Probabilistic Performance Prediction of a Hydrogen-Converted SI Engine Using a Markov-Chain-Wiebe Framework

Abstrak

This study employs a novel Markov-chain modeling framework to analyze the combustion-performance interaction in a hydrogen-fueled spark-ignition engine. The methodology integrates a Wiebe heat-release model within a Markov-chain state-transition framework, where each discrete engine state defines combustion parameters and probabilistic transitions capture cycle-to-cycle variability. Results demonstrate that engine behavior is dominantly governed by combustion phasing, with spark timing exerting primary control over torque, efficiency, and brake-specific fuel consumption (BSFC). Sensitivity analysis confirms spark timing produces the steepest performance gradients, while ignition voltage offers secondary benefits and engine speed exhibits minimal influence. The model reveals a highly nonlinear torque response to spark advance, characterized by a rapid rise culminating in a narrow maximum brake torque (MBT) plateau at 8°–10° BTDC, corresponding to a distinct BSFC minimum. Significant data scatter underscores the stochastic nature of hydrogen combustion, arising from multidomain interactions between air-fuel ratio, ignition strength, and phasing. The Markov-chain approach successfully captures this coupled deterministic-probabilistic behavior, highlighting the critical need for precise spark-timing control to optimize performance in hydrogen applications.

Kata kunci

• Hydrogen internal combustion engine
• Combustion phasing
• Spark ignition
• Markov-chain
• Performance Prediction