Abstract

The move towards using permanently-installed sensors for structural integrity assessment is driven by their ability to detect changes in real-time. Such information can be of great value when evaluating the structural integrity and remnant useful life of the component. However, evaluating the detection capabilities of a permanently-installed sensor is often challenging. The sensitivity of a sensor to an initiating defect is dependent on its relative position to the defect, and there is often an associated uncertainty in where the damage may occur. Naturally, there is a need to optimally compromise between sensitivity and spatial coverage when choosing the appropriate sensor and sensor parameters for each specific application. In order to evaluate the detection capabilities of a given permanently-installed sensor, its spatial sensitivity would have to be evaluated in conjunction with the spatial uncertainty in damage initiation location. In this research, a finite element approach that incorporates the probabilistic nature of damage initiation is used to map the spatial probability of fatigue damage initiation. A model-assisted probability of detection (PoD) approach is used to map the spatial sensitivity of the sensor. These two maps are then combined to evaluate the overall PoD of the sensor for detecting the initiation of a defect in the monitored component. To illustrate the approach, the detection capabilities of two permanently-installed sensors are compared in two example situations. The two sensors are a guided ultrasonic wave sensor representing a system with lower sensitivity but higher volume coverage, and a bulk wave ultrasonic sensor representing a system with higher sensitivity but lower volume coverage. The two examples evaluated are based on detecting the initiation of a fatigue crack in a rectangular beam under fatigue bending. In the first example, the beam is under three-point bending, representing a case where the area over which damage may initiate is small, whereas in the second example, the beam is under four-point fatigue bending, representing a case where the area over which damage may initiate is large. The outcome is a methodology for assessing the PoD of permanently-installed sensors. The methodology may be used for evaluating the efficacy of a monitoring system or be used for optimizing monitoring system design parameters. It will therefore help the adoption of Structural Health Monitoring.

How to Cite:

Leung, M. S., Corcoran, J. . & Cawley, P. ., (2019) “A structural integrity informed approach to evaluating the probability of detection obtained with permanently-installed sensors”, Review of Progress in Quantitative Nondestructive Evaluation .