This paper is concerned with the use of the sharply notched cylindrical specimen as an index of plane-strain fracture toughness in quality assurance of aluminum alloy products. Specifically, information is presented that relates to the use of the ASTM Tentative Method for Sharp-Notch Tension Testing with Cylindrical Specimens (E 602-76T) in quality assurance programs. The first part of this paper describes the results of an investigation into the influence of fundamental testing variables on the sharp notch strength of several high-strength aluminum alloys. The results indicate that variations in the notch root radius and eccentricity of loading (expressed in terms of the percent bending in a verification specimen) within the range permitted by the Tentative Method can contribute significantly to the scatter observed in relations between the sharp notch to yield strength ratio (NYR) and K_Ic. The results also show that the upper limit of K_Ic beyond which the NYR loses useful sensitivity to further increases in K_Ic decreases with decreasing specimen size (diameter). It appears that the notch strength of the smaller of the two specimens (½ and $1\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$16$}\right.$ in. diameter; 13 and 27 mm diameter) specified in the Tentative Method will have rather limited application as an index of K_Ic for the tougher high-strength aluminum alloys. However, the upper limit of 1.3 presently placed on the NYR appears to be overly conservative for high-strength aluminum alloys.

The second part of the paper describes the statistical analysis of correlations between the NYR and K_Ic for various lots of 2124-T851 aluminum alloy plate. The purpose of the analysis was to demonstrate how the sharp-notch cylindrical specimen could be used in a quality assurance program for high-strength aluminum alloy products based on a minimum acceptable value of K_Ic. The results indicate that the NYR from the larger of the two specimens specified in the Tentative Method provides a better correlation with K_Ic than does the NYR from the smaller specimen. A modified regression analysis is introduced which establishes tighter tolerance limits for the correlations than can be obtained using conventional procedures. The consequence is an improvement in the cost effectiveness of quality control procedures using the sharply notched cylindrical specimen. A review of existing data shows that crack orientation and product thickness can influence the correlations but that for practical purposes of quality assurance, correlations based on the T-L orientation will ensure that the minimum value of K_Ic is exceeded in all three orientations. Thickness effects can be handled by establishing separate correlations depending on whether the plate product is greater or less than 4 in. thick. Employing the modified regression analysis, a simple quality assurance plan for fracture toughness guarantee of aluminum alloy products was developed and shown to be cost effective based on available data for the aluminum alloy 2124-T851.