|
Bridge sensors send stress data to Drexel researchers
Nov 16, 2009 (The Philadelphia Inquirer - McClatchy-Tribune Information Services via COMTEX) --
When heavy trucks rumble across the Burlington-Bristol Bridge, new sensors detect the stresses they put on the 2,300-foot span and instantly transmit the information to computers in Philadelphia.
Drexel University researchers in the city then use the stream of data to develop computer models that display the bridge's condition and its potential problems.
Such measures, known to engineers as structural health monitoring, are commonly done for a limited period and confined to academic research. But Drexel and its partner, the Burlington County Bridge Commission, plan to use the system continuously to help the agency better allocate its resources for bridge maintenance.
Around the country, engineers are researching uses of "smart technology" to advance the safety and efficiency of bridges, while also grappling with how to make sense of the vast information such devices convey.
In Minnesota, for example, engineers are analyzing data from hundreds of sensors on the new Interstate 35 bridge erected last year in Minneapolis to replace the span that collapsed in August 2007, killing 13 people.
Others are exploring wireless sensors, which were used in a pilot project on the Golden Gate Bridge. At the University of Michigan, wireless sensors under development would harvest energy from bridge vibrations. Other projects there include a "sensing skin" that would be painted on parts of a bridge to detect cracks and corrosion.
The data from sensors are "invaluable to get what's happening inside the structure," said Khaled Mahmoud, chairman of the Bridge Engineering Association in New York City. The information, he said, helps not only engineers but also bridge owners who must allocate limited funds for maintenance.
Drexel began working with the bridge commission following the I-35 bridge collapse, when the agency reached out to the university for guidance on bridge monitoring.
The commission takes in just under $30 million a year from $2 tolls on the Tacony-Palmyra and Burlington-Bristol Bridges. The spans are about 80 years old.
In coming months, the grid deck of the Tacony-Palmyra will be replaced and equipped with strain gauges -- metal strips that measure stress -- and accelerometers, which measure vibration and are used in such products as BlackBerrys and air bags.
Ten miles up the river, the commission will replace the bearings on the Burlington-Bristol and outfit them with load cells, cylinders about 2 inches high and 6 inches in diameter with a ball at the top. They will transmit data on how load is being transferred across the bridge.
Both replacement projects were in the works when the bridge commission decided to outfit the new parts with sensors.
Such devices are not new technology, but their use in bridges has provided new ways to monitor their performance.
"We have plenty of sensors, and we can get good-quality data," said Franklin Moon, an engineering professor at Drexel who is on the project. "It's how do we interpret that data to help owners make decisions more effectively and stretch their dollars?"
Already under way is the placement of strain gauges on a metal component that spans the top of the Burlington-Bristol. The devices send information on bridge stress 24 hours a day in real time to computers in the University City building of Pennoni Associates, which holds a contract as the commission's resident engineer and has cleared office space for Drexel doctoral and postdoctoral students to work on the project.
Reviewing all the data -- which appear on computer screens in the way an EKG measures a heartbeat -- is neither practical nor useful. The team has established thresholds that, when triggered, allow the data to be saved for later analysis.
Those can then be used to "calibrate" computer models of the bridges so that the models are based increasingly on real information rather than assumptions. Moon said his research looks at how to generate tens of thousands of computer bridge models in an automated way, which would give the engineers a clearer picture of what a bridge is telling them than just one or several models could provide.
Eventually, Pennoni -- whose founder, C.R. "Chuck" Pennoni, is interim president of Drexel -- plans to establish a control center at the bridge commission's main building in Palmyra.
John Fisher, a professor of engineering at Lehigh University, said there's no need for long-term, real-time monitoring of many structures if they are well-designed and there is no evidence of distress.
"It has to be rationalized because, one, there's a cost associated with it, and two, why make measurements if they're not needed?" he said.
The use of continuous, real-time structural monitoring is not widespread on the nation's bridges.
As computer technology exploded in the 1990s, sensors became cheaper, more advanced, and widely available, but engineers didn't make a good case for what bridge owners could get out of them, said Zee Duron, chair of the engineering department at Harvey Mudd College in Claremont, Calif.
"Engineers have got to get smarter about what the sensors are telling us, and we've got to get more politically astute in terms of how we take that information and turn that into economic and public policy that actually improves the infrastructure of the United States," he said.
Bridge commission officials, however, described an interest in taking the initiative to use the latest research, rather than waiting for it to come to them.
"It certainly is an investment, and it's something that [bridge owners] need to see the cost benefit of, and until they're convinced of the business case, they may not take advantage of it," said Moon, whose university has a contract with the commission capped at $400,000.
At the University of Minnesota, engineers are using information from sensors to develop more accurate and sophisticated 3-D computer models of the I-35 bridge that can look at how damage might manifest itself.
Sensors in the deck of the bridge are designed to detect corrosion -- part of a system that "helps you be ahead of the game, and it helps you replace the deck before you have a problem," said Cathy French, a University of Minnesota civil-engineering professor leading the project.
Yet one challenge in monitoring long-standing bridges is understanding stresses that are inherent in the structure, according to French. She pointed out that one cause of the I-35 collapse was that its gusset plates were too thin -- a design error committed a half-century ago and not necessarily detectable if sensors had been attached.
There is still broad agreement that such systems should not replace the in-person inspections required every two years by the Federal Highway Administration, which says the technology's long-term value has yet to be demonstrated.
"The commercially available systems that are being marketed to bridge owners today produce huge amounts of data, but limited useful decision-making information," spokeswoman Nancy Singer wrote in an e-mail, on behalf of the agency's bridge experts.
"They require a high level of expertise to interpret the data," she said, "and detailed analytical models to assess the importance of it."
Contact staff writer Maya Rao at 856-779-3220 or mrao@phillynews.com.
To see more of The Philadelphia Inquirer, or to subscribe to the newspaper, go
to http://www.philly.com/inquirer. Copyright (c) 2009, The Philadelphia Inquirer
Distributed by McClatchy-Tribune Information Services. For reprints, email
tmsreprints@permissionsgroup.com, call 800-374-7985 or 847-635-6550, send a fax
to 847-635-6968, or write to The Permissions Group Inc., 1247 Milwaukee Ave.,
Suite 303, Glenview, IL 60025, USA.
[ Back To it.tmcnet.com's Homepage ]
|