Towards an Automated Crack Monitoring using Distributed Fiber Optic Sensors

Abstract

The high spatial resolution of distributed fiber optic sensors enables quasi‐continuous strain measurements, which makes it a promising technology for structural health monitoring. From this quasi‐continuous strain data, location and width of cracks can be estimated. For major benefit to infrastructure maintenance, this crack width calculation is need to be automated. A framework for crack width estimation is presented, streamlining the different approaches for concrete and steel sensors, mainly differing in the compensation of tension stiffening. Relevant parameters for the analysis are identified and their influence is studied. Two sensors – a concrete embedded one and a reinforcement attached one – are used to estimate crack width using the presented framework. The results are compared to digital image correlation. Accurate estimates are achieved. Tension stiffening showed to have a negligible small influence on the estimated crack widths. The concrete embedded sensors have a higher sensitivity and are recommended for crack width monitoring.

Crack Monitoring on Concrete Structures using Robust Distributed Fiber Optic Sensors

“…Further background information on crack width calculation using steel and concrete embedded DFOS and a sensitivity analysis of the different influencing parameters can be found in [35]. The authors provide an open source Python framework that enables a semi-automated calculation of crack widths.…”

“…For automated crack detection, a peak finding algorithm is used. To define whether a peak is a crack, the term prominence is introduced, which describes how far a peak stands out from its surrounding [35]. Due to the low noise level, the prominence for automated crack detection could be set to 100 µm/m.…”