TWI Industrial Member Report Summary 36/1977
By M S Kamath and J D Harrison
Background
This report is a quantitative study of some aspects of the problem of ductile tearing during a fracture toughness test. Part A of the report deals mainly with compliance based crack extension measurements in relation to single specimen deltaR/JR testing. It is shown that existing theoretical relationships can underestimate the maximum tear length quite significantly. On the basis of test data covering a range of steels, thicknesses and temperatures, an empirical relation is proposed to predict the tear length from the final offloading clip-gauge compliance. The equation may be used both in conventional COD testing and in plotting COD-R curves.
In Part B of the report, JR curves have been constructed for X-65 and HY130 steels using the Sumpter and Turner analysis as the computational basis. A mean load construction procedure instead of the limit load was found to give more accurate J values, especially at increasing tear lengths. The JR curves were used to predict fracture toughness - in the case of the X-65 steel, critical defect sizes from KIc (based on Ji) and the COD design curve (based on deltai) were in agreement, while in the HY130 steel, Ji underestimated the material KIc. Similar Ji estimates on HY130 from six other laboratories showed large variance, with a straight line curve-fit for the JR curve consistently giving the lower bound Ji. Finally, the relation J = 2 sigma Ydelta overestimated the measured value of J in all cases, irrespective of the extent of tearing, if a constant rotational factor (r) of 0.45 is assumed. However, results on HY130 show that the measured rotational factor is nearer to 0.33.