PFI-TT: An artificial intelligence system for prediction of wind hazards in civil engineering applications

Award ID: 2140723

The broader impact/commercial potential of this Partnerships for Innovation - Technology Translation (PFI-TT) project is the creation of technology for leveraging performance-based design to mitigate the massive economic loss and societal disruption caused by damage to buildings and other structures from extreme windstorms, e.g. hurricanes. Performance-based design represents a paradigm shift that is recognized by the engineering community as the future of the profession. To accelerate this fundamental shift for buildings subject to extreme windstorms, this project will create a first-of-its-kind technology for the implementation of performance-based wind design by leveraging recent breakthroughs in computational modeling of inelastic dynamic systems and uncertainty propagation, therefore resulting in tools that explicitly ensure (1) enhanced resilience of buildings and other structures to future windstorms and (2) the stakeholder need for cost effectiveness. This will create transformative advances on the profession of civil and structural engineering through arming the designers of buildings and other structures subject to extreme winds with a means to implement advanced performance-based wind design in a new process. The project will also train a postdoctoral researcher and undergraduate students in innovation and entrepreneurship in civil engineering technology. This project will create prototype technology for carrying out advanced performance-based wind design through the explicit evaluation of building performance at all hazard intensities, including at ultimate load levels where significant damage is expected, while systematically treating uncertainty through reliability. The technology will provide, for the first time, a tangible means for design engineers to evaluate the system-level inelastic performance of buildings and other structures subject to extreme winds through the limit states of dynamic shakedown and/or incipient collapse. This will enable the assessment of building systems against failure due to ratcheting, low cycle fatigue, and/or instantaneous plastic collapse, the evaluation of which is recognized as key to achieving greater resilience against extreme windstorms at reduced costs to stakeholders. By advancing stress resultant plasticity and hybrid computing, uncertainty propagation schemes will enable direct evaluation of reliability at the performance objectives of dynamic shakedown and incipient collapse, metrics at the very core of advanced performance-based wind design. To enable user interaction with the technology, a comprehensive object-oriented interactive graphics interface will be developed that possesses extensible and maintainable characteristics, therefore ensuring the results of this project will endure well beyond its conclusion. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Program Director: Mara E. Schindelholz