With respect to the irradiation environment there are four major variables that determine the duration of the transient. The first three are related to each other: irradiation temperature, temperature gradi­ents, and temperature history. The fourth is strongly synergistic with temperature and is the dpa rate, which will be covered in the following section.

Some temperature histories, especially when gradually falling from one temperature to a lower temperature, produce a shorter transient compared to that of either the starting or final temperature, primarily because such histories tend to accelerate the radiation-induced formation of nickel and silicon-rich phases, especially that of the g phase.1’136 Formation of these phases usually precedes swelling.1

Strong gradients in temperature across thin fuel cladding have also been shown to accelerate the onset of swelling, producing more swelling than what iso­thermal irradiation would produce at either the upper or lower cladding temperature.137,138 The exact cause is unknown but it was speculated that the stress gradients associated with strong tempera­ture gradients might be a contributing factor.

For isothermal irradiation the temperature is an important determinant of the transient duration, not only because it directly impacts diffusion and void nucleation, but also because of its influence on phase stability and phase evolution. However, over the wide range of temperatures experienced in fast reactors, temperature has no effect on the posttransient steady-state swelling rate of 300 series stainless steels at ~1% per dpa.

However, it is frequently assumed that at constant dpa rate there is a peak swelling temperature or peak swelling rate as a function of temperature for swelling of austenitic steels. This persistent misperception is a consequence of the historical use of fast reactors. All ofthe early data on swelling was derived from small fast reactor cores such as EBR-II and DFR, which have strongly peaked dpa rate profiles, both axially and radially. Later studies conducted in larger cores such as that of FFTF showed that assuming a temperature — dependent steady-state swelling rate was incorrect. More careful analyses of other data from these smaller cores also support this point of view.

Neutron irradiation of RPV steels results in an in­crease in the yield stress (sY) and an upward shift in the DBTT. The increase in the DBTT is generally determined from the shift in the Charpy impact curve at a specific energy level, typically 41J (AT41j). Historically, most data on effects of irradiation on the DBTT arose from irradiation and testing of Charpy specimens, although frequently increases in yield stress were also reported. The specimen geometries in both cases were convenient for irradiations in the restricted space available for either surveillance capsules1 or rigs inserted into the cores of material test reactors (MTRs).1 More recently, with the advent of the Master Curve tech — nique,16 there has been greater interest in acquiring

data directly on fracture toughness (see, e. g., Viehrig and Lucon17) (see also Chapter 4.14, Fracture Tough­ness Master Curve of bcc Steels). Correlations have been established between the different measures; for example, see Sokolov and Nanstad18 for experimental data on the relationship between Charpy and static fracture toughness shifts and between Charpy and yield strength shifts. Williams et a/.19 have published hardness/Charpy shift correlations for A533B plate and weld.

From the early test programs in the 1960s to the present day, great reliance has been placed on measur­ing the shift in the Charpy transition from lower energy to the upper shelf energy (USE) as a means of determining the effect of irradiation damage on bulk mechanical properties. Figure 1 shows an example for an RPV forging; it can be seen that not only is there a shift in the transition temperature (usually measured at the 41 j level), but there is also a drop in the upper shelf. Note that the decrease in the USE is accompa­nied by a decrease in the slope of the Charpy curve in the transition region; this can create issues with esti­mating the Charpy shift if the upper shelf level approaches the indexing level. The greatest attention has been given to the shift in the transition region.

aIn the LWR, community fluence (or equivalently dose) is generally expressed in terms of neutron exposure (e. g., ncmT2, E> 1 MeV); this reflects the fact that the neutron spectrum does not vary significantly from location to location or plant to plant. In contrast, in the United Kingdom a key feature of Magnox reactors is that the neutron spectrum varies significantly with location, and displacements per atom, dpa, is a more appropriate measure.

The change in mechanical properties is required as a function of irradiation fluence (dose), dose rate (flux), irradiation temperature, steel type and composition, product form (plate, forging, or weld), and material heat treatment. Insight into the ranges of the irradia­tion variables of interest can be seen from Table 4. At the design stage of the early reactors, an allowance for the potential irradiation-induced embrittlement was included. The magnitude of this allowance was based on the judgment of potential irradiation effects. At the time of the design of Magnox reactors, the shift allowance for SMA welds was ^40 °C.20 Such allow­ances were frequently exceeded once data relevant to plant conditions became available.

The intent of this section is not to provide an exhaustive description of the exact mechanical prop­erty response of individual steels to specific irradiation conditions; rather the intent is to identify the main parameters that control the radiation response of fer­ritic pressure vessel steels and to give an indication of the potential changes in mechanical properties that may occur through typical in-service conditions and lifetimes.

The fluence dependence of RPV embrittlement has been studied since experimental programs were initiated in the 1950s. Most data available refer to the case in which embrittlement is determined by irradiation-induced hardening (rather than nonhar­dening embrittlement — see Section 4.05.4) at tem­peratures greater than 150 °C and less than 300 °C. Indeed, the available data are dominated by experi­ments that investigated the embrittlement of MnMoNi steels at an irradiation temperature of 270-295 °C. It should also be recognized that after over 50 years of experimentation, a significant quan­tity of data has been obtained. For example, the US surveillance database arising from BWR and PWR reactors now encompasses ^800 individual data points on the embrittlement observed from various steels irradiated at a range of fluences (and a restricted range of irradiation temperatures).21,22 Data on the results of French surveillance programs have also been published.23 In parallel, well- controlled experiments in MTRs, frequently making use of steels with well-controlled compositional varia­tions, have been performed. These have provided incredibly valuable information that has helped develop an understanding of the radiation damage processes in RPV steels. Indeed, early irradiation pro­grams by Odette and coworkers,24 and Hawthorne and coworkers,25 focused on a restricted fluence range but on a number of steels with well-controlled composi­tional variation. Other notable programs were the irra­diations performed as part of the Heavy-Section Steel Irradiation (HSSI) program at Oak Ridge (see, e. g., Taboda etal26 and Nanstad and Bergen27) and a num­ber of IAEA coordinated programs (see, e. g., Interna­tional Atomic Energy Agency28). The most recent program by Odette and coworkers at the Ford Nuclear Reactor, University of Michigan, focuses on a range of fluxes, fluences, and material compositions. The resulting IVAR irradiation database encompasses 57 alloys that were irradiated at three different fluxes and three different temperatures over a range of flu­ences, giving in total 1537 different alloy/irradiation conditions. Irradiations were at 270, 290, and 310°C, fluxes between 5 x 1010 and 1 x 1012 ncm~2s_1, and fluences between 0.004 x 1019 and 4 x 1019ncm~2, E > 1 MeV (see Heatherly et al.15 and Eason et al29). The irradiation temperature of the irradiation rig was restrained to <±5 °C and the uncertainty in fluence was ~±7%. (It should be noted that there are well — defined techniques for establishing the neutron spectra in specific locations in reactors and for evaluating the resultant damage in the irradiated material; see Heatherly etal.15 and ASTM Standard 185 on ‘Practice for Conducting Surveillance Tests for Light-Water Cooled Nuclear Power Reactor Vessels’ and related ASTM Standards for a discussion of these techniques.) It is to be noted that typical MTR dose rates are x 1012—1 x 1013ncm~2, E > 1 MeV s-1, that is, one or two orders of magnitude higher than the PWR surveillance dose rates given in Table 4.

Overall, it was found in all RPV steels of interest that embrittlement increases with increasing fluence, but the rate of embrittlement may decrease (with increasing fluence). Furthermore, embrittlement does not saturate in the fluence range of interest to power reactor applications. These trends are illustrated in Figure 2 for Magnox CMn steel SMA weld transition shifts and for a MnMoNi plate HSST-02.

Irradiations in the 1960s demonstrated that com­position was a major factor controlling the response of the ferritic low alloy steels employed in operating reactors. By the mid-1960s, it was thought likely that residual elements in steels could be responsible for much of the observed scatter in the irradiation embrittlement response.3 , The work suggested that reducing the residual element content of A302-B steel would markedly improve the resistance

to irradiation embrittlement at typical service tem­peratures of 550 oF (288 oC).

In a pioneering set of studies, Hawthorne and coworkers undertook studies that explored the effect of material composition on irradiation embrittlement in a systematic manner. Studies were undertaken on materials irradiated at a controlled temperature of 288 °C in the Union Carbide Research Reactor (UCRR) and in the light water-cooled and moder­ated test reactor, UBR, at the Buffalo Materials Research Center.25 A series of small (150 kg) labora­tory melts were produced to the nominal plate steel specifications using pure elements. These were then split, generally into three blocks, two of which were remelted and selected residual element additions added, while the third was kept to provide a low residual element reference. Each steel was also com­pared to material obtained from normal commercial production.

Potapovs and Hawthorne33 demonstrated that additions of Cu, and Cu and Ni, to a laboratory melt containing low level of residuals greatly increased the observed embrittlement (see Table 5). This must be regarded as a landmark paper in the understanding of the factors that control radiation damage in RPV steels. Note that it was ^15 years before the underlying mechanisms were elucidated (see Section 4.05.4). The effect of different Cu and Ni levels in steels irradiated as part of US surveil­lance programs is illustrated in Figure 3. The effect of increased levels of Cu in steels of the same Ni level and the effect of increased levels of Ni at constant Cu level is clear (data taken from Eason et a/.33).

0

0 1X1019 2X1019 3X1019 4X1019 5X1019 6X1019 7X1019

(b) Fluence n cm-2 (E >1MeV

Figure 3 Charpy shift (AT41 J (°C)) for (a) a US weld and a US forging containing 0.25 and 0.06 wt% Cu, respectively, and (b) US welds and a US plate containing ~0.2 wt% Cu and varying levels of Ni.

observation from Figure 4 is that tin additions (0.023% vs. <0.004% Sn) to the high phosphorus alloy did not affect embrittlement. They further established that an arsenic content of 0.035% does not have an observable effect on the radiation sensi­tivity of plates containing ^0.020% P and <0.18% Cu. Alloying with 0.50% chromium also did not alter the radiation resistance of plates having a copper content of ^0.30% Cu.

In the United Kingdom, there was considerable interest in the effect of irradiation temperature because of the wide variation of temperatures in Magnox pressure vessels. Barton et al34 reported the irradiation temperature, the dose rate, and the dependencies of the yield stress increase for CMn steels following irradiation in the PLUTO MTR. The work focused on EN2, a Si-killed mild steel and an Al-killed grain-size-controlled mild steel. The Cu contents of the steels were 0.14, 0.18, and

0.13%, respectively, by weight. Irradiation tempera­tures were reported to have been controlled to ±1 °C34 and the fast neutron doses were restricted to a maximum of 2.5 x 1017ncm-2 (as measured by Ni monitors). The results between 100 and 350 °C exhibited a simple linear dependence of yield strength increase as a function of irradiation temper­ature. Jones and Williams35 carried out an analysis of the Barton data34 and another dataset on the irradia­tion temperature dependence of similar steels from Grounes,36 and pointed out that the combined data form a homogeneous data set with relatively little variability. The least squares regression is

FT(As) = 1.869 — 4.57 x 10-3 T [1]

This parameter has been important for the correla­tion of data from similar steels irradiated at different irradiation temperatures.2

Understanding the effect of flux, or dose rate, on radiation damage of ferritic steels has proved particularly important in the formulation of mecha­nistically derived DDRs. There was particularly strong interest in the United Kingdom in under­standing the effects of flux on bulk properties as Magnox RPVs operated within a range of fluxes.2 A number of experimental investigations have examined the dose rate dependence of hardening.37 In the absence of precipitation effects, no influ­ence of dose rate on irradiation hardening has been detected. Data obtained from over five orders of magnitude change in dose rate for C-Mn plate steels at an irradiation temperature of 200 °C, typical of Magnox applications,37 demonstrated no dependence on dose rate.

However, there is agreement that there is a strong effect of flux on the embrittlement of Cu-containing steels.38 In these steels, it was found that at doses before the ‘saturation’ of embrittlement (see Section 4.05.4) the rate of embrittlement with fluence increased with decreasing dose rate. Williams etal. studied the effect of dose rate on the embrittlement in low Ni welds at preplateau doses19 (1 x 1019ncm-2, E> 1MeV is approximately 1.5 x 10-2dpa). They reported the irradiation-induced shift in the 41 J transition temper­ature of a number of Mn-Mo-Ni SMA welds after irradiation in MTRs at dose rates between 6 x 10-10 and 2 x 10-8dpas-1 and doses of less than ^30mdpa. It was observed that for the welds SD and SL, in which the Cu levels are low (<0.15 wt%), there is no apparent effect of dose rate (Figure 5). At higher copper levels (0.56 wt% for weld SH, 0.36% for SG, and 0.24% for SF), there was a marked effect of dose rate at low

Figure 5 Comparison between results obtained at different dose rates and in different irradiations of quenched and tempered low alloy welds (DIDO, HERALD, and OSIRIS are all MTRs in which samples were irradiated). Reproduced from Williams, T. J.; Ellis, D.; Swan, D. I.; etal. In Proceedings of the 2nd International Symposium on Environmental Degradation of Materials in Nuclear Power Systems — Water Reactors, Sept 1985; ANS: Monterey,

CA, 1986.

doses (<0.01 dpa), where low dose rate produced a significantly higher shift (Figure 5). At higher doses, no such effect was observed.

In studies where both Charpy specimens and ten­sile samples were irradiated, it was found that radia­tion damage in RPV steels generally caused an increase in both Charpy shift and yield strength. This is significant, as irradiation can also lower the fracture strength of materials (see Section 4.05.4) and cause nonhardening embrittlement. As will be shown, none of the DDRs developed for MnMoNi steels have found it necessary to account for the embrittlement due to nonhardening embrittlement. It is only in the case of CMn steels that evidence has been found for nonhardening embrittlement in irradiated SMA welds.2 There have been investiga­tions of nonhardening embrittlement in experiments mounted in MTRs; for example, McElroy eta/.39 and Nanstad eta/.40 investigated nonhardening embrittle­ment of simulated coarse-grained heat-affected zones (CGHAZ). A review of intergranular embrittlement in RPV steels can be found in English et a/.41

4.05.3.1 Summary

In contrast to the assumptions made at the design stage, it was found that radiation damage in ferritic RPV steels could cause a significant change in bulk properties, primarily an increase in DBTT (as man­ifested by an increase in Charpy 41 J level) and an increase in yield strength. Evidence of radiation

damage causing nonhardening embrittlement was also found in CMn steels.

There are proven mechanical test techniques for determining the change in bulk properties and large irradiation programs have been performed. These have made use of both surveillance programs and MTRs. The level of the measured embrittlement depends on the fluence, flux, and irradiation temper­ature. The most important discovery was the sensi­tivity to steel composition, in particular a strong dependence of embrittlement on the levels of Cu, Ni, and P.

4.07.4.1 Elastic Modulus of Monolithic SiC

Figure 13 summarizes the irradiation temperature dependence of the elastic modulus change. Irradia­tion generally reduces modulus to a greater extent for lower temperature irradiation. The modulus reduction becomes negligible when irradiation temperature reaches or exceeds 1273 K. There seems to be an indistinct stage between 1073 and 1273 K. As expected, the elastic modulus trends with ‘point defect swelling’ of SiC. Point defect swelling is an isotropic volume

expansion that is believed to occur by lattice relaxation due to accumulated isolated point defects and small point defect clusters during irradiation at tempera­tures where vacancies are not readily mobile. In SiC irradiated in the point defect swelling regime, a fairly good agreement between dimensional expansion and lattice spacing has been confirmed by X-ray diffrac- tometry studies. In contrast, the data in the nonsatur­able swelling regime is somewhat limited, although the data suggest that there is little reduction in elastic modulus in spite of the swelling being relatively large. However, in this regime, the defects responsible for swelling are voids and other relatively large defects, which would have less of an effect on elastic modulus as compared to point defects.

An estimation of the influence of lattice relaxa­tion on elastic modulus was attempted using the Tersoff potential.54 The result predicted a linear lattice
swelling of 1% causing approximately 10% reduction in elastic modulus (Figure 14). The predicted elastic modulus change could be varied by more than 10% depending on a selection of interatomic potential, with the Tersoff potential giving a relatively high sensitivity of modulus to the interatomic distance among various empirical interatomic potential functions. Therefore, the measured elastic modulus changes observed in this experiment are generally greater than the theoreti­cal prediction. It is noted that the methods applied for generating the data of Figure 14 are various and of differing quality. Typically, elastic modulus as measured by nanoindentation, which sometimes is the only alternative for miniature specimens, tends to give widely scattered and less reliable data than the mechan­ical or sonic modulus methods. Nonetheless, it is clear that the lattice expansion is a major cause of the irradiation-induced elastic modulus reduction in SiC.

‘Hot cracking’ can also be primarily mechanical in nature; the restraint, constraint, and geometry of the weld act to pull apart the weld metal at temperatures near the solidus. This type of cracking may be trans­granular or interdendritic and is favored by mechan­ical notches and partial penetration weld joints. Note that the hot tear shown in Figure 2 (bottom) was the only crack present in that multipass weld, illustrating the dominant effect of the notch at the weld root.

Liquid at the time of straining

….. 1…………

Solid at the time of straining

Solidification

■ r і

cracks

——————

Figure 6 Illustration of solidification-versus liquation-type cracking in a transvarestraint sample of Alloy 625 tested at 10% strain. Solidification cracks (bottom right) form on-cooling in the mushy zone behind the solid/liquid interface. Liquation cracks (upper left) are in the partially melted zone (PMZ) and/or heat-affected zone adjacent to the autogenous weld bead.

The rate of change of a material property can be related to displacement rate of carbon atoms (dpa s— ). However, it is not possible to directly measure dpas—1 in graphite, but dpas—1 can be related to the reactor flux. The flux depends on reactor design, and varies with position in the reactor core.

Neutron flux is a measure of the neutron popula­tion and speed in a reactor. In a reactor, neutrons move at a variety of speeds in randomly orientated directions. Neutron flux is defined as the product of the number of neutrons per unit volume moving at a given speed, as given by eqn [6] below.

number number /cm

f ——— — = « ————- 3- v — [6]

cm2 s cm3 s

Table 5 Ratio of graphite damage to nickel flux as measured in PLUTO
However, as there is a spectrum of neutrons, with many velocities, this is not a useful unit for the material scientist. Therefore, integrated flux is used over a range of energies E1 to E2 as given by eqn [7].
Position Ratio fd/fNi C4 — inside fuel element stainless steel thimble 0.518 D3 — inside fuel element stainless steel thimble 0.468 C4 — inside fuel element aluminum thimble 0.507 D4 — empty fuel element 0.564
f	[7]

Modified from Bell, J.; Bridge, H.; Cottrell, A.; Greenough, G.; Reynolds, W.; Simmons, J. Philos. Trans. R. Soc. Lond. A Math. Phys. Sci. 1962, 254(1043), 361-395.

In this way, a measure of neutron damage at any position within a structural component can be defined as follows. For a material such as graphite,

the damaging power (displacement rate), f d, can be expressed as an integrated flux as given in eqn [8].	energy E1 to produce a recoil atom with energy E2, and v(E2) is a ‘damage function’ giving the number of atoms displaced from their lattice site by recoil energy E2. The carbon displacement rate, f ds, at a standard position in DIDO is 5.25 x 10-8 dpa. The derivation of the damage function (Figure 7) is on the basis of billiard ball mechanics, energy losses to the lattice due to impacts, and to forces associated with excitation of the lattice. The early Kinchin and Pease17 form of the dam­age function was found to underestimate damage in graphite. To give greater dpa, it was recommended that ‘Lc’ was artificially increased, but this was not satisfactory. The Thompson and Wright18 damage function was used in the official definition of EDNF. However, the Norgett eta/.19 damage function is used in most modern reactor physics codes and it has been recently shown that there is little difference in the calculation of graphite damage using either of these latter two functions.20,21 It is assumed that the ratio of dpa to nickel flux (fds/fs) at the standard position, which is equal to 1313 x 10-24 dpa (n cm-2 s-1)-1, can be equated to the same ratio fd/fNi in the reactor of interest as given by eqn [11]:
1 C(E)f(E)dE 0
fd	[8]
where f (E) is the neutron flux with energies from E to E + dE and C(E) is a function to describe the ability of neutrons to displace carbon atoms.
4.11.5.4.1 DIDO equivalent flux At the standard position in a hollow fuel element, the nickel flux, fs, can be defined by eqn [9].
f	[9]
0
1 where J f s(E)sNi(E)dE is the integral of neutron 0 flux multiplied by the nickel cross-section at the standard position in DIDO, and S0 is the average nickel cross-section for energies > 1 MeV, which is equal to 0.107 barn. The value of fs at this position is 4 x 1013 n cm-2 s-1. The carbon displacement rate can be calculated using eqn [10].
1313 x 10 24dpa(ncm 2s 1) 1 [11]
f(E1)s(E1, E2)v(E2)dE1dE2
[10]
where f (E1) is the flux of neutrons with energy E1, This value was derived using the Thompson and s(E1, E2) is the cross-section for a neutron with Wright damage function and an early flux spectrum
Figure 7 Comparison of various damage function models that describe the number of displaced atoms versus energy of primary knock-on atom.

for the standard position in DIDO. Hence, the EDNF or f d can be calculated at the position of interest. The equivalent DIDO nickel dose (fluence) (EDND) is derived by integrating EDNF over time, as given in eqn [12]:

fd(t )dt

0

Table 6 compares the calculated and measured graphite damage rates in various systems using the Thompson and Wright model.

Finally, for those wishing to try and reproduce damage in graphite using ion beams, Table 7 gives the energies, cross-sections, and mean number of displacement for various particles.

4.11.5.3 Energy Above 0.18 MeV

Dahl and Yoshikawa22 noted that for energies above 0.065 MeV, eqn [13] was reasonably independent of reactor spectrum under consideration:

f(E )ff(E)v(E)dE

f(£>£0 = 1 [13]

f(E)dE

E1

Equation [13] is the integral of graphite displace­ment for a position in the particular reactor of interest, divided by the integral of flux from E1 (0.065 MeV in this case) to infinity at the same position. Table 8 gives this ratio for two other values of E1.

(E)f(E, t)dE

fG = 1 — 1 [14]

(E)w(E)dE/ w(E)dE

0 d 0

Equation [14] is essentially graphite dpa divided by a normalized fission flux. A similar unit is defined by Simmons4 in his book. However, the use of this unit was never taken up for general use.

4.11.5.7 Fluence Conversion Factors

Table 9 gives the conversion factor from other units to EDND. The following should be noted:

• EDND is a definition,

• Calder equivalent dose and other units relating damage to fuel ratings are approximate,

• BEPO equivalent dose is a thermal unit and should be avoided,

• Energies above En are a good approximation,

• dpa is directly proportional to EDND.

4.11.5.8 Irradiation Annealing and EDT

The reasoning behind the use of equivalent DIDO temperature (EDT) is that if two specimens are irra­diated to the same fluence over two different time periods, the specimen irradiated faster will contain the most irradiation damage. The reasoning is that the spec­imen irradiated at the slower rate would have a longer time available to allow for ‘annealing’ out of defects caused by fast neutron damage as outlined below.

The rate of accumulation of damage dC/dt can be described by eqn [15].

where ф is the flux, E is the activation energy, T is temperature (K), and k is Boltzmann’s constant. Equating the damage rate for two identical samples at different flux levels ф1 and ф2 and different tem­peratures T1 and T2,

ф1 = ф2 4-щ) e4-щ)

Rearranging this gives the EDT relationship:

1___ 1 = k і (ф)2

01 02 E n ф )1

The term on the left is the difference in the recipro­cal of the temperatures in the two systems (tempera­ture has traditionally been given the symbol ‘0 ’ when used in this context) and the term on the right con­tains the natural log ofthe damage flux (or fluence) in the two systems divided by each other. In practice, the activation energy E is an empirical constant.

The use of EDT has recently been investigated24 at temperatures above 300 °C. The authors concluded that the use of EDT was inappropriate (Figure 8). However, below 300 °C, there was some evidence of the applicability,15 but at these lower temperatures there is little reliable data. Therefore, the use of the EDT concept is not recommended for modern graph­ite moderated reactors where the graphite is usually irradiated above 300 °C.

In the United Kingdom, to extend the creep law to higher fluence and to account for radiolytic oxidation, the following approach was taken by the UKAEA. Creep strain, decr/dg, was assumed to be defined by

"df = a(-T)Y texp(~bg)] + b(r )Ef N

where a(T), b(T), and b are temperature-dependent functions equal to 1.0, 0.23, and 4.0, respectively in the AGR and Magnox temperature ranges, and a and g are stress and irradiation fluence, respectively. The need for the temperature dependence outside this range was defined by data for HTRs obtained in the United States and Russia (Figure 63).

The ‘creep modulus’ in the UKAEA model was defined as

Ec = EoSE{ox [56]

where E0 is the unirradiated SYM and S is the irradi­ation temperature — and fluence-dependent struc­ture term derived from the irradiated modulus data (Section 4.11.17.4). To account for radiolytic weight loss, E[ox] is a modulus weight loss term defined as E/E0 = exp(—lx) where l is an empirical constant equal to about 4.0 and x is the fractional weight loss.

There are no rigorous observational data to underpin this model other than a few data points for preoxidized graphite given in Figure 64. It should be noted that in Figure 64, the only two

Longitudinal creep strain (%)

(b) Longitudinal creep strain (%)

1E-24

о c о (D

Russian graphite American graphite (EGCR) American graphite (CGB)

1E-25

100 200 300 400 500 600 Temperature (^C)

700 800 900 1000

samples with a significant amount of weight loss are irradiated to a relatively low fluence. Similarly, there are some, even less convincing, data on creep sam­ples initially irradiated to high fast neuron fluence before loading.92

The main criticisms of the UKAEA creep model, for inert conditions, is that it gives a very poor fit to the high fluence creep data obtained in Germany and the United States, as discussed in the next section.

4.11.20.6.1 German and US creep model

This model was devised for helium-cooled HTR applications where radiolytic oxidation was of no concern. The form of some of the US and German data is given below (Figure 65). There appears to be a difference between tension and compression at high fluence. However, this is the only data that shows this and it is not clear if it is a real effect.

It was assumed that microstructural changes at medium to high fluence would modify the creep rate and account for the shape of these curves. It was assumed that this could be accounted for by modifying the secondary creep coefficient in the UK creep law by the following expression:

/ s A

ecr(secondary) K

where K is the UK creep coefficient, AV/V0 is the change in volume (which is a function of fluence and
irradiation temperature), (AV/V0)m is the volume change at volumetric ‘turnaround,’ and m is a graphite grade and temperature-dependent variable.

It is widely agreed77,100 that the irradiation-induced hardening in zirconium alloys results mainly, as for many other metals, in the creation of a high density of small point-defect clusters that act as obstacles for dislocation glide. As described earlier, the point — defect clusters in zirconium alloys consist mainly of small prismatic loops, with Burgers vector lying in the (a) direction and the habit plane close to the prismatic plane of the hcp crystal lattice. Several authors have discussed that dislocations interact with irradiation-induced dislocation loops through their long-range stress field106,107 and also through contact interactions, which can lead to junction creation that are strong obstacles to dislocation motion.108-110 Several authors have investigated in more detail the junction formation between dis­locations and loops in zirconium alloys. Particularly, Carpenter111 has considered the mechanism

proposed by Foreman and Sharp109 and he applied it to the prismatic glide in zirconium alloys. He has shown that an edge dislocation gliding in the pris­matic plane that is pinned by a loop can annihilate the loop. More recently, it has been discussed that the junctions between the loops and the dislocations gliding in the basal plane are always glissile, whereas they are sessile when the dislocations glide in the prismatic plane.1 , This phenomenon could then

lead to a lower hardening of the basal slip system compared to the other slip systems. Lately, MD com — putations114 have been undertaken in order to gain a better understanding of the interaction mechanisms between dislocation and loops in zirconium alloys. It is shown that all the slip systems are not affected in the same way by the presence of the (a) type loops, the basal slip system being less hardened than the prismatic slip system, for instance.

The advanced ferritic and ferritic-martensitic steels of current interest have evolved5 from their prede­cessors, the creep-resistant ferritic steels, over nearly a century. The first of the series was the carbon and C-Mn steels with a limited application to about 523 K. Subsequent developments through different levels of chromium, molybdenum have increased the high temperature limit to 873, leading to the current ferritic and ferritic-martensitic steels, that is, the 9-12% Cr-Mo steels. In addition to being economi­cally attractive, easy control of microstructure using simple heat treatments is possible in this family of steels, resulting in desired mechanical properties.

The propensity to retain different forms of bcc ferrite, that is, ferrite or martensite or a mixture at room temperature in Cr-Mo steels, depends crucially on the alloying elements. Extent of the phase field traversed by an alloy on heating also depends on the amount of chromium, silicon, molybdenum, vana­dium, and carbon in the steel. The combined effect of all the elements can be represented by the net chromium equivalent, based on the effect of the aus­tenite and ferrite stabilizing elements. A typical pseu­dobinary phase diagram6 is shown in Figure 1(a). Increase in chromium equivalent by addition offerrite stabilizers or V or Nb would shift the Fe-9Cr alloy into the duplex phase field at the normalizing temperature. The phase field at the normalizing tem­perature and the decomposition mode7-9 of high temperature austenite (Figure 1(b)) dictate the result­ing microstructure at room temperature and hence, the type of steel. Accordingly, the 9CrMo family of steels can either be martensitic (9Cr-1Mo (EM10) or stabi­lized 9Cr-1MoVNb (T91)), ferritic (12Cr-1MoVW (HT9)) or ferritic-martensitic (9Cr-2Mo-V-Nb (EM12)) steel. The stabilized variety of 9-12 CrMo steels could result10 in improved strength and delayed grain coarsening due to the uniform distribution of fine niobium or vanadium carbides or carbonitrides.

The transformation temperatures and the kinetics of phase transformations depend strongly on the composition of the steels. Sixteen different 9Cr steels have been studied11,12 and the results, which provide the required thermodynamic database are shown in Figure 2, with respect to the dependence of melting point, Ms temperature and the continuous heating transformation diagrams. The constitution and the kinetics of transformations dictate microstructure and the properties.

In the early stages, the oxidation resistance and creep strength were of prime importance, since the Cr-Mo steels were developed4 for thermal power stations. In addition to the major constituent phases discussed above, the minor carbides which form at temperatures less than 1100 K, dictate the long term industrial performance of the steels. Evaluation of tensile and creep properties of Cr-Mo steels exposed

to elevated temperature for prolonged durations have been extensively studied.5,13,14 The following trends

were established: The optimized initial alloy compo­sition considered was 9Cr, W-2Mo = 3, Si = 0.5, with C, B, V, Nb, and Ta in small amounts. Higher chro­mium content has two effects: it increases the hard — enability leading to the formation of martensite and also promotes the formation of 8-ferrite thereby reducing the toughness. A reduction in the chromium

content lowers the oxidation resistance. If W + Mo concentration is kept <3%, creep strength will reduce, while higher amount promotes the formation of 8-ferrite and brittle Fe2Mo Laves phase. The addi­tion or partial replacement of molybdenum with tungsten and boron increased the stability of M23C6, and slowed down the kinetics of recovery. Lower nickel introduced 8-ferrite, while its increase reduces creep strength. When Si is less than 0.3%, oxidation resistance gets lowered, while higher silicon content led to agglomeration of carbides with an increased amount of 8-ferrite. On similar lines, the composition of all other elements could also be optimized, based on structure-property correlation studies.

The components of the steam generators are often subjected to repeated thermal stresses as a result of temperature gradients that occur on heating and cool­ing during start-ups and shutdowns or during
variations in operating conditions. Steady state operation in between start-up and shutdown or transients would produce creep effects. Therefore the low cycle fatigue (LCF) and creep-fatigue inter­action assume15 importance in the safe life design approach of steam generator components. The alloy exhibited a decrease in fatigue life with increasing temperature, thus limiting its upper limit oftempera — ture up to about 773 K.

The joining technologies of Cr-Mo steels have been well investigated.16,17 One of the major pro­blems during welding of ferritic steels has been the formation of 8-ferrite, if the amount of ferrite stabi­lizers is high. The partial substitution of Mo with W enables austenite stabilization and hence reduces the tendency to form 8-ferrite. In fact, there needs to be an intricate balance between austenite and ferrite stabilizing elements in 9-12Cr-Mo steels.

This would ensure a satisfactory solidification process with a fully austenitic structure. Additionally, this enables easier hot workability during primary proces­sing and tubemaking, without losing high tempera­ture creep resistance. The formation of 8-ferrite reduces toughness due to the notch sensitivity, pro­motes solidification cracking and embrittlement due to sigma-phase precipitation and reduces the creep ductility at elevated temperatures of operation. Other problems relate to solidification cracking, hydrogen cracking, and reheat cracking, which have been exten­sively studied.18 The Type IV cracking in ferritic steel weldments and the brittle layer formation in the dis­similar welds are discussed in detail later.

It was pointed out in Section 4.05.4.1 that the segre­gation of certain impurities to grain boundaries could cause nonhardening embrittlement. This phenome­non has received less attention than the hardening from the production of small clusters. Several reliable techniques (AES, FEGSTEM, and atom probe) exist with which grain boundary segregation may be not only observed, but also quantified,41 and there have been a number of critical studies that have both measured and modeled the segregation ofimpu — rity elements under irradiation.39,101-105 (Extensive experimental programs on long-term aging have per­mitted the accumulation of segregation data on a variety of model alloys and steels. It has been possible to interpret these data in terms of the simple McLean theory of equilibrium segregation (McLean, D. Grain Boundaries in Metals; Clarendon Press: Oxford, 1957). The success of the McLean model in describing segregation in these alloys and steels indicates that segregation is generally thermodynamically con­trolled, and defect gradients have no effect.)

The segregation of P and C to grain boundaries in irradiated materials has received greatest atten — tion.41,100 Overall P segregation increases with irradi­ation dose in all of the model alloys and steel types examined. The rate of P segregation under irradia­tion appears quite variable, both in different classes of steel and within a given class. It is possible that P segregation under irradiation is slower in welds than in the CGHAZ microstructure, because of the presence of additional traps for P in the welds. Other causes of variability are less consistently observed. The behavior of C is less consistent. In the model alloys and the CMn steels, grain boundary C gener­ally decreases with fluence, but in the MnMoNi steels C segregation may either increase or decrease. Desegregation of C appears more likely to be related to carbide precipitation in these materials with rela­tively high free C than merely to trapping of C at matrix defects.

Quantifying the data has been attempted in sev­eral cases.9 , 0 The majority of models indicate that P is dragged to grain boundaries during radiation by the flux of irradiation-induced defects to sinks. Consis­tency between the models and data need not necessarily confirm the validity of the model, as all have adjustable parameters, and no data set is large enough or coherent enough to test the models with much stringency.

Importantly, a conclusion from the European Commission 5th Framework PISA programme was ‘‘On the basis of the observations made here and else­where, it appears unlikely that nonhardening embrit­tlement will influence RPV condition during normal operation for homogeneous MnMoNi steels (i. e., A508 Class 3, A533B, 22NiMoCr37) of <0.02 wt% P.’’101

4.05.4.2 Summary

The formation of matrix defects has been established as an important contribution to RPV embrittlement. The studies of low Cu steels have established that they are produced continuously during irradiation but the nature of these defects is still uncertain. The experimental data provide strong evidence for clus­ters that are sensitive to positrons; and for PA being a useful technique for studying MD. Although micro­voids have been identified in model alloys from posi­tron lifetime studies, none has been identified in commercial steels from such studies. It is important to note that there is increasing evidence that vacancy clusters are not responsible for the observed harden­ing; it has been inferred that interstitial clusters are responsible for the observed hardening in RPV steels. Postirradiation annealing has been shown to be a pow­erful means of investigating the nature of the matrix defects further. A major development has been char­acterizing MD as being due to two components; first, SMD and second, at high fluxes, UMD. UMD are matrix defects which, although thermally unstable at the irradiation temperature, are frozen into the micro­structure during the cooldown after irradiation.

4.08.3.3.1 Continuous cooling transformation diagram

The preparation of a CCT (continuous cooling trans­formation) diagram is essential to the microstructure

control of9Cr-ODS steels. Figure 12 exhibits a CCT diagram that was experimentally constructed for 9Cr-ODS steel.21 The minimum cooling rate for the matrix phase in order to fully transform to martensite is extremely higher in 9Cr-ODS steel (solid circular symbol) than in mechanically milled EM10 (open diamond symbol) that does not contain added Y2O3. Residual ferrite plays an important role in the process of continuous cooling transformation. The minimum cooling rate is known to increase with a decrease in the size ofprior austenite (g) grains. This smaller size of prior g grains provides more nucleation sites (grain boundaries) for a g-a-phase transformation, so that a higher cooling rate is required to enable steel with small prior g grains to fully transform to a. The presence of residual ferrite restricts the growth of g grains; the prior grain size of residual ferrite-containing steel is roughly 5 pm, thus increasing the minimum cooling rate to produce a full martensite matrix.

In steel that does not contain residual ferrite and the mechanically milled EM10, the size of the prior g grains is roughly 10 pm and 35 pm, respectively. The results shown in Figure 12 can be explained by the relationship between the size of prior g grains and the minimum cooling rate.21 As for the normal­izing heat treatment used in commercial furnaces, the cooling rate would be roughly 3000 °Ch~ , so that

Time from 800 °C, t (s)

Cladding tube At intermediate heat Cold rolling

treatment (pilger mill)

At final heat treatment

Figure 13 Cladding tube manufacturing process developed for 9Cr-ODS steel.

the matrix phase of 9Cr-ODS steel cladding consists of residual ferrite, martensite, and a small amount of transformed ferrite from the g-phase.

4.08.3.3.2 Manufacturing process

9Cr-ODS steels are promising materials to enable fast reactor fuel cladding to realize a high burnup of 200GWdt-1 at 700 °C, since they have superior radiation resistance and high temperature strength. Figure 13 shows a series of manufacturing processes of fuel cladding that is 8.5 mm in diameter by 0.5 mm in thickness by 2 m in length. The element powders and yttria powder are mechanically alloyed for 48 h in an argon gas atmosphere using an attrition type ball mill with a capacity of 10 kg batch. The mechanically alloyed powders are sealed in hollow-shaped cans and degassed at 400 °C in a 0.1 Pa vacuum for 2 h. The hollow shape of the bars is consolidated by hot — extrusion at an elevated temperature of 1150 °C to the dimensions of 32 mm in outer diameter, 5.5 mm in wall-thickness, and 4 m in length. After machining to the precise dimensions, claddings are produced at their final dimension (8.5 mm in outer diameter, 0.5 mm in thickness, and 2 m in length) by four-pass rolling with about a 50% reduction ratio on each pass by using a pilger mill.

Without heat treatment, it is too difficult to man­ufacture cladding for ODS steels by the cold-rolling process. Using the CCT diagram of 9Cr-0.13C-2W — 0.2Ti-0.35Y2O3, as shown in Figure 12, a cooling

rate of about 150 K h-1 was applied to the intermedi­ate heat treatment in order to induce the ferrite phase at room temperature without martensite transforma­tion. This phase has a lower degree of hardness. Hard­ened cladding due to the accumulation of cold deformation can be sufficiently softened by this inter­mediate heat treatment, and cold rolling can then be continued with the softened ferrite structure. Figure 14 represents the typical hardness change of 9Cr-ODS steel in the process of cladding manufacturing by repeated cold rolling and intermediate heat treatment. The elongated grain structure induced by the fourth cold rolling can ultimately be made into equi-axed grain structure by the final heat treatment, which
consists of normalizing at 1050 °C for 1 h, followed by tempering at 800 °C for 1 h.