By means of charging large thick specimens with hydrogen, we investigated the effects of specimen size on hydrogen-embrittlement cracking. Crack extension in a hydrogen-charged 3.5T-CT specimen extended over a longer duration than was the case for a 1.0T-CT specimen. However, the values of the lower-bound threshold stress intensity factor (KIH) for 1.0T-CT and 3.5T-CT specimens were similar to one another when these were determined with a short-term rising load (dK/dt=0.005 MPa•m1/2•s-1). We conducted numerical analysis on the hydrogen diffusion and accumulation around a crack tip, taking into consideration the hydrogen distribution in the specimen. This analysis demonstrated that the maximum hydrogen concentration for cracking can be reached under the conditions present during a short-term rising load test (dK/dt=0.005 MPa•m1/2/s). Thus, the results of the numerical analysis confirm that a minimum value of KIH equivalent to that of a heavy section steel can be obtained with a small fracture mechanics specimen. We also attempted to explain long-term crack extension characteristics, taking into consideration hydrogen dissipation from a specimen. The analysis predicts that when the mean hydrogen concentration falls below a certain level (e.g., about 1.6 ppm under certain assumptions), the value of KIH increases significantly. This increase in KIH occurs because when the necessary stress intensity factor for cracking increases as a result of a decrease in the mean hydrogen concentration, the gradient of the maximum hydrostatic stress distribution becomes moderate, especially when the applied stress intensity factor is more than about 48 MPa•m1/2. Finally, we propose a method for the prediction of the long-term crack extension behavior of a large thick specimen; the method takes into consideration the hydrogen dissipation curve and the effect on KIH of a decrease in the mean hydrogen concentration.

The life of creep crack growth for W strengthened 9-12%Cr steel is sensitive to the alloying additions and material structures, such as lath martensitic structure and grain size caused by inhomogeneous cooling rate in steel ingot during the manufacturing stage, which results in the large scattering of experimental data from the law of creep crack growth life. In this paper, creep crack growth tests were conducted using W strengthened 9-12%Cr steels with various contents of alloying additions and the dimensions of micro-nano structures. The effects of the composition of alloying additions and material structures on the life of creep crack growth for W strengthened 9-12%Cr steel were clarified.

The corrosion fatigue tests were carried out for Cr-Mo steel (SCM440) by varying the stress rising time (tR), holding time (tH) and descending time (tD). In the present paper, for the case of symmetrical stress wave (tR=tH), the constitutive equation of corrosion fatigue crack growth rate, da/dN|CF, was derived explicitly in terms of tR and tH as follows:da/dN|CF=Dda/dN|air, D=1/1-dtRetHfΔKbexp{a(lnΔK)2+c} (1)where da/dN|air is the fatigue crack growth rate in air, ΔK is the range of stress intensity factor, a, b, c, e, and f are constants. The acceleration coefficient with holding time D(=[D]tH) is this equation was derived by the following method; First, the acceleration coefficient under the condition without holding time, [D]tH=0 was obtained as:[D]tH=0=a(lnΔK)2+blnΔK+c (2)Next, under the condition with holding time, the holding time effect was estimated by the following equation, [D]tH-[D]tH=0/[D]tH=dtRetHf (3)The experimental results of corrosion fatigue crack growth rate under various stress rising and holding time conditions were found to be well expressed by eq. (1). On the other hand, this equation does not satisfy the critical condition that da/dN should be proportional to 1/tH for tH→∞, which is characteristics of stress corrosion cracking. Thus a more satisfied constitutive equation was derived as the equation of the following type: da/dN=F0(ΔK, tR, 2tR+tH). From this equation, it can be found that the corrosion fatigue crack growth rate is affected by f=1/(2tR+tH)) which concerns the time dependent effect and tR which concerns the fatigue effect under the symmetrical stress wave condition.Furthermore, it was deduced that the mechanism of the tR process is controlled by the fatigue effect coupled with the time dependent effect controlling the tH process, while the tD process is controlled mainly by the time dependent effect although it includes some fatigue effect.