The little balls of fire Paul Ronney saw in April 1997 were so round, so perfect, so nearly motionless that they shocked him, even though he had been looking for them— even though he had sent an apparatus into orbit on the space shuttle just to create them. The apparatus was an aluminum cylinder one foot across and filled with gas, around 5 percent of which was hydrogen; clear quartz windows enabled Ronney's cameras to record the goings-on inside. The first time he let a spark jump through that chamber, he got flame balls a third of an inch in diameter. They drifted slowly around the chamber like little UFOs, avoiding one another but also the walls, and never going out— until Ronney's preprogrammed apparatus snuffed them after eight minutes. He had never expected them to last that long. "It was kind of disappointing," says Ronney, a combustion physicist at the University of Southern California in Los Angeles. "Like when a 100-year-old man gets hit by a bus and you wonder, 'How long could he have lived?' " Ronney's orbiting flame balls are the otherworldly ideal of a phenomenon— fire— that is hideously complex here on Earth. Gravity, wind, the chemical intricacy of most fuel— all those things are unimportant when you're burning hydrogen, one of the simplest combustibles, inside an airtight cylinder on the space shuttle. Neither Ronney nor any other scientist can fully explain his flame balls. But they are a lot closer to doing that than they are, say, to predicting in advance how efficiently and cleanly a new V-6 engine will burn its fuel— a question of some importance to the automotive industry. They are even closer to understanding flame balls than they are to understanding a thing as simple as a burning candle. Like most fires, a candle is a system for making hydrocarbon molecules react with oxygen to produce heat and light, as well as carbon dioxide and water. "It's really quite a mess," says Howard Ross, a senior researcher at NASA. "There are literally thousands of reactions that go on from the moment the fuel vapor is produced and leaves the wick to the time it actually burns and produces CO2 and water." As melted wax rises through the wick, its long hydrocarbon molecules are vaporized and cracked apart by heat emanating from the flame. Some of the fragments migrate outward. Some are transformed into ring-shaped molecules called polycyclic aromatic hydrocarbons; those clump together, forming large particles of soot, which drift upward and are burned, or escape from the top of the flame as smoke. Most of the heat— but not the light— is released at the surface of the flame, where fuel vapor diffusing out from the wick meets oxygen diffusing in from the surrounding air. The various carbon compounds and the oxygen molecule O2 have weaker bonds but more potential energy than do CO2 and H2O. When the carbon and oxygen combine to form CO2 and water, some of the energy difference is released as invisible infrared radiation, or heat. The visible light of a candle flame is caused by two different processes: incandescence and chemiluminescence. The bright yellow light in the tongue of the flame comes from incandescent soot particles. The faint blue light around the bottom comes from furiously vibrating CH and C2, still on their way to being burned. An ordinary fire doesn't sit passively waiting for oxygen to diffuse toward it. Instead, it helps itself. As air heated by the flame rises, cool oxygen-rich air flows in at the bottom. (A forest fire can generate winds of more than 100 miles per hour.) Therein lies the origin, says physicist Marc Thuillard of Cerberus, a Swiss division of Siemens that manufactures smoke and flame detectors, of one of fire's most mesmerizing characteristics: its flickers. A well-made candle usually doesn't flicker unless it's buffeted by external air currents. But what combustion folk call a pool fire— which is anything from a burning pool of hydrocarbon spilled at a chemical plant to a skillet full of crêpes suzette flambé— will almost always flicker, all on its own, even in a still room. In fact, if the fuel is distributed in a regular way, the burning pool will pulse as regularly as a clock. Without gravity or wind to shape them, flames, ignited in a hydrogen-filled cylinder on the space shuttle, form into long-burning balls a third of an inch in diameter. Photo by Paul Ronney/NASAPicture a small, round, shallow dish of ethanol on the floor of Thuillard's flameproof lab, which is a mostly empty concrete cube 30 feet high and 30 feet across. A colleague of Thuillard's holds a Bic lighter to the ethanol, igniting it with a low whoosh. What's happening in that flickering flame, Thuillard explains as we watch it, is that the cool air rushing over the flame is creating a wave on it, like wind blowing over the sea. The flame wave travels from the edge of the pool toward the center, where it becomes a flame mushroom that billows upward and outward— and then the next cycle of the flicker begins. When there are gusty winds fanning the fire, or when it's a heap of burning logs in your fireplace instead of a pool of ethanol, the flicker becomes irregular. But Thuillard has found that, either way, flames still contain hidden mathematical regularities that make it possible to distinguish them from any other light source, such as sunlight that flickers on a wall because it is irregularly blocked by blowing foliage outside. In the past couple of years Cerberus has sold tens of thousands of a new flame detector that incorporates this insight of Thuillard's. So far, he says, no one has reported a false alarm. To show off its capabilities, he and his colleagues place the detector on one side of the lab, with a bright lightbulb and a flashing orange alarm light directly in front of it; the detector remains quiet. With those distractions still glaring, Thuillard's colleague lights a two-inch dish of ethanol 25 feet away. As waves of flame start to pulse up from the dish, the detector starts an insistent bleat. Flickering in a flame or undulations on the sea are both examples of gravity waves, which occur at the boundary between liquids or gases of different density— in the case of flames, the boundary between dense cool air that wants to sink, because of gravity, and hot air that wants to rise. In the microgravity of space, flames presumably don't flicker in the same way. Until Howard Ross had astronauts light candles on the shuttle and the Russian space station, some people even questioned whether flames would get enough oxygen to burn in space at all. Because hot air doesn't rise in space, they pointed out, cold air won't flow in to feed a flame. In Ross's experiment, the candles had spherical flames so faint as to be almost invisible and left behind floating balls of liquid wax. The record for the faintest and strangest flames, though, seems to belong to Ronney. His flame balls— in which hydrogen and oxygen diffuse toward each other in stillness and harmony to create perfect spheres of light— put out only one to two watts, about a fiftieth as much power as a birthday candle. They do so regardless of the size of the gas chamber or the pressure inside. "If all this sounds a little weird," Ronney says, "you're not alone. Even when I talk about flame balls to colleagues who are combustion experts, their first response is, 'I don't believe it.' " This year NASA is giving Ronney a chance to fly his experiment one more time. He says flame balls are to combustion physics what the fruit fly is to genetics. If we could crack the principles underlying what he calls "the simplest possible flame," we would have a better shot at designing leaner-burning, less-polluting internal combustion engines. And at the very least, we might understand candles better.
For a thorough description of Marc Thuillard's intelligent fire detector, complete with diagrams and photos, see www.siemens.com/FuI/en/zeitschrift/ archiv/Heft1_99/ artikel03/index.html. For more about fire in space, see NASA's "Candle Flames in Microgravity" Web page: zeta.lerc.nasa.gov/expr/candle.htm