Glen C. Martin

An Experimental Investigation of the Origin of Increased NOx Emissions When Fueling a Heavy-Duty Compression-Ignition Engine with Soy Biodiesel

It is generally accepted that emissions of nitrogen oxides (NOx) increase as the volume fraction of biodiesel increases in blends with conventional diesel fuel. While many mechanisms based on biodiesel effects on in- cylinder processes have been proposed to explain this observation, a clear understanding of the relative importance of each has remained elusive. To gain further insight into the cause(s) of the biodiesel NOx increase, experiments were conducted in a single- cylinder version of a heavy-duty diesel engine with extensive optical access to the combustion chamber. The engine was operated using two biodiesel fuels and two hydrocarbon reference fuels, over a wide range of loads, and using undiluted air as well as air diluted with simulated exhaust gas recirculation. Measurements were made of cylinder pressure, spatially integrated natural luminosity (a measure of radiative heat transfer), engine-out emissions of NOx and smoke, flame lift-off length, actual start of injection, ignition delay, and efficiency. Adiabatic flame temperatures for the test fuels and a surrogate #2 diesel fuel also were computed at representative diesel-engine conditions. Results suggest that the biodiesel NOx increase is not quantitatively determined by a change in a single fuel property, but rather is the result of a number of coupled mechanisms whose effects may tend to reinforce or cancel one another under different conditions, depending on specific combustion and fuel characteristics. Nevertheless, charge-gas mixtures that are closer to stoichiometric at ignition and in the standing premixed autoignition zone near the flame lift- off length appear to be key factors in helping to explain the biodiesel NOx increase under all conditions. These differences are expected to lead to higher local and average in-cylinder temperatures, lower radiative heat losses, and a shorter, more-advanced combustion event, all of which would be expected to increase thermal NOx emissions. Differences in prompt NO formation and species concentrations resulting from fuel and jet-structure changes also may play important roles.

An Experimental Investigation of In-Cylinder Processes Under Dual-Injection Conditions in a DI Diesel Engine

Fuel-injection schedules that use two injection events per cycle (dual-injection approaches) have the potential to simultaneously attenuate engine-out soot and NO x emissions. The extent to which these benefits are due to enhanced mixing, low-temperature combustion modes, altered combustion phasing, or other factors is not fully understood. A traditional single-injection, an early-injection-only, and two dual-injection cases are studied using a suite of imaging diagnostics including spray visualization, natural luminosity imaging, and planar laser-induced fluorescence (PLIF) imaging of nitric oxide (NO). These data, coupled with heat-release and efficiency analyses, are used to enhance understanding of the in-cylinder processes that lead to the observed emissions reductions. Results show that combustion of the early-injected fuel occurs in two phases: a cool-flame phase characterized by very weak chemiluminescence, followed by a premixed-burn phase characterized by localized regions of bright soot incandescence. Combustion of the early-injected fuel liberates only a fraction of its chemical energy. Spray visualization images show that this low combustion efficiency could be due at least in part to liquid fuel penetrating to and wetting in-cylinder surfaces, but NO PLIF images of the early-injection-only case also show strong interferences from unburned fuel vapor and/or condensed fuel droplets, suggesting that incomplete bulk-gas combustion and quenching in crevices also may play roles. The traditional single-injection case produced the highest NO PLIF signal levels. Both dual-injection cases reduced NO PLIF signal levels, with the reduction being most dramatic for the retarded-main-injection case.

Using Carbon-14 Isotope Tracing to Investigate Molecular Structure Effects of the Oxygenate Dibutyl Maleate on Soot Emissions from a DI Diesel Engine

The effect of oxygenate molecular structure on soot emissions from a DI diesel engine was examined using carbon-14 ({sup 14}C) isotope tracing. Carbon atoms in three distinct chemical structures within the diesel oxygenate dibutyl maleate (DBM) were labeled with {sup 14}C. The {sup 14}C from the labeled DBM was then detected in engine-out particulate matter (PM), in-cylinder deposits, and CO{sub 2} emissions using accelerator mass spectrometry (AMS). The results indicate that molecular structure plays an important role in determining whether a specific carbon atom either does or does not form soot. Chemical-kinetic modeling results indicate that structures that produce CO{sub 2} directly from the fuel are less effective at reducing soot than structures that produce CO before producing CO{sub 2}. Because they can follow individual carbon atoms through a real combustion process, {sup 14}C isotope tracing studies help strengthen the connection between actual engine emissions and chemical-kinetic models of combustion and soot formation/oxidation processes.

The Influence of Charge-Gas Dilution and Temperature on DI Diesel Combustion Processes Using a Short-Ignition-Delay, Oxygenated Fuel

The influence of nitrogen dilution and charge-gas temperature on in-cylinder combustion processes and engine-out NO x and smoke emissions was investigated in an optically accessible heavy-duty Dl diesel engine using a high-cetane-number, oxygenated fuel. Engine-out measurements of NO x and smoke emissions and in-cylinder images of natural luminosity were obtained for charge-gas oxygen concentrations from 9% to 21% and TDC charge-gas temperatures of 680 and 880 K. Charge-gas temperature was found to have a significant influence on engine-out NO x emissions, but NO x emissions levels less than 0.2 g/ihp-hr were achieved at both the 680 and 880 K charge-gas temperatures within the investigated range of oxygen concentrations. An indicated engine-out NO x emissions level of 0.09 g/ihp-hr at 18 bar IMEP was achieved using charge-gas dilution and 3.0 bar intake boost pressure. The proportion of NO 2 to NO emissions increased with decreasing oxygen concentration, with NO 2 reaching 81% of NO x emissions at an oxygen concentration of 12%. The increasing fraction of NO 2 with decreasing oxygen concentration is attributed to increased quenching of NO 2 -to-NO reactions due to decreasing flame temperatures. Flame lift-off lengths were measured using in-cylinder images of natural luminosity. The measured flame lift-off lengths and estimated charge-gas conditions were used to determine the local mixture stoichiometry at the flame lift-off length. The results show that soot incandescence can be negligible for fuel-rich local mixture stoichiometries that would result in intense soot incandescence under undiluted operating conditions. It is hypothesized that flame temperatures and/or residence times are too small for soot inception under highly dilute charge-gas conditions. Reduced flame temperatures also potentially explain the low measured NO x emissions levels.

Two-photon nitric oxide laser-induced fluorescence measurements in a diesel engine

A two-photon nitric oxide (NO) laser-induced fluorescence (LIF) technique was developed and applied to study in-cylinder diesel combustion. The technique prevents many problems associated with in-cylinder, single-photon NO planar-laser-induced fluorescence measurements, including fluorescence interference from the Schumann-Runge bands of hot O2, absorption of a UV excitation beam by in-cylinder gases, and difficulty in rejecting scattered laser light while simultaneously attempting to maximize fluorescence signal collection. Verification that the signal resulted from NO was provided by tuning of the laser to a vibrational off-resonance wavelength that showed near-zero signal levels, which resulted from either fluorescence or interference at in-cylinder pressures of as much as 20 bar. The two-photon NO LIF signal showed good qualitative agreement with NO exhaust-gas measurements obtained over a wide range of engine loads.