APPLICABILITY OF SUB-CHARPY-SIZE BEND AND IMPACT SPECIMENS FOR ESTIMATION OF FRACTURE TOUGHNESS IN THE TRANSITION REGION

摘要

As part of a test program to introduce small specimen test techniques, fracture toughness tests have been performed with 20% side-grooved small (1010-55 mm) and "ultra-small" (3-4-27, 5-5-27 and 510-55 mm) specimens. The objectives were to evaluate the measuring capacity of small specimens and to verify the applicability of the statistical cleavage fracture model of VTT for such test data. Two pressure vessel steels were used in the study. Most of the tests were performed in the transition range and ended at cleavage fracture, but some tests were also conducted in the temperature range of ductile tearing. According to the measured data, the 100 MPa1/2m reference temperatures (To), determined even with 3-4 mm specimens, are in good agreement with those of the larger specimens. The statistical specimen size correction only was applied in the analyses, i.e., without any consideration of specimen constraint. The measurements made in the ductile range indicate that practically similar fracture resistance curves can be obtained with the 510 mm and larger specimens up to the J-level of 500 kN/m. Also, correlations have been developed, by applying a fracture mechanics approach, for estimating the Charpy transition temperatures of ferritic structural steels from small specimen test data. The correlations were verified by analyzing test data produced with Charpy (ISO-V) and KLST-type subsize specimens (3-4-27 mm). Two energy criterion combinations were applied for determining transition temperatures, i.e., the criteria based on the area of the specimen ligament, being in accordance with the master curve definition, and those derived from an empirical mean relationship found between the upper-shelf energies measured with ISO-V and KLST specimens for various steels. The results show that both of these criteria can be used to give a satisfactory correlation between the transition temperatures and that the standard deviations of the correlations are approximately equal, i.e., around 15°C. A method is proposed for estimating the Charpy upper-shelf energy from subsize specimen data.