Directionally solidified Rene 80 (DS R80) and polycrystalline Inconel 738(IN 738) Superalloys were tested in thermo-mechanical fatigue (TMF) over the temperature range of 500-900°C and plastic strain range from 0.1 to 0.8% using a DSI Gleeble thermal simulator. Thermo-mechanical testing was carried out on the parent material (baseline) in the conventional solution treated and aged condition (STA), as well as gas tungsten arc welded (GTAW) with an IN-738 filler, followed by solution treatment and ageing. Comparison of the baseline alloy microstructure with that of the welded and heat treated alloy showed that varying crack initiation mechanisms, notably oxidation by stress assisted grain boundary oxidation, grain boundary MC carbides fatigue crack initiation, fatigue crack initiation from sample surfaces, crack initiation from weld defects and creep deformation were operating, leading to different “weakest link” and failure initiation points. The observations from this study show that the repaired samples had extra crack initiation sites not present in the baseline, which accounted for their occasional poor fatigue life. These defects include lack of fusion between the weld and the base metal, fusion zone cracking, and heat affected zone microfissures.

Directionally solidified Rene 80 (DS R80) and polycrystalline Inconel 738(IN 738) Superalloys were tested in thermo-mechanical fatigue (TMF) over the temperature range of 500-900°C and plastic strain range from 0.1 to 0.8% using a DSI Gleeble thermal simulator. Thermo-mechanical testing was carried out on the parent material (baseline) in the conventional solution treated and aged condition (STA), as well as gas tungsten arc welded (GTAW) with an IN-738 filler, followed by solution treatment and ageing. Comparison of the baseline alloy microstructure with that of the welded and heat treated alloy showed that varying crack initiation mechanisms, notably oxidation by stress assisted grain boundary oxidation, grain boundary MC carbides fatigue crack initiation, fatigue crack initiation from sample surfaces, crack initiation from weld defects and creep deformation were operating, leading to different “weakest link” and failure initiation points. The observations from this study show that the repaired samples had extra crack initiation sites not present in the baseline, which accounted for their occasional poor fatigue life. These defects include lack of fusion between the weld and the base metal, fusion zone cracking, and heat affected zone microfissures.

Welding, as one of the most effective joining methods for metals, has been extensively applied in engineering usage for a long time. When cracks occur in the vicinity of weldments, weld repairs are frequently considered for crack repair to extend service life. In order to evaluate to what extent the weld repair has improved the fatigue life of a cracked welded structure, it is necessary to be able to determine the residual life of the cracked welded joint, as well as the life of the weld repaired joint. Both these assessments require that the fatigue crack growth data be available. The determination of crack propagation rates of welded and weld repaired structures is thus of paramount importance to implement a damage tolerant approach to structural life extension. However, since most studies on welded joints so far have concentrated on fatigue life evaluation, at the present time only limited information is available on crack propagation rates in welded joints, and virtually none on fatigue behaviour and crack propagation in weld repaired joints. This thesis has focused on examination of fatigue and crack propagation behaviour in as welded and weld repaired aluminium alloy 5083, a weldable marine grade alloy extensively used in construction of high speed ferries and aerospace structures. Crack growth rates were measured during constant amplitude fatigue testing on unwelded, as-welded and weld repaired specimens of 5083-H321 aluminium alloy. A 3-D finite element analysis was conducted to determine the stress intensity factors for different lengths of crack taking into account the three-dimensional nature of the weld profile. The effects of crack closure due to weld residual stresses were evaluated by taking measurements of the crack opening displacements and utilised to determine the effective stress intensity factors for each condition. Metallurgical examinations and fractography of the fracture surface were conducted using an optical microscope and SEM. It was found that crack growth rates in welded plates are of the same order of magnitude as those of parent material when effective stress intensity factors were applied. However weld repaired plates exhibit higher crack growth rates compared to those of unwelded and once-only welded plates.