Modified West Jefferson Burst Test for Assessment of Brittle Fracture Arrest in Thick-Wall TMCP Line-Pipe Steel With High Charpy Energy

Abstract

Three partial gas pipe burst tests were conducted to assess the brittle-to-ductile transition temperature and brittle fracture arrestability of a heavy-walled TMCP line-pipe steel. This steel had a very high Charpy energy (400 J) which is typical of many modern line-pipe steels. In standard pressed-notch DWTT specimen tests this material exhibited abnormal fracture appearance (ductile fracture from the pressed notch prior to brittle fracture starting) that occurs with many high Charpy energy steels. Such behavior gives an invalid test by API RP 5L3, which makes the transition temperature difficult to determine. The first burst test was conducted in a manner that is typical of a traditional West Jefferson (partial gas vessel) burst tests. The crack was initiated in the center of the cooled vessel (with a partial air gap), but an unusual result occurred. In this test a ductile fracture just barely started from each crack tip, but one of the endcaps blew off. The pipe rocketed into the wall of a containment building. The opposite endcap impacted the wall of the building and brittle fractures started there with one coming back to the center of the vessel. The implication from this test was that perhaps initiation of the brittle fracture in the base metal gives different results than if the initial crack came from a brittle location. The second burst test used a modified West-Jefferson Burst Test procedure. The modification involved cutting a short length of pipe at the center of the vessel and rotating the seam weld to the line of crack propagation. The HAZ of the axial seam weld had a higher dynamic transition temperature. The initiation flaw was across one of the center girth welds so that one side of the initial through-wall crack had the crack tip in the base metal while the other side initiated in the seam weld HAZ. On the base metal side, the crack had about 220 mm of crack growth before reaching steady-state shear area, i.e., the shear area gradually decreased as the fracture speed was increasing. On the other side, a brittle fracture was started in the HAZ as expected, and once it crossed the other central girth weld into the base metal, the fracture immediately transformed to a lower shear area percent. These results along with those from the first burst test suggest that the DWTT specimen should have a brittle weld metal in the starter notch region to ensure the arrestability of the material. The final burst test was at a warmer temperature. There was a short length of crack propagation with higher shear area percent, which quickly turned to ductile fracture and arrested. In addition various modified DWTTs were conducted and results were analyzed using an alternative brittle fracture arrest criterion to predict pipe brittle fracture arrestability.