Whether you venture onto a few wooden planks over a trout stream, a steel colossus over a swift river or a concrete viaduct carrying bumper–to–bumper commuters, you trust the beams and girders to hold you up. This act of faith, made daily by millions of motorists on U.S. highways, was shaken last summer when a steel truss bridge in Minneapolis plunged into the Mississippi River during rush hour. As media coverage raged and pundits called for reform, Oregon Governor Ted Kulongoski ordered an immediate inspection of 34 similar bridges across the state.
Meanwhile, just 30 miles south of the statehouse, some of the world’s most advanced studies in bridge science were in full tilt. At OSU, researchers in multiple disciplines (civil and structural engineering, ocean and coastal engineering, computer modeling) are investigating destructive forces and possible countermeasures. Human impacts — the loads exerted by cars and trucks, as well as the occasional collision by a boat or barge — comprise just one set of challenges. Equally critical to bridge safety are the myriad processes of nature. Most are routine: currents, tides, wind, erosion, salt air, sub–zero winters, simmering summers. Others are rare but often devastating: floods, hurricanes, earthquakes, tsunamis.
For guidance on bridge evaluation, repair and replacement, as well as design for worst–case scenarios, state and federal transportation officials have turned to OSU.
High off the ground, a guy in a hardhat sits at the controls of a 35–ton yellow crane. As though born to the task, he pushes and pulls the levers, maneuvering the 100–foot hydraulic boom into position over a 40,000–pound concrete beam. Workers grab the bulky hook dangling from the boom and attach it to the massive slab. They give the thumbs–up. With a deafening roar that makes earplugs standard equipment here, the crane hefts the load and swings it into position. Construction site? No, engineering lab.
At OSU’s Structural Engineering Research Laboratory, experimental precision depends on tools that pound, lift, shake and cut: diesel–powered machines, hydraulic rams, welding torches, rebar benders and shears (trade name, Rodchompers). The guy at the crane’s controls, an engineering professor studying the physics of bridges, reveals that his supply lists (which recently included Arctic parkas for his crew of graduate students) have raised a few eyebrows with Research Office accountants.
“I may be the only structural engineering professor in the U.S. who’s a certified hydraulic crane operator,” says a grinning Christopher Higgins as he climbs down from the cab.
Pointing toward the concrete girder, now encased in a cage of steel columns and rods resembling a giant Erector Set, Higgins projects an almost paternal air. “This guy,” he says proudly, “is our Goliath.” An intermediate bridge support called a “bent cap,” the kind that sits mid–river to support bridges with long spans, carries the integrity of the whole structure. As Higgins explains, “If it fails, you can lose the whole bridge.”
“We’re doing some things that no lab in the world has ever done before. We built a moving load simulator that can actually roll, acting like a truck traveling across full-size girders.”
As part of Higgins’ comprehensive research program on concrete bridge components funded by the Oregon Department of Transportation (ODOT) and the Federal Highway Administration (FHA), Goliath will undergo a series of strength and rehabilitation experiments inside the steel cage, the structural–engineering equivalent of a test tube. The futures of the 155,000 U.S. bridges rated “structurally deficient” or “functionally obsolete” by the FHA could depend on the findings.
To reduce risks of catastrophic collapses on our highways, OSU researchers have taken bridge experiments to a whole new level: life–size. Historically, most studies have been done on miniature replicas. Many models are only a fraction of the size of the real structure, says Higgins, a professor in the School of Civil and Construction Engineering. Trouble is, tests on these scaled–down versions have inherent limitations. A pencil–thin wooden beam, for instance, doesn’t act like a two–ton timber, no matter how carefully you design the experiment. That’s because the physical properties of wood, concrete and steel differ geometrically with size. So do the forces that impinge on them.
To get around this problem, Higgins tests bridge components that are as big as the ones holding up the phalanx of ramps and overpasses that crisscross every major city in the country. Access to real–size data lets engineers correct the assumptions and interpolations that plague analytical models built on sub–size experiments. When we think of bridges, behemoths come to mind, like Portland’s I–5 Marquam Bridge, which curves dramatically to a knee–weakening summit high above the Willamette River. But the structures Higgins typically deals with are not “the striking or soaring long–span kind that grab people’s attention,” he says. Rather, Higgins focuses on the mundane and unsung, the “bread and butter” of the highway system, “the ones you cross under and over without even realizing it.”
Whether they are aesthetic masterpieces or unlovely chunks of pure functionality, bridges are being asked to withstand the ceaseless crush of ever–bigger, ever–heavier and ever–more–numerous vehicles. So Higgins and his students punish their experimental girders (bent caps like Goliath, along with smaller T–shaped girders called
“The forces that threaten bridges are not always visible to the naked eye, like a rusty beam or a semi–truck, or to a weather satellite, like a windstorm or a hurricane. Instead, they are distant, invisible and unpredictable.”
T–beams) with mega–forces and maxi–stressors. They pound them with hydraulic cylinders, pummel them with tons of rolling force, and subject them to extremes of heat and cold, down to minus 13 degrees Fahrenheit (hence the need for Arctic parkas). They even use sound to detect invisible defects. By listening to acoustic emissions, the researchers can analyze internal noise sources and pinpoint structural weaknesses.
“We’re doing some things that no lab in the world has ever done before,” says Higgins. “For instance, we built a moving load simulator that can actually roll, acting like a truck traveling across full–size girders. We found that a moving load affects the bridge structure differently than a single load pushing at one spot. The internal stresses change as the load moves across.”
More than 70 of these full–size T–beams, the workhorses of concrete bridges, have been subjected to loads as heavy as 500,000 pounds (the equivalent of about 100 SUVs) in the OSU lab. The idea is to make them fail, determine how to predict that failure and then figure out how best to fix them. “I’m all about existing structures, how to squeeze more life out of them, figure out how much strength is left in them and quantify the risks associated with them,” says Higgins.
As greater and greater force is applied, hairline cracks form at the concrete surface like networks of varicose veins. But causing a 26,000– to 40,000–pound hunk of reinforced concrete to crack is no mean feat. The researchers do it in several ways. To simulate the movement of continuous, everyday traffic (what engineers call “high–cycle fatigue”), the researchers apply millions of repeated bounces to the T–beams with a hydraulic cylinder. To imitate the impact of heavily loaded triple–tractor–trailer rigs (“low–cycle fatigue”), they apply a half–million pounds of downward pressure (a million pounds for the massive Goliath). To test the effects of temperature and shrinkage on strength, they pull the girders lengthwise, using as much as a quarter–million pounds of force.
Once cracked, some of the beams are mended. The purpose is to test the performance of both novel and existing repair techniques. Some are applied internally, others externally. The researchers inject epoxy and insert steel rods. They wrap cracked beams in sheets of a polymer, a composite material reinforced with carbon fibers originally developed for aerospace applications, that bonds to the surface and restricts the cracking like, he says, “a Band–Aid across a cut.” The lab–induced fissures mimic the fatigue cracks that inspectors have found on some 500 of Oregon’s 1,800 concrete bridges, most of which date from the 1950s when President Dwight D. Eisenhower launched the Interstate Highway System. In 2000, ODOT hired Higgins and OSU’s multidisciplinary Kiewit Center for Infrastructure and Transportation to help it assess the state–owned spans. The result of the $1.5 million study was another milestone for Higgins and his team, one that produced a more accurate bridge assessment process and has already saved up to a half–billion dollars for the state. The breakthrough was twofold: better prediction of load capacity for existing bridge components and the development of a load–rating tool known as a “load factor” (a number that statistically represents the expected loading on the bridge). The research has produced the first state–specific load factors in the nation. By using actual traffic data in place of generic figures, the new load factors have brought unprecedented precision and specificity to Oregon’s bridge rating process. “Site–specific load factors are more refined because they are characteristic of a particular bridge site, route or jurisdiction,” wrote former graduate research assistant Jordan Pelphrey (who now designs and fabricates bridges for Knife River in Harrisburg, Oregon) in a paper coauthored with Higgins. “They reflect the actual truck traffic and likely maximum loadings over the exposure period.”
“Based on Professor Higgins’ research, we were ableto reduce the number of bridges that were required to be replaced or repaired,” explains Bruce Johnson, state bridge engineer at ODOT.
In September, Higgins submitted written testimony to the U.S. House Committee on Science and Technology, calling on Congress to create “a national research center focused on safety evaluation of existing bridges that draws on expertise from across the country.” Such a center, modeled after the National Science Foundation’s Earthquake Engineering Research Center, would be “a logical and fruitful” nexus of university research and federal support, Higgins told the committee.
The forces that threaten bridges are not always visible to the naked eye, like a rusty beam or a semi–truck, or to a weather satellite, like a windstorm or a hurricane. Instead, they are distant, invisible and unpredictable. On the seafloor deep beneath the Pacific is a 600–mile seam that bubbles and broods, unseen except by the giant clams and worms inhabiting the superheated, sulfuric waters. At this Cascadia subduction zone off the West Coast of North America, the Earth’s crust is slipping, millimeter by millimeter, beneath the crumpled edge of the continent. Pressure is building, inexorably. When this pressure next releases, as it does every few centuries, the violent quake it unleashes will most likely be followed by a train of water roaring toward shore at 600 miles an hour, inundating communities from Canada to northern California. At Newport on the central Oregon coast, tsunami warnings posted along the beach advise people to head for the hills when sirens blare, a graphic reminder of the offshore fault that could rupture at any moment.
Coastal bridges are vulnerable.
So while Higgins studies load stresses, other OSU engineers investigate wave forces. The cataclysmic 2004 tsunami that killed 230,000 people in Indonesia and neighboring nations caused Oregon highway officials to take a new look at bridge vulnerability along Highway 101. ODOT hired OSU to do a case study of the Spencer Creek Bridge on Oregon’s main artery between Newport and Depot Bay.
Using blueprints of the bridge under construction at Spencer Creek, OSU engineer Solomon Yim ran simulations of three Cascadia quake scenarios on a stateof–the–art supercomputer. The professor of structural and ocean engineering, in collaboration with scientists at the University of Hawaii, used principles of fluid–structure interaction to estimate wave loads on the bridge design for each scenario.
“Although the inundation for two of the three scenarios is generally small because of the steep mountain slopes along the coastline,” Yim says, “the third scenario could send floodwater deep into valleys and basins between mountain ridges, possibly as much as a mile up the Spencer Creek basin.”
Yim stresses, however, that these results are preliminary and that specific recommendations for changes in bridge design are premature. Future supercomputer simulations and large–scale experiments in OSU’s wave lab will lead to new design guidelines for tsunami–resistant structures down the road.
Making sure bridges can stand up to nature’s most fearsome forces is the aim of yet another OSU investigation, this one undertaken by engineering professor Daniel Cox and funded by the Oregon Transportation Research and Education Consortium. In the same cavernous building that houses the Structural Engineering Research Lab, Cox is studying the 2004 failure of Florida’s Escambia Bay Bridge during Hurricane Ivan. The I–10 bridge, whose design is typical of those on the southeastern coast, lost its superstructure (the highway deck) when the storm surge and waves washed over it.
“This is a first–of–its–kind test,” says Cox, who directs the O.H. Hinsdale Wave Research Laboratory at OSU. “No one else has simulated hurricane–force waves on a large–scale physical model of an actual highway bridge.” A veritable pincushion of electronic sensors, the concreteand– steel model will undergo the surging forces of life–like waves in OSU’s flume, North America’s longest hydraulic wave tank. When the data on horizontal and vertical loads, impact pressures and wave conditions are collected and analyzed, engineers will be better equipped to design hurricane–proof bridges to safeguard Gulf and East Coast residents, already braced for the next killer storm.
Trust in Trusses
Gusts, however, don’t have to be hurricane–force to wreak havoc. The day a brand–new bridge in Washington state began to buck like a bronco, the 42–mile–per–hour winds were whipping around wildly but were well short of hurricane velocity. The year was 1940, and the 2,800–foot span — opened to traffic just four months before — had quickly earned the nickname Galloping Gertie for its rollercoasterlike motion.
Engineers were studying ways to stabilize the bridge. But they never got the chance. On that blustery July morning, as Gertie twisted like a corkscrew high above the Tacoma Narrows, motorists abandoned their cars and crawled to safety on hands and knees moments before the bridge broke apart. They watched as their vehicles (along with one hapless cocker spaniel named Tubby) plummeted into Puget Sound.
To this day, the wreckage of that engineering disaster rusts at the bottom of the narrows. The story of Galloping Gertie, legendary in the Pacific Northwest (you can see eyewitness film footage at www.pbs.org/wgbh/nova/bridge/ tacoma3.html) is a cautionary tale known to every civil and structural engineering student.
ODOT engineer Gary Bowling has inspected thousands of bridges. He knows better than most what can go wrong and what’s at stake. “When you’re driving along the highway, you’re putting your faith in people doing their job — the engineers, the inspectors, the maintenance workers,” he says. “Surface hazards like potholes are visible and easy to avoid. You can drive around them. But bridge hazards tend to be hidden. I have yet to see a businessman or a soccer mom stop their car before crossing a bridge and get out to examine the substructure for signs of corrosion or faulty design.”
Whether the focus is on fixing old structures or building new ones, on mitigating traffic loads or withstanding natural forces, OSU’s research has one overarching goal: making bridges worthy of the public trust.