Open Access Open Access  Restricted Access Subscription Access

Efficient Guided Wave SHM Baseline Capture and Selection

A. J. CROXFORD, O. PUTIKIS, P. D. WILCOX

Abstract


Guided wave structural health monitoring is generally accomplished using baseline subtraction. This typically takes the form of acquiring a large baseline set over a long period to capture all conditions the structure will experience. In the case of many structures this range is unknown or it is simply impractical to collect a baseline set before use. The requirement for many baselines is complicated by the time that may be required to capture an individual baseline. This paper addresses these problems from two directions. Firstly by altering the physical data capture method the time taken to acquire signals can be reduced by an order of magnitude. Using a coded excitation signal, specifically a chirp, does this. This signal puts in more energy to the structure in line with the length of the chirp. In doing this, the effect on signal to random noise is the same as performing a greater number of averages. Thus the same performance can be achieved with a single firing rather than the use of multiple averages. To address the problem of acquiring a relevant baseline set to correct for environmental changes a method is proposed to add baselines to a data set based on their difference to any previously acquired dataset. This has the effect of adding a recorded value to the dataset if the value has changed more than some specified minimum difference. In doing this the baseline set grows rapidly initially but as the set covers the conditions the structure is exposed to this rate slows. It is shown that this approach can be used to develop a baseline set online without defining it before operation. In addition it is shown that this does not affect the ability to detect a damage event.

Full Text:

PDF