Issue34

K.L. Yuan et alii, Frattura ed Integrità Strutturale, 34 (2015) 476-486; DOI: 10.3221/IGF-ESIS.34.53 477 parameter. Ultrasonic impact treatment consists of ultrasonic waves and mechanical impacts. Very limited studies [12-14] have demonstrated that the ultrasonic vibration superimposed by UIT on metals has indeed an acoustic softening effect on material properties undergoing deformation. Normally the experimental methods for residual stress measurements need considerable cost, time and skillful technique. It is practically impossible to obtain full field measurement of residual stress. For this reason, the residual stress induced by UIT has been recently evaluated by some studies using finite element (FE) analysis. There are mainly two groups of simulation methods for UIT process: quasi-static implicit and dynamic explicit methods. The former is modelled as pressing the pin at the weld toe to one prescribed depth by displacement or load control approaches, rather than peening [15-16]. In the latter [17-18], the modelled pin impacts to a symmetry-cell model that is widely used in shot peening simulation [19], however, the welding residual stress is not taken into account. It is found that the predicted in-depth residual stress and groove depth could not consist well with measurements together [15-16], because in the above mentioned simulations the acoustic softening effect has not been introduced, which plays an important role in the mechanism of UIT. Therefore in this study, one novel 3D prediction approach including thermal-mechanical welding simulation, dynamic elastic-plastic FE analysis of UIT-process, and an evaluation of surface fatigue crack growth for cruciform weld joint has been proposed. The actual process parameters and ultrasonic induced material softening, which is appropriately adjusted to fit experimental results, are considered. The predicted residual stresses distributions, treated weld toe shape and fatigue strength are compared with published test data [3, 20]. T HE UIT PROCESS s depicted in Fig.1, the ultrasonic impact treatment method is based on conversion of harmonic oscillations of the ultrasonic transducer into impact pulses at the treated surface. In order to efficiently transfer the energy into the work piece, the installed cylindrical pins can freely move in a gap between the waveguide end and treated surface. This kind of high frequency impacts of the pin in the combination with the ultrasonic stress waves transmitted into the work piece through the pin contacting the treated surface is called as ultrasonic impact [12]. Figure 1 : Mechanism of UIT [12]. Model of ultrasonic impact In order to consider ultrasonic impact phenomenon in numerical simulation, one simplified model based on the description by Statnikov [12, 21] as illustrated in Fig.2 is proposed. During one impact period T im , it comprises both the ultrasonic impact period t 1 and pause between impacts t 2 . In this work, the t 1 is assumed as 1 millisecond, in the case of impact frequency f im of 100Hz and the relation t 1 / T im =0.1, which is in the range of ultrasonic impact time from hundreds of microseconds to units of milliseconds measured by Statnikov [12]. The oscillations of pin at a frequency of 27 kHz, which mainly cause the plastic deformation, will recur about 30 times during reboundless contact phase up to 1 millisecond. To construct UIT FE model, the impact velocity of pin during reboundless period is essential. For the impact velocity, it can be assumed here that all the ultrasonic impact on the treated surface will occur at the same velocity and an average A