Influence of structure on the resistance of 65G steel to fatigue crack growth

Abstract

(230-mm diameter) with a center crack, rectangular (width 25 mm) with an edge crack for axial tensile testing, and rectangular compact (width 81 mm) for off-center tensile testing~ The line of action of the forces always coincided with the rolling direction and the through crack developed perpendicular to it. To obtain different structures, the disk samples were normalized or hardened in oil from 810~ and tempered at 180, 320, 380, 440 and 5800C. After heat treatment the samples were ground to a thickness of 3 mm. Data on the microstructure and the mechanical properties in short-term static tension of standard flat samples is shown on Table i. To study crack growth in steel of a given structure from one to four samples were used and in individual cases up to seven rectangular samples, which were cut from test disks. The tests were made using the known method [2, 3] in a machine with an inertial load exciter [4] in laboratory air with a relative humidity of 40-60% using an almost from-zero (R = 0.05) sinusoidal tensile cycle with a loading frequency of 16-20 Hz and at high crack growth rate (more than i0-6 m/cycle), because of the difficulty in measuring its length, also at a frequency of 0.017 Hz on a machine with rigid loading [5]. In the i0"-i0 -" m/cycles range of rates the tests were repeated at the two frequencies in order to be convinced of their weak influence on the results. The rate of development of the original crack in growth of it to an initial length of 2 mm from an initiating notch was always greater than 10-9 m/cycle despite the low loads and the use of various methods easing origin of it [6]. Therefore, for investigations of crack resistance at lower rates the load was reduced in steps with observance of existing recommendations [2, 7]. The sample was tested in stages during which all parameters of the load cycle were maintained constant. In each stage the range of the stress intensity factor ~K = KmaX-Kminin the disk samples was constant if the crack length did not exceed 95 mm and in the remalning cases increased with crack length. The rate of crack development v in the stages with a constant AK was calculated from five to eight successive measurements of its length using the method of least squares [2] while with an increasing AK the value of v was determined as the ratio of the increase in the crack between two successive measurements to the corresondlng number of cycles. For determination of the characteristics of cyclic crack resistance and an analytical description of the full fatigue failure curves, use was made of the equation [8]