Key words: Earthquake ground motions, tall buildings, interface between earthquake science and engineering
The tremendous advancements in computing technologies have enabled realistic nonlinear dynamic analysis of structures that were not possible just a few years ago. The refinement of structural models results in enhanced accuracy and complexity to capture important structural behavior such as structural collapse, cumulative damage and "in-cycle" strength and stiffness degradation. At the same time, denser instrumentations provide more earthquake event recordings that contribute to growing ground motion databases of empirical observations (e.g., PEER NGA database). Improved understanding of the physical process of earthquake rupture and wave propagation, coupled with high performance computing (HPC), also allows the advancement of earthquake simulations (e.g., CyberShake and SCEC Broadband Platform). Seismic performance evaluation can benefit tremendously from HPC-facilitated advancements in structural modeling and seismic hazard characterization, utilizing databases of both recorded and simulated ground motions.
Figure: Schematic illustration of the future of ground motion selection from the combined ground motion database to match the intensity measures of interest for engineering applications.
This study is motivated by a growing interest in tall buildings in seismically active regions, including major cities in the western US, Japan, China and other Pacific rim countries, as well as the availability of simulations for engineering use. It aims to validate earthquake simulations from an engineering perspective through assessment of structural performance of tall buildings subjected to recorded and simulated ground motions. The study will compare the use of simulations to the use of recorded ground motions, and explore the effect of spectral shape and shaking duration on structural response. Insights regarding the use of simulations for nonlinear dynamic analysis of tall buildings will be provided.
Principal Investigator: Ting Lin (Marquette University)
Graduate Researcher: Matt Thomas (Marquette University)
Key words: Probabilistic hazard analysis, climate change, sea-level rise, uncertainties
The release of the Intergovernmental Panel on Climate Change Fourth Assessment Report (IPCC AR4) sent a cautionary message of significant future anthropogenic warming-induced sea-level rise (SLR), as a result of thermal expansion of the oceans as well as ice melt from glaciers and ice sheets. More troubling, satellite observations and tide gauge measurements over the past twenty years indicate an even higher rate of SLR than predicted by IPCC AR4. As SLR knowledge advances, numerous researchers construct and update Sea-Level Rise Prediction Models, which estimate different SLR over different time frames based on various greenhouse gas emission scenarios. Significant uncertainties, both aleatory and epistemic, exist in the SLR predictions. However, engineering and policy decisions regarding affected coastal infrastructure, populations, and ecosystems need to be made in response to SLR prediction.
This project proposes a new analytical framework, termed PROBABILISTIC SEA-LEVEL RISE HAZARD ANALYSIS (PSLRHA), that accounts for aleatory uncertainties from emission scenarios and epistemic uncertainties from prediction models. This probabilistic framework integrates the current knowledge of potential SLR for informed engineering and policy decisions. The output of the PSLRHA framework could be a Global Sea-Level Rise Hazard Map (GSLRHM) based on exceedance probabilities, similar to that of the United States National Seismic Hazard Map. This map can be used for Performance-Based Sea-Level Rise Engineering (PBSLRE) over geographically distributed locations in the long run as well as cost-benefit analysis for various mitigation and adaptation strategies. Ultimately, informed decisions are enabled through the proposed probabilistic framework that incorporates SLR hazard based on multiple scenarios, from global and regional/local projections by multiple modelers, for the time frame and design/policy of interest.
Key words: Performance-based engineering, ground motion selection, response spectra, nonlinear dynamic analysis, Conditional Spectrum
GROUND MOTION SELECTION is the bridge between seismic hazard and structural response, the first two components in Performance-Based Earthquake Engineering. It determines ground motion input for a structure at a specific site for nonlinear dynamic response history analysis. As nonlinear dynamic analysis becomes more common in research and practice, there is an increased need for clear guidance on appropriate ground motion selection methods. Ground motion selection provides a significant basis for conclusions regarding structural safety, since ground motion uncertainty contributes considerably to uncertainty in structural analysis output. In order to select representative ground motions to effectively assess structural performance, it is important to understand which ground motion properties have the greatest effect on structural response. Ting's research evaluates current practice and develops new tools for ground motion selection.
A MORE RIGOROUS ground motion selection methodology will carefully examine the aleatory uncertainties from ground motion parameters (Effect of ground motion selection criteria on structural dynamic analysis in code-based procedures), incorporate the epistemic uncertainties from multiple Ground Motion Prediction Models (Conditional Spectrum computation incorporating multiple causal earthquakes and ground motion prediction models), make adaptive changes to ground motions at various intensity levels (Adaptive Incremental Dynamic Analysis), and potentially use the Conditional Spectrum as the new target spectrum (Effects of target spectrum and conditioning period on structural reliability calculations).
Cover photo (NIST, 2011): Illustration of example Conditional Spectrum for Palo Alto, California, anchored for 2% in 50 year motions at T* = 2.6s.
Figure 2e (Lin et al., 2013a): Conditional standard deviation spectra for Sa(0.2s) with 10% probability of exceedance in 50 years at Bissell, California.
Figure 4e (Lin et al., 2013b): Response spectrum comparison of ground motions selected to match the Conditional Spectra at T* = 0.85s (in solid lines) versus corresponding ground motion hazard curves (in dashed lines).
Figure 1a (Lin et al., 2013c): Statistics of structural responses from intensity-based assessments of a 20-story perimeter frame using the Conditional Spectra with varying conditioning periods at Palo Alto, California.