Fatigue and Fracture Assessment of Butt Welded Joints and Thermal Cut Edges under Axial and Bending Loads

Abstract

The aim of this work is to provide a research for fatigue of structures applied on container vessels. Considerable increase on size and capacity of container ships has led to the utilization of thicker plates and higher tensile materials. For instance, it is highly observed on upper flange, especially hatch corners, and also for butt welds. This work intends to verify the influence of internal flaws on butt welded joints for different loading types on thick plate; and to check the consequences on fatigue life of thermal cut edges considering distinct in service edge treatments, materials, loading type and plate thickness. For the first part, experimental fatigue tests previously carried out by DNV GL for several butt welded joints specimens with thickness of 50 mm made of high tensile material with pre-established internal defect were considered for analysis. In each specimen, the lack of fusion was measured by ultrasonic test and, after cyclic loading test, the S-N curve is obtained. Linear elastic fracture mechanics approach is used to simulate one selected specimen and the results are compared. Parameters of Paris-Erdogan equation are taken from well-known references. Differences between pure axial and pure bending loads are verified, including a comparison with literature. Embedded elliptical crack model with same area of the actual internal lack of fusion can represent the expected life of the specimen. It is shown that IIW2008 parameters are much more conservative than the parameters taken from DNV GL butt welded joints lab tests; also that the specimen subjected to pure bending load presents higher life than when only axial acting load is applied. On the other hand, a deeper research for fatigue enhancement due to bending needs to be carried out for internal defects. Second part of this work presents a finite element analysis of specimens with distinct thicknesses and different edge treatments established after thermal cut procedure is adopted. Analytical calculation for no edge treatment is determined and kept as reference value for axial and bending cases. Stress concentration and maximum stresses influenced by various edge treatments are determined for pure axial and pure bending loads, including a comparison between the different edge treatments. Later on, fracture mechanics analysis is performed providing a reliable outcome about the fatigue life for different crack models that are considered to simulate actual cracks evidenced on previous tests and references, along with a small initial crack size that was carried out for axial and bending loads. Parameters for base material are obtained from notable reference and results from this analysis will be used by DNV GL in the future, in which the simulation results achieved herein will be confronted with the experimental tests. Comparison between axial and bending is once more performed for each crack model. Lastly, a real case scenario investigated by fracture mechanics based on GL report, standards and guidelines considering variable loads using straight line spectrum is also included. No considerable difference was observed for notch stress on the specimen due to edge treatment. Simulation for corner crack model presented more impact in fatigue life of materials than semi-elliptical surface crack model; for same material and crack model, thicker specimens achieved higher fatigue life; and high tensile materials results are not completely agreeing with formulation based on classification societies standards, thus further analysis is proposed. Influence on lifetime of container ship’s structures due to different crack sizes and models is verified for variable loads, leading to better consideration on inspection and maintenance. Finally, conclusions are summarized for both parts of this research and future works are recommended, along with all references that were used on the elaboration of this master thesis.