Issue34

K.L. Yuan et alii, Frattura ed Integrità Strutturale, 34 (2015) 476-486; DOI: 10.3221/IGF-ESIS.34.53 482 when the resistance necessary for plastic deformation is decreased due to acoustic softening, and the parameter, η=40%, is used in the following UIT simulation. This result is comparable to the measured yield strength reduction of 45% in stainless steel specimen under direct ultrasonic energy 600 Watt [14], while the consumed power of 27 kHz-type UIT equipment is 600-1200 Watt [1]. The internal residual stress profiles during 30 times of pin’s impinging (one ultrasonic impact without translation along weld line) are shown in Fig.10. Near the plate surface the compressive residual stress layer appears. The surface compressive residual stress gradually increases with the peening, but this increase is saturated after 30 times of continuous peening. Having introduced the yield strength reduction due to acoustic softening, the compressive residual stress layer is deepened approximately by 2mm, while the maximum value decreases in comparison with that without acoustic softening (see Fig.10b). a) b) c) Figure 9 : Determination of acoustic softening by comparing the treated weld toe shape; (a) simulated with η=0%, (b) simulated with η=40%, (c) measured by Suzuki et al. [20]. a) b) Figure 10 : Evolution of in-depth transverse residual stress distributions during one ultrasonic impact; (a) simulated without acoustic softening, (b) simulated with η=40%. Residual stress distribution in the depth direction of As-weld and UIT joints In practice, the peening tool is travelling along the weld line, until the original shape of the weld toe disappears and the smooth groove is formed. The transverse residual stress σ x acting perpendicular to the welding line is considered as the driving or resistance force for the surface crack at the weld toe. Therefore, considering the movement of the pin, the transverse residual stress distribution and weld toe shape before and after UIT are illustrated in Fig.11. It has been confirmed that the tensile welding residual stress at the weld toe has been changed to significant compressive residual stress. The step-by-step movement of 0.4 mm provide sufficient overlapping of the impacts. In order to verify the accuracy of the finite element simulation, the residual stress distributions in the depth direction are compared with those measured by X-ray and neutron diffraction methods, respectively [20]. As show in Fig.12, there exist good agreements between the experimental and numerical results. Slight over-estimation of UIT compressive residual stress on the surface