April  2008

Earthshaking Research

A nationwide network of engineering researchers is improving
understanding of seismic events, the built environment, and how to improve safety when the ground starts to rumble.

BY EVA KAPLAN-LEISERSON

Of all the natural disasters, an earthquake is one of the most hazardous and devastating. According to one estimate, more than one million people have lost their lives in the 20th century due to earthquakes worldwide. One in China killed 250,000 people in 1976. While earthquakes in the U.S. cause relatively few deaths—220 in 15 quakes since 1965—they can cause serious economic losses. In California, the 1989 Loma Prieta and 1994 Northridge earthquakes caused $5.9 billion and $30 billion in damage, respectively.

For earthquake engineering researchers, the challenges are many. They can't predict when an earthquake will occur, and their observations come after the fact, if they're granted access to a site or if debris hasn't already been cleaned up. Plus, data from tests on small-scale models is often not scalable to full-size structures.

Since 2004, however, a unique collaborative network has greatly enhanced the field of earthquake engineering. Today, the network provides researchers from around the world access to state-of-the-art experimental facilities and data to model and learn about earthquakes' effects.

The George E. Brown Jr. Network for Earthquake Engineering Simulation (NEES) was funded by the National Science Foundation and authorized under the National Earthquake Hazards Reduction Program. Named after the late Congressman who was a strong advocate of federal support of earthquake engineering research, NEES is made up of 15 university equipment sites that share facilities and data and enable collaboration over the ultra high-speed, high-performance Internet2 network.

The network's mission is "to enable collaboration and transformative research to reduce seismic risk by providing world-class community infrastructure."

Steve McCabe, P.E., CEO of the NEES Consortium, which manages the network, explains that it wasn't economically feasible to put large-scale testing facilities at many different university labs. So through a competitive proposal process, NSF identified the best sites to host equipment and then provided open access to researchers. "[The facility] becomes a community asset," he says.

Some of the facilities were preexisting at the universities but were expanded with NSF funds; others were built specifically for NEES. The equipment falls into several main categories: shake tables (or earthquake simulators), geotechnical centrifuges, a tsunami wave basin, and facilities for large-scale laboratory experimentation and field experimentation and modeling.

The network enables experimentation and simulation of how buildings, bridges, utility systems, coastal regions, and geomaterials perform during seismic events. The equipment at a particular site is open to faculty and students from any institution, and NSF accepts proposals for research projects, which are funded annually.

Those who provide their own funding, such as industry representatives, can also use the equipment for testing. In addition, NSF encourages research projects to include participation from practitioners in the field and other stakeholders.

Whereas in the past, researchers were limited to using what existed at their own institutions, NEES "opens up the possibility for people who are very smart and have good ideas to use state-of-the-art facilities," says Andre Filiatrault, P.E., professor in the department of civil, structural, and environmental engineering at the University at Buffalo and director of the Structural Engineering and Earthquake Simulation Laboratory, which is one of the 15 NEES sites. "It's a revolutionary concept."

Also groundbreaking is the information technology infrastructure that links the sites. It enables several key components of NEES. One is the centralized repositories that archive data. Such data storage is required for NSF-funded NEES projects, and it enables researchers around the world to access and use the information.

The IT infrastructure also enables virtual meetings, test observation from remote locations via cameras, and performing experiments via the network.

Tests are driven by computers and actuators, explains Roberto Leon, P.E., president of the NEES Board of Directors and professor at the School of Civil and Environmental Engineering at the Georgia Institute of Technology. "You can sit at a computer terminal and enter the load history that you want applied to a test specimen," he says. "You could be sitting 2,000 miles away and do exactly the same thing."

The technology allows researchers to link one site to another, explains McCabe, in order to remotely connect and control experiments at one or more locations from a third location. That enables much larger scale experiments. "It's a large-scale virtual laboratory," he says. "It represents the future as far as large-scale testing goes."

And NEES has changed the research culture as well. Says McCabe, "We're going from an isolated group of maybe one or two research teams [working on a project] into a larger scale operation that involves multiple faculty members, multiple schools, and also the practitioning community, the federal or state agencies, just a whole one or two orders of magnitude more participation from different angles."

At the equipment site hosted by the University at Buffalo, there are two large shake tables. Their hydraulic actuators are linked to an electronic control system that enables reproduction of earthquake motion, explains Filiatrault. Once an earthquake occurs anywhere in the world, as long as there are instruments on the ground to measure the ground's movement, equipment can reproduce that earthquake in the university's lab.

Each table can accommodate 50 tons of payload, or 100 tons if they are linked and used in unison. When researchers want to test a structure that is too large for the shake tables, the university and other NEES sites do hybrid testing. That enables part of the structure to be modeled in the computer and part to be tested in the lab. They can then feed data on how the structure behaved into the computer model.

A NEES project at the University at Buffalo that received a lot of media attention was the testing, in 2006, of a full-scale residential light frame wood building. At 1,800 square feet, it was complete with furniture, partitions, and a car in the garage. The aims were twofold: create a database to verify computer models of similar type buildings and examine the effect of wall finishes on the structure's performance.

Although the testing was done at the University at Buffalo, the host institution for the project was Colorado State University. Texas A&M University, Rensselaer Polytechnic Institute, and the University of Delaware were also involved.

Manufacturers of building materials donated more than $50,000 in supplies in order to determine how their products would perform in an earthquake. A dozen video cameras recorded the shaking of the building on the tables, and 250 sensors inside the house gathered information about the behavior of building components.

The test has already begun to generate data on how to make wood-frame houses and buildings safer for occupants during earthquakes. In 2009, the new seismic design processes will be validated with a six-story wood-frame building shipped to Miki City, Japan, and tested on the world's largest shake table.

NEES is helping to shorten the time between research results and resulting changes in design standards and building codes, explains McCabe. Many of the new versions of building codes affected by the research won't come out until 2010 or 2011, but Leon says that many NEES researchers are involved in code writing.

An important project that the University at Buffalo has coming up is one of the three "grand challenges" NSF is currently funding within NEES's research portfolio. It will be led by the University of Nevada, Reno, and will look at the behavior of a building's nonstructural components, such as the ceiling, piping, and partitions, during an earthquake. That project will help determine failure criteria and limit states as well as generate a testing database to compare models against, Filiatrault explains.

In October 2007, University at Buffalo dedicated its nonstructural component simulator, a two-story shake table that simulates motion not on the ground, but on the floors above. Few shake tables in the U.S. can do that, Filiatrault says.

Another grand challenge project, starting in July, will partner University at Buffalo with Georgia Tech to examine the behavior of crane structures in ports during earthquakes. The project is inspired by the Kobe, Japan, earthquake in 1995, which caused huge damage to the city's port and economy. This year researchers will perform a small-scale test to examine how structures behave; in 2009 they will test a crane at close to full-scale on the two main shake tables.

That grand challenge will bring together not only researchers and practitioners in container port design and operation, but also economists and social scientists to examine the larger issues. According to Leon, NEES can make a real impact "by looking at the seismic problem from a system level, making this more understandable to the public and public officials." That will hopefully lead to more mitigation and planning, he says.

Across the country, at Oregon State University, the experiments are quite different. NEES's Tsunami Research Facility examines the phenomenon that's often sparked by underwater earthquakes.

Two large wave basins at the university contain actuators that can create a wide variety of wave conditions. The three-dimensional wave basin is 160 feet long and can be programmed to create random waves that move independently of one another. The two-dimensional wave flume, 342 feet long, enables the isolation of certain behaviors and their study in larger scale.

Oregon State's wave basins were already in operation when NEES started, but the NSF funding was used to quadruple the volume of the three-dimensional basin, explains Solomon Yim, professor in the School of Civil and Construction Engineering and principal investigator at the facility.
The university recently received another $1.2 million NSF grant to build a new wave maker that will enable the creation of waves with greater velocity. That will be in operation by January 2009.

Since NEES began its work in 2004, other institutions have used the basins to study wave impacts on residential housing and coastal cities, among other projects. The facilities can also be used to study storm surges. According to Yim, Oregon State's facilities have contributed to research on the Indian Ocean tsunami in 2004 and Hurricane Katrina in 2005. For example, Oregon State researchers are conducting tests on the effects of the storm surge and storm wave during Hurricane Katrina on coastal structures such as highway bridges.

The Oregon State Wave Research Facility is the only one of its kind in the country and, according to Yim, "a national treasure." The facility receives 3,000–5,000 visitors per year, including high school and junior high school classes in and around Oregon.

Education, outreach, and training are other key factors of the NEES model. According to Joy Pauschke, P.E., program director for NEES at the National Science Foundation, NSF encourages the integration of research and education. Thus, NEES research projects supported by NSF each have an educational component, ranging from developing graduate or undergraduate course modules to K–12 activities.

In addition, part of the operations award to the NEES Consortium pays for annual training workshops at NEES facilities to help potential users become familiar with the equipment and their capabilities. The organization also sponsors a variety of Web seminars with practitioners and university researchers in which participants can see data streams and video from actual tests.
Also offered by the NEES Consortium is the Research Experiences for Undergraduates summer program that involves 15–20 undergraduate students each summer in a 10-week internship, in which they receive a faculty advisor and join an NEES research team.

But undergraduate and graduate students work throughout the year on projects as well. NEES is a good opportunity for students to get involved in cutting-edge research, points out Filiatrault.

In 2006, President Bush announced the American Competitiveness Initiative in his State of the Union address. The initiative aims to maintain U.S. leadership in science and technology, and one of its goals is to promote advances in materials science and engineering that will lead to better structural performance during earthquakes, hurricanes, and other hazardous events.

Pauschke points out that NEES research does just that. Filiatrault puts the importance of NEES another way: "It's completely changed the field of earthquake engineering in the U.S."