A Procedure for Drop-Tower Testing of Shallow-Cracked, Single-Edge Notched Bend Specimens

Unacceptably pessimistic predictions of structural reserve capacity often result from the use of conventional initiation fracture-toughness values (for example, J_Ic, δ_c) measured using deeply cracked bend specimens. Conversely, favorable comparison between predictions and structural fracture behavior have been reported when the predictions are based on toughness values measured in such a manner that the crack-tip constraint and the loading rate closely match those experienced in service. As a result, considerable attention has been paid recently to the elevated fracture initiation resistance of cracks shallower than the 0.45 to 0.75 crack length/width (a/W) used in standardized tests. However, standardized procedures for estimating the fracture initiation resistance of shallow cracks, particularly at the high loading rates characteristic of certain severe service conditions, have yet to be developed. In this investigation, techniques useful for estimating the load, load-line displacement, and the time of crack initiation during impact tests of single-edge notched bend specimens having shallow fatigue cracks (a/ W ≈ 0.1) were developed and validated. The bending stress distribution across the specimen midway between the support and loading points, determined based on strain measurements and a uniaxial stress-strain relationship, was used to estimate the load imparted to the specimen by the impactor. Load-line displacement was estimated using four noncontacting transducers positioned along the underside of the bend specimen, whereas the time of crack initiation was inferred from strain readings taken from the elastically loaded region behind the fatigue crack tip. The applicability of these procedures was demonstrated at an impact loading rate of 4.88 m/s, whereas control experiments at quasi-static loading rates were used to demonstrate the accuracy of these new procedures for both low- and high-strain-hardening steel alloys.