A combustion model of a hydrogen–methane–blended fuel for internal combustion
engines is developed and validated. Mixed fuels include hydrogen–methane,
octane–methanol, and octane–ethanol blends.
To address the complex dependencies of laminar flame speed of
hydrogen–methane–blended fuel on temperature, pressure, equivalence ratio, and
exhaust gas recirculation (EGR) ratio, a machine learning–based model was
constructed. Gaussian process interpolation and polynomial extrapolation were
employed to create a comprehensive laminar flame speed map. Additionally, two
flame-quenching models, wall quenching and turbulent flame stretching, were
introduced to predict unburned hydrocarbons. NOx emissions were
estimated using the extended Zel’dovich mechanism. The accuracy of these models
was verified by comparing numerical simulations with experimental data from
single-cylinder engine experiments. Results showed strong agreement for cylinder
pressure, heat release rates, and emissions across various hydrogen ratios and
engine operating conditions. Across all investigated cases, the model reproduced
combustion duration (CA10–90) within ±2.2°CA, with an error ≤11%. Notably, the
machine learning–based laminar flame speed model demonstrated high accuracy,
even at elevated temperatures and pressures, without requiring additional
parameter tuning for turbulence flame model. This study highlights the highly
accurate modeling techniques for simulating the combustion of renewable
hydrogen–methane blends. The results in this study will contribute to the
development of more efficient, lower emission internal combustion engines, and
support the transition to sustainable vehicle technology.