The quake faker: GWU 'shake table' helps engineering students brace for nature's wrath

By Karen Houppert
Sunday, November 7, 2010; W30

Trying to find the high-tech research building that houses George Washington University's driving simulator laboratory in Ashburn can be challenging. There is construction along Route 7, a new overpass and limited signage. MapQuest is behind the curve; directions have been rendered obsolete.

The temptation to pick up a cellphone and call the folks at the simulator lab -- "Where the h*** are you guys?" -- is severe.

Don't do it!

That's the advice of graduate students Johann Moreau, 26, and Simon Bonnevie, 26, who are studying transportation safety at GWU and can rattle off the statistics from memory: You are four times more likely to crash if using a hand-held device. You're 20 times more likely to crash if you're texting. But these engineering students also understand the danger in a more visceral way. A few months back, Bonnevie swerved into oncoming traffic while texting and almost killed two pedestrians crossing the road.

But then, this was on the university's driving simulator.

The blue Buick Regal, its engine removed and a slew of wires connecting it to a nearby bank of computers, sits in a warehouse-like laboratory. On the large screen in front of the car, images of a road winding through suburbia flicker by. Moreau and Bonnevie, along with their classmates, conducted a series of tests on each other last year to find out how texting, talking on a cell, encountering external distractions (such as construction) and internal distractions (such as chatty passengers passing maps) affect driving safety.


This is just one of the ways GWU is using high-tech simulators to conduct research that eventually will help save lives.

Down the hall from the Buick is the "shake table," as students call the earthquake simulator. The table -- a $1 million, 10-by-10-foot metal structure that moves in six directions -- replicates earthquakes and allows engineering students to test construction materials to see how they hold up under tremors of varying strength. Last year, students built a 90 percent scale model of a bridge's concrete columns to test how earthquakes affect such highway overpass supports.

"If I say, 'the concrete is crushed and the column deformed,' these are just words," explains Pedro Silva, an associate professor in GWU's Civil and Environmental Engineering department. "But if they can watch it fall apart, and if the bar fractures, they can witness it and feel it and then explain it mathematically."

At the moment, Silva is gearing up for his next project with graduate students: building a scale model of a Haitian home using material and construction techniques employed by builders there. Using the shake table to create powerful tremors, Silva will watch how the Haitian house falls. The table moves as much as 8 inches in any direction and is built on 170 tons of concrete that dip 25 feet to 30 feet below to help keep the table's vibrations from shaking the building itself.

"We will study the rubble and how the house collapses so that fire or search-and-rescue teams can see what area a person is mostly to survive in," he says. "That way, if they have only a few minutes to go in, they will know where they should look first." After this first experiment, Silva and his students will rebuild the house, retrofitting it with materials that will make it more structurally sound and efficient. "Not stronger," Silva cautions, explaining that this term can be misconstrued. "With earthquakes, you want to build so that the house is more flexible." They will then reactivate the shake table to test how well the altered structure withstands a quake.

For Silva, what makes this project so challenging -- and interesting -- is that neither he nor his students can work in isolation but have to think within the social and economic context of Haiti. In other words, it won't do any good for Silva's research to simply identify the best material for reinforcing houses in Haiti; if it is costly or hard to come by, the upgrades won't happen.

To that end, Silva also is partnering with other GWU departments and universities in the Caribbean to identify realistic options. At GWU, students in the engineering management department will analyze trash from Haitian landfills to see what salvageable materials might be recycled and cheaply redeployed to reinforce existing structures. They are casting a wide net. Plastics? Metals? Fruits? An abundance of any of these things might prove useful. Silva is feeling optimistic about mangoes. A factory to weave the discarded fibrous tissue might put out useful reinforcing material, he speculates, or maybe it could be baled into bricks?

As Silva dreams, his colleague down the hall, professor Azim Eskandarian, would probably nudge him from his reverie. Eskandarian, who directs the Center for Intelligent Systems Research, has been using simulation technology to examine the effects of driver distraction and drowsiness. He discovered that it's fairly simple to discern when a driver is getting sleepy. You don't need wires to detect drooping eyelids or other high-tech devices; the steering wheel tells the whole story. Turns out a fatigued driver overcompensates with the steering wheel, zigzagging more than he might ordinarily because it takes him longer to notice if he is drifting toward the center line and then, because his reaction time is slower, turns the wheel more sharply to compensate.

"People make micro-corrections in steering the whole time they're driving," Eskandarian says. "But they make much larger corrections when they're tired, and we can easily measure that difference."

There are approximately 80,000 crashes a year as a result of drowsiness, and equipping cars with steering sensors could save lives, Eskandarian says.

"Maybe there could be a warning signal -- auditory beeping or visual flashes -- to alert the driver that his steering is getting more erratic," he says. The technology already exists and is used in lane detection devices that beep if a driver goes out of a lane without using the turn signal, or in cars that read the speed limit and alert motorists who have exceeded it.

And perhaps the next generation of researchers -- Eskandarian's students, for example -- will use the driving simulator to develop a warning system for those who text and drive. The impetus is there. "More than 90 percent of my students who texted and drove got into accidents," he says.

Karen Houppert is a contributor to the Magazine. She can be reached at

© 2010 The Washington Post Company