• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.6 no.3
• /
• pp.53-62
• /
• 1986

For the research on the fatigue fracture behavior in the welded joints of steel structures, base metal specimens and welded ones were selected, and the direct fatigue tests were carried out. Thereafter, fatigue-life (S-N) curves, plastic strain-number of cycles (${\varepsilon}_p$-N) curve, the extrapolated fatigue-life (${\varepsilon}_p$-$N_c$) curve, and da/dN-${\Delta}K$ curves were plotted. By these results the followings were obtained. It was shown that the ratio of fatigue strength at $2{\times}10^6$ cycles of the welded specimen to that of the base metal one was 0.6, and that 0.72 for the base metal and 0.65 for the welded one were the ratio of fatigue strength at $2{\times}10^6$ cycles to yielding stress. The S-N curve for the welded specimen was separated into two sections, the low gradient section and the steep section. As this result, it was shown that the more stress became to reduce, the more the reduction of fatigue strength became to be great. It was shown that fatigue strength at $2{\times}10^6$ cycles from this case was about 83 % of that from the S-N curve plotted with one section. It was thought that the reason was that weld flaw acted greatly on the fatigue strength within the low stress range. It was shown that at the instart of crack initiation plastic strain increased abrupt1y in the case of the welded specimen more than the case of the base metal specimen, and increased abruptly in the upper stress range in both cases. It was shown that the experimental constant ${\alpha}$, 0.42, in the base metal nearly accorded with Manson-Coffin's result, but this made a great difference with the case in the welded specimen. It was thought that it was due to the abrupt change of plastic strain and the influence of weld flaw.