Browsing by Author "Arffman, Pentti"

Dissimilar metal welds(DMW) in nuclear power plants have been identified to be prone to failure by cracking. Generally cracks grow straight, unless a weaker microstructure is adjacent. In DMWs cracks have propagated to the fusion line. Fracture toughness of weldments has been reported to sink in the same location. It has been shown that hardness can be used to predict strength of a homogeneous material. In this thesis, such model was applied to DMWs. On the other hand, fracture toughness and crack propagation was studied for both ductile and brittle specimens. Two DMWs were studied, MU1 and NI1. They shared the same SS base metal and Alloy 52 weld metal. Furthermore the LAS base metals were very similar: 18MND5 and SA508 for MU1 and NI1, respectively. NI1 specimens were studied in three different ageing conditions. Some of the data was reutilized from previous projects. Vickers hardness was measured over the LAS fusion line. To be able to utilize hardness measurements without knowledge of their distribution, the data was bootstrapped. The means of the bootstrapped medians provided estimates of the average hardness and its deviation. Three tensile specimens were tested from four different locations. Tensile results were compared to strength predictions from hardness. The VTT model developed for homogeneous materials predicted strength from hardness inadequately in its original form. However, once the parameters were varied, predictions agreed with measurements. Ductile specimens were made from both MU1 and NI1. Temperatures for testing were 300°C and 20°C for MU1 and NI1 respectively. Their fracture toughness was determined with JIc, which is defined in ASTM E1820. Crack path was studied with either cross-cutting or profilometry. In the case of profilometry, the specimens were profiled from three different locations and the mean of results was used. Brittle specimens were all from NI1. ASTM E1921 determines cleavage fracture toughness Jc. Test temperatures were between −80°C and −130°C. Crack path was studied with profilometry. Profiling location was at cleavage nucleation point. The point was determined with scanning electron microscope. The accuracy of profilometry itself was confirmed by comparing the results to specimen cross-cuts. Linear correlation between fracture toughnesses and crack jumps towards fusion line was established successfully for most datasets. Ageing of the specimens reduced the fracture toughness of brittle specimens, but not of ductile ones. Instead, ductile fracture toughness increased with test temperature.