving the continuity, momentum, and energy equations for the two-dimensional film that flows over a three-dimensional surface. The major physical processes considered in the model include mass and momentum contributions to the film due to spray drop impingement, splashing effects, various shear forces, piston acceleration, dynamic pressure effects, gravity driven flow, conduction, and convective heat and mass transfer. In order to adequately represent the drop interaction process, impingement regimes and post-impingement behavior have been modeled using experimental data and mass, momentum and energy conservation constraints. The regimes modeled for spray-film interaction are stick, rebound, spread, and splash. In addition, modified wall functions for evaporating wavy films are provided and tested. The fuel film model is validated through a series of comparisons to experimental data for secondary droplet velocities, secondary droplet sizes, spray radius, spray height, film thickness, film spreading radius, and percentage of fuel adhered to the surface.

Premixed flame effects on turbulence and pressure-related terms

Abstract Direct numerical simulations (DNS) were carried out for premixed, turbulent flames. Heat release effects are accounted for by inclusion of variable density. The simulated flames are thin in the sense that the reaction progress variable is bi-modal and consistent with BML theory. The DNS data were used for detailed study of flame effects on turbulence within the turbulent flame brush by examining the turbulent kinetic energy budget. The flame effects on turbulent kinetic energy were found to depend strongly on the heat release. Both mean and fluctuating pressure terms were found to be the main factors responsible for increases in turbulent kinetic energy. The main sinks for turbulence are dissipation and mean dilatation. Pressure diffusion was found to dominate the other turbulent kinetic energy diffusion terms. A model was developed for pressure dilatation that matches the DNS results very closely. The model indicates that pressure dilatation will remain an important source of turbulence even as heat release increases.

Dynamic One-Equation Nonviscosity Large-Eddy Simulation Model

Anew approach for a nonviscosity large-eddy simulation (LES)subgrid stress model is presented.Theapproach uses a scaling that is provided by the subgrid kinetic energy and a tensor coefe cient that is obtained from the dynamic modeling approach, hence, a dynamic structure model. Mathematical and conceptual issues motivating the development of this new model are explored. Attention is focused on dynamic modeling approaches. The basic equations that originate in dynamic modeling approaches are Fredholm integral equations of the second kind. These equations have solvability requirements that have not been previously addressed in the context of LES models. These conditionsare examined for traditional dynamic Smagorinksy modeling, that is, zero-equation approaches, and one-equation subgrid models. It is shown that standard approaches do not always satisfy the integral equation solvability condition. It is also shown that traditional LES models that use the resolved scale strain rate to estimatethesubgrid stressesscalepoorly with e lterlevel, leading to signie cant errorsin themodeling of the subgrid scale stress. The poor scaling in traditional LES approaches can result in not only weak models, but can also cause nonrealizability of the subgrid stresses. A better scaling based on the subgrid kinetic energy is proposed that leads to a new one-equation nonviscosity model that does satisfy the solvability conditions and appears to maintain realizability. Both integral and algebraic formulations of the new one-equation nonviscosity model are presented. The resolved and subgrid kinetic energies are shown to compare well to a direct numerical simulation of decaying isotropic turbulence.

Application of artificial neural networks in engine modelling

Abstract A study was conducted to develop an accurate simulation tool with a small computer resource footprint for engine design. The modelling approach uses artificial neural networks (ANNs) based on multilayer perceptrons (MLPs). The ANN is used to represent engine in-cylinder processes by training the ANN to approximate computational fluid dynamics (CFD) simulation results of the engine. The ANN approach was applied to model a turbo-charged direct injection (DI) diesel engine over a wide range of operating conditions. Seven primary diesel engine control parameters were varied over their possible ranges: engine speed, engine load, start of injection, injection pressure, mass in the first injection pulse of a split injection, boost pressure and exhaust gas recirculation (EGR). The model output includes five quantities: cylinder pressure, cylinder temperature, cylinder wall heat transfer, NOX emission and soot emission (the elemental carbon fraction of particulate matters). The cylinder pressure, temperature and heat transfer are crank angle resolved, while the emissions are cycle resolved. In total 71 engine steady state operating conditions were simulated with CFD, and five MLPs were trained individually to approximate the five engine output parameters as a function of the seven engine control parameters. The testing results showed that the five trained MLPs achieved satisfactory capabilities of predicting engine responses and representing the characteristics of the engine over a wide range of operating conditions. Additionally, the ANN modelling accuracy was improved by incorporating prior knowledge into the ANN design and using a committee of networks instead of the best single network to make predictions.

Large eddy simulation modelling of spray-induced turbulence effects

Abstract This study is focused on the development of spray models for large eddy simulations (LESs) of high injection pressure diesel sprays. Basic governing equations are numerically solved using the Lagrangian—Eulerian approach for liquid and gas phases respectively, in the KIVA-3V code. Non-reacting simulations of evaporating spray are performed using a dynamic structure LES model in which the subgrid stress tensor is modelled with a non-viscosity tensor coefficient. This is a one equation based model, in which an extra transport equation for the subgrid kinetic energy (k) is solved. Subgrid scale energy exchange between droplets and the gas phase is identified as an important mechanism to capture the correct scaling of k, which eventually feeds back to both the spray droplets and the gas phase turbulent mixing. Hence, to account for spray-induced gas turbulence, a LES spray source model for k is developed. This model requires subfilter scale velocities, which are obtained by defiltering the filtered velocity field available from LES. Results are compared with corresponding RANS results and available experimental data for various physical spray variables such as spray penetration, gas phase penetration, droplet distribution, etc. The LES spray model was found to be important in predicting the correct liquid spray evolution and responds correctly to varying parameters such as nozzle size and ambient density.