Sunday, January 1, 2012

Seasonal stalactites: The sharp end of icicle science

Power-line engineers and Hollywood CGI specialists would love to know more about icicles – but answers are surprisingly elusive
Icicles can be a menace to power lines and a sore point for passers-by <i>(Image: Tomitheos Photo Image Media Artwork/Flickr/Getty Images)</i>
Icicles can be a menace to power lines and a sore point for passers-by (Image: Tomitheos Photo Image Media Artwork/Flickr/Getty Images)

IT IS not often Stephen Morris helps save a life - he is a physicist, after all, not a physician. But when an architect telephoned him in his office at the University of Toronto, Canada, last year, with a potentially lethal problem, his advice was to the point.
The architect's problem was icicles. He had designed a building whose windowsills accumulated snow in bad weather. Worried about a passer-by being engulfed by a sudden avalanche and suing, he had installed heaters on the windowsills. Consequently meltwater was dripping off the sill and forming enormous icicles that loomed dagger-like overhead. The architect was still worried, and with good reason. Falling icicles reportedly killed five and injured 150 in St Petersburg, Russia, last winter.
Fortunately, Morris understands icicles. In fact, he is the world's leading expert on their formation. Perhaps that's a matter of location. "I live in a cold country," he says. Although Toronto's location on Lake Ontario gives it one of Canada's balmier climates, the surrounding area can seem like one big ice slab in winter.
There is more at stake with icicles than just the risk of an impaling. In January 1998, an ice storm destroyed much of the power infrastructure in the Canadian province of Quebec. Pylons buckled under the weight of the ice and some power lines snapped, leaving millions of homes and businesses without electricity for days. "The cost was estimated at more than 20 billion Canadian dollars," says Masoud Farzaneh, head of the International Research Centre on Atmospheric Icing and Power Network Engineering based at the University of Quebec in Chicoutimi.
Morris's contributions to limiting such costs emerge principally from a large, styrofoam-insulated box standing in the middle of his lab. "We call it the rotisserie," he says. Inside the box a downward-pointing wooden spike attached to a turntable rotates at the stately rate of 12 revolutions per hour. A nozzle drips water onto the wide base of the spike, while an air fan sits in each of the box's corners, blowing cold air at it. The whole contraption is cooled by antifreeze that circulates through the box's walls.
This is enough for Morris and his graduate student Antony Chen to strip away layers of assumption about icicles that have accumulated over the years. Take the established picture that they grow with a "self-similar" shape: that the ratio of an icicle's length to its circumference is always the same. According to this idea, an icicle's precise shape depends on local factors such as the difference between the icicle's surface temperature (0°C), the ambient air temperature (significantly colder), and the rate at which this falls as you move away from the icicle's surface.
Not so, say Morris and Chen. Their experiments show that self-similarity appears in only a small fraction of icicles, when conditions - air movement and water purity, for instance - are just so. "Things that you wouldn't think matter turn out to matter a great deal," says Morris. Only when the fans in the rotisserie gave the air a gentle stir, for example, did a classic icicle with a single tip form; with still air, the tip tended to split in two (Physical Review E, vol 83, p 026307).
It is unlikely that engineers will be able to do much about air movement around power lines, of course. But knowing the conditions under which ice will build up fastest could be a first step towards designing equipment and systems that are more robust. "Icicles are the simplest thing you can think of in this class of problem," Morris says.
In another of his experiments, Morris tackled an idea about icicles floated by Farzaneh and his colleague Kazuto Ueno, now at Kyushu University in Japan. Many natural icicles have a rippled surface, as if composed of a narrowing succession of rings. Theories of icicle growth had traditionally had nothing to say about such features, but Farzaneh and Ueno suggested surface tension between the freezing water and the surrounding air was the culprit: the higher the surface tension, the fewer ripples on the icicles (Physics of Fluids, vol 22, p 017102).
That seemed reasonable. Natural icicles form from dirty water that sits on roofs and in guttering, which has a low surface tension and so should form ripples. Sure enough, when Morris performed his rotisserie experiments with distilled water, no ripples formed. But there was a sting in the tip: when he used tap water, also a pretty pure sort of water, ripples suddenly appeared.
Jerome Neufeld, a theoretical geophysicist at the University of Cambridge, in the less frozen, but damp east of England, confesses himself baffled by this. "The level of impurity even for tap water should be so low, it's hard to conceive of how it should affect the experiment - but it does seem to," he says. It certainly suggests surface tension is not the primary factor.
Neufeld is particularly interested because he performs Morris's experiments in reverse: he studies how icicles melt. Working with Cambridge colleagues Raymond Goldstein and Grae Worster, he left a cylindrical block of ice in a warm room and watched how its shape changed over time, exploring how things such as the rate of heat transfer to and from the block and its own material consistency changed matters (Journal of Fluid Mechanics, vol 647, p 287).
That might seem akin to watching paint dry, but it could prove important. Understanding how layers of air insulate the surface of glaciers, for example, is vital to making accurate estimates of how fast they will melt - and sea levels will rise - as the Earth warms under its blanket of greenhouse gases.
There is another, more unexpected area in which icicle research is followed with interest: Hollywood. You might think that ice is only a problem there when bar staff can't find any. But in an age where many films owe more to CGI wizardry than sharp camera work, requests for realistic simulations of melting, freezing, crystal growth and so on come up "more often than you might at first expect", says Christopher Batty, who researches computer graphics at Columbia University in New York City. For instance, Batty was part of a team asked to create a Kryptonian crystal spaceship for the film Superman Returns. "Some of the internal decor and controls of the ship were actually supposed to appear to grow organically," he says.
The scene was ultimately deleted - which was disappointing, but the work goes on. "If we could grow icicles on the computer, in a physical way, then we can create virtual models of chasms, caves, ice storms, shut-down airports and so on in days instead of months," says Batty's colleague Eitan Grinspun.
That could now be on the cards: Batty, Grinspun and Morris have just teamed up in a new collaboration to master ice-making for the big and small screen once and for all. It's the next problem for Morris to solve, after he gave the architect the benefit of his accumulated wisdom. "I told him to stop heating the windowsills; you have to just leave the snow," Morris says. "It didn't turn out to be a very profound solution."

Michael Brooks is the author of Free Radicals: The secret anarchy of science (Profile, 2011)

No comments:

Post a Comment