Flames do something odd in space: they form tiny almost-invisible balls that might reveal the secrets of combustion here on Earth.

Listen to this story via streaming audio, a downloadable file

It happened in 1984 when Ronney, a combustion researcher, was at the NASA Glenn Research Center's Microgravity Drop Tower in Ohio. He pressed a button and sent a can of burning hydrogen falling down a 90 ft. shaft. For 2.2 seconds it plummeted, freely falling and weightless, with a 16mm movie camera recording the action. Ronney knew that flames did strange things in low gravity--that's why he was doing the experiment--but he wasn't prepared for what he saw in the film room later.

Right: Looking down the shaft of the Glenn Research Center's 2.2 second Microgravity Drop Tower. [more]

The flames had broken apart into tiny balls that moved around like UFOs. "I thought I had done something wrong," he recalled. Some of his colleagues didn't believe him when he described the experiment. Indeed, "it was ridiculous. No one had ever seen anything like it."

But the flame balls were real--later experiments proved it.

"Flame balls are the weakest flames we have," says Ronney. "Compared to a birthday candle's 50 to 100 watts, a flame ball produces only 1 to 2 watts of thermal power. They burn using very little fuel. It's almost as if a hydrogen-burning flame's last line of defense as it approaches extinction is to draw itself into a simple ball."

Ronney, who is now an engineering professor at the University of Southern California, believes that flame balls will help him and others crack the unsolved mysteries of burning. Considering that combustion powers our automobiles, generates our electricity, and heats our homes, there's much about it we don't understand. "For example," he says, "a moderate amount of turbulence makes a flame burn faster, but too much turbulence extinguishes it." No one knows why.

Below: Tiny flame balls that form in low gravity are hard to see. These were filmed in the dark by a low-light video camera onboard the space shuttle Columbia in 1997.

Flames are hard to understand because they are complicated. In an ordinary candle flame, for example, thousands of chemical reactions take place. Hydrocarbon molecules from the wick are vaporized and cracked apart by heat. They combine with oxygen to produce light, heat, carbon dioxide and water. Some of the hydrocarbon fragments form ring-shaped molecules called polycyclic aromatic hydrocarbons and, eventually, soot. Soot particles can themselves burn or simply drift away as smoke. The familiar teardrop shape of the flame is an effect caused by gravity. Hot air rises and draws fresh cool air behind it. This is called buoyancy and is what makes the flame shoot up and flicker.

Flame balls, on the other hand, are simple. The balls form in low gravity where turbulence and buoyancy have little effect. Oxygen and fuel combine in a narrow zone at the surface of the ball, not hither and yon throughout the flame. Once ignited and stabilized, their size remains constant. Unlike ordinary flames, which expand greedily when they need more fuel, flame balls let the oxygen and fuel come to them. Finally, the fact that flame balls are spherical reduces their dimension to one: the radius of the flame itself.

"Flame balls are to combustion scientists what fruit flies are to geneticists," says Ronney. "It's not that we want more fruit flies, or flame balls, but they provide a simple model for testing hypotheses and checking computer models."

Right: A schematic diagram of a flame ball. Credit: Paul Ronney.

One of many mysteries about fire is the way weak flames go out before their fuel is totally exhausted. It puzzles physicists and vexes automakers who want to build clean, efficient "lean-burning" engines that run on fuel-air mixtures with low fuel concentrations--much like a flame ball. Ronney believes that studying one (flame balls) will help us with the other (cars).

Here on Earth, researchers can't study flame balls for long. A typical plunge down the drop tower lasts only 2 seconds. So, working with NASA scientist Karen Weiland and others at the Glenn Research Center, Ronney designed the Structure of Flame Balls at Low Lewis-number (SOFBALL) experiment. It's a sealed chamber where flame balls flying onboard the space shuttle can burn for a long time.

SOFBALL orbited Earth for the first time in 1997 on shuttle Columbia--and it produced some surprises.

Computer models had predicted the flame balls would be small and either extinguish or drift into the chamber walls in a few minutes. Instead they were two to three times larger than predicted and burned for over 8 minutes until the experimental system automatically extinguished them. Furthermore, although the flames were large, they were the weakest ever seen--emitting little more than 1 watt of thermal power.

"We knew then that we still had lots to learn about weak combustion," recalls Ronney.

Left: A candle flame on Earth (left) and onboard the space shuttle (right). [more]

The experiment, upgraded and re-named SOFBALL-2, will soon fly again. It's slated for launch onboard space shuttle Columbia (STS-107) in late 2002 or 2003. During the mission, flame balls will be allowed to burn for 25 to 167 minutes. Instruments will monitor their temperature, brightness, heat loss, and the composition of their gaseous byproducts. Because flame balls are so sensitive to motion, the shuttle will drift during the experiments instead of using its reaction control thrusters to maintain position.

Because this research is so fundamental, it touches on many aspects of combustion: lean-burning engines for cars and airplanes; explosion hazards in mine shafts and chemical plants; emissions from cars and coal-burning plants; arson investigations. The list is long ... and it doesn't stop on Earth.

Below: Astronaut Janice Voss (the sister of co-author Linda Voss) monitors a combustion experiment onboard shuttle Columbia in 1997.

Flames act differently in space, so fire safety is also different. If you see a fire on Earth, you might run over and stomp it out or use a fire extinguisher. In orbit, rushing over and stomping on a flame might accelerate combustion, at least temporarily, because you are creating an airflow that did not exist before. Flames in low-gravity tend to spread slowly, so stomping might cause a flame to jump to something else when it wouldn't have otherwise. Furthermore, flame balls are stealthy: they give off no smoke and little or no visible light. It's hard to extinguish something you can't find. What happens if a loose flame ball runs into something? Will it ignite? SOFBALL-2 could answer many such questions.

SOFBALL will also set the stage for longer-term experiments aboard the International Space Station inside the Fluids and Combustion Facility--yet to be installed in the US lab module. That's a long way from Ohio, where Ronney discovered flame balls in 1984. But he says it's worth the trip to find out how else "those ridiculous little flame balls" might surprise us.

Web Links

NASA's Office of Biological and Physical Research -- supports studies of basic physics for the benefit of humans in space and on Earth

SOFBALL will fly again onboard the STS-107 research mission in late 2002 or 2003.

History: Ronney didn't know it in 1984, but flame balls (or something like them) had been predicted in 1944 by Soviet physicist Yakov Zeldovich. Zeldovich thought the phenomenon, like balancing a pen on its point, was possible in theory but too unstable to exist in reality. And indeed flame balls can exist only within a very narrow window of conditions, one of the requirements being microgravity, which means they would not have been observed under normal conditions on Earth.

more NASA combustion research links: Microgravity Combustion website (NASA/GRC); SOFBALL (NASA/GRC); Combustion Module Home Page (NASA/GRC)

Movies of flame balls from the SOFBALL experiment onboard space shuttle Columbia in 1997: movie #1 (1 MB); movie #2 (1 MB).