HAZ Toughness: Realistic Testing for Pipeline Integrity

Fracture toughness testing of the heat-affected zone (HAZ) of linepipe seam welds is a requirement for most pipeline projects. Occasionally, low individual values can be measured in the HAZ and these have been attributed to, among other factors, the statistical nature of the HAZ and the associated probability of encountering local brittle zones. The structural significance of these outliers has remained a subject of debate between linepipe users and manufacturers [1], especially as their low significance can be demonstrated via large-scale structurally-representative tests [2–3]. To circumvent the higher cost of such large-scale testing, constraint-corrected fracture toughness testing can be used such that the conditions in the small-scale test more closely reflect those in service. However, there is little consistency between the many test and application codes in terms of how such tests should be carried out, and what steps are required to demonstrate that the measured toughness is structurally representative. Furthermore, the level of benefit to be obtained cannot be easily predicted. In the current study, a range of fracture mechanics tests was conducted on the HAZ of the longitudinal seam weld of a grade X65 U-O-E SAW pipe. Varying degrees of constraint, scale and loading mode were evaluated to establish the characteristic toughness of the HAZ in a statistical manner, with over fifty specimens tested in total. The specimens tested included notched bend (SENB) and tension (SENT) designs as well as surface notched tension (SNT), all with varying crack depth. The range of specimen and loading types, when compared with the requirements of the various relevant standards, highlighted the contradictory nature of current standards. The toughness established for each set of specific test conditions was used in a theoretical Engineering Critical Assessment (ECA) assuming various levels of applied stress, residual stress and flaw size inputs. The wide range of conclusions that would have been reached based on the small-scale toughness tests carried out under varying levels of constraint was easily demonstrated via the ECA. The conflicting requirements of several testing and application standards for longitudinal welds should be addressed, and their consistency with current approaches for girth welds improved. The study also shows that a single-parameter fracture criterion is an insufficient indicator of real HAZ toughness and constraint (metallurgical and geometrical) level must also be considered. The use of standard deeply-notched CTOD specimens, representing high constraint, gives a highly pessimistic view of seam weld integrity, especially when subsequently combined with an ECA.