Abstract

This paper presents a study on the structural hot spot stress method and the master curve method (Battelle) for evaluating the fatigue life of a T‐shaped welded joint. Both methods are based on the concept of structural stress, which is usually determined using finite element models. Using different types of shell finite element models, the results obtained by the two methods were compared to experimental data, verifying significant differences between the studied methods. The loading level had a great deal of influence over the accuracy of both methodologies, but on the other hand the hot spot method revealed to be more precise for the analyzed joint. The main differential for the Batelle method was the correction of the stress amplitude into an equivalent structural stress parameter, since, without this correction, the stresses on the weld toe were very similar between the Battelle and the hot spot methods. Comparing the master curve for the fatigue behavior of the Battelle method to the fatigue curve of hot spot method, with both curves expressed in terms of the nominal stress applied in the experiment, it was verified that the hot spot method is conservative. Among the different modeling techniques used, only one provided results that are not in agreement with the others. Using the Battelle method, the behavior of the bending stress component varied considerably according to the employed numerical model, indicating that can be a greater variability in the expected fatigue life for joints under more complex loading configuration.