Historically, the influence of differences in dpa rate across small cores was perceived as an effect of tem­perature on swelling rate rather than a flux effect, primarily because it was difficult to separate the influence ofdpa, dpa rate, and temperature in limited data fields from small cores. While it was recognized for many years that there was some effect of dpa rate to determine the transient duration, until rather recently the full strength of the rate effect was underappreciated.

The new appreciation for the strong influence of dpa rate arises from two categories of studies con­ducted over the past decade. The first type involved direct single variable comparisons of the effect of dpa rate on swelling. The second category involved the examination of components irradiated at very low dpa rates and often at temperatures below the previ­ously perceived lower limit of swelling.

4.02.8.3.5.1 Category I of dpa rate effects

Three examples of the first category of dpa rate studies are presented here. The first experiment by Garner and coworkers involved the examination (density change and microscopy) of five unfueled hexagonal subassemblies constructed of a single heat of annealed AISI 304 stainless steel irradiated for many years in the reflector rows 8, 9, 10, and blanket row 14 of the EBR-II fast reactor.139,140 These com­ponents were chosen because they were made of the same steel used to construct the baffle-former-barrel assembly of PWR internals and the hexagonal sub­assemblies spanned the full range of dpa rates and temperatures found in the most swelling-vulnerable parts of the PWR baffle-former assembly.

The EBR-II experiment isolated the effect of dpa rate by concentrating on a limited range of tempera­tures (373-388 °C), but covering a very large range of dpa rates (0.06-3.8 x 10~7dpas_1), with no sig­nificant difference in He/dpa ratio. The data in Figure 52 clearly shows that the transient regime of swelling is progressively shortened as the dpa rate decreases, such that only 10 dpa is required to reach 1% swelling in row 14. In a previous publication it was shown that 30-50 dpa were required to exceed 1% swelling when data were collected at these tempera­tures from rows 2 to 4 inside the EBR-II core at higher dpa rates.141 In this experiment the swelling rates at the highest doses reached are still far from the 1% per dpa known to be a characteristic of this alloy (Figure 53).

Voids and carbide precipitates were found in all examined specimens with swelling ranging as high as 2.8%. Examples of the void microstructure and its sensitivity to dpa rate are shown in Figure 54.1 2 Universally, it was found that lower dpa rates at a given temperature increased the swelling.

The second series of experiments were reported by Okita and coworkers and involved simple model alloys, ternary Fe15Cr16Ni and quaternary Fe15Cr16Ni-0.25Ti, with very low levels of other solutes.143-145 These alloys have no possibility to be involved in segregation-induced precipitation of Ni-rich phases, so any dependence on dpa rate must involve the evolution only of voids, loops, and dislocations.

These simple austenitic alloys were irradiated in the FFTF fast reactor with actively controlled tem­peratures near 400 °C at seven different dpa rates. Measurement techniques used were density change

and microscopy. Multiple specimens were irradiated side-by-side and the measured swelling was remark­ably reproducible.

Figure 55 shows swelling for five of the seven dpa rates where there was a progressive shortening of the transient regime as the dpa rate decreased. At the lower two dpa rates (not shown here) the transient regime had decreased to less than 1 dpa. Most importantly, the steady-state swelling rate appeared to be approach­ing or to have reached 1% per dpa at all seven dpa rates. The most illuminating observation came from the microscopy, however, showing that the
microstructural feature most prominently associated with attaining the steady-state swelling rate was the loss of all Frank loops and the establishment of a glissile rather than sessile dislocation structure.

In a companion experiment the ternary Fe15Cr16Ni alloy was irradiated over a range of temperatures using nickel ions at three much higher dpa rates; it was shown that while voids can nucleate in a highly sessile microstructure, they cannot grow at a high rate.146 Most importantly, it was confirmed that increases in dpa rate led to a progressive decrease in swelling even in sessile networks.

Whereas most void swelling models focus on the rate dependence of void nucleation and growth, Okita showed by microscopy that the effect of dpa rate was most strongly manifested in the Frank loop population. High dpa rates produced a high density of loops of smaller size, while low dpa rates produced fewer loops at larger size. The latter ensemble is more prone to unfaulting, the first step toward pro­ducing a glissile microstructure. Denser ensembles at smaller sizes resisted unfaulting for a longer period. Thus the dependence of transient duration on dpa rate arose primarily from its influence on the stability against loop unfaulting.

In the third series of experiments, Budylkin prepared two experimental alloy series to be irra­diated in very similar neutron spectra in both the BOR-60 and BN-350 fast reactors at nearly identical temperatures and dpa levels.147 The first four-alloy series was Fe-16Cr-15Ni-3Mo-0.6Nb-0.6Mn-0.06C — 0.008P but varying in silicon content from 0.4 to 1.2 wt%. The second three-alloy series contained the 0.63% silicon variant from the first series and two other alloys where 0.15% titanium either was added to or replaced the 0.6% Nb.

The irradiations proceeded at 5.06 x 10~7dpas_1 and 480 °C in BOR-60 and at 1.58 x 10~6dpas_1 and 490 °C in BN-350. Thus there was approximately a factor of three difference in dpa rate. As shown in Figure 56, significantly higher swelling was uni­formly observed in the lower flux irradiation in BOR-60, regardless of alloy composition.