Richard W. Anderson

Prediction of spark kernel development in constant volume combustion

Abstract Combustion initiation is studied in atmospheric pressure propane-air mixtures in a constant volume bomb with a high speed (10,000 fps) laser schlieren system. The spark current and voltage waveforms are simultaneously recorded for later model input. A phenomenological model for early flame kernel development is presented which accounts for the initial, breakdown generated, spark kernel and its subsequent growth. The kernel growth is initially controlled by the breakdown process and the subsequent electrical power input. A new, spark power induced, mass entrainment term is shown to model this initially rapid volume increase adequately while later growth is mainly dominated by diffusion. Results and model comparions are presented for the effects of power input, spark energy, and equivalence ratio.

Spark ignition of propane-air mixtures near the minimum ignition energy: Part I. An experimental study

Kernel growth from a spark in propane-air mixtures at atmospheric pressure is studied in a constant volume bomb with a high-speed laser schlieren system. The spark current and voltage waveforms of an inductive ignition source are simultaneously recorded with the photographic recordings. The temporal growth of the measured equivalent radii at conditions near the minimum ignition energy shows the existence of a critical radius and the influence of the critical radius on kernel development. In addition, it is shown that the net spark power for ignition can be estimated using data from minimum ignition energy, electrode fall energy losses, and spark calorimetry experiments. These results are used in Part II to develop a model for kernel growth.

Speciated Hydrocarbon Emissions from the Combustion of Single Component Fuels. I. Effect of Fuel Structure

Speciated hydrocarbon emissions data have been collected for six single-component fuels run in a laboratory pulse flame combustor (PFC). The six fuels include n-heptane, isooctane (2, 2, 4-trimethylpentane), cyclohexane, 1-hexene, toluene, and methyl-t-butyl ether (MTBE: an oxygenated fuel extender). Combustion of non-aromatic fuels in the PFC (at a fuel/air equivalence ratio of Φ = 1.02) produced low levels of unburned fuel and high yields of methane and olefins (> 75 percent combined) irrespective of the molecular structure of the fuel. In contrast, hydrocarbon emissions from toluene combustion in the PFC were comprised predominantly of unburned fuel (72 percent). With the PFC, low levels of 1, 3-butadiene (0.3-1.8 percent) were observed from all the fuels except MTBE, for which no measurable level (<0.2 percent) was detected; low levels of benzene were observed from isooctane, heptane, and 1-hexene, but significant levels (9 percent) from cyclohexane and toluene. No measurable amount of benzene (< 0.2 ...

Spark Ignition of Propane-Air Mixtures Near the Minimum Ignition Energy: Part II. A Model Development

Abstract A model is developed to simulate the kernel growth observed in the exprimental study of Part I. Kernel growth, described as a two-step process, initially involves a blast wave over a negligible short time followed by a diffusive growth with an electrical input power. The diffusive growth is formulated by an integral approach involving temperature dependent overall reaction kinetics and electrode heat loss. The model predicts the kernel growth reasonably well with the measured spark power input. It predicts both ignition and nonignition kernel growth. The existence of a critical radius is also demonstrated. In addition, dimensional analyses are given to clarify the physical aspects of the critical radius and the characteristic radius of the blast wave.

Spark Anemometry of Bulk Gas Velocity at the Plug Gap of a Firing Engine

The objective of the present work was to investigate a rapid method of obtaining the convection velocity of the bulk gas near the spark plug gap of a firing engine at the time of ignition. To accomplish this, a simple model was developed which utilized both the secondary current and voltage signals, from a conventional spark discharge. The model assumed the spark path was elongated in a rectangular U-shape by the flow. Based on experimentally measured electrical signals, the mean convection velocity was computed. The convection velocity calculated by the model first needed calibration which was accomplished with a bench test that used a hot wire anemometer. The technique has a weak correlation at low velocities of 1--2 m/s, but correlates well as higher velocities up to 15 m/s. Although the accuracy of prediction by the technique is moderate, it is shown to be suitable for rapidly studying the bulk flow velocity ear the plug gap in an operating engine without modification of the combustion system. It is also shown to favorably compare with data taken with a fiber optic equipped spark plug.

Turbulent scales in a fan-stirred combustion bomb

Turbulent scales were investigated in a fan-stirred combustion bomb. Two-dimensional particle velocimetry measurements were used to evaluate flow fields and scales for the Reynolds number range Rel = 37 - 711, the subscript l referring to the integral scale. The Taylor scale λ is correlated with the electrical power input Π for the fans producing turbulence. It is found that the integral scale remains unchanged and λΠ1/6 = constant for Rel > 100. When electrical power input is assumed to be a measure for turbulent production, dimensional arguments based on a volume-averaged large-scale estimate for viscous dissipation lead to the foregoing correlation for homogeneous turbulence.