The master curve approach specified within the new ASTM E1921 Test Standard is a significant advance in defining an indexing temperature, T-0, the median fracture toughness, and associated failure probability bounds for ferritic steels in the ductile-to-brittle transition regime. An objective in developing: this standard test procedure has been that it should, if possible, allow the use of precracked Charpy-size specimens to measure T-0 and the associated fracture performance. However, the supporting technical basis document for E1921 [NUREG/CR-5504, US Nuclear Regulatory Commission, Washington, DC, November 1998] presents no experimental evidence which demonstrates the accuracy of T-0 estimates from precracked Charpy specimens. There was also little independent experimental verification presented [NUREG/GR-5504, US Nuclear Regulatory Commission, Washington, DC, November 1998] to support the limiting deformation criterion (M = 30) contained in the standard. Computational and analytical support [NUREG/CR-5504, US Nuclear Regulatory Commission, Washington, DC, November 1998] of the deformation criterion for precracked Charpy specimens is also lacking. The most applicable recent computational work by Ruggieri and Dodds [Engng. Frac. Mech. 60 (1) (1998) 19] supports a deformation criterion that is almost twice as stringent as the E1921 standard to ensure that constraint loss does not occur. Historically, constraint loss has been evaluated far individual specimens, while E1921 implicitly only requires that constraint loss does not significantly alter the median toughness. This work experimentally evaluates the use of precracked Charpy specimens to estimate T-0 by comparing results obtained on both precracked. Charpy and 1T (or larger) fracture toughness measurements. The data demonstrate that the Charpy-size specimens tend to have a non-conservative bias, even when the data fall within the allowable E1921 deformation criterion. This effect may be a function of the degree of crack tip constraint loss, which strongly depends on the material flow properties. An attempt is made to correct for any constraint loss in the Charpy-size specimen and remove any bias in the measured T-0 using a simple constant stressed volume cleavage failure criterion [Int. J. Frac. 74 (1995) 131]. While this method does decrease the difference between T-0 values measured using Charpy and conventional 1T specimens, more detailed cleavage initiation models which can be independently calibrated will be required to rigorously account for any constraint differences between Charpy and larger specimens. (C) 2001 Published by Elsevier Science Ltd.
