Fatigue crack propagation monitoring method for single-sided notched aluminum alloy test piece based on FBG sensor

Abstract

The invention discloses a fatigue crack propagation monitoring method for a single-sided notched aluminum alloy test piece based on a FBG sensor. The method adopts a distributed fiber Bragg grating sensor network to sense fatigue crack propagation in the structure, extracts a central wavelength offset of the FBG sensor as a characteristic parameter, establishes a fatigue crack length prediction relationship model by using the variation of the central wavelength offset of the FBG sensor with different positions as the crack propagation, and realizes the fatigue crack propagation length prediction. The fatigue crack propagation monitoring method in the invention realizes the prediction of the crack propagation position by calculating the central wavelength offset of the FBG sensor, and has the characteristics such as fast speed, perceptual intuition and no need for a large amount of prior knowledge.

Description

technical field [0001] The invention relates to the technical field of fatigue damage monitoring for structural health monitoring, in particular to a method for monitoring fatigue crack growth of a single-side notched aluminum alloy specimen based on an FBG sensor. Background technique [0002] At present, aluminum alloys account for the vast majority of structural materials used in aerospace vehicles, and in complex mechanical structures such as aluminum alloy spacecraft and aircraft, the key structures and components of the machinery are subject to external heat or force for a long time. The action of various alternating fatigue loads can easily cause structural wear and failure, resulting in fatigue damage and fracture. Fatigue crack is the initial state form of fatigue damage fracture. Although aerospace vehicles can continue to be used when there are tiny cracks, their performance will be reduced. If they are not monitored and maintained for a long time, it will eventu...

AI Technical Summary

Problems solved by technology

[0004] At present, in the monitoring and prediction of fatigue crack damage, neural networks and support vector machines are mostly used to train the network with a large number of sample data. Although the accuracy is high, because it is based on large samples, the generalization performance is not good, real-time performance and poor usability

Images