SciTech

CEE professor awarded $1.6M

Civil and environmental engineering professor Jacobo Bielak was awarded $1.6 million by the National Science Foundation (NSF) PetaApps program, to put in effect his extensive research on ground motion that occurs during earthquakes.

Bielak was honored with this generous grant in light of his efforts to develop earthquake simulations that will help scientists and engineers understand the impact of earthquakes in urban areas. The project is a collaborative effort between the departments of civil and environmental engineering and computer science.

The unique combination of the two disciplines involves using large-scale computers to formulate equations that illustrate the ground motion generated by an earthquake.

“These simulations will provide unprecedented detailed knowledge of how an urban system performs in a large earthquake and what is needed for improving disaster planning and preparation,” Bielak stated in a Carnegie Mellon press release.

Bielak is working in cooperation with David R. O’Hallaron, associate professor of computer science and electrical and computer engineering, and the Southern California Earthquake Center (SCEC) to form these earthquake simulations. The team of scientists and researchers also includes doctoral students Leonardo Ramirez-Guzman and Ricardo Taborda.

The earthquake simulation is a model of the abrupt exertion of energy and displacement of the fault. The precipitance of this action leads to the movement of seismic waves through the Earth.

“The simulation of this phenomenon as a whole in a computer, sequentially, or in a supercomputer in parallel, is the basis of our research,” Taborda stated in an e-mail. Simulating data sequentially refers to organizing and processing data in a consecutive order, while simulating it in a supercomputer in parallel implies that data is being modeled side by side with other data.

After winning the Analytics Challenge Award last year at SC06 — the annual international conference on high-performance computing, networking, storage, and analysis — the team moved on to merge its research plans with those of the other teams that are part of the proposal, such as University of California Berkeley, UC San Diego, and UC Davis.

The NSF grant will provide the team with resources to generate 3-D models of the ground movement that can simulate the impact of possible earthquakes on the surrounding areas, Bielak stated in the Carnegie Mellon press release.

This new earthquake research leaps beyond the potential of current software and hardware facilities.

“The large-scale simulation of earthquakes will be created using unique software that can be used in very large computers. So we are trying to use it in the largest computer available. It has thousands of processors and displays a finer resolution,” said Ramirez-Guzman.

Whereas in the past Bielak and his team had solely focused on simulating the earthquake and the apparent motion of the seismic waves through a large section of the Earth, they now intend to take into consideration buildings, bridges, and other infrastructural aspects of the affected region.

“The objective is to generate an equation that describes the phenomena, and [to] predict the areas that will experience larger displacement in the ground,” said Ramirez-Guzman. “We want to know how the ground shakes, and applying the specialized software to the model will help us develop our research further.”

Ramirez-Guzman also said that based on the history of past calamities in a particular geographic region, the energy likely to be released can be assessed without detailed information about the fault line of earthquakes.

According to Taborda, this can be accomplished by an “in-house developed method” that will divide the process into two distinct phases. The first phase involves creating a large simulation that includes the geological fault that drives the earthquake. In the second phase, this data is used to view structures such as buildings, roads, and pipelines in smaller sub-regions in greater detail.

The importance of incorporating structures and the non-linearity of the soil is that it changes the original ground motion that would be obtained without their presence.

“This would eventually help us to a more real assessment of what the potential impact of a moderate or large earthquake could have in an earthquake-prone region such as southern California and the greater Los Angeles (basin) area,” Taborda stated in an e-mail.

Such a realistic estimation will enable civil engineers and seismologists to make more informed decisions about necessary precautions prior to natural disasters such as earthquakes.

According to the Carnegie Mellon press release, the researchers will also join forces with the Pittsburgh Supercomputing Center to apply some of the specific “algorithms and simulation structures” to get an exact idea of the measures needed to be taken to insure public safety throughout an earthquake.

By planning ahead, Bielak and his team of researchers and scientists may emerge successful in reducing much of the destruction that occurs during a natural catastrophe.