Shaking up EAPS
Read about how Professors Oral Buyukozturk (CEE), Nafi Toksoz (EAPS) and Mehmet Celebi of the U.S. Geological Survey, fitted the Green Building with a seismic monitoring system and what they found out about the tallest building in Cambridge.
Read this story at MIT CEE
The seismic monitoring of buildings is particularly important in high-population urban areas like Greater Boston. While Massachusetts’ seismic building codes are adapted from California’s, the geological conditions in the two states are very different. The soft soil of land reclaimed from the Charles River on which some areas of Boston and Cambridge are built could make structures here more vulnerable to damage from earthquakes of small magnitude, particularly if the frequency of the seismic activity matches the fundamental frequency of a structure or site. But baseline measurements of buildings in these cities in response to environmental conditions — which could help predict their response to a high-magnitude earthquake — are generally not available.
Professor Oral Buyukozturk, working with Professor Nafi Toksoz of MIT’s Department of Earth, Atmospheric and Planetary Sciences and Mehmet Celebi of the U.S. Geological Service, outfitted the 20-story Green Building on the MIT campus, the tallest building in Cambridge, with a seismic monitoring system. Designed by I.M. Pei and built in the early 1960s, the building stands 80 meters above ground, has a one-story basement, measures 14.6 meters on the east-west ends and 34 meters on the north-south sides, is made of reinforced concrete and has a foundation of thick concrete beams reinforced by rebar. Structural characteristics that make it a good subject include shear-resistant walls of concrete integrated into the outer walls of the windowless east-west ends to prevent in-plane movement; an open ground floor; and asymmetry created by having elevators at only one end of the building and heavy meteorological research equipment on the roof.
A building can move by deformation (shearing between floors, bending or torsion) or the entire structure can rotate or shift translationally. To detect deformation, the team deployed 36 accelerometer sensors at points where the building is most likely to move. Structures are impacted by both the magnitude of a seismic wave and its frequency. The site of the Green Building is known to have fundamental frequency of 1.5 hertz. Data provided by the sensors allows the researchers to compute the velocity of movement and the vibration frequency of the building.
Using the accelerometer array, the team observed bending, shear and torsional deformations, as well as some of the building’s more unique structural behaviors. One novel finding is that the building undergoes base-rocking in the narrow direction. That is, the foundation makes a rigid-body rotation about the east-west axis, and the corresponding piles are pulled and pushed in the ground. Soft soil conditions and the stiff exterior walls likely cause this motion, which is rarely observed under low-amplitude vibrations. They also detected a strong twisting motion along the building’s height. Closer analysis of the data revealed the structure’s resonant frequencies: 0.68 hertz in the east-west direction, 0.75 hertz in the north-south direction and 1.5 hertz in torsion.
This study demonstrates that the deployed accelerometer system can be used to calculate interstory drifts and bending, shear and torsional deformations, which are important parameters for characterizing a structure’s dynamic performance. These values, coupled with the capabilities of the accelerometer system, make it possible to assess the Green Building’s ability to withstand strong-motion earthquakes — an unknown for most buildings in New England. Using the building’s fundamental frequencies and ambient vibrations also allows the team to create computer simulation models to predict the impact of seismic and other environmental events.
A paper on this work by Oral Buyukozturk, Nafi Toksoz and Mehmet Celebi will appear in the September issue of Earthquake Spectra. Graduate student Peter Adam Trocha and others are now working with Buyukozturk and Professor Eduardo Kausel to create a number of computer simulation models.
Download this issue as a pdf.