Engineering: A new generation of “smart” bridges use sensors to detect structural problems and warn of impending danger
WHEN an eight-lane steel-truss-arch bridge across the Mississippi River in Minneapolis collapsed during the evening rush hour on August 1st 2007, 13 people were killed and 145 were injured. There had been no warning. The bridge was 40 years old but had a life expectancy of 50 years. The central span suddenly gave way after the gusset plates that connected the steel beams buckled and fractured, dropping the bridge into the river.
In the wake of the catastrophe, there were calls to harness technology to avoid similar mishaps. The St Anthony Falls bridge, which opened on September 18th 2008 and replaces the collapsed structure, should do just that. It has an embedded early-warning system made of hundreds of sensors. They include wire and fibre-optic strain and displacement gauges, accelerometers, potentiometers and corrosion sensors that have been built into the span to monitor it for structural weaknesses, such as corroded concrete and overly strained joints.
On top of this, temperature sensors embedded in the tarmac activate a system that sprays antifreeze on the road when it gets too cold, and a traffic-monitoring system alerts the Minnesota Department of Transportation to divert traffic in the event of an accident or overcrowding. The cost of all this technology was around $1m, less than 1% of the $234m it cost to build the bridge.
The new Minneapolis bridge joins a handful of “smart” bridges that have built-in sensors to monitor their health. Another example is the six-lane Charilaos Trikoupis bridge in Greece, which spans the Gulf of Corinth, linking the town of Rio on the Peloponnese peninsula to Antirrio on the mainland. This 3km-long bridge, which was opened in 2004, has roughly 300 sensors that alert its operators if an earthquake or high winds warrant it being shut to traffic, as well as monitoring its overall health. These sensors have already detected some abnormal vibrations in the cables holding the bridge, which led engineers to install additional weights as dampeners.
The next generation of sensors to monitor bridge health will be even more sophisticated. For one thing, they will be wireless, which will make installing them a lot cheaper.
Jerome Lynch of the University of Michigan, Ann Arbor, is the chief researcher on a project intended to help design the next generation of monitoring systems for bridges. He and his colleagues are looking at how to make a cement-based sensing skin that can detect excessive strain in bridges. Individual sensors, says Dr Lynch, are not ideal because the initial cracks in a bridge may not occur at the point the sensor is placed. A continuous skin would solve this problem. He is also exploring a paint-like substance made of carbon nanotubes that can be painted onto bridges to detect corrosion and cracks. Since carbon nanotubes conduct electricity, sending a current through the paint would help engineers to detect structural weakness through changes in the paint’s electrical properties.
The researchers are also developing sensors that could be placed on vehicles that regularly cross a bridge, such as city buses and police cars. These could measure how the bridge responds to the vehicle moving across it, and report any suspicious changes.
Some civil engineers are sceptical about whether such instrumentation is warranted. Emin Aktan, director of the Intelligent Infrastructure and Transport Safety Institute at Drexel University in Philadelphia, points out that although the sensors generate a huge amount of data, civil engineers simply do not know what happened in the weeks and days before a given bridge failed. It will take a couple of decades to arrive at a point when bridge operators can use such data intelligently, he predicts. Meanwhile, the Obama administration’s stimulus plan has earmarked $27 billion for building and repairing roads and bridges. Just 1% of that would pay for a lot of sensors.
Source of Information : The Economist 2009-09-05