From case-specific to general tuning: A structured, system-level framework for scalable RCCI simulation

Multi-zone models (MZMs) are widely used for system-level simulation of reactivity-controlled compression ignition (RCCI) engines. However, their predictive robustness remains strongly dependent on CFD-informed fuel stratification inputs, case-specific calibration and fuel-dependent chemical kinetic mechanisms. This limits their transferability across engine platforms and operating conditions. This study addresses these limitations by developing a structured, CFD-independent reduced-order predictive framework for scalable engine-level RCCI simulation. Implemented within the University of Vaasa’s advanced thermo-kinetic multi-zone model (UVATZ), the methodology integrates (i) systematic reactivity scaling to compensate for chemical kinetic deficiencies without disturbing thermodynamic boundary conditions, and (ii) a generalised beta-distribution-based high-reactivity fuel stratification model which eliminates reliance on CFD-derived mixture profiles. The framework is encapsulated in a modular solver architecture and uses a structured calibration procedure requiring only limited test-bench data to achieve predictive capability. The new submodels were assessed for scalability across different chemical kinetic mechanisms, and their physical consistency was verified against CFD spray simulations. The complete framework was experimentally validated using 40 operating points from a prototype Wärtsilä W31 medium-speed dual-fuel engine operating in RCCI mode. Key performance metrics (maximum pressure and indicated mean effective pressure) were predicted within ± 6%, without case-specific tuning. CA50 errors were below 5 CAD (±1 CAD standard deviation). Model deviations remained within measured cycle-to-cycle variability, and simulation time did not exceed three minutes per cycle, comparing favourably against best-in-class multi-zone models which rely on calibration to a specific operating point.