Evaluation of BFRP strengthening and repairing effects on concrete beams using DIC and YOLO-v5 object detection algorithm

Abstract

In this study the effectiveness of using Basalt Fiber Reinforced Polymer (BFRP) for strengthening and repairing concrete beams after pre-damage was investigated. Four-point bending tests were conducted on concrete beams, and their strain nephograms were analyzed using 3D-DIC (Digital Image Correlation) to assess the BFRP reinforcement and repair effects on concrete. A damage indicator model based on stress-displacement curves was proposed to quantify concrete damage and evaluate the BFRP strengthening effect. Additionally, the YOLO-v5 object detection algorithm was employed to detect surface damage on concrete specimens, and its results were cross-verified with the damage indicator. The study reveals that BFRP reinforcement primarily affects the concrete beams beyond the elastic limit, enhancing their load-bearing capacity by 37.91% and improving ductility. The proposed damage indicator proves successful in quantifying damage in BFRP-reinforced concrete beams, distinguishing the transition points between elastic, cracking, and failure stages. The YOLO-v5 algorithm accurately detects damage in strain nephograms, providing sensitive and timely identification of concrete damage. Results proves that the offset ratio between the damage identification results of YOLO-v5 and the experimental stress-displacement curve data is between − 3.03% and 8.66%. The proposed damage indicator and YOLO-v5 detection offer valuable tools for evaluating damage and monitoring the health of concrete structures in engineering applications.