Erik Anderson

MIT/WHOI Joint Program, Applied Ocean Physics and Engineering

Applied Ocean Physics & Engineering
Advisors: Mark Grosenbaugh, Wade McGillis, Mike Triantafyllou
Research interest: Hydrodynamics, energetics, and optimization of swimming in marine organisms with application to marine vehicle propulsion and marine ecology.
Expected graduation: May, 2004

A dolphin should not be able to swim. So said Cambridge University zoologist James Gray in the 1930s. The friction caused by water moving over a dolphin's skin, he said, should be like swimming in cold molasses. But dolphins obviously can swim, and Erik Anderson wants to find out why. The paradox springs from a conflict between idealized models and a reality that is difficult to observe and measure. In trying to calculate drag-the force that slows the movement of an object through air or water-Gray could only use formulas and observations based on the flow of water over rigid, static bodies. The physical and mechanical models fell short of the biological reality, and Gray and other scientists were stuck with the implausible conclusion that the force a dolphin or fish needed to overcome drag was greater than the force their muscles could generate.

But the body of a swimming dolphin or fish is hardly rigid or static. So instead of modeling the forces of drag, Erik Anderson has combined 21st century lasers, robots, and cameras with old-fashioned attention to microscopic detail to build an experiment that measures the forces. "I want to know how a real fish swims," says Anderson, a student in the MIT/WHOI Joint Program. "I want to know their trade secrets."

The Accidental Tourist

In the summer of 1997, Anderson had an encounter that would change his scientific life. As a master's degree candidate from St. Francis Xavier University (Nova Scotia), he visited WHOI for a study of the fluid dynamics of squid motion. Anderson and colleagues had traveled to Woods Hole because their subjects were abundant and accessible-the squid Loligo pealei migrates through the waters off the WHOI pier every spring-and because the Coastal Research Laboratory had a large flume close to the delicate animal's habitat.

One day while shooting high-speed, high-resolution images of Loligo swimming in the flume, Anderson was approached by a man who-without introducing himself-started asking questions about the experiment. "I thought he was a tourist and I was very busy," he recalls, "so I brushed him off with simple answers."

A week later, Anderson attended a scientific presentation by WHOI Associate Scientist Mark Grosenbaugh of the Applied Ocean Physics & Engineering Department. Anderson was curious to meet Grosenbaugh, whose papers he had read. When Grosenbaugh stood up to give his seminar, Anderson did a double take: Grosenbaugh had been his "tourist." "He was not a guy I should have been getting rid of," Anderson says.

Despite the awkward introduction, Anderson began talking with Grosenbaugh and WHOI Associate Scientist Wade McGillis (AOP&E). They are now his thesis advisors. "When they told me about the fish-drag problem, it sounded like a logical step after the squid work," Anderson recalls. "I was especially drawn to the challenge of engineering an experiment."

Within a year, Anderson was admitted to the MIT/WHOI Joint Program and began conducting experiments on a robotic tuna... and hatching plans for one of the most sophisticated studies of fish propulsion ever devised.

Swimming Against the Stream
Fish swim against two fluid forces: form drag and friction drag. Anderson compares the physics to riding a bicycle. "When you cycle through air-which is also a 'fluid' medium-you feel pressure on your chest and suction from a wake at your back," Anderson notes. "The pressure and suction are the source of form drag. Friction drag is caused by the air sliding over your skin."

In the 1970s, Japanese researchers who observed fluid flowing over a waving rubber mat found that form drag may be almost negligible on a swimming fish. They demonstrated that as a surface undulates, fast-moving fluid is drawn closer to the surface, creating a smaller wake and reducing suction. The streamlined shape of most fish also results in reduced pressure in the front.

That leaves drag from friction, a force that has been modeled and studied in idealized experiments, but not measured directly on freely swimming fish. Anderson decided to use modern technology to zoom in on the "boundary layer," the critical zone around a fish where the effects of friction have a measurable impact on the flow of water. This layer is usually less than one centimeter thick.

The experiment seems simple enough: photograph the water flowing through the boundary layer and across a fish's skin. From such pictures, Anderson can calculate the friction drag and visualize the shape of the fluid flow, much as engineers study the flow of air over airplane wings and cars. But as Anderson explains, "It was a challenging engineering problem to build a data-taking apparatus that could accurately measure friction drag."

"No one's been able to measure water's speed in the boundary layer of fish before," says George Lauder, who studies the flow around aquatic animals at Harvard University. "It's extremely hard to do, and Anderson's system is the first to be able to accomplish this."

How does Anderson see water moving? He puts microscopic, silver-coated beads into the flume as a tracer, then illuminates them with a sheet of light from a laser. To capture an image of water within 12 millimeters (0.47 inches) of a fish's skin, Anderson points a high-resolution digital camera at the sheet of laser light from beneath the flume. Another camera captures the same angle, but with a field of view ten times larger, in order to get a general picture of the flow inside and outside the boundary layer. A third camera snaps a side view, recording the position where the laser impinges on the fish.

The three cameras and laser are synchronized in "dual pulses," taking pairs of freeze-frames two milliseconds apart. Lit by the laser, the beads appear as tiny white dots against the photo's black background. The "image pairs" generated by any of the cameras are nearly identical photos, but close inspection shows that the beads have moved slightly from one exposure to the next. With the help of several sophisticated computer programs that Anderson wrote himself, he can measure the distance the beads travel, effectively measuring how fast the water in the boundary layer traveled in millimeters per millisecond.

Anderson creates a map of water speed and direction along the entire length of a fish. From measurements of how water velocity changes with distance from the fish, he calculates friction drag at the skin. With that data, Anderson can generate a complete picture of the "shape" of the boundary layer and the friction drag along a wriggling fish.

In the first year of his study, Anderson waited days, sometimes weeks, for the fish to swim through his 1-centimeter-square field of view. He surfaced, bleary-eyed, from his experiments with just ten seconds of usable image data. With experience and increased funding from the National Science Foundation, the Office of Naval Research, and WHOI's Ocean Ventures Fund, Anderson's system has evolved from fixed cameras on stands to cameras and the laser gliding on the motorized rails of an 800-pound robot.

"Now, I just sit in front of the tank with the joystick, training the laser on the fish," Anderson says. These days, the heavy lifting comes from sifting through the mountains of data. Each camera captures hundreds of gigabytes of images per two-hour experiment.

Data isn't the only thing Anderson catches with his work. An avid angler who spent childhood weekends fishing with his dad, he collects his own experimental "volunteers" (eel, mackerel, dogfish, bluefish, and scup) in Vineyard Sound and Cape Cod Bay. "Having to fish as part of my research is hilarious," Anderson says. "All my fishing buddies just laugh when they hear I get paid to go fishing."

Something Fishy About Drag
Ever since Gray calculated that dolphins shouldn't be able to swim, there has been a controversy in the field of biomechanics: Is drag more intense on a wriggling fish than on a rigid body like a submarine? James Lighthill, another Cambridge scientist and the mathematical genius who helped design the supersonic Concorde jet, predicted in the 1970s that the drag on a real fish would be up to five times greater than on a stiff mock-up.

After observing real swimmers and crunching the data, Anderson saw what Lighthill had predicted. "The friction drag on scup and dogfish was higher than on a rigid body," Anderson says. Essentially, the friction on a swimming fish is higher than that on a coasting fish. "But we also calculated that a scup has two to twenty times the strength needed to overcome this drag." He reported his findings in a 2001 paper in The Journal of Experimental Biology, a publication that James Gray edited for 20 years.

Anderson suggests that his most important observation is that the historical concept of "drag reduction" has been improperly defined when it comes to swimming fish. "A true study of swimming fish should compare the drag of one swimming motion compared to another, not a swimming fish to a rigid model," says Anderson. "A rigid eel goes nowhere, so it doesn't make any sense to compare its drag to a swimming eel, unless you are studying swimming versus gliding.... or trying to decide whether vehicles should be rigid or fish-like.

"It's possible that while a swimming fish has higher friction drag than a rigid, fish-shaped model, a real fish might tune its motion to have lower drag than another, similarly shaped fish," he adds. "This is the more interesting question: What slight differences in swimming motion and body structure result in enhanced efficiency, stability, or maneuverability?"

Much of the time, the key to minimizing friction drag is to keep water flowing smoothly over the fish's skin and to reduce surface area. "For small and slow fish, the boundary layer is often laminar or smooth," Anderson notes. Some fish appear to have natural drag-reducing strategies, such as slots where they can tuck their fins for a sleeker profile.

But sometimes the opposite is true. "For some large and fast fish, the boundary layer sometimes appears to be turbulent," Anderson says. The rough skin of sharks, for instance, may actually stir up the boundary layer. Such turbulence, although it increases friction drag, can decrease form drag by a greater amount. That's because an eddying boundary layer produces a smaller wake and decreased suction at the tail end. (The same principle is at work in the dimples on a golf ball.) Sharks have a method to counteract friction as well, using tiny riblets along their surfaces to organize the turbulent flow in a way that decreases friction.

Anderson's research "has implications for the design of more energy efficient underwater vehicles," says Lauder. "His findings might help show whether future vehicles should be flexible like a fish or rigid like a submarine." Along those lines, boat builders have adopted some of nature's ideas by using riblet technology on the hulls of racing boats.

As Anderson finishes his Ph.D. dissertation on fish motion, he is already daydreaming about how to put Gray's dolphin paradox to the test. He'd like to measure the drag on a live dolphin. If he can find the resources and adapt the technology, he will put his camera and laser system into a watertight backpack and snap shots of the dolphin's boundary layer as it swims. That's assuming that a dolphin actually can swim...

Originally published: April 1, 2003