Hydrogen conversion of a marine internal combustion engine : combustion and tribological analysis
Loading...
Date
2026-10-15
Authors
Γαλατσίδας, Παναγιώτης
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Over the last five years, the maritime sector has undergone major advancements in its approach to adopting alternative fuel options to power its global fleets. The introduction of stricter IMO rules for greenhouse gas emissions, coupled with the increasing eco-conscious agenda among industry peers, pushes the narrative of a greener marine propulsion future. Incorporating propulsion strategies centered on pure hydrogen presents its advantages as an ultimate zero emission candidate. Further evaluation of its feasibility and potential for mass adoption is needed within the array of fuel options available. Hydrogen Internal Combustion Engines, or in short H2ICEs, leverage existing IC infrastructure and are viewed as a cost-effective solution for immediate decarbonization of the field. One major merit justifying additional maritime research is mainly related to the ship’s storage capacity to accommodate the larger volumes mandated by hydrogen’s low volumetric energy density, compared to that of other land-based options like passenger cars or heavy duty vehicles. Specifically, large medium-speed engine options found in Ro-Pax ferries or cruise liners stand out as the most promising platform to run on 100% H2, because they operate on fixed routes between specific ports without needing to carry months of fuel, and could refuel more regularly and by extension reliably.
In this study, the four-stroke, single-acting, medium-speed marine engine MAN 48/60B is simulated in ANSYS Forte under two fuel conditions, 100% diesel and 100% hydrogen. Later, its first piston ring is analyzed tribologically in ANSYS Fluent for both cases. As its name suggests it has a 480 mm bore and 600 mm stroke. The engine runs at 100% load, or MCR, with a specific fuel oil consumption of 186 g/kWh. The diesel setup relies on direct injection and compression ignition, whereas hydrogen is strategically switched to spark ignition.
MAN B&W 48/60B operates at a rated speed of 514 RPM. It features improved fuel efficiency and a 14% increase in specific power output, specifically 1,200 kW/cylinder. Optimized for the purpose of limiting soot and nitrogen oxide emissions. To support this performance, it utilizes a Miller combustion cycle with IVC and EVO being -203° and +117° ATDC respectively, an elevated compression ratio equal to 15.3:1, and a double turbocharging system.
The maximum cylinder pressure, also known as firing pressure, is in the range of 190-200 bar, and moreover the fuel injection pressure is specified at 1,600 bar. Mean piston speed is calculated at 10.28 m/s. The mechanism by which the fuel is delivered relies on a cam-driven jerk pump system with a single central injector. So, the computational mesh is simplified leveraging the combustion chamber’s periodic symmetry. As literature indicates for the part MAN 4860B22104M the nozzle is configured with 14 holes, at a total included spray angle of 162°.
For the first part, the research aims to calibrate the ANSYS Forte diesel model utilizing engine specifications, and by extension combustion data. Extensive parametric runs are needed to validate the diesel variant. Afterwards, the hydrogen case is merely established on simulation data retrieved from the baseline model, with few proper adjustments. Due to the engine’s large bore and stroke, computational resources and simulation time are critical parameters to consider, despite the closed-cycle nature of the system.
In this context, the principle of geometric similarity in marine engines is introduced, achieving a reduction of the cylinder dimensions to one third of their original size. P-law methodology is selected, thus, MAN 48/60B engine is proportionally scaled down to a 16 cm bore, 20 cm stroke. In particular, this scaling principle dictates that Mean Piston Speed and Injection Pressure should be maintained invariant. Therefore, in order to achieve this, it is necessary to triple the engine rotational speed to N=1,542 RPM, from the initial value of N=514 RPM.
In the baseline diesel run, the 1/3 compression ignition scaled model has a bore of 160 mm, a displacement of 4,021 cm^3, a compression ratio of CR=15.30. The downscaled engine burns per cycle 0.536 g of n-heptane, with physical properties of n-dodecane, and a lower heating value of LHV=44.23 kJ/g. Now, the full scale engine has a Gross Indicated Power of 1,175.4 kW, Gross Indicated Work of 274.5 kJ, and IMEP of 2.528 MPa. Furthermore, Gross Indicated Thermal Efficiency is equal to 42.88% and combustion is complete. The peak of firing pressure is at 286.2 bar and the mean temperature peak stands at 1,592.6 K. The maximum pressure rise rate is 2.10 MPa/° CA. These results are attributed to the ignition delay that follows the P-law speed scaling. Of the 651.3 kJ of chemical energy released, 130.1 kJ or about 20% is transferred to the walls. The exhaust composition is quantified by a NOX emission index of 134.0 g/kg_fuel, an emission index for CO of 0.082g/kg_fuel and low emission indices for UHC and soot, 0.052 g/kg_fuel and 0.025 g/kg_fuel correspondingly.
In the hydrogen variant, the same geometry is employed as a lean spark-ignition platform with an air excess ratio of λ=2.5. The consumption rate per cycle is 0.197 g with a lower heating value of LHV=120 kJ/g. This fuel variant for the full-scale engine version increases the Gross Indicated Power to 1,271.7 kW and a Gross Indicated Work of 296.8 kJ. More importantly, the Gross Indicated Thermal Efficiency corresponds to 46.50%. Regardless of the increased power output, the mechanical loading of the piston ring pack is significantly reduced. Specifically, pressure peak drops to 224.4 bar and happens later at 14° ATDC, whilst the maximum pressure rise rate is 0.52 MPa/° CA, about one quarter of the diesel value. The hydrogen pressure trace shows two combustion peaks, a first around 14° ATDC that corresponds to the end of the main premixed combustion. And a second lower shoulder peak around 20° ATDC, where lean hydrogen that is pushed out of the squish and crevice areas repressurizes the cylinder as the piston is moving down, burning late. The mean temperature rises higher at around 1,900 K, as befits the quick premixed heat release. For the purpose of the tribological work, the combustion is virtually clean. The NOX emission index is reduced by 97% and all the other exhaust emissions are completely eliminated.
The tribological performance of the top piston ring/cylinder liner, or PRCL configuration of the downscaled model was the subject of a quasi-2-D investigation. This was performed via a combined approach of an analytical mixed-lubrication model with CFD cross validation. In detail, 1-D Reynolds equation, solved under the half-Sommerfeld condition for the hydrodynamic pressure, was coupled with the Greenwood-Tripp asperity contact model. Afterwards, it was validated with a thin-film CFD model developed in ANSYS Fluent. The simulation included a single-phase lubricant, SAE-40 grade oil under isothermal conditions at 120℃. It features a density and a dynamic viscosity of ρ=815 kg/m^3 and η=0.0061 Pa⋅s respectively. The top piston ring was represented by a parabolic profile, assigned a composite surface roughness of σ=0.25 μm, an elastic tension of p_el=0.30 MPa and a boundary friction coefficient of μ_b=0.17. Four characteristic crank angles were chosen for each fuel variant, specifically -45° ATDC, the firing pressure peak +4° ATDC for diesel and +14° ATDC for hydrogen, and +30° ATDC,+75° ATDC where the piston’s velocity is increasing markedly. The instantaneous cylinder pressure at each of the aforementioned CAs was extracted directly from the thermodynamic output of Ansys Forte. Excluding the crank angles near the TDC, only at the other three hydrodynamic stations was the CFD method applied and agreed with the analytical one within an 8.3% deviation. That is because at TDC the ring velocity is equal to zero, the oil film is thinning and the Ansys Fluent model does not account for asperity to asperity contacts. Hence, calculations from the analytical model are more accurate. To be more specific, the equilibrium of the force balance was closed with an error far below 0.1%. Scaled to full engine size by a factor of three accounting for the circumference, and the need to simulate a physically realistic ring by maintaining the axial ring height, the minimum film thickness at firing pressure CAs, rises from h_min= 0.103 μm for diesel to h_min= 0.273 μm for hydrogen. The hydrodynamic load share remains unchanged, W_h=266 kN against W_h=254 kN. This highlights the tribological advantages of operating on hydrogen at the hydrodynamic stations. Yet, as far as the calculation of the total ring friction force is concerned the 200 bar rated diesel engine performs better than hydrogen at every specific crank angle.
Description
Keywords
Hydrogen conversion of a marine ICE, Hydrogen internal combustion engine (H2ICE), Four-stroke medium-speed marine ICE, MAN B&W 48/60B, International Maritime Organization (IMO), Alternative marine fuels, Maritime decarbonization, Computantional fluid dynamics (CFD), ANSYS Forte, Closed-cycle simulation, Geometric similarity scaling for marine diesel engines (Power-Law), ANSYS Fluent, In-cylinder combustion, Diesel compression-ignition, Hydrogen spark-ignition, Pollutant emissions, Nitrogen oxide (NOx) emissions, Lower heating value (LHV), Gross indicated thermal efficiency, Top piston ring-cylinder liner tribology, Minimum lubricant film thickness, Hydrodynamic lubrication, Greenwood-Tripp asperity contact model, 1-D Reynolds equation, Analytical tribology model (Python), Ring load balance, Total friction force