To accurately evaluate the fracture behaviors and fatigue life of welded structures, it is necessary to consider welding residual stress in addition to the stress produced by the applied loads. Characteristic tensor, which has a proportional relationship to stress intensity factor (SIF), is one new approach to evaluate stress singularity caused by a crack. Characteristic tensor is the tensor which is easily calculated by integrating stresses in a simple sphere zone around a crack tip using coarse mesh without special limitation in the mesh division, even under general case like mixed-mode loading conditions. In this study, the stress singularity in the welded plate is evaluated by characteristic tensor and its effectiveness with the residual stress field is verified. Firstly, the residual stress distribution in butt welded plates is calculated by finite element method. Secondly, the singular stress field is obtained by introducing a straight single crack in the welded plates. A crack with different lengths and the crack tip located at different residual stress states are considered in reproducing the stress singularity. In each case, SIF is estimated from the characteristic tensor calculated from stress distribution nearby a crack tip and compared to those from the green function or virtual crack extension method. Additionally, the stress singularity caused by the external load is evaluated in the same manner. The estimated SIFs with coarse mesh are in good agreement with the conventional methods. The results suggest that characteristic tensor provides a practical approach to evaluating stress singularity in welding
In order to investigate the effects of residual stress modification on the results of crack-tip opening displacement (CTOD) tests in heavy thick welded joints, experiments using a submerged-arc welding (SAW) joint of an EH47 steel plate with the thickness of 70mm and residual stress analyses by the finite element method (FEM) were performed. It was confirmed that appropriate reversed bend conditions improved fatigue precrack front shapes, and the low ratio of the fatigue precrack length to the size of the compressive plastic zone induced by reversed bending contributed to the fatigue crack shape improvement. Although the local compression with a large diameter platen was effective in improving the fatigue precrack front shape, small diameter platens were not effective even if they applied to wide ranges of specimens. In addition, specimens treated by the reversed bending showed higher critical CTOD values than those treated by local compression. According to the FEM analysis, the reversed bending introduced tensile residual stress at the mechanical notch root, although the initial residual stress distribution of the welded joint still remained a little away from the notch root. On the other hand, the initial residual stress was almost removed by the local compression with a large diameter platen. It was presumed that the initial residual stress distribution affected fatigue precrack front shapes and the critical CTOD in welded joints. The reversed bending is very useful if appropriate conditions can be used.
Identification of the relationship between fatigue crack propagation rate (da/dN) and stress intensity factor range (∆K) is inevitable to apply the fracture mechanics approach to assess the growth of fatigue crack growth. The relationship between da/dN and ∆K is widely applied to evaluate fatigue crack propagation behavior. To evaluate the fatigue crack growth history under variable loading history, it is necessary to replace ∆K to the effective stress intensity factor range, which can quantitatively consider fatigue crack opening and closing behavior.
∆Keff proposed by Elber is well known as the effective stress intensity factor range, but even if ∆Keff is applied to evaluate the fatigue crack propagation behavior, a threshold value ((∆Keff)th)was occurred. On the other hand, it is known that fatigue cracks propagate even at ∆Keff below (∆Keff)th under variable loading history. This implies that ∆Keff is an insufficient parameter to describe fatigue crack propagation behavior.
∆KRPG, which has a close relationship with the cyclic plastic behavior in the vicinity of the crack tip proposed by Toyosada and Niwa, can give the solution of the problem of ∆Keff, and even under complicated variable loading history including multiple frequency components.
To apply the fatigue crack propagation law with ∆KRPG as a parameter, it is necessary to experimentally measure the RPG load and identify the propagation law constants (∆KRPG=C(∆KRPG)m) C and m.
A conventional method for identifying RPG loads requires the superposition of the hysteresis loop near the crack tip and its reversal loop as measured by the unloading elastic compliance method. However, advanced skills in this method, such as understanding the characteristics of measurement errors associated with loops, is required.
In this study, we propose a method to automatically measure the RPG load equivalent to that of an expert, and compare the automatic measurement results by this method with the past conventional measurement results under multiple materials and loading conditions, and validate the automatic measuring method.
Two-dimensional axisymmetric simulation was carried out to investigate heat source characteristics during an arc spot welding with a constricted nozzle. As a result, it was clarified that a part of the welding current flowed from a cooling nozzle which was an anode to a tungsten cathode directly. The temperature increase of the cathode tip was suppressed by the decrease of the Joule heating at the tip because of this separation of the current path. Comparing temperature and heat input density distributions on a base metal surface with a conventional tungsten inert gas (TIG) welding, there was no difference in the temperature on the central axis of the tungsten cathode, but the peak of the heat input density during the arc spot welding was lower than that in the TIG welding. In addition, the heat input range was narrower than that in the TIG welding by a cooling nozzle contacting the base metal. It was also clarified that the heat input to the base metal surface and the temperature increase of the tungsten electrode were suppressed by the nozzle. Inner gas flowing from the constricted nozzle cooled the outer edge of the arc plasma, suggesting that it contributed to the long lifetime of the tungsten electrode.