Numerical investigation of a hydrogen-based internal combustion engine

Methods

The project combines experimental investigations with multi-scale numerical simulations to study hydrogen combustion phenomena in internal combustion engines. High-fidelity three-dimensional CFD simulations were conducted using the CONVERGE software framework to resolve in-cylinder flow, mixture formation, ignition, and combustion processes. Both Reynolds-Averaged Navier–Stokes (RANS) and Large-Eddy Simulation (LES) approaches were applied to capture turbulence–chemistry interactions across a wide range of engine operating conditions.

For hydrogen spark-ignition combustion, the STEP2 combustion model was implemented and validated to capture the influence of thermodiffusive instabilities on turbulent flame propagation. Key combustion metrics such as in-cylinder pressure, heat release rate, burn duration, and maximum pressure timing were evaluated across different engine speeds, loads, air–fuel ratios, and injection timings.

Hot-surface-induced pre-ignition was investigated through combined experiments and numerical simulations, including zero-dimensional ignition delay calculations and detailed CFD analyses of ignition near heated surfaces. For diesel engines, phase-resolved heat-release analysis was performed using a triple-Wiebe decomposition to separate premixed, diffusion, and burnout combustion phases.

The simulations involve millions of computational cells and long transient simulations over multiple engine cycles, making HPC resources essential for performing large parametric studies and resolving the relevant combustion physics.

Results

The simulations and experiments provided new insights into hydrogen combustion behavior under engine-relevant conditions. For hydrogen spark-ignition engines, the STEP2 model demonstrated accurate prediction of key combustion parameters such as maximum in-cylinder pressure, burn duration, and heat-release rates across a wide range of operating conditions. The results show that thermodiffusive instabilities significantly enhance local flame propagation. Under ultra-lean conditions, the effective turbulent flame speed can increase by up to a factor of five compared with cases without TDI effects. This mechanism enables stable combustion even at high excess air ratios and therefore provides a potential pathway toward improved efficiency.

The investigation of hot-surface-induced pre-ignition revealed strong differences between hydrogen and reference fuels such as methanol. Hydrogen exhibits an abrupt transition to full pre-ignition once a critical hot-spot temperature is reached, whereas methanol shows a gradual transition. Numerical ignition delay calculations reproduced these trends and explained the pressure-dependent ignition behavior of hydrogen mixtures.

For diesel engines, high-fidelity CFD simulations showed that mixture formation and combustion chamber geometry strongly influence the late diffusive combustion phase. Hydrogen enrichment and oxygenated fuels were found to shorten the diffusion phase and reduce soot and unburned hydrocarbon emissions while improving combustion efficiency.

Discussion

The results demonstrate that hydrogen combustion behavior in engines is governed by complex interactions between turbulence, chemical kinetics, and mixture formation. Thermodiffusive instabilities represent an important mechanism that can significantly enhance flame propagation under lean conditions, providing a potential optimization lever for hydrogen spark-ignition engines. Accurately capturing these effects requires advanced combustion models and high-resolution CFD simulations, highlighting the importance of HPC resources.

The study of hot-surface-induced pre-ignition emphasizes the challenges associated with hydrogen ignition control. Unlike other fuels, hydrogen exhibits highly non-linear ignition behavior, which makes hot-surface ignition unsuitable for compression-ignition concepts. Instead, mitigation strategies must focus on controlling mixture distribution, injection timing, and flow structures to avoid premature ignition.

In diesel engines, hydrogen co-firing shows promising potential for reducing soot emissions and enabling low-carbon operation, although the interaction between hydrogen and diesel combustion requires careful optimization. Future work will extend the simulations to larger parameter spaces using RANS and LES approaches and will incorporate advanced combustion and soot models to further improve predictive accuracy.

Overall, the project provides validated modeling tools and mechanistic insights that support the development of efficient, safe, and low-emission hydrogen combustion engines.

Additional Project Information

DFG classification: 404-03 Fluid Mechanics
Software: CONVERGE CFD
Cluster: CLAIX

Publications

Hongchao Chu, Vignesh Varatharajan, Benjamin Pehlivanlar, Terence Lehmann, Lukas Berger, Dominik Golc, Stefania Esposito, Thomas L. Howarth, Emanuele B. Porcelli, Davide Laera, Marco Günther, Michael Gauding, Joachim Beeckmann, Stefan Pischinger, Heinz Pitsch,
An extended G -equation formulation for simulating thermodiffusively unstable hydrogen flames,
https://dx.doi.org/10.1016/j.proci.2025.105945, 2025

Dominik Golc, Stefania Esposito, Mohammad Khosravi, Joachim Beeckmann, Heinz Pitsch,
Investigation of hot-spot-induced pre-ignition in methanol- and hydrogen-fueled spark-ignition engines,
https://dx.doi.org/10.1016/j.proci.2025.105932, 2025

Stefania Esposito, Enrica Malfi, Massimiliano De Felice, Dominik Golc, Joachim
Beeckmann, Heinz Pitsch, Vincenzo De Bellis,
Methanol fuelling of a spark-ignition engine: experiments and 0D/1D
predictive modelling,
Peer-reviewed publication, 2024

Aranya Dan, Stefania Esposito, B Pehlivanlar, Marco Günther, Stefan Pischinger and Heinz Pitsch,
Injection modeling and combustion simulation of a direct-injection hydrogen engine,
Peer-reviewed publication, March 2023

Thesis

Jonas Sieben,
Experimental and simulative investigation of surface-induced pre-ignition of hydrogen in combustion engines,
Master Thesis, 2025

Björn Brand,
An Experimental and Computational Approach to Modeling Flame Speeds in RANS Simulations to Predict Thermodiffusive Instabilities in Hydrogen Combustion Engines,
Master thesis, 2025

Vincenzo Scaringi,
Unsteady RANS simulations of a hydrogen-fuelled single-cylinder piston engine,
Master thesis, 2024

Vignesh Varatharajan,
A consistent G-equation formulation for simulations of thermodiffusive unstable premixed hydrogen flames,
Master thesis, 2024