Onboard dry-cell electrolysis for hydrogen-enriched internal combustion engines: Multiphysics modeling and experimental validation

This work develops and experimentally assesses a multiphysics model of a dry-cell oxyhydrogen (HHO) electrolyzer considered as the hydrogen-production component of an onboard waste-heat-recovery concept. Dedicated experiments were conducted on a four-cell dry-cell electrolyzer operated with 0.1055 mol.L−1 KOH under a 0-30 V stepwise protocol, with the quasi-steady model assessment performed over plateaus spanning 21-30 V and measured HHO flow rates up to 1.9 L.min−1. The baseline model reproduced the overall dependence of HHO production on electrical loading but retained a systematic positive discrepancy of approximately 0.09-0.14 L.min−1 under quasi-steady conditions, with local time-resolved differences reaching approximately 0.25 L.min−1. The productive-current representation was therefore revised by separating electrical, transport, species-transfer, and chemical-consumption effects according to their physical roles. A nonlinear current-efficiency relation coupled with Faraday’s law was used to determine the quasi-steady HHO production, while the remaining time-resolved intra-plateau variation was represented through a pressure-coupled gray-box formulation using measured current, electrolyte temperature, and HHO pressure. Using one parameter set determined over plateaus 3-11, the revised formulation reduced the systematic operating-point overprediction of the baseline model and represented part of the measured intra-plateau variation over the investigated domain. The electrolyzer analysis was complemented by a crank-angle-resolved thermodynamic model of an internal combustion engine under prescribed HHO-enrichment scenarios, including temperature- and composition-dependent properties, finite-duration apparent heat release, wall heat transfer, and mechanical losses. At ψ = 0.35, the model yielded relative brake-efficiency increases of 9.11% for RON98, 9.36% for RON95, 9.28% for E10, and 16.39% for E85 compared with ψ = 0. These engine results are model-based, remain distinct from the experimental assessment of the electrolyzer, and should not be interpreted as a net-energy assessment of the complete onboard conversion chain.