Optimizing Channels for Non-dispersive Infrared Sensors for Fuel Property Prediction

The variability of jet fuels, amplified by the increasing use of Sustainable Aviation Fuels (SAFs), presents significant challenges for ensuring consistent engine performance and operational efficiency. Key properties such as Derived Cetane Number (DCN) and density directly impact engine control, fuel-air mixing, and combustion stability. DCN, which measures ignition propensity, is critical for optimizing parameters like injection timing and pressure, especially in engines operating with SAFs that exhibit a wider property range than conventional fuels. Similarly, fuel density influences energy content, atomization, and emissions, making its accurate measurement essential for fuel loading and flight performance. Traditional methods for measuring DCN and density, such as the Ignition Quality Tester (IQT) and oscillating U-tube method, rely on large, benchtop systems unsuitable for real-time or onboard applications. To address this limitation, this dissertation explores the development of miniaturized, real-time fuel property sensors using Non-Dispersive Infrared (NDIR) spectroscopy. Advances in Filter Array Detector Array (FADA) technology enable NDIR systems to achieve narrow-channel infrared measurements, detecting key functional groups responsible for DCN and density. This research focuses on optimizing spectral channel selection to enhance prediction accuracy while adhering to size, weight, and power (SWaP) constraints. The proposed system integrates NDIR-based sensors with feed-forward engine control models to ensure efficient and adaptable operation across diverse fuel types, including SAFs. This work provides a foundation for compact, high-performance fuel sensors that enable next-generation aviation engines to operate reliably in a variable and multi-fuel environment.