A probabilistic study of welding residual stresses distribution and their contribution to the fatigue life

Journal
Engineering failure analysis  
Article Number
104787
Citation
Engineering Failure Analysis (118): 104787 (2020-12)
Publisher DOI
10.1016/j.engfailanal.2020.104787
Scopus ID
2-s2.0-85089837744
Welding is a joining process that is associated with heating cycles which leads to considerable change in local material microstructure and the formation of high welding Residual stresses (RS) in the welded joint. Residual stresses can have a detrimental effect on the fatigue strength of welded joints. In this paper, previously published data from measurements of residual stresses in various types of welded joints are compiled. In total, more than 100 test results are studied covering steels with yield strengths between 307 MPa and 1050 MPa in different welded details (butt joints, longitudinal and transverse attachments, cruciform joints, as well as K-joints) with varying thicknesses. The collected data is used to study the distribution of welding residual stresses (regardless of the welding parameters) at weld toe and through the thickness of the welded plate. Probabilistic analysis is then used to arrive at a model that represents the value and distribution of residual stresses in welded joints. This model is used to predict and explain the scatter in fatigue test data from recent fatigue testing of welded samples.
Subjects
Compressive residual stresses
Probability density function
Residual stresses shapes
Tensile residual stresses
Welding