Computational fluid dynamics and Machine learning-based Piston-Bowl optimization for Energy-Assisted compression ignition of low cetane number sustainable aviation fuel blends

Abstract

Hybrid-electric powertrains have been gaining tremendous interest to propel unmanned aerial vehicles (UAVs) for Advanced Air Mobility. Present UAV powertrains are designed with internal combustion engines capable of combusting single fuel types, which have significant restrictions on fuel ignition quality. However, sustainable aviation fuels (SAFs) can exhibit wide variance in ignitability with cetane numbers (CNs) ranging from 15 to greater than 50. Coupled with high-speed and high-altitude engine operating conditions, low CN jet fuels can pose a serious challenge to achieving reliable, clean, and fuel-flexible ignition. Energy-Assisted Compression Ignition (EACI) has demonstrated superior ignition-controlling capability with jet fuels due to the precise thermal energy deposition achieved from the continuously operating high-temperature Ignition-Assistant (IA). However, the current literature highlights enhancement needs in the operational capabilities of EACI strategy with low CN jet fuels. Therefore, this paper aims to demonstrate improved EACI characteristics of low and varying ignition-quality SAF blends used in high-altitude aerial conditions through a combined computational fluid dynamic (CFD) and machine learning (ML)-based piston-bowl optimization methodology. In this work, a CFD-enabled Design-of-experiments approach, covering nine piston-bowl and three operational parameters, is used to develop a Gaussian Process Regression (GPR)-based ML modeling routine. Further, a merit-function-driven piston-bowl optimization is performed using the GPR model for improved EACI operations with CN 30 SAF blend. The combined CFD and GPR methodology helped screen more than 20 million piston designs to find the optimum. The study found that the piston optimization allowed consistent improvements in EACI operability for CN 30 to CN 48 jet fuels with maximum reductions in ignition delays of 17% and CA50 (50% mass-burn fraction crank angle) of 22% for CN 30 SAF blend compared to the stock piston. The optimized piston-bowl also helped reduce IA input power requirements by 20–28%, thus improving the reliability and capabilities of EACI to achieve multi-jet fuel operation.