Search

Damage tolerant structures subjected to fatigue loading require inspection at regular intervals for detection and reporting of possible damages such as fatigue cracks. SHM systems integrated into those structures consisting of a sensor network combined with data analysis algorithms can further enhance the effectiveness of monitoring either by automating the inspection process or by adaptively changing the inspection intervals. To implement such a SHM system, a network of ultrasonic guided wave actuators and sensors including the electronics will have to be deployed to or integrated into the structure. Design of such a system should consider the probable location of damages that have to be monitored and the ways to enhance the sensitivity of detecting those damages. When a damage approaches its tolerable size limit, it should be detected with a higher accuracy compared to when the damage is first detected at an earlier stage and smaller size. Several probabilistic factors can alter the nature of damage and hence can affect the damage detection sensitivity as a consequence of factors like the specific sensor network algorithms applied, the type of damage and the zone of inspection. This is one of the major challenges to achieve reliable monitoring of some evolving wide spread fatigue damage using an integrated SHM system. In this paper, a specific example of fatigue crack initiation from the rivet holes in a stiffened structural component has been simulated considering a probabilistic sequence of fatigue loading. A scheme of determining the probability of damage growth up to a specific size is studied considering the progressive changes in the load path and randomness in the loading sequence. Our present effort is to develop a scheme of modeling and analysis that uses information regarding variability in the loading and relates those to the probability of a critical damage. Detailed understanding of these relationships is considered useful in arriving at design specifications for an integrated SHM system for improving the system’s reliability