[4A4] Using ultrasonic waves to characterise fatigue state in steel components
G Sarris¹, S Haslinger², P Huthwaite¹ and M Lowe¹
¹Imperial College London, UK
²University of Liverpool, UK
The cyclic pressure and temperature of water and steam that travels through the pipes in steel components in power generating plants causes fatigue damage. The state of this fatigue is difficult to assess due to the rapid changes in amplitude and period of the loading, which is determined by the power output of the plant at a given time. Therefore, this poses the need to develop a method capable of accurately characterising the fatigue state of such components, to reduce the inspection and maintenance costs and to ensure their safe operation. Ultrasound is a potential solution to this issue as it has already been shown that the presence of a fatigue zone will cause an ultrasonic wave to travel more slowly in its vicinity. We have assessed the performance of three types of wave in terms of their sensitivity to the presence of fatigue: compression, shear and Rayleigh waves. We produced compression wavespeed C-scans on fatigued components, which were able to reveal the presence of the fatigue zone; however, the changes in propagation speed were less than 0.5% compared with the healthy region propagation. Using electromagnetic acoustic transducers (EMATs), we were able to produce analogous shear wavespeed C-scans, where the observed changes in speed more than doubled. Finally, the use of Rayleigh waves, which propagate on the surface of a material where fatigue is concentrated, was able to amplify the changes in speed by 10× compared with through-thickness compression waves. The implementation of each of these techniques was subsequently further studied for both flat and cylindrical geometries. Additionally, using a stiffness reduction method and the finite element software package Pogo, we were able to derive a method to accurately predict the time-of-flight of compression and Rayleigh waves in fatigued steel components. The results agreed well with corresponding experimental measurements.
