<![CDATA[Liquefied petroleum gas (LPG) fuel in modern direct-injection spark-ignition (DISI) engines must be modeled carefully to predict combustion behavior. In this work, we reformulate a student project into a research manuscript by comparing isenthalpic (Joule–Thomson) versus isentropic (ideal adiabatic) expansions of liquid LPG (propane surrogate) during injection. Using REFPROP thermophysical data and MATLAB simulations, we vary fuel rail pressures (45–100 bar) and fuel temperatures (30–85 °C) to determine critical flow properties at the injector throat (Mach 1 conditions). The choking point is identified by iterating pressure drop until the Mach number reaches unity in either a single-phase or two-phase region. We compute the resulting flashing ratio (liquid volume to vapor volume) for each model. Our results show that fuel temperature has a far greater effect on the speed-of-sound drop than rail pressure across all models, with higher temperatures yielding smaller acoustic drops. Nearly all cases produce flashing ratios Rp>1 (indicating significant vaporization), except under the second isenthalpic model where Rp falls below unity. Notably, the isentropic, first-isenthalpic, and isothermal models best reproduce a reference spray flash pattern, but their flashing ratios are very similar. Thus, we cannot definitively rank one model superior. Our analysis highlights that isentropic expansion yields a larger temperature drop than isenthalpic throttling, consistent with thermodynamic theory. The isentropic and first isenthalpic models predict almost identical choked-flow velocities and speed-of-sound behavior, whereas deviations appear only under the nonideal (second isenthalpic) cases. In summary, this modeling confirms that choosing a flash expansion assumption has only a subtle effect on predicted LPG fueling, provided the two leading models are considered.]]>
