There are other components in the core which, while they do not directly contain fission products, do affect the ability of the cladding to do so and have a major effect on the ability of the nuclear plant operators to con­trol the reactor as the plant ages. These components include the structural materials in the fuel assembly such as the top and bottom nozzles, springs, fasteners, grid structures, skeleton, and the control members (rods in a PWR and blades in a BWR). In a BWR there are also water channels that chan­nel the coolant flow. Many of these structures are currently made of various zirconium alloys and some are of stainless or high alloy steels. The inside components of the PWR control rods are Ag-In-Cd alloy while BWR con­trol blades contain B4C. The research areas of interest are:

• Understanding and predicting the corrosion of the current alloys in their respective BWR and PWR environments.

• Development of new materials such as SiC composites for PWR control rod cladding and BWR water channels.

• Development of new control rod materials including gray rods (con­trol rods with much lower thermal neutron absorption materials, such as tungsten, than control rods but higher than steel or zirconium) for PWRs.

Identification of the parameters allowing the control of the degradation process is an essential part of the AMP development. Some parameters indicate the evolution of degradation directly, for example the wall thick­ness of piping. The water chemistry parameters can be used as indirect con­trolling parameters of all internal surface corrosion mechanisms.

Definition of the method for the detection of ageing effects

Most of the postulated ageing effects can be detected during the execution of the current programmes of the plant, as follows:

• Non-destructive testing performed in the context of in-service inspec­tion programmes

• Visual inspections performed in the frame of maintenance programmes

• Visual structural inspections

• Walk-down inspections.

As described in the foregoing sections on fracture, the energy absorbed before failure is an indication of the toughness of a material. Exposure of the material to neutron fluxes reduces the toughness value and increases the transition temperature (RTndt), below which the material fails in a brittle manner (absorbing very low energy) (Fig. 1.5); this is commonly referred to as radiation embrittlement. The extent to which the RTndt is raised from the initial value is an indication of the degradation. A simple model devel­oped using the concentrations of copper and nickel, and the value of flux and fluence could yield a fair prediction for the shift in RTndt.31 The most severely affected region in a RPV is the belt region where fast neutron bom­bardment is at a maximum. Presence of copper and phosphorous are found to promote the embrittlement effect while that of nickel acting alone is unclear,32 although in combination with copper, the effect of embrittlement due to nickel is increased. The role of other elements like Mo, Mn, As, Cr and Sn is not completely understood. The change in DBTT may be evalu­ated using Cottrell brittle fracture theory and is given in terms of friction and source hardening terms:

1 + (2Ф 2ayl c,) (do^ ^Ф)

ADBTT = Ac,-, Ac, = aGb Ф = aGb Ф. [1.23]

1 + ( )(d dr) ‘ • ^ [ ]

Thus, one needs to know the influence of neutron irradiation and test temperature on both the friction and source hardening terms to evaluate

the effect of neutron radiation exposure on DBTT or RTndt. However, a lack of information on the fluence dependence of as as well as other fac­tors makes it rather difficult to apply the above formula in predicting the changes in radiation embrittlement of nuclear pressure vessels. Radiation embrittlement of ferritic steels is far more complex than using simple frac­ture theory. The question now arises as to the superimposed effect of DSA on fracture toughness and whether DSA adds to the increase in DBTT (Fig. 1.23). Jung and Murty33 examined the effect of neutron irradiation on the elastic-plastic crack initiation fracture toughness (/IC) using an unload­ing compliance method before and after radiation exposure. Figure 1.24a illustrates the effect of DSA on decreased fracture toughness in the DSA regime which happens to be in the upper shelf region where we note a dip in toughness. After irradiation the dip appeared at a slightly higher tem­perature (Fig. 1.24b) illustrating the fact that neutron radiation exposure results in reduced amounts of interstitial C and N atoms in solution and that radiation does not eliminate DSA but postpones its occurrence to higher temperatures. These synergistic effects of radiation-produced defects and interstitials are sensitive to the composition of the steels.34

Pull tube examinations of 92 tubes from some Korean nuclear power plants have been carried out since 1989 (Hwang et al., 2007). The tubes had dif­ferent types of failures such as pitting, ODSCC, primary water stress cor­rosion cracking (PWSCC) and intergranular attack (IGA). A new type of ‘PWSCC’crack was found during the ISI carried out after the chemical clean­ing in 1990, and 22 tubes in SG A and 26 tubes in SG B had to be sleeved.

The pitting of plant A was related to high copper dissolved from condenser material, chloride and high levels of dissolved oxygen. Transgranular SCC of plant B seemed to be related to lead compounds. ODSCC and IGA in plant A were related to a caustic environment in the crevices. PWSCC in plant A and plant C originated from the inherent characteristics of the materi­als, which were not properly thermally treated (Hwang, 2003). After failure analysis, the performance of non-destructive testing was evaluated based on destructive metallographic examination, and some counter measures, such as material change, inhibitor injection, molar ratio control and temperature reduction operation, were suggested.

In a typical case of high cycle fatigue, a complete 360-degree break occurred in the cold leg side tube in Row 9 of the North Anna Unit 1 plant in the United States on 15 July 1987. The case was explained as follows (Shah and MacDonald, 1993):

1 An anti-vibration bar (AVB) was not installed around row 9.

2 A small dent was found in the tube. It opened due to mean stress as the fatigue strength of the material had dropped.

3 An uneven AVB was installed around the troubled tube, which caused sectional high speed coolant flow.

4 High amplitude and deteriorated fatigue strength caused fatigue destruction.

After this accident in North Anna Unit 1 plant, the US Nuclear Regulatory Commission (USNRC) ordered an inspection of the power plants that showed potential fatigue destruction due to denting around the TSP and fast sectional flow. Except for the U-bend area, well-installed TSP structures have caused no high cycle fatigue.

It is said that in the mechanical ageing progress, fretting, wear and thin­ning are caused by the vibration between the tube and tube support struc­ture (TSP and AVB). But thinning occurs where there is no flow-induced vibration, so it is difficult to say that tube vibration is a cause of thinning, Only in certain cases can we can say that thinning derives from pure corro­sion wastage. There are many factors influencing fretting, wear and thinning including the distance between the tube and support plate, coolant flow, oxide film formation and corrosion product accumulation. Of these, friction of the same side causes fretting and a large vibration causes wear. When a combination of vibration and corrosion predominates, thinning results.

One example of low temperature PWSCC is the stress corrosion cracking which was detected in the tube of OTSG in the Three Mile Island (TMI-1) plant in 1981. Most of the cracks were circumferential, and were found mostly in the HAZ of the weld or Top of Tubesheet (TTS) of the expansion part. The tube of the lower defective zone was repaired using the explosive expansion method, and the unrepaired tube was plugged. The tube of the OTSG made by B&W was sensitized, heat treated and carbides were created at the grain boundary. The tube was SCC-sensitive if it was exposed to acid because of the tensile stress in the material from the manufacturing process.

In France, cracks have been found in the divider plate of a steam gener­ator. As of the end of 2007, defects were found in the divider plates of ten steam generators. Various inspections have highlighted the fact that these defects are located in the stub of the hot branch, with no signs of significant evolution, either by fatigue or corrosion.

In Japan in 1976, leakage occurred from the U-bend of the steam genera­tor row 1 of the Takahama Unit 1. It was assumed that the crack was caused by the PWSCC due to plastic deformation of the tube. The deformation, which was located between the U-bend and the tube, had been created by passing a ball mandrel through the tube, or had developed by the curving process during tube manufacture. This area usually has high residual stress. PWSCC of the U-bend also occurred in Ohi Unit 1 and Mihama Unit 2. Further, another leakage occurred in a small radius U-bend in Ohi Unit 2 in 1994. In this area, the ovality was larger, relatively speaking.

Since 1982, PWSCC of the tubesheet zone has been detected in many power plants. In Mihama Unit 3 and Ohi Units 1 and 2, PWSCC was found in both hard rolled areas and expansion transitions (made using full depth expansion) in tubesheet. PWSCC was detected at expansion transitions with part depth rolling in tubesheet at Takahama Unit 1 and Mihama Unit 2. PWSCC in the expansion transition occurs due to high residual stress in the zone where materials are mechanically rolled. This is caused by insufficient expansion during mechanical rolling over by uneven tubesheet holes. Such PWSCC of 600 MA pipe can be removed by replacing the steam generator with one with 690 TT tubes.

In a case of PWSCC of 600 TT tubes in the Japanese Kansai power plant, some cracks were found by ECT on Alloy 600 TT in tubesheet region of three power plants since 1999. Inspection was carried out on the damaged tubes and PWSCC was proven as a result. In three power plants (Sendai Unit 1, Takahama Units 3 and 4) the depth was expanded by full depth mechanical rolling after full depth hydraulic expansion. Cracks were found in the upper part of transitions to the hydraulic expansion area, indicat­ing that cracks are located near the area expanded by the mechanical roll. But cracks did not occur where there was hydraulic expansion transition. Inspection of Takahama Unit 4 found the diameter of the tube hole to be sectionally large. It is considered that the oval shape was made by polishing the eccentricity of the tube hole during manufacturing. Cracks occurred in the zone where oval shaped holes were present. From a mock-up experi­ment, high residual stress was observed at the zone where there were tube holes that were irregular in shape, and mechanical rolling had been carried out. It is thought that mechanical rolling causes PWSCC where there are irregularly shaped tube holes.

Dedicated ageing management programmes already control some of the processes addressed by the time-limited ageing analyses presented above, for example the process of settlement of the main building and erosion — corrosion of piping wall. The results of the above analyses show that only a few non-compliances or lifetime-limiting cases have been found and all of them can be managed by the extension/amendment of the existing ageing management programmes and/or other plant programmes. For example, in relation to RPV and internals the stud joints fixing the polygon mantle to the core basket are the critical structures from the point of view of irradia­tion-assisted stress corrosion cracking and void swelling. In order to man­age these mechanisms, review and extension of the present programmes are ongoing. Regarding operational limits and conditions for injection into the pressurizer, the margin to allowable stresses is minimal and the number of allowable cycles is rather small; consequently, the number of cycles should be monitored. It was also found that during certain heat-up and cool-down processes the averaging intervals of the temperature measurements have to be modified at certain components. With respect to the containment civil structures the existing ageing management programme should be extended for managing the change of material properties of heavy concrete structures and for the corrosion of the steel liner on a heavy concrete surface.

8.5 Plant programmes credited for long-term operation

Review of the existing plant programmes can qualify these programmes as

adequate for ageing management. For example, the following programmes

can be classified as AMPs or part of AMP:

• Preventive and predictive maintenance programme can be considered to be a part of AMP because it is one of the solutions for ageing mitiga­tion and because AM requires information on preventive maintenance of SCs that is carried out

• In-service inspection programme

• Functional Testing Programme — for active components if they are in the scope of AM.

From the vast reactor operating experience it is noted that the cause of fuel failures in terms of the number of units with leakers has decreased over the years (Fig. 1.33) and the US nuclear industry has been tending toward 100% no-leakers. As of July 2010, more than 90% of units in the

1.33 Fuel leaker causes in (a) US PWRs and (b) world-wide BWRs.90

United States were failure-free.90 A recent review of the fuel failures in LWRs by the International Atomic Energy Agency (IAEA) revealed the following causes^1 crud/corrosion, debris, PCI/SCC, grid-to-rod fretting (GTRF), fuel handling, fabrication with some ‘unknown’ where the cause could not be pinpointed. Interestingly GTRF was seen to be the major issue confronting the PWR industry, accounting for around 50% of the total fuel failures. Figure 1.34a and 1.34b summarize the fuel leak causes in PWRs in the United States and BWRs across the world, respectively. Debris (31%), crud/corrosion (32%) and unknown (24%) accounted for

1.34 Fuel leaker causes in (a) PWRs and (b) BWRs.92

around 87% of fuel failures in BWRs while a relatively low proportion (9%) was noted to be due to PCI/SCC. Many of these issues are dealt in detail in Part II of this volume.

VVER-440/213 containments

The reduced pressure containment of VVER-440/213 is made of reinforced concrete and the steel liner ensures its leak tightness. Therefore, the basic concern is the effect of ageing on the containment leak-tightness. The leak rates of the VVER-440/213 containment, allowed by the design and justified by the regular integral tests, is equal to 14.7%/day at the post large-break LOCA, when the design internal containment pressure equals 2.4 MPa. It is clearly higher at some plants than what is allowable for Western NPP containments. Therefore, the goal of the VVER operators is to improve the leak tightness. (It should be noted that comparison with Western NPP containments is not straightforward. This is because, in connection with the design basis accidents, the pressure suppression system tends to cause pres­sures below atmospheric, rather than overpressure, at the time period when the atmosphere of the containment has its highest contents of radioactive aerosols, and when the potential for radioactive releases would thus be the highest.)

Containment leakage has a complex origin. Investigations carried out at the Paks and Bochunice NPPs, almost from the time of start-up tests, show that the poor sealing of doors and hatches mainly cause the containment leakage and thus the leakage is a maintenance problem rather than an age­ing issue.

Some VVER plants are built on relatively soft soil. Geodetic control of the settlement of the main building of these plants was started during construc­tion and it is periodically performed. The phenomenon might be a concern when there is uneven settlement, that is the differential movement causes unacceptable additional deformation of the structures. Experience shows that the differential movement may cause cracks in non-structural masonry walls. Another concern might be if the non-uniform settlement results in non-allowed tilting of the RPV vertical axis, which would cause problems for control rod drive mechanisms (CRDMs). The operating experience and analysis of settlement with extrapolation to extended operational lifetime is discussed for the Paks NPP (Katona et al, 2009a).

Operational experience is that ageing of neither the reinforced concrete load bearing structure nor the liner would limit the LTO of the VVER — 440/213 plants.

The second line of defense for nuclear plants is the primary system bound­ary in any LWR. The primary system for a PWR consists of the reactor vessel, the primary side piping, steam generators, pressurizer and coolant pumps. For a BWR, the primary system consists of the reactor vessel, pip­ing, steam turbine, condenser and coolant feed pumps. In both cases, main­taining the integrity of this boundary is crucial to maintaining the cooling functions for the reactor and therefore its long-term ability to prevent the release of fission products into the environment. Loss of integrity of this boundary leads to the classic LOCA (or small break LOCA) and the possi­bility of eventual destruction of the core due to the inability of the reactor to provide continued cooling to remove the still considerable heat from the decay of residual fission products (about 7 MWt after 2 min for a 1000 MWe reactor). Once breached and after the ability to provide cooling to the core is lost, the loss of integrity of the primary system leads to the release of fission products from the core to the reactor containment build­ing. Maintaining the integrity of the primary system is therefore considered a key safety (and regulatory) issue.

Of equal importance is the maintenance of clean heat transfer surfaces. Heat transfer is the reason why nuclear plants exist — ultimately to produce electric power for sale. Heat transfer issues, usually due to deposit forma­tion, lead to a temperature rise on the surface of the fuel and on the sec­ondary side of the PWR steam generator. Issues other than heat transfer are the build-up of boron in the fuel deposits in a PWR and of radioactive materials on both BWR and PWR fuels. The general response by PWRs has been cleaner chemistry (less Al, Si, and Ca in the primary water) and tighter pH specifications. Lately both BWRs and PWRs have explored the addition of materials such as Pt and Zn to maintain the integrity of system components.

The factors involved in maintaining this boundary intact have evolved over the last 50 years of reactor operation. Initially, the discipline of main­taining nuclear plants was viewed by both the utilities and the vendors as being similar to that of coal fired boilers. Due to the build-up of solids on fuel rods, primary side water specifications for PWRs and BWRs saw a drop in allowed levels of dissolved solids in the order of a factor of 1000 to low ppb levels. In the 1960s and early 1970s secondary side cooling water chemistry for PWRs was similar to that of any coal fired boiler which used phosphate chemistry. As deposit build-ups on the secondary side of steam generators occurred in PWRs (multiple tons in the early 1970s), problems began to emerge with the cracking of tubes at the support plates and U-bends of the steam generator tubes which necessitated the mass plugging of tubes in highly radioactive environments and the derating of some PWRs. It was then realized that different standards of water cleanliness would be required and all volatile chemistry was introduced, air leaks in condensers were repaired (and brass condensers were replaced with stainless steel ones), and feed­water specifications increased dramatically. Meanwhile cracking in coolant piping and in the welds of steam dryers (Fig. 19.4), bowing of water chan­nels in BWRs (which would hinder the insertion of the control blades) also began to reinforce the notion that the care and feeding of nuclear plants required a very different approach than the care of fossil fired plants where boiler tubes might be routinely replaced every ten years with no worries about radioactivity.

Recently, concerns have increased for the integrity of the primary ves­sels and baffle bolts for PWRs due to radiation hardening and loss of ten­sile strength (these bolts hold the flow baffles together on the outside of the core, see Fig. 9.5). Cracked welds around the penetration tubes through PWR reactor vessel heads have allowed internal corrosion of the alloy steel below the weld overlay material resulting in a potential small break LOCA situation (NRC, 2008a). In BWRs, breaking of welds due to excessive vibra­tion from the large steam flow as well as boiling within the core and crack­ing of the coolant piping continue to be an issue.

A key integrity issue for PWRs is the reactor coolant pump seal. This is a component that must have a minimal flow since it is a leak path for primary coolant from the primary coolant system. This has been a long-standing issue that comes and goes for various plants; for some plants is the highest potential issue for a small break LOCA.

Over the past 40 years of LWR operations, these issues have led to a new emphasis on the importance of effective chemistry control; research into new additives such as zinc for nuclear service; the need for better design methods for components specifically to take into account vibration issues in both the core area and the primary side piping; and the introduction of new alloys that can withstand the rigors of very clean water. Examples of the alloy changes are the use of the 690 and 800 alloys to replace the 600 series alloys first used in steam generator tubes and the use of enhanced alloy steels for steam generator support tubes to replace the low alloy car­bon steels that were previously used.

However, replacement of failed or deteriorating components is not always a feasible option and replacement is rarely cheap. For instance, a broken component such as a coolant injection nozzle deep inside a BWR reactor, or cracked baffle bolts inside the PWR reactor, cannot require the replacement of the vessel. Since these issues are occurring in highly radioactive and some­times hard to access components, remote methods of repair (such as under­water laser welding) as well as methods for access (tethered robotic welders) have been developed. Replacement of steam generators costs $40 million to $50 million each (and there are two to four in every PWR). Replacement of reactor pressure vessel heads cost about $20 million each (NEI, 2010).

Finally, there is the design of the components. As those who take Six-Sigma® courses always learn, the root of most problems is in the initial engineering. For example BWRs have, for many years, had an issue with cracking of welds in the steam separator (see Fig. 9.4). While not directly a safety issue, this problem has led to replacement of the steam dryers, and continued cracking in the welds has been observed. Recently, modeling tools that provide a more accurate prediction of the vibration stresses affecting these welds have been developed and deployed on much more powerful computers. This capability has allowed better designs for the steam dryers to be made and installed which, so far at least, have led to elimination of this issue. Of course, better modeling tools alone cannot accomplish better designs. Better understanding of materials in the relevant chemistry and radiation environment of the LWR is needed. This is especially true for the understanding of environmentally assisted crack growth which presents the potential for catastrophic failure of key boundary components (NRC, 2008b). This understanding can then be translated into more phenomeno­logical models for use in modeling components and perhaps even successful prediction of the response of these components to conditions outside those which were used to originally develop the models.

Another broader issue is the standards to which pressure vessels are designed. The current fleet of reactors was designed to the ASME Boiler and Pressure Vessel Code, Section III, in effect in the 1970s. This code was based on laboratory environment testing and not on tests using actual operating conditions (including radiation) (Majumdar, 2011). In the past 40 years, there has been enough data collected in LWR conditions that illus­trates a need for changes in these codes, but more importantly, how actual LWR operating conditions have affected the current fleet. It is important to know what changes need to be made in current plants to keep them operat­ing safely for the next 20, 40, or more years. Since the replacement of these plants would require an investment of over $500 billion, there is a huge eco­nomic driver for this work.

Much work remains to be done in this area including:

• Understanding of the effects of chemistry on the development of cor­rosion and especially SCC in piping, equipment, fasteners (bolts) and welds both inside and outside the core area. The likely result of better understanding of the chemistry will be the development of new addi­tives for use in the primary and secondary systems of LWRs to prevent corrosion and cracking, the build-up of radioactive materials, and the maintenance of clean heat transfer surfaces.

• Long-term effects of radiation, temperature and pressure cycling, and chemistry on the integrity of the reactor vessel and its component mate­rials. This also applies to the other components in the primary system, though radiation effects will be much less important.

• In-place maintenance methods (for instance, in-place annealing) that can be used to reverse the effect of irradiation on materials

• Maintenance methods for remote repair of components such as under­water welding and robotics

• Development of new alloys (for instance alloy 690 for steam generator tubes) or new materials (such as SiC composites) that can be used to eliminate potential issues due to radiation and chemistry

• Development of new materials for pump seals capable of withstanding high pressure drops, high wear, and primary system water chemistry

• Development of modeling tools to more accurately predict the stresses that components undergo during operation, and to design components that perform as well or better than the current components. Part of this effort should be on developing the phenomenological models required to predict the performance of current materials after having been in ser­vice for 40 or more years. This capability will be required for license renewals of current plants.

• Development of monitoring tools for in-use components that will pro­vide sufficient warning to plant operators of impending maintenance and repair issues so that appropriate steps can be taken during sched­uled outages, minimizing the potential for failure generated accidents and disruptions of the electrical supply.

A definition of the methods for monitoring, trending and condition eval­uation is the fifth step in the development of the AMPs. For example, the monitoring of the trend of fast neutron fluence absorption in the critical components of the reactor pressure vessel is one of the most important indi­rect ageing management elements. The monitoring of load cycles defined during design and of their parameters belongs to the ageing management of fatigue degradation mechanism. The monitoring of the number and growth of crack-indications found during material inspections and visual inspec­tions in the frame of in-service inspection can be assigned to each local deg­radation phenomenon. The monitoring and trending of the value of wall thickness reduction could be taken into account in the case of degradation forms with general material loss. In the case of heat exchangers, the moni­toring of the number of plugged tubes can also be considered as an element of the ageing management programme.

As we noted earlier in Fig. 1.12 voids and precipitates are generated in mate­rials such as stainless steel during neutron irradiation resulting in reduced density or increased volume known as radiation swelling that, in turn, leads to dimensional changes even in the absence of external stresses. Zirconium

alloys on the other hand resist void formation albeit undergoing stress-free radiation growth due to the inherent crystallographic texture with preferred crystallographic orientation developed during the thermo-mechanical pro­cessing of thin-walled tubing used to clad nuclear fuel (UO2). Radiation exposure of single crystal Zr exhibits elongated a-axis with decreased c-axis thereby the single crystal becomes short and fat (Fig. 1.25a) mainly due to the formation of interstitial <a> loops on prism (jlOlo}) planes albeit the volume is unchanged. Cladding tubes typically exhibit preferred orientations

1.25 ( a) Effect of neutron radiation exposure on a Zr-single crystal leading to decreased c-axis (vertical) and increased a-axis (horizontal); (b) preferred orientation (texture) in a typical Zircaloy cladding tube.

or textures such that the c-axis of the grains are mainly oriented at 30° from the radial (thickness) direction towards the hoop (transverse) direction as shown in Fig. 1.25b. This results in small, positive strains along the axial and hoop directions equal to the contractile strain along the radial or thickness direction such as that the total sum is zero:

£z + £g + £r = 0. [1.24]

Since all the radiation-induced strains are relatively small, no observable changes occur in the diameter (~9.5 mm) and thickness (~0.56 mm) while measurable changes occur along the axial direction of the ~3.9 m long clad­ding tubes. This lengthening of the cladding tubes due to radiation exposure is known as radiation growth with no change in volume; this is in contrast to void swelling where dimensional changes are accompanied by increased volume. Radiation growth of Zircaloy cladding leads to rod bow and, in cases where a gradient in neutron flux and/or texture along the cladding tubes exist, then the entire assembly can bow; more details may be found in the later chapters on Zr-alloys.

Deformation due to creep occurs in the presence of external stress during irradiation. At relatively low temperatures, where thermal creep is negligible,

radiation may induce creep due to the increased concentration of point defects during irradiation which enhances diffusion which is thus known as radiation-induced creep. At high temperatures where thermal creep can take place, radiation enhances creep due to increased defect concentration and is referred to as radiation-enhanced creep. In an extremely simplified way one may express the vacancy concentration as due to thermal and radi­ation so that the creep-rate equation (Equation [1.8]) becomes

£ = ADcC, D = ^a2vD (C* + Cirr)e~&‘/RT, C* = e_Q/and C” — dpa.

6

[1.25a]

In general, however, the radiation component of the creep rate is seen to be temperature insensitive and proportional to the flux and stress:

4г= Вфс, [1.25b]

and with very little primary creep so that the strain due to radiation creep is given by

£irr= B0ct. [1.25c]

In-reactor results are often sensitive to the neutron spectrum and are very scattered to unequivocally describe the stress and time dependences. However, irradiation creep is not just the thermal creep imposed with high defect density; in the former the interstitial and vacancy loops that form during irradiation play a major role in the creep mechanism while in the latter the creep rate increases with temperature. Two mechanisms are pro­posed to explain the irradiation creep phenomenon: (a) in a stress-induced preferential absorption (SIPA),35 extra planes of atoms accumulate on crystal planes so as to produce creep strain in the direction of the applied stress, whereas (b) stress-induced preferential nucleation (SIPN) assumes that nucleation of loops is preferred on planes with a high resolved normal stress.3637 Both of these mechanisms assume that the growth or formation of loops occurs at a favourable orientation with respect to applied stress and causes macroscopic strain. Irradiation generated point defects (vacan­cies and self interstitials) migrate to different sinks like dislocations and grain boundaries, in order to reduce the energy of the system, and do so in a preferential manner due to the anisotropy of the zirconium crystal lattice. Because of the diffusional anisotropy, interstitial atoms tend to migrate to dislocations lying on prism planes and to grain boundaries oriented paral­lel to prism planes while vacancies drift preferentially to dislocations lying on basal planes and to boundaries parallel to basal planes. This gives rise to elongation in one direction and contraction in the other. The creep rate can be controlled by suitable alloying additions and modifying the texture of the zirconium matrix such that the dislocation glide rate is reduced. The complex behaviour was modelled by Nichols38 using dislocation climb-glide processes as a function of stress and neutron radiation dose. There is an extensive literature on radiation creep of different materials and the reader is referred to the literature for more details.