Original paper(Vol.45 No.3 pp.340)

Method for predicting service load from the width and height of striation

Furukawa Katsuaki; Murakami Yukitaka; Nishida Shin-ichi

Abstract:In order to predict service load from fatigue fracture surface, the use of striation spacing (s) and the ratio of striation height (H) to s, H/s is proposed. It is well known that s is correlated with the crack growth rate da/dN. However, since the crack growth curve da/dn-DK depends on the stress ratio R(=Pmin/Pmax=Kmin/Kmax), it is not possible to determine the minimum service load Pmin and the maximum load Pmax only from da/dN-DK or s-DK relationship. In the previous paper, the authors found that there existed one-to-one correspondence between H/s and the stress ratio R for 2017-T4 aluminum alloy and SUS304. Thus, the key to evaluate the stress ratio R in service load is an accurate determination of H/s. the method used in the previous paper to measure H of striations is composed of several procedures such as cutting, embedding in plastics and finally sectioning specimens by a micrometer (a special cutting machine). In this paper, an improved method of sectioning the fracture surface directly by the microtome is developed. By the new method, an increased number of accurate data of H/s become available and the histogram indicating the scatter of H/s values specific to a definite R ratio can be obtained. Three materials including 2017-T4 aluminum alloy, SUS304 and a high Mn steel were investigated to study the effect of material properties on the relationship between H/s and R. It was found that the mean value of the upper twenty data of 100 measured values of H/s was found to be a unique function of R for all three materials. This result implies, with a high degree of plausibility, the existence of one-to-one correspondence between H/s and R regardless of materials. Thus, the master curve (H/s-R relationship) obtained in this study may be used for the failure analysis of various structures of different materials.

Key Words:fatigue, fracture surface, stress ratio, service load