EDOUARD BRANLY would have been amazed. Who could have predicted that his experiments with electricity would defy explanation for well over a hundred years? In 1890, working as a physicist at the Catholic University in Paris, Branly flashed electrical sparks into a heap of metal filings hooked up to a circuit. To his amazement, its resistance plunged by several orders of magnitude. Tapping the glass container in which the filings were held restored the resistance to its original level.
Nobody could explain this weird behaviour, but that didn’t stop the Italian inventor Guglielmo Marconi exploiting the effect a few years later. At the time, Branly’s filings – whose resistance changed in response to radio waves and flashes of light as well as sparks – were the only practical way to detect radio signals. They kicked off the telecommunications era. But engineers soon replaced his device with more sensitive wire aerials, and no one ever got round to explaining the bizarre electrical behaviour of the filings.
Until this year, that is, when a group of physicists at the University of Liège in Belgium became curious. “It seemed like magic,” says Marcel Ausloos, who led the research. Ausloos and his colleagues Nicolas Vandewalle and Stephane Dorbolo now say they have solved the 100-year-old mystery, and their solution could change the way we think about all grainy materials, including powders, sands, cereals and filings. The apparently unpredictable properties of heaped grains perplex pharmacists, farmers and physicists, and are a nightmare for manufacturers who deal with grainy materials. They can stop the flow of cornflakes from a bag, even when the mouth of the packet is wide open. They rupture grain silos, bring sugar factories to a grinding halt and make it impossible for pharmaceutical companies to make tiny pills simply by mixing piles of powder.
Advertisement
What makes grainy materials such a pain in the neck? Unlike solids and liquids, whose properties are governed by regular, predictable forces between their constituent molecules, the properties of granular media are governed by irregular contacts between grains. Each particle has a slightly different shape and roughness, which makes some grains more sticky than others. Where contact is good between them, a whole necklace of fused, consecutive grains can suddenly form.
These necklaces, known as “arches”, snake through granular piles within a few seconds. As arches build up in flowing grains, they can have a number of unforeseen consequences. They can span the width of a pipe and cause a jam that is hard to locate and holds up production. Or they can transmit the weight of a whole pile of grains away from the bottom of a vessel and towards the sides, which can crack. “Silos do fail all the time, so this is not just a theoretical idea,” says Arshad Kudrolli, a granular physicist at Clark University in Worcester, Massachusetts.
Ausloos and his colleagues try to solve problems like these by studying the mechanical properties of grains, but five years ago they became curious about their electrical properties, too, and started playing with some metal filings they had picked up from a local recycling factory. Then, in 2000, a colleague told them about Branly’s experiment and they decided to recreate it. Rather than using iron filings, they used millimetre-sized lead beads coated with a patina of insulating lead oxide. These do not conduct electricity as well as iron filings, so Ausloos hoped that whatever mysterious mechanism was causing the Branly effect would be slowed down, making it easier to measure.
He connected a pile of the beads to a circuit with a highly sensitive voltmeter, pounded the beads with electrical sparks, and measured how the resistance of the pile changed. Rather than plunging suddenly, as Branly had found, the resistance of the lead beads decreased quickly at first but then gradually over about 20 seconds, bottoming off at about 80 per cent of the original resistance. Branly’s filings probably did the same thing, but did it so quickly that he mistakenly thought the change was instantaneous.
The pattern of decrease suggests an explanation for the effect. Electromagnetic waves emitted by the electrical spark can melt the points of contact between individual beads, allowing them to solder together. But this doesn’t happen all at once. Instead, some beads in the circuit happen to be closer together, so the contact between them is easier to solder. As soon as two of them have been soldered, the resistance between them drops and more current reaches the next bead. This increase makes the nearest contact more likely to melt next.
So a chain of soldering passes through the material rather like the “growth of a tree”, says Dorbolo. As the effect propagates through the pile of beads, a series of conducting paths is set up and resistance plummets. The decrease in resistance slows down and eventually bottoms out when the paths span the heap of filings.
This theory not only accounts for the way resistance falls, but also explains why tapping the container restores the original resistance: it cracks the delicate soldered contacts and returns the pile of metal beads to its disconnected state. Explaining exactly how a path of “microsoldering” evolves would help to solve the problem that has dogged the field of granular physics. The same chains of force that block grain silos and cornflake boxes determine which metal beads solder together first, because the contact will be better where they are pressing on each other, so the soldering will happen more easily there. “Arches are the easy way for current to flow,” says Dorbolo. So by replacing non-conducting grains with conducting replicas, researchers could use electrical measurements to study how blockages form.
When the group published its work in April (Physical Review E, vol 67, p 040302), it was the first time anyone had shown a link between the mechanical and electrical properties of grains. But not everyone was convinced. Granular physicist Bob Behringer of Duke University in Durham, North Carolina, says that there is no proof that current cannot leap between particles that are not intimately connected by soldered arches. He would like to see how the resistance varies as stress is applied to the system to prove the relationship between force chains and conductivity.
But others are excited by the prospect of using resistance measurements as a tool to probe the flow of grains. Kudrolli says it might be possible to measure the tendency of arches to form in piles of powder by measuring the electrical resistance of the pile. The more arches that span a pile, the faster its resistance will drop when it is exposed to flashes of light.
A tool like this could make a huge difference to factory workers and farmers. At the moment, they simply don’t know when a grain-piping system is about to give way or jam. But electrical measurements could help. Even though grainy materials such as cornflakes and wheat do not conduct electricity, Dorbolo believes that moisture in the air between the flakes could transmit a detectable electric current. If he is right, a drop in the resistance of the flowing grain could warn that arches are building up and the flow could jam. Workers could stir the flow to break up the arches before they cause trouble.
Tony Royal of Jenike & Johanson, a consultancy in Westford, Massachusetts, that solves granular problems in the pharmaceutical and food industries, warns that more work is needed before resistance measurements can predict jams in factories. But he believes the extra insights that come from being able to look inside flowing grain will also lead to better models of granular flow. One day these models could be used to predict whether a system is likely to jam even before it is built.
If electrical measurements live up to their billing, then instead of modelling flowing grains as fluids – which is what physicists do at present – scientists will begin to crack their chaotic and irregular behaviour. Branly’s filings could soon be back in business.