Research On The Stress State Of Dynamic Fracture Specimen And Its Effect On The J-integral

In recent years, the dynamic fracture test under the higher loading rate by Hopkinson bar has been paid more attention. Generally, in the dynamic fracture test, it is believed that the stress condition of dynamic fracture specimen is the three-point bending. And the analysis of the dynamic stress intensity factor is the based on the three-point bending under static condition. Yet it is well known that, in the static fracture test, the specimen has been in contact with supports and in a state of three-point bending. That is because the response of the specimen is synchronized with the loading rate. However, in the dynamic fracture test, the specimen always bounces off supports due to the higher loading rate. Thus, the state of dynamic fracture specimen is no longer a three-point bending stress state in the usual way. Nevertheless, there is no consistent understanding about the phenomenon and laws for the loss of contact. In this paper, the above problem is more systematically studied, and the main results are as follows.1) When the specimen dimensions and span are designed by the current three-point bending test method in the dynamic fracture test by Hopkinson bar, there is loss of contact between the specimen and supports. The one-point bending of the Timoshenko beam model has been used to analyze the specimen stress state. The result also indicates that the existence of the loss of contact between the specimen and supports.2) The results of the theoretical calculation for the displacement-time curve of nominal support-point show that the bouncing displacement amplitude decreases along with the decrease of the span until the span approaches to a critical value. After that, the specimen is in contact with supports, i.e. in the three-point bending stress state. And the linear relationship between the nominal support point and the crack-to-width ratio has been gained by the first-order approximation about the Timoshenko beam model, which can be used to judge the contact state between the specimen and supports under the loading process. Based on this analysis, the true three-point bending critical condition has been presented: S/L≤0.72 Thus, for a standard Charpy specimen, when its span is less than 0.72L, the specimen keeps contact with supports during loading process. Otherwise, there is the phenomenon of the loss of contact.3) The contact state law between the specimen and supports has been studied which is influenced by some factors such as the impact velocity, specimen dimensions. It is found that the influence of the impact velocity on the contact state is not obvious. And the span, specimen length and width have effect on the contact state and these influences have the same tendency. That is to say, the time for loss of contact decreases with the decrease of the span and the increase of the dimensions of specimen. When the span and specimen dimensions achieve a certain value, the time for loss of contact stabilizes at a certain level, i.e. the specimen is in contact with supports. The concept of critical state has been put forward and the critical criterion has been observed. On the basis of the above-mentioned law, the criterion of the structure factors has been established, which reflects the contact state between the specimen and supports. S/W＝1.17＋0.45L/W The criterion can be used to judge the contact state scientifically with various sets of specimen dimensions and span.4) The three-point bending test on the apparatus of Hopkinson bar has been simulated using the ANSYS software. The simulation results show that the phenomenon of the loss of contact in experiments is verified by the support-point displacement-time curve. And the dynamic J-integral of the specimen crack-tip has been calculated under different contact states. The results indicate that the stress state of specimen has obvious effect on the dynamic J-integral. And the calculation results by the finite element and spring-mass model have been analyzed for the Charpy specimen under the span of 80mm. The J-integral for these methods have an obvious difference, which have an error of about 20%at the crack initiation. If the stress intensity factor for the specimen which loses contact with supports has been calculated by the three-point bending analysis, the results will have big error. Therefore, for the dynamic fracture specimen, the analysis of the stress intensity factor must be based on the real stress condition during the loading process.