Winds of change: a Danish aerodynamicist has discovered a way to reduce the turbulence that plagues many long-span bridges.

[ILLUSTRATION OMITTED]

When design of Norway's Halogaland suspension bridge begins in earnest later this year, it is likely to be the testing ground for a new development in aerodynamic bridge deck design. By reducing the angle along the edges of the box girder, Cowi wind engineer Allan Larsen believes it will not be necessary to install guide vanes to ensure its stability.

The idea, which came to Larsen while he was sailing, paves the way for safer bridges that also reduce construction and maintenance costs. But its origin goes back more than a decade to the official opening of Denmark's Storebaelt Bridge in 1998.

Larsen recalls how warm spring air passed over the cold waters of the Great Belt, the strait that separates the islands of Zealand and Funen in Denmark. The interaction between air and water created a steady breeze that unexpectedly set the bridge into motion. "I received a call from the bridge's inspection team on the bridge that I better have a look, because the bridge was shifting by up to one metre," says Larsen. The oscillations were caused by wind vortex shedding--the phenomenon that arises when airflow around an object separates--much like an aircraft wing stalls when its angle of attack exceeds its performance envelope.

Bridge designers have been well-acquainted with the phenomenon ever since they began using closed box girder decks. While the box girders simplify the construction process and provide greater torsional rigidity than the lattice girders used on older structures, they are known to create coherent vortices in the wake of the box girder.

"When you have a very light and flexible structure like a suspension bridge or a cablestayed bridge, then the formation of vortices and the flexibility of the structure work together to move the bridge in resonance--sometimes by up to a metre. And that is thoroughly unpleasant to watch or experience," says Larsen.

The collapse of the Tacoma Narrows Bridge shortly after its opening in 1940 was in fact preceded by vortex-induced oscillations and is often erroneously attributed to this process. Almost a decade ago, Larsen's research revealed that the action that finally tore the bridge apart was in fact due to vortices forming in an alternating fashion at the upper and lower upwind edges of the H-shaped plate girder. On the Storebaelt Bridge engineers avoided the same outcome by fitting a series of guide vanes to the bridge girder, smoothing the air flow around the lower corners of the box girder and reducing the creation of vortices.

A year after opening the Storebaelt Bridge, Larsen and Cowi were approached by the Norwegian Road Directorate for a similar problem on the newly-completed Ostery Bridge. Wind tunnel tests with a model of the girder of this bridge proved that the guide vane solution could also be applied to this bridge although with a somewhat different design than that on the Storebaelt Bridge. Now, ten years down the road, the Osteroy Bridge remains tamed and without a trace of vibrations.

But although guide vanes cured the problem, Larsen wasn't satisfied. "They take away the worst movements, but they're a bit of a patchwork solution," he says. He continued to contemplate the wind vortex shedding problem while working on other major bridge projects. But more than a decade passed before he found a solution in a very unexpected way. Last summer, Larsen was trying to trim the jib on his sailboat while sailing off the coast of Denmark. He noticed that the wool tufts attached to his jib--often referred to as 'tell-tales'-were streamlined when the sail formed an angle of about 15[degrees] with the wind.

With the [empty set] resund Bridge in the distance, a moment of insight struck him as his thoughts drifted to the problem of wind vortex shedding. Larsen explains: "It occurred to me that the sails were telling me what to do: be sure that the wind flow is smooth along the surface and then your problem will be cured." Back on dry land, Larsen began the preliminary sketches for an experiment to test his idea. He proposed building five models of bridge box girders to the same design but with a range of angles from 12[degrees] to 26[degrees] at the lower corners of the box.

With financial support from the Cowi Foundation, Larsen travelled to the aerodynamic laboratory at the National Research Council in Ottawa, Canada. Using the laboratory's wind tunnels and a spring-supported model set up, he was able to simulate the cross-wind conditions that often produce wind vortex shedding and to map out the wind speed ranges in which oscillatory cross-wind movements of the deck models occurred. A hot-wire probe mounted in the wake of the bridge deck models was used for detecting the frequency of vortex shedding, thus verifying that the movements were actually due to resonance between vortex shedding and the natural frequencies of the spring suspended model. Refracted laser light passing through smoke injected in the flow revealed the formation of vortices and the images were captured by a high speed camera.

[ILLUSTRATION OMITTED]

[ILLUSTRATION OMITTED]

[ILLUSTRATION OMITTED]

Larsen's experiment confirmed his suspicions, and he found that the angle between the bottom plate and the lower sides of bridge girders played a decisive role in whether air flows evenly over them. The model with an angle of about 15[degrees], for example, was virtually free of vortex formation in the near wake and the oscillations that accompany wind vortex shedding when at resonance with structural frequencies.

Larsen welcomed the chance to return to the laboratory and experiment--"that's something we don't get the chance to do everyday"--and he looks forward to applying the discovery to upcoming bridge designs. "This discovery not only reduces vortex shedding excitation and the accompanied structural motions, but it will also save money. One time I calculated on the back of an envelope that it could save about 7[euro] million per kilometre of span--that's not much compared to the overall cost of a bridge, but it's still money," he says. "Plus the architects and maintenance people will probably like it--it's sleeker and requires less paint work," he adds with a smile.

The source of his idea demonstratess how inspiration can come at the most unexpected moment. Larsen has been sailing since he was eight years old and says it's no coincidence the idea for 'trimming' bridges came to him outside the office, rather than while he was staring at his computer screen. He suggests some of his best thinking is done while navigating his boat; the combination of sun, wind and water encourages the flow of creative thoughts, he opines. And although the idea came to Larsen far from the office, he and Cowi colleagues have already begun to draw benefits from the discovery as they begin initial work on a proposed fixed link between Denmark and Germany over the Fehmarn Belt