20 June 1999

Engineers Set off Mini-Earthquake

Engineering researchers from Brigham Young University and the University of California, San Diego created a custom mini-earthquake in the middle of the San Francisco Bay this week to test exactly what happens to buildings, homes and bridges when the soil beneath them gets a little shakey.

The engineers detonated explosives buried 12 feet deep in the island's sandy soil to simulate a geologic phenomenon called liquefaction, which occurs when an earthquake causes an increase in water pressure, turning once-stable soil into a vat of mud with only a fraction of its former strength. It has occurred during earthquakes in Japan and California - for example, the 1989 tremor that interrupted the World Series.

Areas with loose, saturated, sandy soil, especially near waterways, are at risk for liquefaction during an earthquake. Six state departments of transportation and the Federal Highway administration funded the BYU/UCSD project in hopes that results of the experiment would provide safety guidelines for bridges and buildings in such areas.

"Until now, engineers have had to rely on 30-inch-long models and educated guesses when trying to build earthquake-safe structures in liquefiable soil," says Kyle Rollins, BYU professor of civil engineering and co-principal investigator on the project. "This test provided the first hard data of its type in the world for designers to base their plans on."

Rollins and teamed with teamed with Scott Ashford, assistant professor of geotechnical engineering at UCSD, to piece together an elaborate experiment two years in the making. Sixteen 1-pound explosive charges were buried around nine steel pillars, like those used to anchor bridges and building foundations, which were sunk 42 feet in the ground. With monitoring equipment and sensors, the researchers hoped to document how the columns would react to the sudden surge in water pressure brought on by the explosions.

"To get the information we needed, we could either set up all the sensors and wait for an earthquake, or use the explosives to 'make' an earthquake," says Rollins. "This way we had all the excitement and educational value of an earthquake without any of the death or destruction."

That excitement included heaving ground and the sudden appearance of "sand boils" - small geysers less than a foot high that left behind volcano-like piles of sand - that indicated liquefaction had been successfully achieved.

With the sand boils forming and the ground settling from the blast, the engineers crowded around computer screens in their portable lab, fashioned from a trailer. They watched a hydraulic jack push on the pillars with up to 300,000 pounds of force, simulating the side-to-side movement of a building during an earthquake. All the while, hundreds of sensors on the pillars and in the sand reported the results.

Engineers have typically assumed that liquified soil has little to no strength, Rollins explains. But initial results from this experiment indicate that substantial resistance can be developed once the pillars have shifted 6 to 9 inches. That means engineers need to build bridges and building foundations that can slide side-to-side without crumbling.

In addition to measuring forces brought about by liquefaction, the researchers also tested ways to mitigate effects of the problem. Plastic drain pipes driven into the soil helped reduce water pressure after one blast.

In another test, thick stone columns embedded near the pillars succeeded in compacting the soil enough to prevent liquefaction. During other tests, the engineers realized there might be a more unorthodox solution.

"We found that the soil in our test areas settled anywhere from 4 to 12 inches after each blast, making it less susceptible to liquefaction," Rollins said. "Perhaps blasting soils before building in them might be a good preventative technique."

The researchers chose Treasure Island, located between San Francisco and Oakland, because it has liquefiable soil and is currently not in use since the base closure.

With their field experiments now complete, Ashford and Rollins have returned to their respective universities to analyze the data. The sensors measured the movement of the pillars and the water pressure every 10th of a second, which left the researchers with four 100-megabyte Zip disks full of data to examine.

They will use this information to create design standards engineers can use for building earthquake-resistant foundations safely in liquefiable soil. Similar design parameters for building in clay and regular sand were introduced to the engineering world two decades ago.

With their tests only a few months old, Rollins and Ashford have already been invited to present their findings at the International Conference on Geotechnical Earthquake Engineering and Soil Dynamics, held every four years, in 2001. They will also publish their specific results in engineering journals in the coming months.