1974 年 23 巻 246 号 p. 208-214
In order to estimate fatigue life under long periodic loading, varying temperature, varying load or the combination of all of them, the increase of fatigue damage should be first well understood. Fatigue tests on a Nb containing austenite stainless steel (SUS 347) were conducted at various temperature levels. Some of the tests were interrupted at predetermined lapses of time before failure for the examination of crack growth. The results obtained in this study are as follows:
(1) The temperature levels can be divided into two ranges, each having its own value for the exponential β in the Manson-Coffin rule Δεp·Nfβ=C, and its own mode of stress response with respect to the number of cycles. The surface micro-crack formation and speed of macrocrack propagation in each of them are different from the other. The dividing temperature is about 650°C for SUS 347, and 300°C for 0.16%C steel.
(2) The geometrical mean of the depth of the longest surface crack dmax and the sum total of the depths of all the micro-cracks in the gauge length D are equal to the depth of equivalent crack of which area is equal to the sum total area of transversal sectional areas of micro-cracks per unit gauge length. So it can be the most realistic parameter representing the amount of damage.