TY - GEN A1 - Vacca, Antonino A1 - Bargende, Michael A1 - Chiodi, Marco A1 - Netzer, Corinna A1 - Gern, Maike Sophie A1 - Kauf, Georg Malte A1 - Kulzer, André Casal A1 - Franken, Tim T1 - Analysis of Water Injection Strategies to Exploit the Thermodynamic Effects of Water in Gasoline Engines by Means of a 3D-CFD Virtual Test Bench N2 - CO2 emission constraints taking effect from 2020 lead to further investigations of technologies to lower knock sensitivity of gasoline engines, main limiting factor to increase engine efficiency and thus reduce fuel consumption. Moreover the RDE cycle demands for higher power operation, where fuel enrichment is needed for component protection. To achieve high efficiency, the engine should be run at stoichiometric conditions in order to have better emission control and reduce fuel consumption. Among others, water injection is a promising technology to improve engine combustion efficiency, by mainly reducing knock sensitivity and to keep high conversion rates of the TWC over the whole engine map. The comprehension of multiple thermodynamic effects of water injection through 3D-CFD simulations and their exploitation to enhance the engine combustion efficiency is the main purpose of the analysis. As basis for the research a single cylinder engine derived from a 1l turbocharged 3-cylinders engine is used to evaluate indirect and direct water injection. The entire engine flow field is reproduced and analyzed with 3D-CFD simulations and numerical models are employed to separate the influence of chemical and thermodynamic properties. Measurements are performed with different injectors for indirect/direct water injection in the single-cylinder engine in order to assess water break-up, wall wetting, spray interaction and penetration. Several injection strategies, such as varying start of injection, injection pressure, and water to fuel ratio, are tested at the single-cylinder engine test bench. Detailed gas phase chemistry is employed to link flame front speed with water concentration and knocking occurrence. These results are correlated with the 3D-CFD simulation of mixture formation, in-cylinder flow and water distribution for two different operating points (part load and maximum power) in order to study water behavior, with focus on the evaporation process, in-cylinder pressure and temperature profile, as well as the combustion development, during multiple engine cycles. KW - Water Injection KW - Computational Fluid Dynamics KW - Simulation KW - Virtual Test Bench KW - Thermodynamics Y1 - 2019 UR - https://saemobilus.sae.org/content/2019-24-0102 U6 - https://doi.org/10.4271/2019-24-0102 PB - SAE International CY - Neapel ER - TY - GEN A1 - Franken, Tim A1 - Seidel, Lars A1 - Matrisciano, Andrea A1 - Mauß, Fabian A1 - Kulzer, Andre Casal A1 - Schuerg, Frank T1 - Analysis of the Water Addition Efficiency on Knock Suppression for Different Octane Ratings T2 - SAE World Congress N2 - Water injection can be applied to spark ignited gasoline engines to increase the Knock Limit Spark Advance and improve the thermal efficiency. The Knock Limit Spark Advance potential of 6 °CA to 11 °CA is shown by many research groups for EN228 gasoline fuel using experimental and simulation methods. The influence of water is multi-layered since it reduces the in-cylinder temperature by vaporization and higher heat capacity of the fresh gas, it changes the chemical equilibrium in the end gas and increases the ignition delay and decreases the laminar flame speed. The aim of this work is to extend the analysis of water addition to different octane ratings. The simulation method used for the analysis consists of a detailed reaction scheme for gasoline fuels, the Quasi-Dimensional Stochastic Reactor Model and the Detonation Diagram. The detailed reaction scheme is used to create the dual fuel laminar flame speed and combustion chemistry look-up tables. The Detonation Diagram is used as a novel approach in the Quasi-Dimensional Stochastic Reactor Model to evaluate the auto-ignition characteristic in the end gas and determine if it is a harmless deflagration or developing detonation. First, the Quasi-Dimensional Stochastic Reactor Model is trained for three engine operating points and a RON95 E10 fuel. Its performance is evaluated based on experimental results of a single cylinder research engine. Subsequently, different spark timings and water-fuel ratios are investigated for different Primary Reference Fuels. The results outline that water addition can effectively reduce the strength of auto-ignition in the end gas for different Primary Reference Fuels. Thereby, it can be stated that the reduction of the auto-ignition strength through water addition by 50 – 80 % water-fuel ratio for high octane number fuels corresponds to the spark timing delay of 6 °CA or an increase of research octane number by 10 points. KW - Gasoline KW - Knock KW - Water KW - Engines KW - Combustion KW - Simulation Y1 - 2020 U6 - https://doi.org/10.4271/2020-01-0551 SN - 2688-3627 SN - 0148-7191 ER - TY - GEN A1 - Franken, Tim A1 - Matrisciano, Andrea A1 - Sari, Rafael A1 - Robles, Alvaro Fogue A1 - Monsalve-Serrano, Javier A1 - Pintor, Dario Lopez A1 - Pasternak, Michal A1 - Garcia, Antonio A1 - Mauß, Fabian T1 - Modeling of Reactivity Controlled Compression Ignition Combustion Using a Stochastic Reactor Model Coupled with Detailed Chemistry T2 - SAE technical papers : 15th International Conference on Engines & Vehicles N2 - Advanced combustion concepts such as reactivity controlled compression ignition (RCCI) have been proven to be capable of fundamentally improve the conventional Diesel combustion by mitigating or avoiding the soot-NOx trade-off, while delivering comparable or better thermal efficiency. To further facilitate the development of the RCCI technology, a robust and possibly computationally efficient simulation framework is needed. While many successful studies have been published using 3D-CFD coupled with detailed combustion chemistry solvers, the maturity level of the 0D/1D based software solution offerings is relatively limited. The close interaction between physical and chemical processes challenges the development of predictive numerical tools, particularly when spatial information is not available. The present work discusses a novel stochastic reactor model (SRM) based modeling framework capable of predicting the combustion process and the emission formation in a heavy-duty engine running under RCCI combustion mode. The combination of physical turbulence models, detailed emission formation sub-models and stateof-the-art chemical kinetic mechanisms enables the model to be computationally inexpensive compared to the 3D-CFD approaches. A chemical kinetic mechanism composed of 248 species and 1428 reactions was used to describe the oxidation of gasoline and diesel using a primary reference fuel (PRF)mixture and n-heptane, respectively. The model is compared to operating conditions from a single-cylinder research engine featuring different loads, speeds, EGR and gasoline fuel fractions. The model was found to be capable of reproducing the combustion phasing as well as the emission trends measured on the test bench, at some extent. The proposed modeling approach represents a promising basis towards establishing a comprehensive modeling framework capable of simulating transient operation as well as fuel property sweeps with acceptable accuracy. KW - Stochastic Reactor Models KW - RCCI KW - Chemical Kinetics KW - Low Temperature Combustion Y1 - 2021 UR - https://www.sae.org/publications/technical-papers/content/2021-24-0014/ U6 - https://doi.org/10.4271/2021-24-0014 SN - 0148-7191 SN - 2688-3627 ER - TY - GEN A1 - Franken, Tim A1 - Srivastava, Vivek A1 - Lee, Sung-Yong A1 - Heuser, Benedikt A1 - Shrestha, Krishna Prasad A1 - Seidel, Lars A1 - Mauß, Fabian ED - Xandra, Margot ED - Payri, Raúl ED - Serrano, José Ramón T1 - Numerical Analysis of the Combustion of Diesel, Dimethyl Ether, and Polyoxymethylene Dimethyl Ethers (OMEn, n=1-3) Using Detailed Chemistry T2 - THIESEL 2022 : Conference on Thermo- and Fluid-Dynamics of Clean Propulsion Powerplants, 13th-16th September 2022 : conference proceedings N2 - New types of synthetic fuels are introduced in internal combustion engine applications to achieve carbon-neutral and ultra-low emission combustion. Dimethyl Ether (DME) and Polyoxymethylene Dimethyl Ethers (OMEn) belong to such kind of synthetic fuels. Recently, Shrestha et al. (2022) have developed a novel detailed chemistry model for OMEn (n=1-3) to predict the ignition delay time, laminar flame speed and species formation for various thermodynamic conditions. The detailed chemistry model is applied in the zero dimensional (0D) stochastic reactor model (DI-SRM) to investigate the non-premixed combustion in a 2-liter diesel engine. Further insights in the formation of unburned hydrocarbons (HC), carbon monoxide and nitrogen oxides during the combustion of OMEn fuels are obtained in this work. The combustion and emission formation of DME and OMEn (n=1-3) are investigated and compared to conventional Diesel combustion. The mixture formation is governed by an earlier vaporization of the DME and OMEn fuels, faster homogenization of the respective air-fuel mixture and higher reactivity. At the same injection pressure, the OMEn fuels obtain higher NOx but lower CO and HC emissions. High amounts of aromatics, ethene, methane formaldehyde and formic acid are found within the Diesel exhaust gas. The DME and OMEn exhaust gas contains higher fractions of formaldehyde and formic acid, and fractions of methane, methyl formate and nitromethane. KW - Polyoxymethylene Dimethyl Ethers KW - Stochastic Reactor Model KW - Detailed Chemistry KW - Modelling KW - Emissions Y1 - 2022 UR - https://www.lalibreria.upv.es/portalEd/UpvGEStore/products/p_6328-1-1 SN - 978-84-1396-055-5 U6 - https://doi.org/10.4995/Thiesel.2022.632801 PB - Editorial Universitat Politècnica de València CY - València ER - TY - GEN A1 - Franken, Tim A1 - Shrestha, Krishna Prasad A1 - Seidel, Lars A1 - Mauß, Fabian ED - Sens, Marc T1 - Effect of Gasoline–Ethanol–Water Mixtures on Auto-Ignition in a Spark Ignition Engine T2 - International Conference on Knocking in Gasoline Engines N2 - The climate protection plan of the European Union requires a significant reduction of CO2 emissions from the transportation sector by 2030. Today ethanol is already blended by 10vol-% in gasoline and further increase of the ethanol content to 20vol-% is discussed. During the ethanol production process, distillation and molecular sieving is required to remove the water concentration to achieve high-purity ethanol. However, hydrous ethanol can be beneficial to suppress knock of spark ignition engines. The hygroscopic nature of ethanol can allow to increase the water content in gasoline – water emulsions even more, without adding additional surfactants, and improve the thermal efficiency by optimized combustion phasing, while keeping the system complexity low. Hence, the effect of gasoline – ethanol – water mixtures on the auto-ignition in a single-cylinder spark ignition engine is investigated by using multi-dimensional simulation and detailed chemistry. The gasoline – ethanol mixtures are defined to keep the Research Octane Number constant, while the Motored Octane Number is decreasing. In total five surrogates are defined and investigated: E10 (10vol-% ethanol-in-gasoline), E20, E30, E70 and E100. The water content is determined according to experimentally defined ternary diagrams that evaluated stable gasoline – ethanol – water emulsion at different gasoline – ethanol blending ratios. The auto-ignition modes of the surrogates are analyzed using the diagram, which determines if hotspots are within harmless deflagration or harmful developing detonation regime. The strongest auto-ignition is observed for the E10 surrogate, while increasing ethanol content reduces the surrogate reactivity and increases the resonance parameter. No auto-ignition of the unburnt mixture is observed for the E70 and E100 surrogates. The addition of hydrous ethanol decreased the excitation time of the surrogates, especially at low ethanol content, wherefor the reactivity parameter is significantly increased. The hotspots for E10, E20 and E30 surrogates with hydrous ethanol are found within the developing detonation regime, while hotspots of the E70 surrogate with hydrous ethanol are found in the transition regime. For the hydrous E100 surrogate no auto-ignition is predicted because of reduced temperature of the unburnt mixture due to water vaporization, which outweighs the increased reactivity due to water vapor addition. KW - Knock KW - Gasoline KW - Ethanol KW - Simulation KW - Detailed Chemistry KW - Spark Ignition Y1 - 2022 SN - 978-3-8169-3544-5 U6 - https://doi.org/10.24053/9783816985440 SP - 175 EP - 222 PB - expert CY - Tübingen ER - TY - GEN A1 - Siddareddy, Reddy Babu A1 - Franken, Tim A1 - Pasternak, Michal A1 - Leon de Syniawa, Larisa A1 - Oder, Johannes A1 - Rottengruber, Hermann A1 - Mauß, Fabian T1 - Real-Time Simulation of CNG Engine and After-Treatment System Cold Start. Part 1: Transient Engine-Out Emission Prediction Using a Stochastic Reactor Model T2 - SAE Technical Paper N2 - During cold start of natural gas engines, increased methane and formaldehyde emissions can be released due to flame quenching on cold cylinder walls, misfiring and the catalyst not being fully active at low temperatures. Euro 6 legislation does not regulate methane and formaldehyde emissions. New limits for these two pollutants have been proposed by CLOVE consortium for Euro 7 scenarios. These proposals indicate tougher requirements for aftertreatment systems of natural gas engines. In the present study, a zero-dimensional model for real-time engine-out emission prediction for transient engine cold start is presented. The model incorporates the stochastic reactor model for spark ignition engines and tabulated chemistry. The tabulated chemistry approach allows to account for the physical and chemical properties of natural gas fuels in detail by using a-priori generated laminar flame speed and combustion chemistry look-up tables. The turbulence-chemistry interaction within the combustion chamber is predicted using a K-k turbulence model. The optimum turbulence model parameters are trained by matching the experimental cylinder pressure and engine-out emissions of nine steady-state operating points. Subsequently, the trained engine model is applied for predicting engine-out emissions of a WLTP passenger car engine cold start. The predicted engine-out emissions comprise nitrogen oxide, carbon monoxide, carbon dioxide, unburnt methane, formaldehyde, and hydrogen. The simulation results are validated by comparing to transient engine measurements at different ambient temperatures (-7°C, 0°C, 8°C and 20°C). Additionally, the sensitivity of engine-out emissions towards air-fuel-ratio (λ=1.0 and λ=1.3) and natural gas quality (H-Gas and L-Gas) is investigated. KW - Spark Ignition Engines KW - Gas Engines KW - Alternative Fuel Engines KW - Natural Gas KW - Nitrogen Oxides KW - Cold Start KW - Carbon Monoxide KW - Methane KW - Formaldehyde KW - Simulation KW - Stochastic Reactor Model KW - Tabulated Chemistry Y1 - 2023 U6 - https://doi.org/10.4271/2023-01-0183 SN - 2688-3627 SN - 0148-7191 ER - TY - GEN A1 - Leon de Syniawa, Larisa A1 - Siddareddy, Reddy Babu A1 - Oder, Johannes A1 - Franken, Tim A1 - Günther, Vivien A1 - Rottengruber, Hermann A1 - Mauß, Fabian T1 - Real-Time Simulation of CNG Engine and After-Treatment System Cold Start. Part 2: Tail-Pipe Emissions Prediction Using a Detailed Chemistry Based MOC Model T2 - SAE Technical Report N2 - In contrast to the currently primarily used liquid fuels (diesel and gasoline), methane (CH4) as a fuel offers a high potential for a significant reduction of greenhouse gas emissions (GHG). This advantage can only be used if tailpipe CH4 emissions are reduced to a minimum, since the GHG impact of CH4 in the atmosphere is higher than that of carbon dioxide (CO2). Three-way catalysts (TWC - stoichiometric combustion) and methane oxidation catalysts (MOC - lean combustion) can be used for post-engine CH4 oxidation. Both technologies allow for a nearly complete CH4 conversion to CO2 and water at sufficiently high exhaust temperatures (above the light-off temperature of the catalysts). However, CH4 combustion is facing a huge challenge with the planned introduction of Euro VII emissions standard, where stricter CH4 emission limits and a decrease of the cold start starting temperatures are discussed. The aim of the present study is to develop a reliable kinetic catalyst model for MOC conversion prediction in order to optimize the catalyst design in function of engine operation conditions, by combining the outputs from the predicted transient engine simulations as inputs to the catalyst model. Model development and training has been performed using experimental engine test bench data at stoichiometric conditions as well as engine simulation data and is able to reliably predict the major emissions under a broad range of operating conditions. Cold start (-7°C and +20°C) experiments were performed for a simplified worldwide light vehicle test procedure (WLTP) driving cycle using a prototype gas engine together with a MOC. For the catalyst simulations, a 1-D catalytic converter model was used. The model includes detailed gas and surface chemistry that are computed together with catalyst heat up. In a further step, a virtual transient engine cold start cycle is combined with the MOC model to predict tail-pipe emissions at transient operating conditions. This method allows to perform detailed emission investigations in an early stage of engine prototype development. KW - Exhaust Emissions KW - Tail Pipe Emissions KW - Three Way Catalyst KW - Gas Engines KW - Cold Start KW - Simulation KW - Detailed Chemistry KW - Methane Oxidation Catalyst KW - Methane KW - Co-Simulation KW - Catalysts Y1 - 2023 U6 - https://doi.org/10.4271/2023-01-0364 SN - 2688-3627 SN - 0148-7191 ER - TY - GEN A1 - Siddareddy, Reddy Babu A1 - Franken, Tim A1 - Leon de Syniawa, Larisa A1 - Pasternak, Michal A1 - Prehn, Sascha A1 - Buchholz, Bert A1 - Mauß, Fabian T1 - Simulation of CNG Engine in Agriculture Vehicles. Part 1: Prediction of Cold Start Engine-Out Emissions Using Tabulated Chemistry and Stochastic Reactor Model T2 - SAE Technical Paper N2 - Worldwide, there is the demand to reduce harmful emissions from non-road vehicles to fulfill European Stage V+ and VI (2022, 2024) emission legislation. The rules require significant reductions in nitrogen oxides (NOx), methane (CH4) and formaldehyde (CH2O) emissions from non-road vehicles. Compressed natural gas (CNG) engines with appropriate exhaust aftertreatment systems such as threeway catalytic converter (TWC) can meet these regulations. An issue remains for reducing emissions during the engine cold start where the CNG engine and TWC yet do not reach their optimum operating conditions. The resulting complexity of engine and catalyst calibration can be efficiently supported by numerical models. Hence, it is required to develop accurate simulation models which can predict cold start emissions. This work presents a real-time engine model for transient engine-out emission prediction using tabulated chemistry for CNG. The engine model is based on a stochastic reactor model (SRM) which describes the in-cylinder processes of spark ignition (SI) engines including large-scale and lowscale turbulence, convective heat transfer, turbulent flame propagation and chemistry. Chemistry is described using a tabulated chemistry model which calculates the major exhaust gas emissions of CNG engines such as CO2, NOx, CO, CH4 and CH2O. By best practice, the engine model parameters are optimized by matching the experimental cylinder pressure and engine-out emissions from steady-state operating points. The engine model is trained for a non-road transient cycle (NRTC) cold start at 25°C ambient temperature and validated for a NRTC cold start at 10°C ambient temperature. The trained model is evaluated regarding their feasibility and accuracy predicting transient engineout emissions. KW - CNG engine KW - Cold start KW - Stochastic reactor model KW - Tabulated chemistry KW - Natural gas KW - Driving cycle Y1 - 2023 U6 - https://doi.org/10.4271/2023-24-0006 SN - 0148-7191 SN - 2688-3627 ER - TY - GEN A1 - Leon de Syniawa, Larisa A1 - Siddareddy, Reddy Babu A1 - Prehn, Sascha A1 - Günther, Vivien A1 - Franken, Tim A1 - Buchholz, Bert A1 - Mauß, Fabian T1 - Simulation of CNG Engine in Agriculture Vehicles. Part 2: Coupled Engine and Exhaust Gas Aftertreatment Simulations Using a Detailed TWC Model T2 - SAE Technical Paper N2 - In more or less all aspects of life and in all sectors, there is a generalized global demand to reduce greenhouse gas (GHG) emissions, leading to the tightening and expansion of existing emissions regulations. Currently, non-road engines manufacturers are facing updates such as, among others, US Tier 5 (2028), European Stage V (2019/2020), and China Non-Road Stage IV (in phases between 2023 and 2026). For on-road applications, updates of Euro VII (2025), China VI (2021), and California Low NOx Program (2024) are planned. These new laws demand significant reductions in nitrogen oxides (NOx) and particulate matter (PM) emissions from heavy-duty vehicles. When equipped with an appropriate exhaust aftertreatment system, natural gas engines are a promising technology to meet the new emission standards. Gas engines require an appropriate aftertreatment technology to mitigate additional GHG releases as natural gas engines have challenges with methane (CH4) emissions that have 28 times more global warming potential compared to CO2. Under stoichiometric conditions a three-way catalytic converter (TWC - stoichiometric combustion) can be used to effectively reduce emissions of harmful pollutants such as nitrogen oxides and carbon monoxide (CO) as well as GHG like methane. The aim of the present study is to understand the performance of the catalytic converter in function of the engine operation and coolant temperature in order to optimize the catalyst operating conditions. Different cooling temperatures are chosen as the initial device temperature highly affects the level of warm up emissions such that low coolant temperatures entail high emissions. In order to investigate the catalyst performance, experimental and virtual transient engine emissions are coupled with a TWC model to predict tail-pipe emissions at transient operating conditions. Engine experiments are conducted at two initial engine coolant temperatures (10°C and 25°C) to study the effects on the Non-Road Transient Cycle (NRTC) emissions. Engine simulations of combustion and emissions with acceptable accuracy and with low computational effort are developed using the Stochastic Reactor Model (SRM). Catalyst simulations are performed using a 1D catalytic converter model including detailed gas and surface chemistry. The initial section covers essential aspects including the engine setup, definition of the engine test cycle, and the TWC properties and setup. Subsequently, the study introduces the transient SI-SRM, 1D catalyst model, and kinetic model for the TWC. The TWC model is used for the validation of a NRTC at different coolant temperatures (10°C and 25°C) during engine start. Moving forward, the next section includes the coupling of the TWC model with measured engine emissions. Finally, a virtual engine parameter variation has been performed and coupled with TWC simulations to investigate the performance of the engine beyond the experimental campaign. Various engine operating conditions (lambda variation for this paper) are virtually investigated, and the performance of the engine can be extrapolated. The presented virtual development approach allows comprehensive emission evaluations during the initial stages of engine prototype development KW - CNG KW - Cold start KW - Afterteatment KW - Three-Way Catalyst KW - Surface chemistry KW - Simulation Y1 - 2023 U6 - https://doi.org/10.4271/2023-24-0112 SN - 0148-7191 SN - 2688-3627 ER - TY - GEN A1 - Pasternak, Michał A1 - Siddareddy, Reddy Babu A1 - de Syniawa, Larisa León A1 - Guenther, Vivien A1 - Picerno, Mario A1 - Andert, Jakob A1 - Franken, Tim A1 - Mauss, Fabian A1 - Adamczyk, Wojciech T1 - Plant modelling of engine and aftertreatment systems for X-in-the-loop simulations with detailed chemistry T2 - CONAT 2024 International Congress of Automotive and Transport Engineering. N2 - Use of numerical simulations at early stage of engine and aftertreatment systems development helps in evaluating their different concepts and reducing the need for costly building of prototypes. In this work, we explore the feasibility of fully physical and chemical-based tool-chain for co-simulating engine in-cylinder and aftertreatment processes. Detailed gas-phase reaction kinetics and surface chemistry mechanisms are applied for the modeling of combustion, pollutants formation and aftertreatment, respectively. Engine in-cylinder performance parameters are simulated using a stochastic reactor model and multi-component fuel surrogate. The engine model is coupled with an aftertreatment model capable of simulating diesel oxidation catalyst (DOC), selective catalytic reduction catalyst, lean NOx trap, ammonia slip catalyst, and three-way catalyst. Both the engine and aftertreatment models are embedded within the Simulink framework. They work in co-simulation and are coupled using Functional Mock-up Interface (FMI) technology. The coupled framework acts as a virtual test bench that is developed given its application for X-in-the-Loop (XiL) simulations. The framework can be applied to engine steady state or transient operating conditions. Here, exemplary calculations are performed using a Model-in-the-Loop (MiL) approach. Simulations are conducted under transient conditions of Worldwide Harmonized Light Vehicle Test Cycle for a compression ignition engine coupled with a DOC. The presented framework is considered a first step towards complex engine plant modeling using detailed chemistry for the virtualization of the development of engine, fuels and aftertreatment systems. Y1 - 2024 SN - 978-3-031-77626-7 U6 - https://doi.org/10.1007/978-3-031-77627-4_14 SP - 151 EP - 163 PB - Springer Nature Switzerland CY - Cham ER - TY - GEN A1 - Asgarzade, Rufat A1 - Franken, Tim A1 - Mauss, Fabian ED - Hemaizia, Abdelkader T1 - Multi-objective optimization of oxyfuel gas engine using stochastic engine model and detailed chemistry T2 - SAE technical papers N2 - The energy transition initiatives in Germany’s renown coal mining region Lusatia have driven research into Power-to-X-to-Power technologies, where synthetic fuel is produced from renewably sourced hydrogen and captured CO2, and converted to electricity and heat through oxyfuel combustion. This work investigates the multi-objective optimization of oxyfuel gas engine using a stochastic engine model and detailed chemistry. Exhaust gas recirculation (EGR) rate, initial cylinder temperature and pressure, spark timing, piston bowl radius and depth are selected as design parameters to minimize the exhaust temperature at exhaust valve opening and indicated specific fuel consumption (ISFC) corresponding to oxyfuel operation with different dry and wet EGR rates. The optimization problem is solved for a dry EGR and four wet EGR cases with various CO2/H2O fractions, aiming to achieve comparable performance as in conventional natural gas / air operation, and energy-efficient carbon capture. The case with the lowest humidity (T10deg) had the lowest temperature of 1537 K, while the one with the highest vapor fraction (T70deg) attained the minimum 260 g/kWh ISFC. The superiority of the T10deg case is offset by much higher cooling demand (3.06 kW) for CO2 separation than that for T70deg case (0.81 kW). The constraint for combustion efficiency (>65%) limited the solution space towards high ISFC values, while the constraint for low indicated mean effective pressure (IMEP) (>7 bar) and the constraint for high IMEP (<8 bar) limited the solution space in between the two distinct clusters of feasible designs, and towards high exhaust temperature, respectively. The optimized designs from all the cases could outperform the reference case in terms of IMEP, nevertheless they fell below 31% indicated efficiency, which is associated with stoichiometric combustion. KW - Gas engines KW - Greenhouse gas emissions KW - Exhaust gas recirculation (EGR) KW - Ignition timing KW - Synthetic fuels KW - Fuel consumption KW - Carbon dioxide KW - Combustion and combustion processes KW - Natural gas KW - Engines Y1 - 2025 U6 - https://doi.org/10.4271/2025-01-0529 SN - 0148-7191 PB - SAE International CY - Warrendale, PA, United States ER -