Testing Of Fatigue Damage Of U71Mn Steel Welded Joints By Nonlinear Ultrasonic Technology

As a common jointing method, welding is extensively applied in manufacturing processes related to aviation, high-speed trains, bridges and automobiles, etc. However, high-quality welded joints may suffer from fatigue damages when bearing c yclic loading, thereby threatening the safe use of welded structures. At present, the fatigue prediction methods commonly used can only predict the fatigue life of welded joints based on complex calculations, and cannot directly characterize f atigue damages within the materials. Compared with conventional fatigue prediction methods, nonlinear ultrasonic detection technique is relatively easy to use and can intuitively characterize the fatigue damages of the materials. However, current r esearch on nonlinear ultrasonic mechanism for welded joints and characterizations of the structural fatigue damages remain at a preliminary stage. In particular, a co mplete fatigue life prediction model to be used for practical detection is lacking. Thus the nonlinear ultrasonic response model in the crack growth stage still requires fu rther research. Therefore, the extraction and relative research of nonlinear characte ristic signals in fatigue-damaged areas of welded joints have already become the focus of current research.In this paper, the wave equations of the ultrasonic waves that propagate in di fferent directions inside the cubic crystal were developed based on the ultrasonic wave theory of solids. The suitable wave mode of nonlinear detection was selected to obtain the nonlinear coefficient used in detection. Meanwhile, the relationship between nonlinear coefficient and material damage was derived in order to provide theoretical basis for the application of nonlinear ultrasonic detection techniques on the characterization of fatigue damage.Base on the perspective of dislocation theory, the relationship between nonli near coefficient and material damage as a theoretical foundation was used in this paper. In this stage, the nonlinear effect that were generated when finite-amplitude ultrasonic waves pass through single and multiple dislocations was used to build up the nonlinear effect model between the dislocation structures and the ultrasound waves. Moreover, in combination with the structure morphologies of the fatigue a reas observed by transmission electron microscopy(TEM), a theoretical explanation for how the nonlinear coefficient of U71 Mn steel after low frequency load changes was proposed in this paper. As shown in the results, in the initial stage of fatigue, there was a monotonic relationship between the damage generated after the low frequency load and the nonlinear coefficient. Furthermore, the main reason for the increasing of nonlinear coefficient is the stretch of the dislocation string and the i n-crease of dislocation density in the fatigued area.The interaction between the finite-amplitude ultrasonic waves and the cracks was discussed in this paper, and the generation principle of the second harmonics was explained. By changing the amplitude of the excitation signal or the load on the crack interface, it became possible to study the effect of the cracks at different closed states on the nonlinear coefficient. The results show that higher amplitudes for the finite-amplitude ultrasonic wave result in stronger nonlinear effects produced on the crack interfaces as a result of the tension/compression effect. Furthermore, a larger closed area for the crack results in a stronger nonlinear effect was produced after the ultrasound waves pass through the cracks. Additionally, the appearance of macroscopic cracks increases the reflectivity of the ultrasonic waves and weakens the nonlinear effects. Based on the analysis above, in the m iddle and last stages of fatigue, there was a non-monotonic relationship between the damage and the nonlinear coefficient. The results show that the nonlinear coefficient increases mainly because of the generation and growth of micro-cracks, where as the nonlinear coefficient decreases because of the formation of macro-cracks.In the process of nonlinear detection, if the finite-amplitude waves pass through the fatigued area from different angles of incidence, the nonlinear effects generated will differ as well. In this paper, end face incidence, normal incidence, and oblique incidence methods were employed in order to extract the nonlinear c oefficient in the fatigue-loaded U71 Mn steel specimen. The studies showed that the oblique incidence method can not only guarantee the interaction between the ultrasonic waves and the fatigue-damaged area, but also show better adaptability to different kinds of welded joints, which helps to achieve better nonlinear detection r esults.The fatigue damage of specimens in different cycles of loading can be observed by TEM and metallographic microscope, and then the relationship curve between fatigue loading times N and the nonlinear coefficients β was proposed(i.e. the β~N curve). It was found that the increase of the nonlinear coefficient was caused by the stretch of the dislocation string and the increase of the dislocation density generated by cyclic hardening/softening. Thereafter, the appearance of micro-cracks further causes an increase of the nonlinear coefficient, where as the growth of macro-cracks leads to a decrease of the nonlinear coefficient.The ultrasound signals transmitted through the fatigued area of the welded joints were processed with HHT method. After signal processing, according to the theory of dislocations and cracks, the β~N curve was divided into three stages. In the first stage, the coefficient increases gradually, which corresponds to the initial f atigue process of stress concentration and the increase of dislocation density. In the second stage, the coefficient increases rapidly, which corresponds to the initiation and the stable growth of micro-cracks. In the third stage, the nonlinear coefficient begins to decrease, which corresponds to the formation and growth of the ma cro-cracks in the fatigue process. The residual fatigue life of the U71 Mn steel weld was predicted with the β~N curve obtained by means of the secondary loading method. The results showed that the β~N curve can characterize the entire damage process, which starts from the emergence of dislocations to the unstable propagation of cracks. The curve can also predict the residual fatigue life of the U71 Mn steel weld in certain conditions.In order to lower the costs of nonlinear ultrasonic testing, pulse inversion si gnal processing technology was utilized to simplify the nonlinear ultrasonic inspection system. The results showed that it can effectively extract the peak values of second harmonics in the ultrasonic signals which transmitted to the fatigued area of welded joints without hardware filtering. The β~N curve and nonlinear coefficients of specimens in different cycles of fatigue loading were also obtained with pulse inversion signal processing technology.