Evaluating Prestressed Concrete Beams with Cracks using Machine Learning

Bridge owners face difficult decisions on whether a bridge should be posted, repaired or replaced when prestressed concrete members have shear related cracks due to overloading. The decisions are currently made based on engineering judgment, costly load-testing or time consuming modeling. Guidance is needed to interpret cracks and their impact on shear capacity to avoid overly conservative load ratings and to keep bridges operational, without compromising safety and economy. This project will develop a tool through machine learning to relate cracking to load history of bridge members. Algorithms will be trained using shear test data from the literature, considering material and geometric properties in addition to crack width as an indicator for distress. The outcome will be the advancement of knowledge on shear evaluation and load rating of in-service precast prestressed concrete bridges with visual signs of distress and guidance for repair actions for bridge owners.


    • English


    • Status: Active
    • Sponsor Organizations:

      University of Illinois, Urbana-Champaign

      Department of Civil and Environmental Engineering
      Newmark Civil Engineering Laboratory
      Urbana, IL  United States  61801-2352

      Office of the Assistant Secretary for Research and Technology

      University Transportation Centers Program
      Department of Transportation
      Washington, DC  United States  20590
    • Performing Organizations:

      University at Buffalo - SUNY

      221 Ketter Hall
      Buffalo, NY  United States  14260
    • Principal Investigators:

      Okumus, Pinar

      Khorasani, Negar Elhami

    • Start Date: 20230901
    • Expected Completion Date: 20240831
    • Actual Completion Date: 0
    • USDOT Program: University Transportation Centers

    Subject/Index Terms

    Filing Info

    • Accession Number: 01893424
    • Record Type: Research project
    • Source Agency: Transportation Infrastructure Precast Innovation Center
    • Files: UTC, RIP
    • Created Date: Sep 18 2023 10:04PM