Buildings, tunnels, utility pipelines, and other structures constructed of steel are critical components of the nation's civil infrastructure. As experienced during previous disasters, such as in the 1994 Northridge, California earthquake, steel structures are vulnerable to collapse or failure, resulting in loss of life and property. Cracking (i.e., fracture) in critical parts of these structures is often the reason for their failure. Consequently, designing structures to minimize the risk of fracture is critical for infrastructure safety and operability. However, methods to predict the growth of cracks are not well-developed, especially when fractures occur during severe shaking caused by earthquakes or other extreme loads. By combining expertise from structural engineering, materials science, and computational mechanics, the objectives of this research are to: (1) create new models and computational technologies to accurately simulate this type of cracking, and (2) integrate these models into the design process for civil infrastructure. Undergraduate and graduate students will actively participate in this research. The new knowledge, models, and software products from this research, through transfer to practitioners, will improve the safety and economy of buildings and other civil infrastructure, benefiting the U.S. society and economy.
Nonlinear analysis is an essential technology for the modern performance-based design and construction of buildings and civil infrastructure that are more resilient to earthquakes and other extreme hazards. This research addresses a significant limitation of nonlinear analysis as it pertains to simulating crack propagation in steel structures, under two important situations: (1) ultra-low cycle fatigue (ULCF) loading, which is characterized by few (<20) cycles of large strain amplitude that can occur under earthquakes and other hazards, and (2) low-stress triaxiality, which often occurs with ULCF loading in shear bands or at protruding corners of structural elements. This research will address these challenges through fundamental theoretical model development (a new damage mechanics based constitutive model for ULCF and low-triaxiality), computational methods synthesis, numerical implementation, laboratory testing, and calibration and validation of model implementations. The work will combine models for ductile crack initiation with computational techniques, resulting in validated approaches for simulating crack propagation under ULCF and low-triaxiality. The project will culminate in a campaign to facilitate adoption of the fracture simulation techniques into engineering research and practice, including standardization of simulation and calibration methods required for their realization, and development of open source software. These research products and their adoption into engineering practice will enhance safety and performance of the built environment.
