The use of jet fuels and synthetic aviation fuels (SAFs) in CI engines for aircraft applications presents significant challenges. These fuels have no minimum requirement for cetane number, making conventional compression-ignition engine operation difficult. Furthermore, aircraft operate at high altitudes, where reduced ambient temperatures and pressures result in low in-cylinder gas densities and temperatures, leading to long ignition delays, high rates of pressure rise, cycle-to-cycle instability, and higher potential for misfire and flameout. To operate CI engines under these challenging conditions, an approach to ensure reliable and complete combustion is needed. The local deposition of energy in-cylinder to ignite one or more of the direct-injected fuel jets, termed energy-assisted compression-ignition (EACI), is a potential solution to these problems. The current work investigates different approaches to deposit energy and ignite one or more of the fuel jets during EACI combustion.
The primary objective of this dissertation was to investigate the fundamental physics and operational limitations of three distinct ignition assistance approaches: hot surface (glow plug), inductive discharge (spark plug), and laser-induced plasma ignition. Specifically, this work identifies the limitations of commercial-off-the-shelf (COTS) glow plugs, evaluates the ignition approach requirements necessary to repeatedly ignite high-pressure fuel jets, and performs a direct head-to-head comparison of these three approaches using low-reactivity SAFs.
The experimental campaign was conducted using two platforms. First, a modified 2.0-L 4-cylinder metal engine converted for single-cylinder operation was utilized to study the engine speed and fuel reactivity limits of a state-of-the-art COTS ceramic glow plug. Next, an optically accessible single-cylinder engine featuring a Bowditch piston extension and a UV-grade fused silica window was used to visualize the initial ignition and combustion processes for the three ignition approaches. High-speed OH* chemiluminescence imaging was used to temporally and spatially resolve the ignition and subsequent combustion process for the three approaches.
Results from the single-cylinder metal engine experiments demonstrated the limitations of hot surface ignition using COTS ceramic glow plugs at elevated engine speeds. For single injection operation, as engine speed surpassed 1500 RPM, increased convective cooling reduced glow plug surface temperatures below the 1300 K threshold needed for rapid ignition, reducing the operation range when running low-reactivity fuels (cetane number of 25). The impact of using an EACI split-injection strategy at moderate load and 1200 RPM was also evaluated. The pilot injection ignited by the glow plug decreased the main injection’s ignition delay, facilitating mixing-controlled combustion (MCC) of the main injection and increasing combustion efficiency of low cetane number (CN) jet fuels (CN 17 and 25). Optimizing the pilot-main injection dwell increased the extent of rapid ignition and mix-controlled combustion of the main injection.
The metal engine study results motivated the optical engine investigation of the 3 ignition approaches and the sensitivity of ignition and combustion to different design parameters for those approaches. The first of those to be studied was hot surface ignition using a COTS ceramic glow plug operated at 75 W with a surface temperature of 1450 K. Two glow plug physical locations were studied at 12-mm and 17-mm downstream of the injection orifice. At the 17-mm downstream location, the distance of the glow plug tip from the fuel jet centerline was varied from 0.66 mm to 2.75 mm. The optimal depth for ignition assistance performance was found to be 1.70 mm from the jet centerline. This intermediate protrusion maximized combustion propagation while avoiding misfires.
Inductive discharge ignition was studied using a COTS J-gap spark plug at the location 17 mm downstream of the injector orifice. Spark timing sweeps determined that the spark command should be sent prior to the fuel jet reaching the ignition assistant, which allows the discharge to enter the glow phase with a sustained plasma channel prior to the fuel jet arrival. Ground strap orientation and protrusion studies showed that an unshielded 180-degree ground strap orientation with a protrusion positioning the bottom of the spark plug gap 2.75 mm from the jet centerline yielded the lowest misfire rates and the highest initial heat release rates. Protruding closer to the jet centerline increased misfire rates potentially due to the higher local flow velocities and fuel-rich conditions resulting in slower ignition kernel growth and combustion propagation.
A first-of-its-kind optical engine campaign was conducted using laser-induced plasma ignition under CI-engine conditions. The custom laser probe achieved successful ignition of high-pressure direct-injected fuel jets at very low pulse energies down to 0.21 mJ/pulse. This was made possible by coupling a short focal length aspheric lens with short (2-3 ns) pulse widths to provide high peak laser intensities. Notably, the laser probe demonstrated slightly longer ignition delays than both the glow plug and the spark plug. This increased delay was hypothesized to be due to the nearly instantaneous timescale of laser energy deposition and its small focal volume, potentially rendering the initial ignition kernel susceptible to turbulent fluctuations and quenching. However, pulsing the laser continuously at high frequencies (up to 9.6 kHz) shortened the ignition delay. Secondary laser pulses after ignition are believed to interact with the early ignition kernel through either direct energy interaction or secondary flame combination, accelerating combustion propagation.
Finally, a head-to-head comparison of the three EACI techniques was conducted at matched fuel energies and matched global equivalence ratios at two in-cylinder densities. The comparison revealed that despite significant disparities in physical size and the timescale of energy deposition, all three ignition assistants produced similar main injection combustion events for an EACI split-injection strategy. In split-injection experiments, the total number of main injection jets experiencing rapid ignition was insensitive to minor differences in the pilot injection’s combustion propagation. The tests additionally attempted to isolate the impact of in-cylinder density on combustion propagation. Higher in-cylinder densities were shown to hinder jet-to-jet combustion propagation. At higher densities, fuel jet penetration velocities are lower, potentially limiting the interaction between adjacent jets at the piston bowl wall. Ultimately, this dissertation affirms that plasma-based and hot-surface ignition assistants are viable pathways for operating low-reactivity aviation fuels in CI engines, provided that ignition systems are optimized for the operating conditions of interest and durability issues can be overcome.