Initiation and Arrest of Delayed Hydride Cracking in Zr–2.5Nb Tubes

Abstract

Using Kim’s delayed hydride cracking (DHC) model, this study reanalyzes the critical temperatures for DHC initiation and arrest in Zr–2.5Nb tubes that had previously been investigated with the previous DHC models. At the test temperatures above 180°C, DHC initiation temperatures fell near the terminal solid solubility for precipitation temperatures, requiring some undercooling or ΔT from the terminal solid solubility for dissolution (TSSD) temperatures, and increased toward TSSD with the number of thermal cycles. At the test temperatures below 180°C, DHC initiation occurred at temperatures near TSSD with little ΔT. DHC arrest occurred on heating toward TSSD where the hydrogen concentration difference between the bulk region and a crack tip ΔC decreased to a minimum ΔCmin, under which nucleation of the hydrides was restrained. ΔCmin after the first thermal cycle increased with increasing temperature, demonstrating that nucleation of the hydrides becomes more difficult with increasing temperatures. Different DHC initiation and arrest temperatures with the test temperatures or hydrogen concentrations are discussed in view of a supersaturation of hydrogen (ΔC) for nucleation of hydrides in the zirconium matrix.

Hydride Reorientation and Delayed Hydride Cracking of Spent Fuel Rods in Dry Storage

“…A simple calculation assuming the growth of a crack to be 0.9 times the thickness over 2 years shows that the crack growth rate was at the least 5.9 9 10 À12 m/s, [25] which is over 10 4 times faster than the predicted crack growth rate, 1.6 9 10 À16 m/s, [7] by the old DHC models. In contrast, Kim [25] has shown that DHC at low temperatures below 180°C occurs due to stress-induced hydride phase transformation from c to d at a crack tip, creating a difference in hydrogen concentration between the bulk and a crack tip due to a higher hydrogen solubility of the c-hydride, which is a driving force for DHC, using Kim's DHC model [5,14,18,20,21] and the experimental facts [26][27][28][29] reported, as shown in Figure 11. Given that the c-hydride is formed at low temperatures below 180°C for the slow-cooled zirconium alloys [30] such as spent fuel rods, it is not until the spent fuel rods are cooled to below 180°C that DHC occurs due to stress-induced hydride phase transformation from c to d. According to Einziger's model, [31] it is predicted to take at least 30 years for cladding temperature to completely fall to below 180°C, as shown in Figure 12.…”

“…Figure 8 shows the DHC initiation temperatures with hydrogen concentration of the Zr-2.5Nb specimens that have been determined only after the first thermal cycle. [18] Under stresses, DHC initiation occurred at the TSSP2 or higher temperatures, not at the TSSP1 temperatures. Given the results of Figure 8, nucleation of hydrides that is the first step of DHC must have occurred at higher temperatures prior to reaching the DHC initiation temperatures due to stresses.…”

“…Note that enhanced nucleation of many of the larger radial hydrides due to creep deformation, as shown in Figures 9 and 10, is caused by a shift of the TSSP1 temperature close to or above the TSSD temperature due to the stress: Were it not for the stress effect on the shift of the TSSP1, precipitation of the radially reoriented hydrides would be the same irrespective of the peak temperature and the stress. Regarding precipitation of the radial hydrides only at a crack tip after cooling to 250°C, there is a controversy between Kim's DHC model [5,14,18,20,21] and the old DHC models: [17,[22][23][24] the former says that the stress-induced precipitation of hydrides is the cause of nucleation of radial hydrides, reducing the hydrogen solubility at a crack tip to TSSD, while the latter has assumed that the reoriented hydrides precipitate when stress-enhanced diffusion of hydrogen increases the crack tip hydrogen solubility to the terminal solid solubility for hydride nucleation (TSSP1) or 64 ppm H from 60 ppm H. [23,24] However, if the latter's assumption were true, the reoriented hydrides could not precipitate irrespective of the loading time, because the TSSP1 at any temperatures between 250°C and 380°C is higher than 60 ppm H, resulting in no precipitation of hydrides on cooling to 250°C. Thus, just a few radial hydrides, as shown in Figure 5(c), would be seen only at the crack tip due to the so-called stress-enhanced diffusion of hydrogen irrespective of the loading time, which is in contrast with the results of Figures 5 and 9, where the distribution and size of the radial hydrides is seen to be quite different depending on the loading time and the peak temperature.…”

“…Consequently, it is clear that swelling due to UO 2 oxidation applies tensile stresses on the defective fuel rods, causing crack growth by DHC and hence long axial cracks in the failed spent fuel rods, as observed in failed boiling water reactor fuel rods. [37] Considering that no DHC occurs in isothermal conditions at high temperatures above 180°C, [14,18,20,21,25,38] the long axial cracks by DHC in the failed spent fuel rods will occur only after cooling to below 180°C in dry storage.…”

“…Using Shi's data, [17] the DHC initiation temperatures were determined for the Zr-2.5Nb specimens with the hydrogen concentration of 7 to 86 ppm H [18] that were heated without a load to about 50°C above the TSSD temperature, cooled at 2°C/min to a temperature of 20°C above the TSSD, and then loaded to a stress intensity factor of 12 or 15 MPaÖm followed by stepwise cooling down to the TSSP1 test temperature. Figure 8 shows the DHC initiation temperatures with hydrogen concentration of the Zr-2.5Nb specimens that have been determined only after the first thermal cycle.…”