Refine
Year of publication
Document Type
Language
- English (18)
Keywords
- (CPV) Approach for Engine Applications (1)
- Advanced Predictive (1)
- Assessment of Water Injection (1)
- Detailed Chemistry (1)
- Development of a Computationally Efficient Progress (1)
- Diesel Engine, Stochastic Reactor Modeling, Particlulate Matter (1)
- Genetic Optimization (1)
- Highly fuel-efficient Diesel engines (1)
- Modeling of Diesel Sprays (1)
- Multi-Objective Optimization (1)
Institute
BTU
Highly fuel-efficient Diesel engines, combined with effective exhaust aftertreatment systems, enable an economic and low-emission operation of heavy-duty vehicles. The challenge of its development arises from the present engine complexity, which is expected to increase even more in the future. The approved method of test bench measurements is stretched to its limits, because of the high demand for large parameter variations. The introduction of a physics-based quasi-dimensional stochastic reactor model combined with tabulated chemistry enables the simulation-supported development of these Diesel engines. The stochastic reactor model mimics mixture and temperature inhomogeneities induced by turbulence, direct injection and heat transfer. Thus, it is possible to improve the prediction of NOx emissions compared to common mean-value models. To reduce the number of designs to be evaluated during …
Water injection is investigated for turbocharged spark-ignition engines to reduce knock probability and enable higher engine efficiency. The novel approach of this work is the development of a simulation-based optimization process combining the advantages of detailed chemistry, the stochastic reactor model and genetic optimization to assess water injection. The fast running quasi-dimensional stochastic reactor model with tabulated chemistry accounts for water effects on laminar flame speed and combustion chemistry. The stochastic reactor model is coupled with the Non-dominated Sorting Genetic Algorithm to find an optimum set of operating conditions for high engine efficiency. Subsequently, the feasibility of the simulation-based optimization process is tested for a three-dimensional computational fluid dynamic numerical test case. The newly proposed optimization method predicts a trade-off between fuel efficiency and low knock probability, which highlights the present target conflict for spark-ignition engine development. Overall, the optimization
shows that water injection is beneficial to decrease fuel consumption and knock probability at the same time. The application of the fast running quasi-dimensional stochastic reactor model allows to run large optimization problems with low computational costs. The incorporation with the Non-dominated Sorting Genetic Algorithm shows a well performing multi-objective optimization and an optimized set of engine operating parameters with water injection and high compression ratio is found.
A Computationally Efficient Combustion Progress Variable (CPV) Approach for Engine Applications
(2018)
Advanced Predictive Diesel Combustion Simulation Using Turbulence Model and Stochastic Reactor Model
(2017)
Several models for ignition, combustion and emission formation under diesel engine conditions for multi-dimensional computational fluid dynamics have been proposed in the past. It has been recognized that the use of a reasonably detailed chemistry model improves the combustion and emission prediction especially under low temperature and high exhaust gas recirculation conditions.The coupling of the combustion chemistry and the turbulent flow can be achieved with different assumptions. In this paper we investigate a selection of n-heptane spray experiments published by the Engine Combustion Network (ECN spray H) with three different combustion models: well-stirred reactor model, transient interactive flamelet model and progress variable based conditional moment closure. All models cater for the use of detailed chemistry, while the turbulence-chemistry interaction modeling and the ability to consider local effects differ.The same chemical mechanism is used by all combustion models, which allows a comparison of ignition delay, flame stabilization and flame lift-off length between the experiments and the results from simulations using the different combustion models. The investigated parameters influence the predictions of computational fluid dynamics simulations of diesel engines. This study indicates that the most reasonable behavior with respect to ignition, flame stabilization and flame structure is predicted by the progress variable based conditional moment closure model.
In this work a soot source term tabulation strategy for soot predictions under Diesel engine conditions within the zero-dimensional Direct Injection Stochastic Reactor Model (DI-SRM) framework is presented. The DI-SRM accounts for detailed chemistry, in-homogeneities in the combustion chamber and turbulence-chemistry interactions. The existing implementation [1] was extended with a framework facilitating the use of tabulated soot source terms. The implementation allows now for using soot source terms provided by an online chemistry calculation, and for the use of a pre-calculated flamelet soot source term library. Diesel engine calculations were performed using the same detailed kinetic soot model in both configurations. The chemical mechanism for n-heptane used in this work is taken from Zeuch et al. [2] and consists of 121 species and 973 reactions including PAH and thermal NO chemistry. The engine case presented in [1] is used also for this work. The case is a single-injection part-load passenger car Diesel engine with 27 % EGR fueled with regular Diesel fuel. The two different approaches are analyzed and a detailed comparison is presented for the different soot processes globally and in the mixture fraction space. The contribution of the work presented in this paper is that a method which allows for a direct comparison of soot source terms - calculated online or retrieved from a flamelet table - without any change in the simulation setup has been developed within the SRM framework. It is a unique tool for model development. Our analysis supports our previous conclusion [1] that flamelet soot source terms libraries can be used for multi-dimensional modeling of soot formation in Diesel engines.
The introduction of a physics-based zero-dimensional stochastic reactor model combined with tabulated chemistry enables the simulation-supported development of future compression-ignited engines. The stochastic reactor model mimics mixture and temperature inhomogeneities induced by turbulence, direct injection and heat transfer. Thus, it is possible to improve the prediction of NOx emissions compared to common mean-value models. To reduce the number of designs to be evaluated during the simulation-based multi-objective optimization, genetic algorithms are proven to be an effective tool. Based on an initial set of designs, the algorithm aims to evolve the designs to find the best parameters for the given constraints and objectives. The extension by response surface models improves the prediction of the best possible Pareto Front, while the time of optimization is kept low. This work presents a novel methodology to couple the stochastic reactor model and the Non-dominated Sorting Genetic Algorithm. First, the stochastic reactor model is calibrated for 10 low, medium and high load operating points at various engine speeds. Second, each operating point is optimized to find the lowest fuel consumption and specific NOx emissions. The optimization input parameters are the temperature at intake valve closure, the compression ratio, the start of injection, the injection pressure and exhaust gas recirculation rate. Additionally, it is ensured that the maximum peak cylinder pressure and turbine inlet temperature are not exceeded. This enables a safe operation of the engine and exhaust aftertreatment system under the optimized conditions. Subsequently, a reliability analysis is performed to estimate the effect of off-nominal conditions on the objectives and constraints. The novel multi-objective optimization methodology has proven to deliver reasonable results. The zero-dimensional stochastic reactor model with tabulated chemistry is a fast running physics-based model that allow to run large optimization problems in a short amount of time. The combination with the reliability analysis also strengthens the confidence in the simulation-based optimized engine operation parameters.
The use of complex reaction schemes is accompanied by high computational cost in 3D CFD simulations but is particularly important to predict pollutant emissions in internal combustion engine simulations. One solution to tackle this problem is to solve the chemistry prior the CFD run and store the chemistry information in look-up tables. The approach presented combines pre-tabulated progress variable-based source terms for auto-ignition as well as soot and NOx source terms for emission predictions. The method is coupled to the 3D CFD code CONVERGE v2.4 via user-coding and tested over various speed and load passenger-car Diesel engine conditions. This work includes the comparison between the combustion progress variable (CPV) model and the online chemistry solver in CONVERGE 2.4. Both models are compared by means of combustion and emission parameters. A detailed n-decane/α-methyl-naphthalene mechanism, comprising 189 species, is used for both online and tabulated chemistry simulations. The two chemistry solvers show very good agreement between each other and equally predict trends derived experimentally by means of engine performance parameters as well as soot and NOx engine-out emissions. The CPV model shows a factor 8 speed-up in run-time compared to the online chemistry solver without compromising the accuracy of the solution.
Spray modeling is among the main aspects of mixture formation and combustion in internal combustion engines. It plays a major role in pollutant formation and energy efficiency although adequate modeling is still under development. Strong grid dependence is observed in the droplet-based stochastic spray model commonly used. As an alternative, an interactive model called 'SprayLet' is being developed for spray simulations based on one-dimensional integrated equations for the gas and liquid phases, resulting from cross-sectionally averaging of multi-dimensional transport equations to improve statistical convergence. The formulated one-dimensional cross-section averaged system is solved independently of the CFD program to provide source terms for mass, momentum and heat transfer between the gas and liquid phases. The transport processes take place in a given spray cone where the nozzle exit is automatically resolved. In the 1D program, the conservation equations are for droplet diameter, droplet temperature, as well as for continuity and momentum of the liquid and the gaseous phase are solved. The source terms between the phases are conservatively embedded into the spray region of the CFD program. In CFD program, the transport equations are solved for gas phase only. The SprayLet model is validated using standard Sandia sprays by comparing penetration lengths and fuel mixture fractions with experimental data.
Meeting strict current and future emissions legislation necessitates development of computational tools capable of predicting the behaviour of combustion and emissions with an accuracy sufficient to make correct design decisions while keeping computational cost of the simulations amenable for large-scale design space exploration. While detailed kinetics modelling is increasingly seen as a necessity for accurate simulations, the computational cost can be often prohibitive, prompting interest in simplified approaches allowing fast simulation of reduced mechanisms at coarse grid resolutions appropriate for internal combustion engine simulations in design context. In this study we present a simplified Well-stirred Reactor (WSR) implementation coupled with 3D CFD Ricardo VECTIS solver. A detailed evaluation of benchmark ECN spray problem is presented demonstrating that a single point calibration of such a model using a bulk reaction multiplier approach can provide correct representation of the solution across a wide range of temperatures on grid sizes typically employed for RANS internal combustion engine simulations with tabulated kinetics or zonal combustion models.
In the SAE article 2023-24-0083: SprayLet: One-dimensional interactive cross-sectionally averaged spray model, we formulatet a one-dimensional Spray model in interaction with the surrounding gas phase. We could demonstrate, that the model predicted liquid and gaseous penetration length in good aggreement with ECN spray experiments. In this paper we use this model in engine CFD (CONVERGE CFD) and demonstrate a strong reduction in CPU time (50%). We can show a strong decrease in grid dependency, which allows a further reduction of CPU time (90%). We will present engine CFD simulations, comparing detailed spray with SpayLet simulations. This includes pressure traces, heat release, and emissions.
The use of chemical kinetic mechanisms in computer aided engineering tools for internal combustion engine simulations is of high importance for studying and predicting pollutant formation of conventional and alternative fuels. However, usage of complex reaction schemes is accompanied by high computational cost in 0-D, 1-D and 3-D computational fluid dynamics frameworks. The present work aims to address this challenge and allow broader deployment of detailed chemistry-based simulations, such as in multi-objective engine optimization campaigns. A fast-running tabulated chemistry solver coupled to a 0-D probability density function-based approach for the modelling of compression and spark ignition engine combustion is proposed. A stochastic reactor engine model has been extended with a progress variable-based framework, allowing the use of pre-calculated auto-ignition tables instead of solving the chemical reactions on-the-fly. As a first validation step, the tabulated chemistry-based solver is assessed against the online chemistry solver under constant pressure reactor conditions. Secondly, performance and accuracy targets of the progress variable-based solver are verified using stochastic reactor models under compression and spark ignition engine conditions. Detailed multicomponent mechanisms comprising up to 475 species are employed in both the tabulated and online chemistry simulation campaigns. The proposed progress variable-based solver proved to be in good agreement with the detailed online chemistry one in terms of combustion performance as well as engine-out emission predictions (CO, CO2, NO and unburned hydrocarbons). Concerning computational performances, the newly proposed solver delivers remarkable speed-ups (up to four orders of magnitude) when compared to the online chemistry simulations. In turn, the new solver allows the stochastic reactor model to be computationally competitive with much lower order modeling approaches (i.e., Vibe-based models). It also makes the stochastic reactor model a feasible computer aided engineering framework of choice for multi-objective engine optimization campaigns.