pacific nw transportation consortium region 10 Pactrans

Pilot Study: Learning Fluid-Structure Interaction via Machine Learning

PI: Barbara Simpson (OSU), barbara.simpson@oregonstate.edu, ORCID: 0000-0002-3661-9548

Co PIs: Michael Scott (OSU)

AMOUNT & MATCH: $40,000 from PacTrans; $40,000 Match

PERFORMANCE PERIOD: 8/16/2019 – 8/15/2021

STATUS: Completed

CATEGORIES: Natural Disasters, Bridges

RESEARCH PROJECT HOT SHEET:

UTC PROJECT DOCUMENTATION:

FINAL PROJECT REPORT:

PROJECT DATA:

DESCRIPTION: This proposal addresses the content area, improving mobility of people and goods, particularly ensuring reliable mobility across bridges after tsunami loading. This work also aligns with ongoing interest in tsunami loading on bridges and machine learning applications by the Oregon Department of Transportation and the Pacific Earthquake Engineering Research (PEER) Center.

Although implemented herein for the analysis of bridges, the resulting machine learning framework would be applicable to other computationally-expensive simulations and a larger set of data-driven transportation problems, such as evacuation models, active traffic control, analyzing sensor data, etc.

Implementing faster models that maintain the efficacy of the original data would result in prompt feedback for analysis and design, increased feasibility for parametric applications, and better fragility functions based on CFD/FSI rather than equivalent static analysis.