ABSTRACT Establishing the fracture toughness of real HAZ can be challenging due to the small extension of the potentially brittle zones. Application of testing of weld thermal simulated HAZ microstructures can provide an alternative approach to establish fracture properties of the more brittle microstructures. In this paper CTOD values measured in real and simulated CGHAZ of a 420 MPa steel are compared at -60 and -30ºC. It is shown that for the lower temperature, where the fracture toughness is very low, similar CTOD values are obtained in both the real and simulated HAZ. At the higher temperature the fracture toughness increases more in the real HAZ compared to the simulated one. FE analyses are used to propose an explanation for this behaviour. The results present an interesting first step in the direction of use of weld thermal simulated samples for obtaining relevant CTOD values for used in assessment of real HAZ.
INTRODUCTION The low temperatures encountered in the Arctic poses a special challenge with regards to the use of steel structures. The risk of brittle behaviour in steels and their weldments must be controlled in order to have acceptable structural integrity. Especially weld metal and heat affected zone (HAZ) could be subject to low toughness. These are also areas where defects are most likely to occur. A challenge associated with the characterization of brittle fracture is the large scatter in fracture toughness test results. Application of statistics is necessary to obtain reliable estimates of the toughness. General approaches in this respect have been developed by Beremin (1983) and Wallin (2002), and examples of statistical methods applied to characterize scatter in HAZ fracture can be found in Hauge (1990). Determination of fracture toughness in the HAZ represents a special challenge. The brittle zones may be of limited size and there could be a large gradient of toughness acting over small distances in the material.