Non-Destructive Testing Campaign on a Steel Bridge through Acoustic Emission Sensors and Accelerometers

An in-situ data acquisition campaign both for local structural diagnosis and vibrational system identification was conducted on a steel bowstring girder bridge with a span length of 65m and a width of 16m. The bridge was simply supported, and the deck consisted of a two-cell box girder with transversal diaphragms connected to lateral cantilever beams. These diaphragms and cantilever beams were spaced 3.6m apart. The deck included a central and two lateral reinforced concrete curbs.  NDTs for local structural diagnosis were carried out using acoustic emission (AE) sensors, while accelerometers were employed to evaluate the vibrational properties of the bridge.

With respect to AE acquisitions, finite element (FE) analysis identified critical stress points on the lower plate of the box girder, particularly around the welds near midspan. Due to logistical issues, testing was conducted as close to midspan as possible, specifically between the 3rd and 4th cantilever beams, along the longitudinal weld. Six AE sensors were used in total, with three placed on each side of the weld, arranged in a 20cm x 20cm grid pattern. The AE acquisition was hit-based, meaning only data exceeding a specific amplitude threshold was recorded. The monitoring lasted for 2 hours and 15 minutes. The raw data collected were automatically processed by the DAQ software, which provided the primary data that were analyzed to perform a local structural diagnosis of the weld. Results showed no significant hits exceeding 90dB for durations longer than 30ms, revealing the detail to be safe.

Regarding acceleration measures, accelerometers were disposed at the four quarters of the bridge span on the central concrete curb, recording in one, two or three directions, according to the image provided. The recordings lasted 2h. The data permitted to identify the frequencies and mode shapes of the bridge.

This study was supported by funding from the RELUIS-CSLLPP consortium as part of the DPC-ReLUIS 2022-2024 project.