Structural stress magnification factor for thin plates with welding-induced distortions

No Thumbnail Available
Journal Title
Journal ISSN
Volume Title
Insinööritieteiden korkeakoulu | Master's thesis
Degree programme
Master's Programme in Mechanical Engineering (MEC)
The implication of thin-walled structures has recently gained attention in shipbuilding solutions for large-size constructions. Although this allow big ships to reduce their weight, research has shown that thin-walled structures generally exhibit severe welding-induced imperfections, which considerably affect the state of stress experienced by steel-made, 3-to-4 mm thick plates. Currently, the IIW standards provide a stress magnification factor k_m to estimate the Hot-Spot Structural (HSS) stress for welded components. Such factor considers eccentricity, linear lateral deflection and straightening effects under tensile load. However, its validity is confined to plates with thickness of at least 5 mm, since the k_m formula neglects the initial curvature effect, which is actually critical for the stress-state of thin plates. In this regard, the aim of this thesis is to provide an analytical model for the k_m factor that includes the initial curvature effect in evaluating the Hot-Spot structural stress due to welding-induced distortions. For this purpose, the study used a Second-order, geometrically non-linear, Beam Theory. Based on this theory, a model was designed by describing the welding-induced deflection only along the longitudinal direction of a plate. The study resulted in a concise analytical model for the k_m factor. Comparisons with the Finite Element Analysis (FEA), provided by the solver ABAQUS, validated the model for 4 mm-thick beam and plate strip elements, each of which assuming 27 configurations with different distortions and tensile loads. As the plate strip represents a plate with infinitely-rigid stiffener effect on lateral edges, a second plate element was modeled to simulate non-rigid lateral edges in order to extend the applicability of the model. A maximum underestimation of the k_m factor of about 8% for the second plate was observed. Both Its centerline and peak stress assessment showed mean errors around the 3%. In conclusion, the thesis confirms the dominating role of the welding-induced initial curvature effect in the thin-plate structural assessment and validates a suitable analytical model for the stress magnification factor k_m. Nevertheless, although the analytical model has been validated by FEA, further experimental validation is recommended for future works.
Remes, Heikki
Thesis advisor
Goncalves, Bruno Reinaldo
welding-induced distortions, thin plate, geometrical non-linearities, second-order beam theory, hot-spot structural stress, stress magnification factor
Other note