PEER Research Project Highlight: "Correlation of Ground Motion Duration with Spectral Acceleration and Implications for Expected Bridge Performance"

October 25, 2021

The impact of a PEER funded research project "Correlation of Ground Motion Duration with Spectral Acceleration and Implications for Expected Bridge Performance" is highlighted below. The project Principal Investigator (PI) is Tracy Becker, Assistant Professor, UC Berkeley. The Research Team includes Jennie Watson-Lamprey, CEO and Principal Engineering Seismologist, Slate Geotechnical Consultants and Camila Lopez Ruiz, Graduate Student Researcher, UC Berkeley.

Download the Research Project Highlight which includes the abstract (PDF)

Research Impact

Accurate prediction of seismic performance of highway bridges is paramount for resiliency and the ability to plan for post-earthquake recovery, and these predictions are highly dependent on their input. Ground motion prediction equations (GMPEs) and duration prediction equations describe the distribution of spectral acceleration and duration given a set of predictor variables such as magnitude and distance, but do not address the correlation between spectral acceleration and duration. From a preliminary study, if the correlation is not accounted for, the durations can be overpredicted on the same order as overpredicting the magnitude by one-half a unit. Numerous studies have shown that the duration of the selected ground motions can impact structural response and should be included in selection procedures, yet there are no clear guidelines on how this should be done. When duration is deemed to be important for the structural response, the state of the practice is to check the duration of the selected records against existing duration prediction models, but these models do not take into account the correlation with spectral acceleration, which may lead to a bias.

The research will allow engineers to select more appropriate records for use in analysis and design of structures and will demonstrate the impact of appropriately selecting for duration. Thus, the research will complement previous and ongoing research in bridge and transportation systems. Furthermore, the question of appropriate durations is especially relevant to high importance structures requiring time series analysis that are expected to experience nonlinear response in large events; this includes transportation infrastructure, tall buildings, geostructures, etc. Thus, the research can have broad impact and is of interest to structural engineering firms as well as geotechnical engineering firms.

Research image

Left: Residuals of spectral acceleration at 1 second and D5-95 duration with correlation estimates for ground motion recordings from the NGA-West2 database for Mw>6, Repi<100 km, and Vs30<300 m/s. Right: Model of bridge with input ground motions of different duration.