• Magnetostrictive sensor
• Non Destructive Evaluation
• Ultrasonic Guided Waves

Nondestructive Evaluation – NDE, is an important aspect of the integrity management of industrial plants, where pipe systems are the dominant component. During the last decade Ultrasonic Guided Waves (UGW), have started to be used as a useful instrument for on-stream long range inspection of pipes. Various procedures and systems have been proposed for the generation and detection of UGW. Presently, they are based on piezoelectric (PZT) or magnetostrictive (MT) transducers or electromagnetic acoustic transducers (EMAT).
It is generally known that PZT based systems have elevated diagnostic capacities due to their high transduction efficiency. However, the elevated costs of installation of such devices make their use for long-term monitoring of piping systems quite improbable.
On the other hand, the MT based systems have the advantage of the reduced costs of the composing materials, simplicity of attaching it to the pipe wall and flexibility regarding the diameters of the pipes that can be inspected. Still, its single-element configuration limits the capacity to characterize the detected discontinuities in terms of geometry, thus being unable to distinguish between possible flaws from symmetrical features, normally located on pipes, like welds or flanges. Furthermore, its reduced capability to geometrically characterize flaws makes the classification of their severity particularly difficult.
The improvement of the diagnostic capacity of MT based systems in order to make practically possible and economically convenient its use in monitoring applications is the main purpose of this thesis.
In this dissertation multiple laboratory and field experiments are described and the magnetostrictive technology is evaluated. Furthermore, a new magnetostrictive transducer for UGW acquisition is presented. It allows step-by-step data acquisition around the pipe circumference revealing important information on the geometry and circumferential position of flaws.
The new sensor was validated by computer simulations as well as further laboratory and field tests. The resulting data was used as input for various digital signal processing techniques to describe geometrically the features detected in the acquired signal.
The final results outline the potential of MT based long-range inspection to reach also a good sensitivity and a good defect sizing and classification with respect to conventional techniques, making it an important candidate for monitoring activities for the integrity management of industrial plants.