GPR-Based Fuzzy Model for Bridge Deck Corrosiveness Index

Ground-penetrating radar (GPR) is a rapid technology for evaluating condition of concrete bridge decks subject to rebar corrosion. In this paper, based on a threshold model recently proposed in the literature, a bridge deck corrosiveness index (BDCI), is developed to have an idea where a bridge deck is during its continuous service life and to suggest corresponding maintenance activity. Based on fuzzy set theory, expert opinions were used to calibrate fuzzy membership function for each condition category found by GPR. Then, for a particular bridge deck, area percentages of all condition categories would be utilized to aggregate these functions into a BDCI using weighted fuzzy union (WFU) operation. The benefit of the developed system is twofold. First, it is based on GPR, a more accurate inspection technology. Second, it employs the knowledge provided by bridge community, and, in the meantime, has the capability to deal with fuzzy information associated with expert responses. Using an automated software, the system is illustrated for several concrete bridge decks in North America. Because the case studies show that the developed system is easy to be implemented, it would be an effective tool for transportation agencies in North America where the corrosion of rebar in concrete bridge decks is one of biggest concerns.