Dual-fuel (DF) engines, in which premixed natural gas and air in an open-type combustion chamber is ignited by diesel-fuel pilot sprays, have been more popular for marine use than pre-chamber spark ignition (PCSI) engines because of their superior durability. However, control of ignition and combustion in DF engines is more difficult than in PCSI engines. In this context, this study focuses on the ignition stability of n-heptane pilot-fuel jets injected into a compressed premixed charge of natural gas and air at low-load conditions. To aid understanding of the experimental data, chemical-kinetics simulations were carried out in a simplified engine-environment that provided insight into the chemical effects of methane (CH4) on pilot-fuel ignition. The simulations reveal that CH4 has an effect on both stages of n-heptane autoignition: the small, first-stage, cool-flame-type, low-temperature ignition (LTI) and the larger, second-stage, high-temperature ignition (HTI). As the ratio of pilot-fuel to CH4 entrained into the spray decreases, the initial oxidization of CH4 consumes the OH radicals produced by pilot-fuel decomposition during LTI, thereby inhibiting its progression to HTI. Using imaging diagnostics, the spatial and temporal progression of LTI and HTI in DF combustion are measured in a heavy-duty optical engine, and the imaging data are analyzed to understand the cause of severe fluctuations in ignition timing and combustion completeness at low-load conditions. Images of cool-flame and hydroxyl radical (OH*) chemiluminescence serve as indicators of LTI and HTI, respectively. The cycle-to-cycle and spatial variation in ignition extracted from the imaging data are used as key metrics of comparison. The imaging data indicate that the local concentration of the pilot-fuel and the richness of the surrounding natural-gas air mixture are important for LTI and HTI, but in different ways. In particular, higher injection pressures and shorter injection durations increase the mixing rate, leading to lower concentrations of pilot-fuel more quickly, which can inhibit HTI even as LTI remains relatively robust. Decreasing the injection pressure from 80 MPa to 40 MPa and increasing the injection duration from 500 µs to 760 µs maintained constant pilot-fuel mass, while promoting robust transition from LTI to HTI by effectively slowing the mixing rate. This allows enough residence time for the OH radicals, produced by the two-stage ignition chemistry of the pilot-fuel, to accelerate the transition from LTI to HTI before being consumed by CH4 oxidation. Thus from a practical perspective, for a premixed natural gas fuel–air equivalence-ratio, it is possible to improve the “stability” of the combustion process by solely manipulating the pilot-fuel injection parameters while maintaining constant mass of injected pilot-fuel. This allows for tailoring mixing trajectories to offset changes in fuel ignition chemistry, so as to promote a robust transition from LTI to HTI by changing the balance between the local concentration of the pilot-fuel and richness of the premixed natural gas and air. This could prove to be a valuable tool for combustion design to improve fuel efficiency or reduce noise or perhaps even reduce heat-transfer losses by locating early combustion away from in-cylinder walls.
Although accumulated in-cylinder soot can be measured by various optical techniques, discerning soot formation rates from oxidation rates is more difficult. Various optical measurements have pointed toward ways to affect in-cylinder soot oxidation, but evidence of effects of operational variables on soot formation is less plentiful. The formation of soot and its precursors, including polycyclic aromatic hydrocarbons (PAHs), are strongly dependent on temperature, so factors affecting soot formation may be more evident at low-temperature combustion conditions. Here, in-cylinder PAHs are imaged by planar laser-induced fluorescence (PAH-PLIF) using three different excitation wavelengths of 355, 532, and 633 nm, to probe three different size-classes of PAH from 2-3 to 10+ rings. Simultaneous planar laser-induced incandescence of soot (soot-PLII) using 1064-nm excitation provides complementary imaging of soot formation near inception. To achieve low combustion temperatures at the threshold of PAH and soot formation, the engine operating conditions are highly diluted, with intake-O2 mole-fractions as low as 7.5%. The optical diagnostics show that increasing dilution delays the inception of PAH by over 2.5 ms as the intake-O2 mole-fraction decreases from 15.0% to 9.0%. At 7.5% intake-O2, no large PAH or soot are formed, while the 9.0% intake-O2 condition forms PAH but virtually no detectable soot. Conditions with 10.0% or more intake-O2 form both PAH and soot. For the threshold-sooting condition with 10.0% intake-O2, large PAH typically forms broadly throughout the cross-section of the downstream jets and along the bowl-wall. Soot appears after PAH, and in narrower ribbons in the jet-jet interaction region. These soot ribbons are on the periphery of the PAH, near the diffusion flame, where the highest temperatures are expected. With increasing intake-O2, the delay time between soot and PAH shortens, and soot tends to shift upstream to the jet region prior to wall impingement, though still on the periphery of the PAH. The spatial distributions of PAH and soot overlap slightly under these threshold-sooting conditions, with soot typically surrounding the PAH. This may indicate that temperatures are only high enough for soot formation on the jet periphery, near the diffusion flame. The minimal overlap also suggests that PAHs are rapidly consumed and/or adsorbed when soot is formed. Additionally, increasing the fuel-injection pressure from 533 to 800 and then to 1200 bar increases soot and large PAH formation, which is opposite to the trend for conventional diesel combustion.
The high-level objective of this project is to solve national-s ecurity problems associated with petroleum use, cost, and environmental impacts by enabling more efficient use of natural-gas-fueled internal co mbustion engines. An improved sci ence-base on end-gas autoignition, or "knock," is re quired to support engineering of more efficient engine designs through predictive modeling. An existing optical diesel engine facility is retrofitted for natural gas fueling with laser-spark-ignition c ombustion to provide in- cylinder imaging and pressure data under knocking combustion. Z ero-dimensional chemical-kinetic modeling of aut oignition, adiabatically constr ained by the measured cylinder pressure, isolates the role of autoignition chemistry. OH* chemiluminescence imaging reveals six different c ategories of knock onset that de pend on proximity to engine surfaces and the in-cylinder deflagration. Modeling resu lts show excellent prediction regardless of the knoc k category, thereby validating state-of-the-art kinetic mechanisms. The results also provide guidance for future work t o build a science base on the factors that affect the deflagration rate.