Thaulow, Christian (Norwegian University of Science and Technology, Dept Engineering Design and Materials) | Østby, Erling (SINTEF Materials and Chemistry) | Akselsen, Odd M. (SINTEF Materials and Chemistry) | Lange, Hans Iver (SINTEF Materials and Chemistry) | Åldstedt, Synnøve (SINTEF Materials and Chemistry)
The requirement to structural steels increases as the oil and gas industry goes north. It can be challenging to achieve high fracture toughness in the Heat Affected Zone and the weld metal at very low temperatures, and the present practice with lower bound fracture toughness testing is challenged. In this paper the effect of surface- vs. through thickness notching in fracture mechanics testing of the HAZ of weldments is examined experimentally and with FE (Finite Element) modeling. The standardized test specimen for lower bound fracture toughness testing is the deeply notched through thickness B*2B specimen. The B*B surface notch has, however, always been considered as an alternative to the through thickness notching as a more realistic and practical testing procedure.
In fracture mechanics testing of weldments, low fracture toughness is frequently experienced in the HAZ. The low toughness depends upon the detailed interaction between the crack tip and the embrittled microstructural constituents in the HAZ. It has further become evident that the large scatter often observed in the fracture mechanics testing of weldments not only is due to the distribution and size of these zones with embrittled microstructure, but also largely depends upon the global and local distribution of mechanical properties and specimen geometry, Toyoda M et al (1994) and Thaulow C et al (1994). The standardized test specimen for lower bound fracture mechanics testing is the deeply notched B*2B geometry, Figure 1. This specimen provides the highest constraint and will also have a high probability of hitting a brittle area because the through thickness notch (TTN) will intersect the HAZ of several weld layers, API (2005), British Standard (2005). But in practice there will seldom be through thickness cracks, and the fusion line will usually be slant with respect to the specimen surface.