Performance and emissions characteristics were measured for a part- load operating point using an optically-accessible single-cylinder gasoline research engine equipped with three different exploratory nanosecond repetitively pulse discharge (NRPD) igniters. The three igniters investigated are as follows: 1) a four-prong advanced corona ignition system (ACIS) that produces large ignition volumes from streamer discharges, 2) a barrier discharge igniter (BDI) that generates strong surface plasma along the insulator that completely encases the power electrode, and 3) a J-hook non-resistive nanosecond spark (NRNS) igniter. For select conditions, high-speed imaging (20 kHz) of excited state hydroxyl (OH*) chemiluminescence was performed to measure flame development in-cylinder. An available NRPD pulse generator was used to supply positive direct current (DC) pulses (~ 10 ns pulse width) to each igniter at a fixed 100 kHz frequency. The minimum pulse number (1 - 200) and primary voltage (900 - 1300 V) required to achieve stable ignition was used for each operating point. The pulse generator featured a sense and control system that would interrupt pulse trains if arc transition was detected. For all conditions, engine speed and load were fixed at 1300 revolutions per minute (rpm), and 3.5 bar indicated mean effective pressure (IMEP), respectively. Sweeps of equivalence ratio and stoichiometric charge dilution were separately investigated, with lean and dilute combustion limits identified for each igniter. Results were benchmarked against the engine equipped with a double fine wire spark igniter driven by a high- energy (93 mJ) inductive coil. For ACIS, the NRPD generated high-energy streamers led to faster early flame development due to the larger initial discharge volume. However, ignition timing advance for lean and dilute mixtures was limited by arcing propensity to the injector tip from the closest prong due to the lower in-cylinder density and the time of discharge. Conversely, ignition for the BDI occurred at the interface between the insulator surface and ground where the electric field strengths are strongest. Given the separation between the nearest ground surface and the power electrode by the insulator, arcing into the combustion chamber was never observed. Relative to the inductive coil spark igniter, both BDI and ACIS featured much more repetitive positioning of the early flame kernel. Finally, for NRNS, primary energy utilization was lowest at all equivalence or dilution ratios, while both the lean and dilute limit extension was greatest with this igniter.
In the present study, the performance and emissions characteristics of three low-temperature plasma (LTP) ignition systems were compared to a more conventional strategy that utilized a high-energy coil (93 mJ) inductive spark igniter. All experiments were performed in a single-cylinder, optically accessible, research engine. In total, three different ignition systems were evaluated: (1) an Advanced Corona Ignition System (ACIS) that used radiofrequency (RF) discharges (0.5 - 2.0 ms) to create corona streamer emission into the bulk gas via four-prong electrodes, (2) a Barrier Discharge Igniter (BDI) that used the same RF discharge waveform to produce surface LTP along an electrode encapsulated completely by the insulator, and (3) a Nanosecond Repetitive Pulse Discharge (NRPD) ignition system that used a non-resistor spark plug and positive DC pulses (~10 nanoseconds width) for a fixed frequency of 100 kHz, with the operating voltage-controlled to avoid LTP transition to breakdown. For the LTP ignition systems, pulse energy and duration (or number) were varied to optimize efficiency. A single 1300 revolutions per minute (rpm), 3.5 bar indicated mean effective pressure (IMEP) homogeneous operating point was evaluated. Equivalence ratio (ϕ) sweeps were performed that started at stoichiometric conditions and progressed toward the lean limit. Both the ACIS and NRPD ignition systems extended the lean limit (where the variation of IMEP < 3%) limit (ϕ = 0.65) compared to the inductive spark (ϕ = 0.73). The improvement was attributed to two related factors. For the ACIS, less spark retard was required as compared to spark ignition due to larger initial kernel volumes produced by four distinct plasma streamers that emanate into the bulk gas. For the NRPD ignition system, additional pulses were thought to add expansion energy to the initial kernel. As a result, initial flame propagation was accelerated, which accordingly shortens early burn rates.
The mixed-mode engine combustion strategy where some combination of spark-assisted compression ignition (SACI) and pure advanced compression ignition (ACI) are used at part-load operation with exclusive spark-ignited (SI) combustion used for high power-density conditions has the potential to increase efficiency and decrease pollutant emissions. However, controlling combustion and switching between different modes of mixed-mode operation is inherently challenging. This chapter proposes to use ozone (O3)—a powerful oxidizing chemical agent—to maintain stable and knock-free combustion across the load-speed map. The impact of 0–50 ppm intake seeded O3 on performance, and emissions characteristics was explored in a single-cylinder, optically accessible, research engine operated under lean SACI conditions with two different in-cylinder conditions, (1) partially stratified (double injection—early and late injection) and (2) homogeneous (single early injection). O3 addition promotes end gas auto-ignition by enhancing the gasoline reactivity, which thereby enabled stable auto-ignition with less initial charge heating. Hence O3 addition could stabilize engine combustion relative to similar conditions without O3. The addition of ozone has been found to reduce specific fuel consumption by up to 9%, with an overall improvement in the combustion stability compared to similar conditions without O3. For the lowest loads, the effect of adding O3 was most substantial. Specific NOx emissions also dropped by up to 30% because a higher fraction of the fuel burned was due to auto-ignition of the end gas. Measurement of in-cylinder O3 concentrations using UV light absorption technique showed that rapid decomposition of O3 into molecular (O2) and atomic oxygen (O) concurred with the onset of low-temperature heat release (LTHR). The newly formed O from O3 decomposition initiated fuel hydrogen abstraction reactions responsible for early onset of LTHR. At the beginning of high-temperature heat release (HTHR), end gas temperatures ranged from 840 to 900 K, which is about 200 K cooler than those found in previous studies where intake charge heating or extensive retained residuals were used to preheat the charge. An included analysis indicates that in order to achieve optimal auto-ignition in our engine, the spark deflagration was needed to add 10–40 J of additional thermal energy to the end gas. We have leveraged these results to broaden our understanding of O3 addition to different load-speed conditions that we believe can facilitate multiple modes (SI, ACI, SACI, etc.) of combustion.