ULCF of welded joints under variable multi-axial strains

The phenomenon of ultra low-cycle fatigue (ULCF) and damage of welded joints under multi-axial and variable amplitude strains is complex and not completely understood. This phenomenon has particular significance for safety and environmentally relevant structures (above-ground petroleum tanks, masts for wind turbines, pressure vessels and bridge piers) subjected to extreme deformations during extraordinary seismic or wind gust events. In the case of above ground tanks, severe seismic ground accelerations can cause the tank bottom to uplift. This behavior causes extensive plastic deformation in the region near the welded connection between the tank wall and the bottom plate, which occur with highly variable and stochastically distributed amplitudes. Fracture in the ultra- low cycle fatigue has also been observed in tubular elements such as pipelines, circular mast shafts and bridge piers, leading to collapse of these components and causing human as well as economic losses. Very little information exists concerning the ULCF response of welded connections subjected to multi-axial loading. Confirmed knowledge and new engineering design procedures would allow for an increase in safety of the mentioned structural elements under load scenarios causing extraordinary local strains. Objectives The objectives of the proposed research are the following: Understand the ultra low-cycle fatigue (ULCF) behavior of welded connections under multi-axial strains Study the realistic deformation sequences in typical welded connections under seismic and other extraordinary conditions which are potentially susceptible to ULCF Provide a model capable of estimating crack initiation and propagation of welded joints undergoing ULCF, with a focus on both conventional and high-strength steels. Develop a damage index for failure of welded connections in ULCF Apply the damage index to assess life of assemblies under combined shear and normal strain ranges Apply infrared thermography to check and enhance the understanding and modeling of plastic strain distribution in ULCF. Proposed research The research consists in combined experimental and numerical studies on the ultra low-cycle fatigue (i.e., 10 or less cycles) of two different welded assembly types subjected to multi-axial strains. The experiments will be carried out under constant and variable amplitude deformation histories. The first assembly, the welded wall-to-base connection in storage tanks, of two different thicknesses and materials, will be used to study the influence of mean strain, plate size and steel grade. The second assembly, a tube-to-base connection as found in circular mast shafts and pipes, will be used to study shear to normal strain ranges combinations, as well as steel grade. These assemblies will be subjected to different multi-axial strain sequences. In total, over 150 tests on the two assemblies will be carried out, thus creating a database for validating the developed damage models. Different local continuum void growth cyclic models, damage accumulation functions as well as propagation models used to predict total failure will be compared. Detailed nonlinear finite element analysis (FEA) at structure`s level and local level will be used. The location, evolution and distribution of plastic zones during crack initiation and propagation phenomena will be studied using thermographic infrared (IR) imaging and compared to detailed cyclic FEA computations.

The phenomenon of ultra low-cycle fatigue (ULCF) and damage of welded joints under multi-axial and variable amplitude strains is complex and not completely understood. This phenomenon has particular significance for safety and environmentally relevant structures (above-ground petroleum tanks, masts for wind turbines, pressure vessels and bridge piers) subjected to extreme deformations during extraordinary seismic or wind gust events. In the case of above ground tanks, severe seismic ground accelerations can cause the tank bottom to uplift. This behaviour causes extensive plastic deformation in the region near the welded connection between the tank wall and the bottom plate, which occur with highly variable and stochastically distributed amplitudes. In this research project, strain histories at critical, welded details in large above-ground liquid storage tanks were studied with the help of specific, realistic numerical analysis models. Different modelling strategies could thereby be compared and developed further, in order to accurately reflect the complex behaviour of a liquid-filled, unanchored storage tank under large seismic excitation. The strain histories were then compared to the resistances obtained from large-scale tests on multiaxially loaded welded joints. The results of the project lead to an increase of understanding of the optimal engineering of such details in safety-critical liquid storage tanks under severe seismic loading.