A new advanced combustion mode is introduced, called Thermally Stratified Compression Ignition (TSCI), which uses direct water injection to control both the average temperature and the temperature distribution prior to ignition, thereby providing cycle-to-cycle control over the start and rate of heat release in Low Temperature Combustion (LTC). Experiments were conducted to fundamentally understand the effects of water injection on heat release in LTC. The results show that water injection retards the start of combustion due to the latent heat of vaporization of the injected water. Furthermore, for start of water injection timings between 20 and 70 degrees before top dead center, combustion is significantly elongated compared to without water injection. The 10% to 90% burn duration with 6.6 and 9.0 mg of water per cycle was 77% and 146% longer than without water injection, respectively. Direct water injection reduces the heat release rate by local evaporative cooling that results in a forced thermal stratification. Finally, the load limits with and without water injection were determined experimentally. Without water injection, the load range was 2.3 to 3.6 bar gross IMEP. By using water injection to control heat release, the load range in TSCI was 2.3 to 8.4 bar gross IMEP, which is a range expansion of over 350%. These results demonstrate that direct water injection can provide significant improvements to both controllability and the range of operability of LTC, thereby resolving the major challenges associated with HCCI.

We review developments in the understanding of cycle-to-cycle variability in internal combustion engines, with a focus on spark-ignited and premixed combustion conditions. Much of the research on cyclic variability has focused on stochastic aspects, that is, features that can be modeled as inherently random with no short-term predictability. In some cases, models of this type appear to work very well at describing experimental observations, but the lack of predictability limits control options. Also, even when the statistical properties of the stochastic variations are known, it can be very difficult to discern their underlying physical causes and thus mitigate them. Some recent studies have demonstrated that under some conditions, cyclic combustion variations can have a relatively high degree of low-dimensional deterministic structure, which implies some degree of predictability and potential for real-time control. These deterministic effects are typically more pronounced near critical stability limits (e.g. near tipping points associated with ignition or flame propagation) such during highly dilute fueling or near the onset of homogeneous charge compression ignition. We review recent progress in experimental and analytical characterization of cyclic variability where low-dimensional, deterministic effects have been observed. We describe some theories about the sources of these dynamical features and discuss prospects for interactive control and improved engine designs. Taken as a whole, the research summarized here implies that the deterministic component of cyclic variability will become a pivotal issue (and potential opportunity) as engine manufacturers strive to meet aggressive emissions and fuel economy regulations in the coming decades.

Gasoline direct injection (GDI) engines can offer improved fuel economy and higher performance over their port fuelinjected (PFI) counterparts, and are now appearing in increasingly more U.S. and European vehicles. Small displacement, turbocharged GDI engines are replacing large displacement engines, particularly in light-duty trucks and sport utility vehicles, in order for manufacturers to meet more stringent fuel economy standards. GDI engines typically emit the most particulate matter (PM) during periods of rich operation such as start-up and acceleration, and emissions of air toxics are also more likely during this condition. A 2.0 L GDI engine was operated at lambda of 0.91 at typical loads for acceleration (2600 rpm, 8 bar BMEP) on three different fuels; an 87 anti-knock index (AKI) gasoline (E0), 30% ethanol blended with the 87 AKI fuel (E30), and 48% isobutanol blended with the 87 AKI fuel. E30 was chosen to maximize octane enhancement while minimizing ethanol-blend level and iBu48 was chosen to match the same fuel oxygen level as E30. Particle size and number, organic carbon and elemental carbon (OC/EC), soot HC speciation, and aldehydes and ketones were all analyzed during the experiment. A new method for soot HC speciation is introduced using a direct, thermal desorption/pyrolysis inlet for the gas chromatograph (GC). Results showed high levels of aromatic compounds were present in the PM, including downstream of the catalyst, and the aldehydes were dominated by the alcohol blending.

Properties related to the combustion, stability, and compatibility of blends composed of high-viscosity heavy fuel oil (HFO) and highly acidic pyrolysis bio-oil were determined to assess the utility of bio-oil as a marine fuel. The addition of bio-oil was shown to be fully stable with HFO at blend levels up to 50 mass % for up to 2 weeks. Bio-oil concentrations as low as 5 mass % significantly reduced the viscosity of HFO at 25 and 50 °C. Aging studies at 50 and 90 °C showed that the HFO inhibited the polymerization of bio-oil. The heating value and lubricity showed a linear dependency with bio-oil content, and combustion quality was acceptable for blends containing up to 15% bio-oil. The highly acidic bio-oil was found to be corrosive to carbon steel, 2.25Cr-1Mo steel, and 409 stainless steels, but not 304L and 316L. When blended into HFO at levels less than 19 mass %, no measurable corrosion was observed on any of the steel materials, but a 50 mass % concentration produced low-to-moderate corrosion in the carbon steel, 2.25Cr-1Mo steel, and 409 stainless steel grades. The combination of good blend stability, polymerization inhibition, reduced viscosity, and acceptable compatibility for low blend levels suggests that bio-oils may be suitable for use as a marine fuel.

No abstract