A dual-fidelity framework for multicomponent aviation fuel spray simulations with GC×GC-informed liquid fuel representations

Accurate prediction of multicomponent fuel vaporization is critical for modeling spray combustion in gas turbine combustors and evaluating novel pathways for synthetic aviation turbine fuels (SATFs). We present a new dual-fidelity model for vaporizing liquid aviation fuel sprays. The model, implemented within the PeleLMeX low Mach fluid solver, couples a high-fidelity representation of thermophysical and transport of the liquid phase (provided by FuelLib) with reduced-order gas-phase species representations compatible with chemical kinetic mechanisms. The dual-fidelity framework is validated through non-reacting simulations of conventional aviation fuels in low-temperature and combustor-relevant conditions. Additional non-reacting simulations in an experimental combustor configuration demonstrate improved predictions of liquid spray penetration and spatial variations in vaporized hydrocarbon-family distributions, with notable differences observed in the vaporization of synthetic and conventional jet fuels. The dual-fidelity model is shown to dramatically improve vaporization predictions without an appreciable increase in computational costs. Novelty and significance Reacting simulations of complex aviation fuels, such as those used in commercial aviation, require reduced gas-phase representations to avoid the prohibitive computational cost associated with large chemical mechanisms, which have historically mandated a similarly reduced liquid-fuel representation. This work establishes that preferential vaporization of multicomponent fuels is governed primarily by the evolution of the liquid-phase composition and is weakly dependent on the compositional fidelity of the gas-phase representation. Consequently, accurate prediction of the vapor pressure governing preferential vaporization requires sufficient fidelity to resolve the evolving liquid composition but does not require equivalent fidelity in the gas phase. This principle is demonstrated through a dual-fidelity modeling strategy that decouples liquid- and gas-phase compositional fidelity, allowing high-fidelity liquid compositions to be coupled with reduced gas-phase kinetic mechanisms while preserving mass conservation and computational tractability.