PSI - Issue 22

C.L. Niu et al. / Procedia Structural Integrity 22 (2019) 361–368

362

2

Author name / Structural Integrity Procedia 00 (2019) 000 – 000

1. Introduction There are a lot of complex welded structures in railway vehicles. The main form of structural failure during the service of railway vehicles is the fatigue damage of welded joints [1]. Therefore, the design of welded joints is of great significance to the service safety of railway vehicles. At present, engineers and technicians have carried out a lot of research on fatigue life of welded structures. Kassner simulated the frame in finite element according to the load specified in EN13749. The maximum stress position was the weld position. The fatigue evaluation was carried out using IIW standard [2]. Then the frame is tested on the bench test, and finally the conclusion that the bench test evaluation method is more conservative is obtained. Ju Yingzi used the master S-N curve method of ASME standard to predict the service life of the car body weld, and analysed the advantages of the master S-N method. The accuracy of life calculation has been greatly improved [3]. However, BS EN15085-3 emphasizes that in the design process, on the premise of satisfying the fatigue life, the welded joints must assess the high, medium and low stress levels in order to carry out the follow-up work. The principle of this standard is clear, but it does not provide specific methods for implementation [4-5]. Zhao Wenzhong et al. studied the method of calculating stress factor based on EN15085-3, and gave the basic steps of calculating stress factor based on structural stress[6]. Based on the main standard IIW-2008 and ASME-BPVC-VIII-2:2015 for welding design and fatigue assessment of steel structures in railway industry, this paper studied the method for evaluating stress state grade of welded joints of complex structures, which meets the requirements of EN15085-3 and is operable. The method is programmed and applied in welding structure design of welded frame [7-8]. 2. Assessment Method and Programming of Stress State Grade IIW-2008 is a fatigue assessment method based on nominal stress structure. It is widely used in engineering, but it is limited by joint form and load mode, and sensitive to mesh[9]. ASME-BPVC-VIII-2:2015 is a fatigue assessment method based on structural stress. It is insensitive to mesh without considering the form of welded joints. It can evaluate the fatigue life of welded joints with welding defects more accurately. According to BS EN15085-3 standard, the stress factor is calculated on the premise of satisfying the fatigue life, and then the high (stress factor > 0.9), medium (0.75 < stress factor < 0.9) and low (stress factor < 0.75) stress state grades are determined. According to the two standards, the feasible technical route of stress state grade assessment is obtained. The method of stress state grade assessment is programmed by using C language and APDL language. 2.1 Assessment of Stress State Grade Based on IIW-2008 Nominal Stress Method IIW-2008 provides 80 S-N curves of welded joints of steel structures. The curves include nearly 100 types of welded joints. The fatigue properties of welded joints are evaluated by the maximum principal stress range. The S-N curves have double slope characteristics. In the standard, Fatigue Grade (FAT) is used to calibrate welded joints with the same fatigue sensitivity. The assessment steps of stress state grade are as follows [10]. 1 ） According to the geometry, load-bearing mode and type of welded joint, the weld seam is divided into different FAT grades, and the stress state grade is evaluated on the premise of satisfying the fatigue life. 2 ） Extracting the principal stress and its six stress components of weld joint under different fatigue calculation load cases in integral coordinate system. 3 ） The maximum principal stress of the evaluation position under all fatigue conditions is determined as the maximum principal stress max  , and its corresponding direction cosine ( x n , y n , z n ) is taken as the direction of the basic stress vector; the direction cosine is determined by vector operation, and the stress tensor matrix of the evaluation point is multiplied with the column vector of the basic stress vector to obtain the stress sequence vector  s  (stress vector acting on the plane with normal vector as basic stress vector). Then the stress vector is projected to the direction of the basic stress vector, and  N  is obtained. The calculation is shown in formula (1). The minimum value of evaluation point  N  under all fatigue cases is determined as min  .