Development and validation of a predictive SITurb combustion model for a heavy-duty turbocharged direct-injection hydrogen SI engine

Decarbonizing heavy-duty road transport requires powertrain solutions that provide extended range and high payload capacity while meeting increasingly stringent emissions regulations. Hydrogen internal combustion engines (H₂ICE) have emerged as a promising near-to mid-term alternative, benefiting from established engine manufacturing and service infrastructure. Nevertheless, the rapid flame kinetics and pronounced sensitivity to turbulence of H₂ICE pose significant challenges for conventional non-predictive combustion models typically employed in one-dimensional (1D) simulations. This study develops and validates a physics-based predictive combustion model (SITurb) for a heavy-duty, 14.8-liter, turbocharged, direct injection hydrogen spark ignition (SI) engine using GT-SUITE. A measurement-driven workflow is adopted, utilizing apparent heat release rate profiles derived from Three Pressure Analyses (TPA) to calibrate key SITurb parameters across a broad operating range. The predictive model is benchmarked against a previously calibrated non-predictive SIWiebe baseline under identical experimental boundary conditions. Both modeling approaches replicate map-level performance trends with acceptable accuracy; however, SITurb consistently demonstrates enhanced combustion fidelity. Across the validated operating points, the mean absolute errors are 0.63 % for brake torque and 0.33 % for brake-specific fuel consumption (BSFC). Prediction of peak firing pressure is notably improved, with a mean absolute error of 0.93 bar compared to 2.36 bar for SIWiebe. These results indicate that TPA-informed calibration enhances the physical reliability of 1D predictive combustion modeling, providing a more robust numerical foundation for pressure-limited assessments, durability evaluations, and calibration-focused design studies in heavy-duty hydrogen spark-ignition applications.