On the Unloading Compliance Method of Deriving Single-Specimen R-Curves in Three-Point Bending

Abstract

The application of unloading compliance for deriving single-specimen R-curves in three-point bending has been examined. Various methods, including the use of a double-clip-gage arrangement, are investigated. Finite-element calculations show that the best estimate of crack length is likely to be obtained from the conventional clip gage/load compliance, using a/W ≥ 0.5. The results are compared with those from blunt-notched calibration test specimens and with fatigue-cracked specimens, both with and without side grooves, of two pressure vessel steels. It is found that the initial crack length may be estimated accurately, but that the estimate becomes less precise as tearing occurs. Consequently, predictions of crack extension are rather inaccurate. However, there is considerable scope for improvement in the measurement technique employed, and for the use of correction factors. The use of two clip gages allows good estimates of load-point displacement, crack-tip opening displacement (CTOD) and J to be made, particularly when plastic displacements become significant. It is suggested that the technique could prove valuable in routine measurements of CTOD and J on bend specimens.