| dc.description.abstract | Distortion-induced fatigue cracking is a primary maintenance and structural safety concern in steel bridges built prior to the 1980s in the United States. Manual, hands-on inspections are currently the primary method departments of transportation and other bridge owners use to identify and quantify fatigue cracks. To improve the efficacy of fatigue crack inspections, previous research has proposed and examined numerous fatigue crack detection approaches, including both user-implemented technology and structural health monitoring methods. However, these approaches typically require human presence and active participation at the location of interest or prolonged mechanical contact and continuous monitoring of the structure. This limits the effectiveness and flexibility of these approaches for inspecting a large number of fatigue-susceptible regions found on steel bridges. Recently, vision-based sensing technologies have been explored for applications related to damage detection and health assessment in civil infrastructure, as they offer the benefits of being low cost, noncontact, and deployable without human presence at the specific region of interest. This paper presents a digital image correlation-based methodology developed from in-plane compact fracture specimens for the detection and quantification of fatigue cracks. The effectiveness of the proposed methodology is further evaluated through experimental tests of a fatigue crack on a large-scale steel girder-to-cross-frame connection, similar to the out-of-plane fatigue cracks commonly found on steel highway bridges. Results indicate that the digital image correlation methodology can adequately characterize fatigue cracks, both in-plane and out-of-plane, in terms of crack length. This quantification from a noncontact inspection technology has the potential to lead to future automation of steel highway bridge fatigue inspections. | |
