麻豆传媒

鈥淚 AM an old man now, and when I die and go to heaven there are two matters on which I hope for enlightenment. One is quantum electrodynamics and the other is turbulence in fluids. And about the former, I am really rather optimistic.鈥�

So said the British physicist Horace Lamb in 1932, shortly before his death. And his words continue to amuse and haunt researchers grappling with the hugely complex physics of turbulence. Today, understanding and controlling turbulence remains one of the great, unsolved problems facing modern science.

Yet enlightenment may be closer than ever before. In recent years, scientists have begun to gain clear insights into why turbulence arises and how it creates drag. In many countries, physicists and engineers are now able to control turbulence with cleverly designed surfaces. Exactly how some of these surfaces work remains a mystery, though the work has already knocked some holes in long-held theories. Other researchers are building micromachines that may one day form part of aerodynamic surfaces. These machines will spot microscopic eddies as they arise and stamp them out before they become a problem.

Drag and turbulence are huge obstacles for engineers designing aircraft, ships and even pipelines. Skin friction and the drag it causes accounts for half the resistance that an airliner must overcome in flight. A reduction in drag of only a few per cent could save billions of dollars in fuel costs.

But until recently, engineers have struggled to reduce drag by even fractions of 1 per cent. The new surfaces and micromachines could reduce drag by more than half. The potential benefits and savings are huge.

Describing the way a fluid flows over a surface is a thankless task. Physicists are most interested in the boundary layer, the part of a fluid closest to a surface, because this is the source of the friction that gives rise to turbulence. In regular, smooth flows at relatively slow speeds, physicists assume that an extremely thin layer of fluid in contact with the surface is prevented from moving by friction and so is motionless. The flow above this layer moves more quickly the farther it is from the surface. This type of regular fluid movement is known as laminar flow and engineers designing aerodynamic surfaces dream of achieving it.

But these dreams can rapidly turn to nightmares. As a fluid speeds up, small chaotic eddies begin to form in the boundary layer and this severely disrupts the orderly state of affairs that occurs in laminar flow. The eddies and vortices rapidly grow into large-scale turbulence, turning a quiet laminar flow into a raging maelstrom of chaos.

The turbulence forces fast-flowing fluid from upper layers onto the surface, and often this flow curls back on itself, ending up opposing the flow and increasing frictional forces by orders of magnitude. When this happens, the transition from laminar to turbulent flow is complete, and in Lamb鈥檚 time nothing could be done to reverse or prevent it.

Ironically, things began to change around the time of Lamb鈥檚 death. In the 1930s, Max Kramer, a German-American physicist, wondered how dolphins were able to swim so quickly, outpacing any ship or boat that humans could build. In short bursts, dolphins can swim at up to 50 kilometres per hour. But by assuming that dolphins have a power output similar to other animals, simple calculations show that frictional forces should overpower them long before they reach this speed. How do they do it?

Kramer believed the dolphin鈥檚 secret lay in its skin. He discovered that dolphin skin has unique properties and began to experiment with artificial skins that mimicked them. The trick, he thought, was to create a spongy, pliant material containing fluid-filled ducts. Kramer reasoned that any instabilities generated in the boundary would be transmitted to the surface where they would be damped out. In tests, he claimed that his artificial dolphin skin could reduce drag by up to 50 per cent (鈥淭he dolphins鈥� secret鈥�, 麻豆传媒, 5 May 1960, p1118). But Kramer鈥檚 tests were difficult to reproduce and his work was almost forgotten.

Work on so-called compliant surfaces began again in the 1980s and has come a long way. Physicists now believe that the way compliant surfaces effect fluid flow is far more complicated than Kramer could have known.

鈥淓ddies need energy to grow into full blown turbulence,鈥� says Peter Carpenter, a physicist at the University of Warwick and a leading researcher in the field. 鈥淭he simplest way to think of compliant surfaces is to imagine that they interfere with this transfer of energy.鈥� In practice, compliant surfaces delay the onset of turbulence, so that the flow is laminar over a greater part of a dolphin鈥檚 body.

But building practical compliant surfaces has not proved easy. Changes in the flow cause pressure variations that pinch the skin. 鈥淵ou can see these wrinkles in photographs of dolphins swimming at high speed,鈥� says Carpenter. Just how to prevent these wrinkles is not yet clear. Yet compliant surfaces have huge potential, in theory at least.

Recently, Carpenter predicted that matching the properties of a surface with the speed of the flow over that point could delay the transition from laminar to turbulent flow indefinitely. In practice, this would require a boat or a submarine to be covered in multiple panels with the required properties. 鈥淰ery large drag reductions of up to 80 per cent may be possible,鈥� says Carpenter, adding that more experiments are urgently needed.

While dolphins rely on compliant surfaces for their speed, sharks seem to use a different mechanism for reducing drag. And since fast-moving sharks, such as hammerheads and mako, can cut through deep water at up to 70 kilometres per hour, this mechanism must be at least as effective as compliant surfaces.

Researchers have known for years that fast-swimming sharks are covered with renewable scales, called dermal denticles, that have microscopic ridges. Just what these ridges did was a mystery until the 1980s when specialists in aerodynamics developed a way of smoothing turbulent flow using microscopic surface grooves that are aligned with the movement of fluid. These grooves are called riblets. 鈥淭here are very close connections between riblets and shark skins,鈥� says Kwing-So Choi, a mechanical engineer at the University of Nottingham specialising in drag reduction.

Riblets work by making turbulence more ordered near the surface. 鈥淭hey are like tiny fences that prevent the lateral movement of turbulence across the flow,鈥� says Choi. While any lateral movement is hindered, any flow that is parallel with the riblets continues unabated. 鈥淚f you optimise the size and shape of the riblets, a drag reduction of up to 10 per cent is possible.鈥�

Riblets have already been tested in real applications. Their biggest claim to fame was as a putative contributor to the sailing coup in 1987 when the US won the America鈥檚 Cup, the prestigious yachting trophy. The American boat had a riblet coating on its hull, a development that was later banned by race officials. Other mariners have been slow to follow suit because riblets are easily clogged by microscopic marine debris. A more promising application is in aircraft, where fouling is less of a problem. Such riblets are already in use on some commercial airliners.

Riblets may also help to dissipate heat like cooling fins. They increase the surface area of a body by up to 100 per cent, so it is easy to think that they would naturally increase heat transfer. But the laws of thermodynamics say otherwise. The flow of heat is closely linked with fluid flow, and according to a long-held theory known as the Reynolds analogy, any reduction in turbulence must also reduce the transfer of heat.

But the Reynolds analogy seems to be on its last legs, at least as far as riblets are concerned. 鈥淩iblets seem to defy the analogy,鈥� explains Choi. They change the flow in a way that was not envisaged when the analogy was devised, he argues, and when this happens, any analogy between heat and fluid flow must break down. He is currently experimenting to find out how the heat transfer is affected and what shape riblets must be to optimise drag reduction and heat transfer.

鈥淚n a few months, we will know more,鈥� he says. If Choi can pull it off, riblets may one day be used in everything from cars to aircraft for the dual purpose of reducing drag and dissipating heat.

While the 10 per cent reduction in drag that is possible with riblets is huge for mechanical engineers, it doesn鈥檛 fully explain how sharks can swim at such high speed. One explanation is that they are somehow able to sense turbulence as it occurs and move their scales in a way that reduces it. One piece of evidence rests on the observation that denticles are attached to subskin muscles.

Moreover, sharks are equipped with pressure sensors called pit organs that might be capable of sensing drag. Researchers suggest that a shark鈥檚 nervous system may process such data and orchestrate its subskin muscles to move denticles in a way that reduces drag. 鈥淪harks might have an active drag reduction system,鈥� says Carpenter. 鈥淏ut we simply don鈥檛 know.鈥�

Either way, researchers in America are now developing active drag reducing systems of their own. Chih-Ming Ho, a mechanical and aeronautical engineer at the University of California in Los Angeles, is leading a team of researchers which is developing micromachines that sense the onset of turbulence and react to prevent it. Funded by the US Air Force Office of Scientific Research and the Defense Advanced Research Projects Agency (DARPA), Ho and his team have already built arrays of microelectromechanical sensors (MEMS) and actuators onto single silicon chips.

Each sensor is about the size of the full stop at the end of this sentence. At its heart is a thinner-than-hair polysilicon wire that is electrically heated and kept at a constant temperature. If the air flow over the sensor is smooth, the amount of power required to maintain the temperature remains constant. But turbulence changes the rate of heat transfer, creating a change in the amount of current required to maintain the constant temperature. In a wind tunnel, the researchers have shown that an array of 25 sensors can locate a region on a test surface where turbulence occurs.

The actuators are flaps less than a millimetre across, harbouring a tiny coil of polysilicon and situated in a magnetic field. When a current passes through the coil, the flap deflects upwards. Turbulence brings fast-moving air towards the surface which creates drag, so the actuators push it away again. Putting this into practice is tough. The trick is to process the data from the sensors in real time so that the actuators can act immediately. The big advantage of MEMS is that the intelligence to do this can be built into the chip. 鈥淔or the first time we have been able to integrate the sensors, actuators and computers,鈥� says Ho.

But other challenges lie ahead. 鈥淓verything is difficult in this field. We鈥檙e working at the forefront of science,鈥� he says. For a start, the actuators cannot yet move quickly enough to tackle turbulence in real time. Ho is confident this can be improved, but then he must understand how vortices interact with the actuators before he can decide how best to control them. Only then will he know how much drag reduction will be possible with these devices. 鈥淚t鈥檚 too early to say at the moment.鈥�

Even if Ho can solve these problems, flap-like actuators will not be the best way to reduce turbulence around aircraft skins, as dust and insects would rapidly clog them up. Such a surface would also be very fragile. 鈥淚f you can鈥檛 walk on it, I sincerely doubt that Boeing will want it,鈥� says Ben Anders, a senior research engineer at NASA鈥檚 Langley Research Center in Virginia.

But Ho鈥檚 group is simply proving the concept, argues Kaigham Gabriel, who runs DARPA鈥檚 programme to develop microelectromechanical systems. 鈥淭here would have to be alternative, more robust, ways of affecting the fluid flow, like dimples that pop up,鈥� he says.

Ho鈥檚 group is taking one of the more audacious approaches to air-drag reduction, but his is not the only team investigating active ways of controlling flow. Ari Glezer, an aeronautical engineer at the Georgia Institute of Technology in Atlanta, and his colleagues have made micromachined arrays of individually controllable microjets.

These dot-sized jets consist of a cavity that is open at one end but has a thin diaphragm at the other that flexes when a voltage is applied to it. As the diaphragm flexes it draws fluid into the cavity, and as it rebounds, it ejects fluid in a series of rings in the same way that smokers create smoke rings. Glezer hopes that arrays of microjets will prevent the formation of turbulence in much the same way as Ho鈥檚 microactuators.

Sucking and blowing

Undoubtedly, it will be some time before micromachines are built into passenger aircraft, but other systems are already being tested on experimental aircraft. One way to get rid of turbulence is simply to suck it away. So researchers at Dryden Flight Research Center in Edwards, California, have fitted a sieve-like titanium sleeve riddled with millions of laser-drilled pinholes over the wing of an F-16 fighter aircraft and sucked air from above the wing through the holes (Technology, 19 October 1996, p 23).

During test flights last year, the team established that the technique could successfully reduce drag, although the fan used up more energy than it saved. Such a device could be used to keep the airflow smooth over the wings of the next generation of supersonic passenger planes.

And last month, researchers at Tel Aviv University verified that the opposite technique-of blowing air through a slot into the boundary layer-also helps to maintain a smooth airflow (Technology, 14 December 1996, p 18). NASA plans to carry out further trials of this technology later this year.

With passenger aircraft already benefiting from riblets, and experimental aircraft from injections and extractions of air, active devices such as MEMS may have a tough job competing on cost. But Ho is upbeat. 鈥淢EMS are an enabling technology. I am sure that they will have major effects on many aspects of our lives in the 21st century,鈥� he says.

Whether they will also lead to the enlightenment that Horace Lamb was so reluctant to hope for, nobody can say.