A Predictive 1D Modeling Framework for Reactivity-Controlled Compression Ignition Engines, via a Chemistry-Based, Multizone Combustion Object 2023-24-0001

Chemical-kinetics-based multizone models (MZM) are effective tools for performance-oriented simulations of low-temperature combustion concepts. It demonstrates a better trade-off between simulation speed and predictivity than both high-fidelity computational fluid dynamics (CFD) and low-fidelity data-driven models. This study applies a newly developed MZM, referred to as UVATZ, to simulate reactivity-controlled compression ignition (RCCI) combustion, fueled by natural gas (NG) and diesel. In such a concept, in-cylinder conditions at intake valve closing (IVC) largely define the kinetically dominated combustion predictions. To secure IVC predictions accurately, UVATZ is for the first time coupled with a detailed air/fuel path dynamics model created in commercial engine modeling software (GT-Suite), forming a 1D simulation framework. The direct coupling enables information exchange of initial and boundary conditions between the two models including IVC and EVO thermodynamic state, wall temperature boundary conditions, residual gas fraction, and composition. The closed part of the engine cycle is simulated using UVATZ, which has a representative zone arrangement for reactivity stratification, predictive interzonal mixing. A kinetic mechanism with 54 species and 269 reactions for combustion and emissions. The coupled model is calibrated and validated using experimental data from a single-cylinder research engine, representing the commercial Wärtsilä 31DF series marine, mid-speed engines. Model validation is performed against a selection of six test cases covering fully-premixed RCCI calibration. The results prove good conformance to in-cylinder and airpath pressure signals, and performance quantifiers within 2% error to measurements. Use of a detailed 1D airpath approach over the commonly used 0D plenum models is also justified from the perspective of chemical kinetics based LTC simulations.