Search Results

A previously developed CFD-based optimization tool is utilized to find optimal engine operating conditions with respect to fuel consumption and emissions. The optimization algorithm employed is based on the steepest descent method where an adaptive cost function is minimized along each line search using an effective backtracking strategy. The adaptive cost function is based on the penalty method, where the penalty coefficient is increased after every line search. The parameter space is normalized and, thus, the optimization occurs over the unit cube in higher-dimensional space. The application of this optimization tool is demonstrated for the Sulzer S20, a central-injection, non-road DI diesel engine. The optimization parameters are the start of injection of the two pulses of a split injection system, the duration of each pulse, the exhaust gas recirculation rate, the boost pressure and the compression ratio.

A global optimization method has been developed for an engine simulation code and utilized in the search of optimal fuel injection strategies. This method uses a Lagrange interpolation function which interpolates engine output data generated at the vertices and the intermediate points of the input parameters. This interpolation function is then used to find a global minimum over the entire parameter set, which in turn becomes the starting point of a CFD-based optimization. The CFD optimization is based on a steepest descent method with an adaptive cost function, where the line searches are performed with a fast-converging backtracking algorithm. The adaptive cost function is based on the penalty method, where the penalty coefficient is increased after every line search. The parameter space is normalized and, thus, the optimization occurs over the unit cube in higher-dimensional space.

Optimal fuel injection strategies are obtained with a micro-genetic algorithm and an adaptive gradient method for a nonroad, medium-speed DI diesel engine equipped with a multi-orifice, asynchronous fuel injection system. The gradient optimization utilizes a fast-converging backtracking algorithm and an adaptive cost function which is based on the penalty method, where the penalty coefficient is increased after every line search. The micro-genetic algorithm uses parameter combinations of the best two individuals in each generation until a local convergence is achieved, and then generates a random population to continue the global search. The optimizations have been performed for a two pulse fuel injection strategy where the optimization parameters are the injection timings and the nozzle orifice diameters.

The discrete droplet model is widely used to describe two-phase flows such as high-velocity dense sprays. The interaction between the liquid and the gas phase is modeled via appropriate source terms in the gas phase equations. This approach can lead to a strong dependence of the liquid-gas coupling on the spatial resolution of the gas phase. The liquid-gas coupling requires the computation of source terms using the gas phase properties, and, subsequently, these sources are then distributed onto the gas phase mesh. In this study, a Lagrange polynomial interpolation method has been developed to evaluate the source terms and also to distribute these source terms onto the gas mesh. The focus of this investigation has been on the momentum exchange between the two phases. The Lagrange polynomial interpolation and source term distribution methods are evaluated for non-evaporating sprays using KIVA3 as a modeling platform.

Non-evaporating diesel sprays have been simulated utilizing the ETAB and the WAVE atomization and breakup models and have been compared with experimental data. The experimental penetrations and widths were determined from back-lit spray images and the droplet sizes have been measured by means of a Malvern particle sizer. The model evaluation criteria include the spray penetration, the spray width and the local droplet size. The comparisons have been performed for variations of the injection pressure, the gas density and the fuel viscosity. The fuel nozzle exit velocities used in the simulations have been computed with a special code that considers the effect of in-nozzle cavitation. The simulations showed good overall agreement with experimental data. However, the capabilities of the models to predict the droplet size for different fuels could be improved.

A further development of the ETAB atomization and drop breakup model for high pressure-driven liquid fuel jets, has been developed, tuned and validated. As in the ETAB model, this breakup model reflects a cascade of drop breakups, where the breakup criterion is determined by the Taylor drop oscillator and each breakup event resembles experimentally observed breakup mechanisms. A fragmented liquid core due to inner-nozzle disturbances is achieved by injecting large droplets subject to this breakup cascade. These large droplets are equipped with appropriate initial deformation velocities in order to obtain experimentally observed breakup lengths. In contrast to the ETAB model which consideres only the bag breakup or the stripping breakup mechanism, the new model has been extended to include the catastrophic breakup regime. In addition, a continuity condition on the breakup parameters has lead to the reduction of one model constant.

The objective of this study is the development of a computationally efficient CFD-based tool for finding optimal engine operating conditions with respect to fuel consumption and emissions. The optimization algorithm employed is based on the steepest descent method where an adaptive cost function is minimized along each line search using an effective backtracking strategy. The adaptive cost function is based on the penalty method, where the penalty coefficient is increased after every line search. The parameter space is normalized and, thus, the optimization occurs over the unit cube in higher-dimensional space. The application of this optimization tool is demonstrated for the Sulzer S20, a central-injection, non-road DI diesel engine. The optimization parameters are the start of injection of the two pulses, the duration of each pulse, the duration of the dwell, the exhaust gas recirculation rate and the boost pressure.

A gradient-based optimization tool has been developed and, in conjunction with a CFD code, utilized in the search of optimal fuel injection strategies. The approach taken uses a steepest descent method with an adaptive cost function, where the line search is performed with a fast-converging backtracking algorithm. The adaptive cost function is based on the penalty method, where the penalty coefficient is increased after every line search. The parameter space is normalized and, thus, the optimization occurs over the unit cube in higher-dimensional space. The application of this optimization tool is demonstrated for a non-road version of the Sulzer S20 DI diesel engine which, for these simulations, is equipped with a multi-orifice, asynchronous injection system. This system permits an independent timing of the fuel pulses, and each orifice has its own diameter and injection direction.

This investigation is a continuation of a previous study by these authors in which a Lagrange polynomial interpolation method was developed to evaluate spray source terms and also to distribute the source terms onto the gas mesh; the method was applied to the liquid-gas momentum exchange. For this investigation, the method has been extended to the mass exchange between the liquid and gas phases due to evaporation. The Lagrange polynomial interpolation and source term distribution methods are applied to the liquid-gas mass and momentum exchange and are evaluated for evaporating sprays using KIVA3 as a modeling platform. These methods are compared with the standard “nearest neighbor” method of KIVA3, and experimental data are used to establish their validity. The evaluation criteria used include the liquid and vapor spray penetration, gas velocities and the computational stability.

The objective of this study is the development of a computationally efficient CFD-based tool with the capability of finding optimal engine operating conditions with respect to emissions and fuel consumption. The approach taken uses a conjugate gradient method, where the line search is performed with a backtracking algorithm. The initial backtracking step employs an adaptive step size mechanism which depends on the steepness of the search direction. The engine simulations are performed with a KIVA-3-based code which is equipped with well-established spray, combustion and emission models. The cost function is based on the idea of the penalty method and is minimized over the unit cube in n-dimensional space, which represents the set of normalized injection parameters under investigation. The application of this optimization tool is demonstrated for the Sulzer S20, a central-injection, non-road DI diesel engine.

A KIVA-based CFD tool has been utilized to simulate the effect of a Common-Rail injection system applied to a large, uniflow-scavenged, two-stroke diesel engine. In particular, predictions for variations of injection pressure and injection duration have been validated with experimental data. The computational models have been evaluated according to their predictive capabilities of the combustion behavior reflected by the pressure and heat release rate history and the effects on nitric oxide formation and wall temperature trends. In general, the predicted trends are in good agreement with the experimental observations, thus demonstrating the potential of CFD as a design tool for the development of large diesel engines equipped with Common-Rail injection. Existing deficiencies are identified and can be explained in terms of model limitations, specifically with respect to the description of turbulence and combustion chemistry.

A correction for the turbulence dissipation, based on non-equilibrium turbulence considerations from rapid distortion theory, has been derived and implemented in combination with the RNG k - ε model in a KIVA-based code. This model correction has been tested and compared with the standard RNG k - ε model for the compression and the combustion phase of two heavy duty DI diesel engines. The turbulence behavior in the compression phase shows clear improvements over the standard RNG k - ε model computations. In particular, the macro length scale is consistent with the corresponding time scale and with the turbulent kinetic energy over the entire compression phase. The combustion computations have been performed with the characteristic time combustion model. With this dissipation correction no additional adjustments of the turbulent characteristic time model constant were necessary in order to match experimental cylinder pressures and heat release rates of the two engines.