Volume 3, Issue 5 (May 2006)
Application of the Reference Temperature to the Evaluation of Cleavage Fracture in HSLA-100 Steel
The purpose of this study was to evaluate the fracture behavior of HSLA-100 steel at a temperature of −40°C (−40°F). Tests were conducted on three different compositions and two orientations of HSLA-100. HY-100 was also tested to provide a baseline for comparison. Various fracture-related tests were conducted, including traditional test methods, such as the Charpy V-notch and the Dynamic Tear, and more recent methods, such as ductile crack growth initiation (JIc) and the reference temperature. Transition temperatures from the Charpy tests were all very similar, falling around −129°C (−200°F), although the upper shelf Charpy energy for the HY-100 was lower than the HSLA-100 steels. Dynamic tear energies at −40°C for the four steels ranged from 800 to 1800 J. The average reference temperature of all four steels, measured according to ASTM E 1921, was around −151°C (−240°F), with one composition of HSLA-100 coming in as low as −179°C (−290°F) and another as high as −123°C (−190°F). Fracture toughness tests were conducted according to E 1820 at temperatures of −40, −29, and −18°C. These temperatures were more than 83°C above the highest of the reference temperatures, so cleavage fracture was not expected to occur. The HY-100 specimens exhibited an average initiation toughness of 227 kJ/m2 (1296 lb/in.) and crack growth remained ductile throughout the tests. Even though the HSLA-100 exhibited initiation toughnesses more than twice as high as the HY-100, many of the tests terminated in cleavage fracture after amounts of ductile crack growth varying from 0 to 1.85 mm (0.073 in.). This was an unexpected result based on good Charpy and dynamic tear energies at −40°C, and low reference temperatures. The higher initiation toughness and tearing resistance of HSLA-100 allowed very high stresses to develop in the fracture process zone. As ductile crack growth occurred, the fracture process zone swept through an increasing volume of metal, which increased the probability of cleavage fracture. It is shown that the load required to reach these levels of crack driving force may be above gross section yielding of a typical structure. In this case failure by gross section yielding would occur before fracture. This is demonstrated with the aid of a failure assessment diagram.