TY - GEN A1 - Franken, Tim A1 - Duggan, Alexander A1 - Matrisciano, Andrea A1 - Lehtiniemi, Harry A1 - Borg, Anders A1 - Mauß, Fabian T1 - Multi-Objective Optimization of Fuel Consumption and NO x Emissions with Reliability Analysis Using a Stochastic Reactor Model T2 - SAE Technical Paper N2 - 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. Y1 - 2019 U6 - https://doi.org/10.4271/2019-01-1173 SN - 0148-7191 SN - 2688-3627 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 - Vadivala, Monang A1 - Franken, Tim A1 - Thapa, Ashish A1 - Mauss, Fabian T1 - Evaluation of metamodels for prediction of species concentration and reactor outlet temperature of Sabatier reactor N2 - The production of green gases using Power-to-gas in industry and the energy sector is essential for reducing the carbon footprint. In this process, green hydrogen and carbon dioxide are converted into synthetic methane using nickel (Ni) catalysts. A one-dimensional (1D) model of a Sabatier reactor enables the simulation of transport processes in the porous medium and reaction kinetics on the Ni/Al2O3 catalyst. KW - Metamodel KW - Methane Synthesis KW - Gaussian Processes KW - Neural Network KW - Random Forest KW - Autoencoder KW - Gradient Boosting Y1 - 2025 UR - https://www.researchgate.net/publication/389675809_Evaluation_of_Metamodels_for_Prediction_of_Species_Concentration_and_Reactor_Outlet_Temperature_of_Sabatier_Reactor ER - TY - GEN A1 - Franken, Tim A1 - Mauss, Fabian A1 - Sharma, Saurabh A1 - Brueger, Arnim A1 - Lepka, Marco T1 - Optimization of oxyfuel biogas combustion in combined heat and power plants : a multi-criteria study T2 - 32. Deutscher Flammentag – Paderborn, Germany: 15th – 17th September 2025 N2 - This paper investigates the influence of oxygen addition on the combustion of biogas and biomethane in a combined heat and power plant using numerical methods. A multi-objective optimization platform was established, employing a stochastic engine model with detailed chemistry to predict oxyfuel combustion and emission formation. Additionally, a hybrid optimization algorithm, combining NSGA-II and metamodels, was utilized to conduct the optimization. The optimization results indicate that the lowest indicated specific fuel consumption was achieved with biomethane, while the lowest NOx emissions were attained with biogas. An increase in oxygen addition proved beneficial for reducing specific fuel consumption. However, higher oxygen addition rates resulted in increased NOx emissions. KW - Biogas KW - Optimization KW - Oxyfuel Y1 - 2025 ER - TY - GEN A1 - Asgarzade, Rufat A1 - Franken, Tim A1 - Mauss, Fabian T1 - Experimental investigation of CH4/O2/CO2 mixtures in a single-cylinder spark ignition engine N2 - The Power-to-X-to-Power (P2X2P) technology involves producing synthetic methane from renewable hydrogen and captured CO2, which is then used for cogeneration of electricity and heat through oxyfuel combustion. With the P2X2P energy system demonstrator, NOx-free and carbon neutral heat and electricity generation as well as storage of excess renewable energy are realized. This work presents the experimental investigation of combustion characteristics for CH4/O2/CO2 mixtures in a single cylinder spark ignition engine that is a part of the P2X2P system. Y1 - 2025 ER - TY - GEN A1 - Asgarzade, Rufat A1 - Franken, Tim A1 - Mauss, Fabian T1 - Development of an oxyfuel engine test bench for power-to-X-to-power application T2 - 12th European Combustion Meeting N2 - This work presents the development of an oxyfuel engine test bench which is integrated into a Power-to-X-to-Power energy storage system demonstrator. These storage systems are considered carbon-free because they recirculate carbon without emitting it into the atmosphere. Y1 - 2025 ER - TY - GEN A1 - Franken, Tim A1 - Rachow, Fabian A1 - Charlafti, Evgenia A1 - Flege, Jan Ingo A1 - Jenssen, Martin A1 - Verma, Rakhi A1 - Günther, Vivien A1 - Mauss, Fabian T1 - Numerical investigation of oxy-methane combustion for stationary engines T2 - 40th International Symposium on Combustion N2 - This work presents a numerical investigation of turbulent oxyfuel combustion of methane in a gas engine with passive pre-chamber. The experimental data of a motored operating point at 1600 rpm and natural gas fired operating point at 2450 rpm, 6 bar IMEP and λ=1.5 are provided by TU Freiberg to validate the simulation model. The performance of the detailed chemistry model of Shrestha et al. predicting laminar burning velocity of premixed methane-oxygen flames is evaluated using the experiments of Mouze-Mornettas et al. The detailed chemistry model predicts the laminar flame speed within an accuracy range of ±10% for elevated pressure, temperature, and different equivalence ratios. For predicting the turbulent combustion in the gas engine, a three-dimensional (3D) Large Eddy Simulation (LES) with G Equation model and laminar flame speed look-up tables is used. The chemistry in the unburnt and burnt gas is solved using a constant volume detailed chemistry solver. The 3D LES model shows a good match of the motored and natural gas fired in-cylinder pressure profile. Subsequently the fuel is switched to methane and oxygen is used as oxidizer. The 3D LES results show an increase of maximum cylinder pressure up to 100 bar for λ=1.5, and the turbulent flame regime is shifted towards high Damköhler numbers compared to combustion with air. Diluting the cylinder gas with 50 mole-% CO2 or 65 mole-% H2O shows a significant reduction of peak cylinder pressure, and lower Damköhler and higher Karlovitz numbers compared to methane-oxygen combustion. KW - Oxyfuel KW - Simulation KW - Engines Y1 - 2024 ER -