As discussed in Chapter 4.10, Radiation Effects in Graphite, irradiation-induced dimensional changes in graphite are highly anisotropic, and a strong function ofirradiation temperature and neutron dose (dpa). The temperature range of interest for fusion applications varies from 100 °C in areas well removed from the plasma of experimental devices, to over 1000 °C for the surface of PFCs, which experience appreciable plasma flux, and for future power-producing machines. As described in detail in Chapter 4.10, Radiation Effects in Graphite, the mechanism of graphite irradiation-induced dimensional change is a combi­nation of intra — and intercrystallite effects. Within the crystallites, displacement damage causes an (a)-axis shrinkage (within the basal plane) and a (c)-axis growth (perpendicular to the basal plane.)

The upcoming ITER reactor will be the first fusion reactor to provide a flux of neutrons to pro­duce measurable thermophysical effects to fusion structural materials. Even so, this will be a relatively modest fluence machine, with the maximum fast dose accumulating less than 1 x 1025nm~2 (E > 0.1 MeV), or less than a displacement per atom, over its lifetime. The work of Bonal provides data on the dimensional changes in CFCs, which are expected in this dose range. Specifically, his work11 irradiated 2D and 3D
composites to doses approaching 1 dpa in the tempera­ture range of 610-1030 °C. Figure 7 shows the dimen­sional instability that occurs in these materials in the sub-dpa region, specifically indicating a shrinkage.

The work of Burchell12 in Figure 8 shows the dimensional change behavior of 1, 2, and 3 direc­tional composites for doses somewhat in excess of the ITER lifetime. In this example, solid cylinders were irradiated at 600 °C to doses ranging to 5 dpa and the resulting diameter and length measured. The behavior of each material can be explained by the accepted theory for dimensional change in graphite (Chapter 4.10, Radiation Effects in Graphite) after taking into account the individual fiber architec­tures, and by the observation that a model for fibers describes them as graphite fiber, filaments of circum­ferential or radial basal planes running parallel to the fiber axis. The irradiation-induced dimensional change of such a fiber is therefore a shrinkage in length and a growth in diameter. However, at doses

less than 1 dpa the dimensional change is relatively minor (Figure 7). As the dose is increased, the direc­tion perpendicular to the fiber axis is more or less unchanged while a significant shrinkage along the direction parallel to the fiber axis occurs. At about 2-3 dpa, swelling in the composite occurs in the perpendicular direction. The random fiber composite of Figure 8 has a random orientation of chopped PAN fibers in the plane of the composite. The speci­men diameter shows practically no change perpen­dicular to the fiber axis to about 4.5 dpa, though it exhibits ^2% shrinkage parallel to the fiber axis. The 3D balanced PAN weave fiber has essentially isotropic shrinkage to a dose of ^2 dpa, at which point the diameter of the fibers, and hence the sam­ple, begins to swell.

Also given in the 3D composite plot in Figure 8 is the radiation-induced dimensional change parallel to the fiber axis of an Amoco P55 pitch fiber com­posite. This material was processed in an identical manner to the PAN fiber composite. From the plot, it appears that the pitch fibers, and thus the compos­ite, undergo slightly less shrinkage, possibly due to the higher fiber crystallinity. This hypothesis is also supported by the observation that fibers with higher final heat treatment temperatures tend to exhibit less dimension change13 and it is also consistent with the observation that elevating the heat treatment temper­ature of graphite reduces the irradiation-induced

shrinkage.14

The irradiation-induced dimensional changes are of fundamental importance to the design and perfor­mance of the fusion structure, and even more so of the PFCs. This is due to the need to precisely define the plasma edge. For this reason, it is instructive to look at the irradiation effects at the higher dose and temperature conditions representative of the next — generation fusion power devices. The data shown in Figures 9 and 10 provide higher temperature dimen­sional swelling data for the FMI-222 3D CFC and MKC-1PH 1D CFC, which were model, high ther­mal conductivity CFCs studied in the early phases of the ITER composite development program.15 In Figure 10, the dimensional change of the 1D composite yields substantial swelling perpendicular to the fiber axis and equally impressive shrinkage parallel to the fiber. The FMI-222 of Figure 10, a nearly isotropic orthogonal weave pitch-fiber com­posite with equivalent fiber volume fraction in the x, y, and z directions, undergoes a positive dimensional change (swelling) parallel to the cylindrical axis of the sample, which increased with increasing temper­ature. The magnitude of swelling was in excess of 10% at the highest temperatures studied at the 2 dpa dose level. This is in contrast to the FMI-222 swelling data reported by Burchell12 and Snead,16 also for HFIR irradiation, though at a lower irradiation tempera­ture. Specifically, a contraction of 0.6% is interpolated from the data of Burchell for FMI-222 irradiated

at 600 ° C to an equivalent fluence as the data of Figure 9. Snead16 reports on an 800 °C irradia­tion to a substantially higher dose (7.7 x 1025nm~2) than the Figure 8 dose (~2.4 x 1025nm~2). In this case, the material underwent a contraction of 3.6%

along the length of a bend-bar (2.3 x 6 x 30 mm). It was also noted in this work that the width and thickness direction exhibited swelling. Specifically, swelling parallel to the width direction (6 mm) was 1.4% and swelling parallel to the thickness direc­tion (2.3 mm) was 5.9%. The overall dimensional effects were related to the effect of (measured) gross changes in the dimension of fiber bundles noting that gaps were evident on the surface of the bend bars. Figure 11 shows an example of the top surface of an FMI-222 composite irradiated in the present work to 980 °C, 2.4 dpa. This composite underwent very low swelling. By inspection of the figure the contraction of the fiber tows below the free surface of the sample is evident. However, there is evidence from this micro­graph that some of the fibers (particularly at the tow edge) have not withdrawn into the sample. This is evidence of shear within the fiber bundle as opposed to the tow-matrix interface. This observation is evi­dence of the large stresses that must be building in the composite under irradiation. The fact that the bundles are not failing at the tow-matrix interface also sup­ports the previous finding that, at least in the initial period of gross dimensional change, the load-carrying capacity of the composite has not been degraded. In fact, previous measurement of FMI-222 irradiated to a dose of ^7.7 x 1025nm~2 (E> 0.1 MeV) at 800 °C described a 54% increase in strength.16

Figure 12 shows an scanning electron microscopy (SEM) image comparing the 2 dpa surface of the FMI-222 composite of Figure 11 with cylindrical samples of the same size, also irradiated near 1000 ° C, though at progressively higher doses. Clearly the dimensional instability continues with dose leading to gross changes in the composite.

A development programme has been launched to develop a fast brazing technique to minimize the holding time at high temperature and consequently retain adequate mechanical properties of the CuCrZr alloy. This was achieved by induction brazing in Europe and by fast heating and cooling using an e-beam test facility in the Russian Federation.

Induction brazing tests were done using the only appropriate silver free braze alloy available in the market, the STEMET 1108 procured from the Russian Federation. It was found that this braze alloy had poor wetting properties and the quality of the product was variable. Difficulties were met for brazing Be tiles of dimensions representative of the Be tiles of first-wall panels. A few first-wall mock-ups were fabricated with inductively brazed Be tiles but showed thermal fatigue performance well below HIPped mock-ups, with detachment of Be tiles between 1.5 and 2 MWm~2.162,171 This result has been considered unsatisfactory. The development work on fast brazing equipment has been stopped in Europe and the effort is being concentrated on the development of a new silver — free braze alloy.

The fast brazing development in the Russian Federation has also been done using the STEMET 1108 alloy but fast heating was performed using an e-beam test facility. First-wall mock-ups were heated on the beryllium side by the e-beam and tempera­tures as high as 780 °C were achieved at the Be/CuCrZr joints for a very short time, followed by fast active cooling of the mock-ups, minimizing therefore the production ofbrittle intermetallic com — pounds.172 Good results were achieved on hyper — vapotron type mock-ups, developed at the early stage of the ITER design, for Be coated divertor

129

components.

Breeding blankets for fusion reactors are categorized into solid breeder and liquid breeder concepts. The liquid breeder blankets have certain advantages over the solid breeder blankets such as continuous chemi­cal control of the breeding material including isoto­pic control of Li, impurity control and tritium recovery, and immunity to irradiation effects. How­ever, some issues such as compatibility of the breeder with structural materials are more serious for liquid breeder blankets4. In addition, blanket structure

can be simplified significantly if the liquid breeder functions as the coolant as well (self-cooled liquid breeder blanket). At present, the major candidate liquid breeder materials are Li, Li-Pb, and molten salt Flibe (LiF-BeF2).

In the cases of self-cooled liquid Li and Li-Pb blankets in magnetic confinement fusion systems, the high-speed flow of these materials perpendicular to the strong magnetic field causes an electric current, which then produces an electromagnetic force as a result of interaction with the magnetic field. This force changes the velocity profile in the cooling ducts and acts to retard the coolant flow, leading to what is called a magnetohydrodynamic (MHD) pres­sure drop. This process is schematically shown in Figure 1(a). The MHD pressure drop may result in loss of flow control and mechanical stresses exceeding the allowable limits of the structural materials. The problems arising from the MHD pressure drop are critical feasibility issues for self-cooled liquid metal breeder blanket concepts with metallic structures.

The quantification of the MHD pressure drop requires a rather complex numerical analysis. How­ever, in simple cases such as straight and constant area cross-section flow in conductive ducts with a uniform magnetic field in a traverse direction, the pressure gradient along the flowing direction, dp/dx, is given as follows5:

dp/dx = ksUB2

where s, U, and B are the electrical conductivity, flow velocity of the liquid metal, and magnetic flux den­sity, respectively, and k is a positive function of elec­trical conductivity of the wall. The equation implies that the MHD pressure drop is an issue in the case of high magnetic field and high velocity flow of conduc­tive liquid metals. In the case of a low flow rate such as would occur in a helium-cooled Li-Pb blanket, the MHD pressure drop will not be an issue.

Magnetic field

*

iquid Li flow

Insulator

coating

Figure 1 Schematic illustration of magnetohydrodynamic pressure drop (left) and the role of insulator coating (right).

To reduce the MHD pressure drop, optimization of the coolant flow path by enhancing the flow fraction parallel to the magnetic field may have some effect. However, a more effective way to reduce the MHD pressure drop would be to electrically insulate the coolant flow from the surrounding walls.6 The reduc­tion of MHD pressure drop by an insulator coating is schematically illustrated in Figure 1(b). The require­ments for the coating can be summarized as follows:

1. compatibility with liquid breeder under flowing conditions with a temperature gradient,

2. high electrical resistivity under irradiation,

3. robustness and/or an effective self-healing capability,

4. potential for covering large and complex surfaces, and

5. fundamental requirements for in-vessel materials such as radiation resistance, low activation proper­ties, and low tritium inventories in blanket conditions.

Quantitative evaluation of the required electrical resistance and an allowable crack fraction are subject to overall blanket design including flow channel structures. A recent model calculation showed that the ratio of electrical resistivity of the insulator to the wall needs to be ^106 and crack areal fraction to be ;S 10~6 to maintain the pressure drop within tolerable range, assuming Li wets cracks.7,8

For the Li-Pb blankets, the insulator coating should be a critical issue if a self-cooled Li-Pb blanket with metallic structural materials is to be designed. How­ever, current blanket design options with Li-Pb are (1) helium cooled with slow-flowing Li-Pb, (2) dual­coolant Li-Pb with fast-flowing Li-Pb but electrically insulated from the wall by a SiC/SiC flow channel insert (FCI), or (3) self-cooled Li-Pb using SiC/SiC as the structural material. None of these concepts needs the insulator coating. However, development of a ceramic coating, necessary mostly for tritium perme­ation reduction and possibly for corrosion protection, is still a critical issue.9

4.15.1 Introduction

4.15.1.1 Tritium Breeding

The fusion reaction of tritium and deuterium is con­sidered one of the most suitable options for near­term large-scale fusion power generation, through

D + T! 4He(3.56MeV) + n(14.03MeV)

Deuterium is a hydrogen isotope with an abundance of 1 out of 6500 atoms in seawater, implying virtually boundless resources. Tritium is the next hydrogen isotope, and it is radioactive with a half-life of 12.3 years under emission of a р-particle; it cannot be obtained from natural resources. Therefore, the D-T fuel cycle requires the breeding of tritium from lithium using one of the following reactions:

n + 6Li! T + 4He + 4.78MeV

n + 7Li! T + 4He — 2.47MeV + n

The neutron supplied by the D-T fusion reaction shown above is also the one that provides useful energy. The reaction with 6Li is exothermic, providing a small energy gain; on the other hand, the reaction with 7Li is endothermic but does not consume the neutron, though a more thermalized neutron is released. Natural lithium contains 7.42% 6Li and 92.58% 7Li.

In fact, lithium has been identified as the only viable element to breed tritium. 6Li has a very high cross-section to capture a neutron (see Figure 1), and through the use of isotope enrichment, the effective
6Li density can be raised from the natural 7.42% to about any desired value.

Tritium can bond to microstructural features within metals, including vacancies, interfaces, grain bound­aries, and dislocations. This phenomenon is generally referred to as ‘trapping.’15-18 The trapping of hydro­gen and its isotopes is a thermally governed process with a characteristic energy generally referred to as the trap binding energy Et. This characteristic energy represents the reduction in the energy of the hydro­gen relative to dissolution in the lattice16,19 and can be thought of as the strength of the bond between the hydrogen isotope and the trap site to which it is bound. Oriani16 assumed dynamic equilibrium between the lattice hydrogen and trapped hydrogen

Ут _ $l Et

1 — yT = 1 — yL exp rt)

where Ут is the fraction of trapping sites filled with tritium and 0L is the fraction of the available lattice sites filled with tritium. According to eqn [9], the frac­tion of trap sites that are filled depends sensitively on the binding energy of the trap (Et) and the lattice con­centration of tritium (0L). For example, traps in ferritic steels, which are typically characterized by low lattice concentrations and trap energy <100kJ mol-1, tend to be depopulated at high temperatures (>1000 K).

For materials with strong traps and high lattice con­centration of tritium, trapping can remain active to very high temperatures, particularly if the trap energy is large (>50kJ mol-1). The coverage of trapping sites for low and high energy traps is shown in Figure 1 for two values of K one material with relatively low solubility of hydrogen and the other with high solubility.

The absolute amount of trapped tritium, ct, depends on 0T and the concentration of trap

15

sites, nT :

ct = a»T$T [10]

where a is the number of hydrogen atoms that can occupy the trap site, which we assume is one. If multiple trapping sites exist in the metal, cT is the sum of trapped tritium from each type of trap. A similar expression can be written for the tritium in lattice sites, cL:

Cl = Ь»ьУь [11]

where nL is the concentration of metal atoms and b is the number of lattice sites that hydrogen can occupy per metal atom (which we again assume is one). Substituting eqns [10] and [11] into eqn [9] and recognizing that 0L ^ 1, the ratio of trapped tritium to lattice tritium can be expressed as

cl [cl + nLexp(-£t/RT)]

Therefore, the ratio of trapped tritium to dissolved (lattice) tritium will be large if cl is small and Et is large. Conversely, the amount of trapped tritium will be relatively low in materials that dissolve large amounts of tritium. The transport and distribution of tritium in metals can be significantly affected by trapping of tritium. Oriani16 postulated that diffusion follows the same phenomenological form when hydrogen is trapped; however, the lattice diffusivity (D) is reduced and can be replaced by an effective diffusivity, Deff, in Fick’s first law. Oriani went on to show that the effective diffusivity is proportional to D and is a function of the relative amounts of trapped and lattice hydrogen:

D

1 + —(1 — 0t)

c l

If the amount of trapped tritium (ct) is large relative to the amount of lattice tritium (cl), the effective diffusivity can be several orders of magnitude less than the lattice diffusivity.20 Moreover, the effective diffusivity is a function of the composition of the hydrogen isotopes, depending on the conditions of the test as well as sensitive to the geometry and microstructure of the test specimen. Thus, the intrin­sic diffusivity of the material (D) cannot be measured directly when tritium is being trapped. Equation [13] is the general form of a simplified expression that is commonly used in the literature:

D

«t

1 + exp

«L

Equation [14] does not account for the effect of lattice concentration, and is therefore inadequate when the concentration of tritium is relatively large. For materials with high solubilities of tritium (such as austenitic stainless steels), trapping may not affect transport significantly and Deff « D as shown in Figure 2. For materials with a low solubility and relatively large Et, the effective diffusivity can be substantially reduced compared to the lattice diffu — sivity (Figure 2). The wide variation of reported diffusivity of hydrogen in a-iron at low temperatures is a classic example of the effect of trapping on hydrogen transport2,2 : while the diffusivity of hydro­gen at high temperatures is consistent between studies, the effective diffusivity measured at low tem­peratures is significantly lower (in some cases by orders of magnitude) compared to the Arrhenius relationship established from measurements at ele­vated temperatures. Moreover, the range of reported values of effective diffusivity demonstrates the sensi­tivity of the measurements to experimental technique and test conditions. For these reasons, it is important to be critical of diffusion data that may be affected by trapping and be cautious of extrapolating diffusion data to experimental conditions and temperatures
different from those measured, especially if trapping is not well characterized or the role of trapping is not known.

The thermal fatigue resistance of tungsten is strongly related to its performance as part of current inertially, and future actively cooled components for applica­tion in magnetic fusion devices. The functional requirements these components have to fulfill are listed in Section 4.17.2.

State of the art inertially cooled components include W coatings on graphite, CFC, and TZM.12,13,46,47,54,140,207,208 These concepts are used or are going to be used in the large operating toka — maks, for example, AUG and JET. During thermal loading, they mainly suffer from the problem of high interfacial stresses as a result of the CTE difference between the W coating and the substrate. Further­more, interfacial reaction products and their poten­tial reduced power handling capability have to be taken into account.

Besides coatings, recent development of an iner­tially cooled bulk tungsten divertor for JET20,123 showed that under thermal fatigue loads the W quality is of minor importance for the integrity of the compo­nent. The major issue for the tungsten PFM was found to be the necessary shadowing of the plasma-loaded surface to avoid overheating and melting at tile edges as a result of the shallow angle of the incident plasma. This was realized by surface shaping.209

In the design of actively cooled components, tungsten is joined to a water-cooled Cu-based heat sink (ITER) or He-cooled steel or W-based heat sinks (e. g., DEMO, ARIES-CS). Direct cooling of the tungsten armor should be avoided, particularly without castellation, as the induced stresses might cause catastrophic material failure with subsequent water or He-leakage.124 Therefore, the only perfor­mance requirements are a sufficiently good surface quality to reduce possible crack initiation points and therefore suitable fabrication and surface finish­ing technologies,1 , , 1 the chemical compatibility

with the heat sink and, if present, the joining interface material, and the cyclic stability of the joint(s). The latter is influenced by the temperature gradient applied during steady state heat loads, the difference of the CTEs, the quality of the joining process and, perhaps most important for reducing induced stres­ses, the tile size, or the dimensions of the castellated segments (see Section 4.17.3.2.4).

Smaller tile sizes significantly improve the stress situation at the interface, and also at the top surface of the PFM. This has to be taken into account when comparing the thermal fatigue results of various kinds of components and the response of different grades and alloys of W, as shown by Makhankov et at., where smaller tile sizes resulted in little or no crack formation. Furthermore, variations in the size of the component investigated can often explain the contradictory results presented in the literature that show good behavior of a material in one test while it fails in another. However, there are limitations to the minimum size of tiles and a compromise between operational and economical needs has to be made.

Despite the fact that design and manufacturing technique seems to be more important than the mechanical properties and the microstructure of the particular W grade or alloy, the latter should still be considered. Similar to thermal shock results, the risk of delamination parallel to the loaded surface at the interface90 or anywhere in the bulk material has to be minimized. Therefore, the grain orientation of the PFM microstructure should be perpendicular to the loaded surface, although this still bears the risk of crack formation toward the cooling structure.108 To avoid subsequent water or He-leakage in case of crack propagation into the heat sink material, particularly in the He-cooled divertor design (see Section 4.17.4.2.2), suitable material and design solutions still have to be found. Furthermore, sha­dowing of adjacent tiles similar to the JET bulk W divertor has not yet been included in the design of the actively cooled components.

Carbon and graphite materials have enjoyed consid­erable success as PFMs in current tokamaks because

of their low atomic number, high thermal shock resis­tance, and favorable properties. However, their use is not without significant issues, and their application in next-generation fusion energy devices is by no means certain. Significant among the issues related to carbon and graphite PFMs are neutron irradiation damage, which causes significant dimensional change and degrades the thermal conductivity resulting in increased PFC surface temperatures; physical sputtering, chemical erosion, and radiation enhanced sublimation, which cause surface material loss to the plasma and redeposition of carbon with tritium; and tritium inventory, which constitutes both a safety problem and an economic impediment to the use of graphite. Joining of CFC to heat sinks has witnessed significant development in the past few decades, which has resulted in good performing designs for near-term test machines such as ITER. The high heat loads and surface temperatures that result after plasma disruptions are also problematic for carbons. However, the same high temperatures make the use of Be, which has a significantly lower melting tem­perature, very unlikely.

Next-generation machines will impose increas­ingly greater thermal loads on their PFCs. High thermal conductivity CFC materials may offer a
solution for the high heat loads, but further research is needed to overcome the problems noted above and to secure the place of carbon materials in the future of fusion power reactors.

At some locations (mostly at the upper inboard region and the lower region of the first wall), the Be armored PFCs must withstand a certain number of ‘slow’ thermal transients resulting from loss-of — control of plasma position during VDEs. Typical parameters for these events are 60 MJ m~2 over 0.3 s. In contrast to thermal quench disruptions, VDEs lead not only to significant erosion or melting, but also to high heat fluxes and a subsequent temper­ature increase at the armor/heat sink interfaces that can result in a failure of the armor/heat sink joints.217 As a matter of fact, because of their short duration (<10ms), ELMs and the thermal-quench phase of a disruption have no significant thermal effects on structural materials and coolant channels. In con­trast, plasma instabilities such as VDEs (duration 100-300 ms), and runaway electron impact, in addi­tion to causing severe surface melting and erosion, can result in substantial bulk damage to these com­ponents. Elevated temperatures and high thermal stresses in the structure can seriously degrade the integrity of the interface bonding and burnout the coolant channels. Runaway electrons (up to many megaelectron volt) penetrate many centimeters of beryllium and directly heat underlying metal struc­tures, potentially damaging coolant channels.218-220

The erosion due to VDEs in a device like ITER has been modeled by various authors,2 1 whereas the results of analysis carried out to quantify the effects on PFCs resulting from runaway electrons can be found in Raffray et al222 As an example, Figure 23 shows the surface temperature of a 5 mm copper

Copper substrate

(Carbon, beryllium, or tungsten coating on
copper structure)

Figure 23 Interface copper surface temperature rise during a vertical displacement event for different surface coating materials. Reproduced with permission from Federici, G.; Skinner, C. H.; Brooks, J. N.; etal. Plasma-material interactions in current tokamaks and their implications for next-step fusion reactors. Nucl. Fusion 2001,41,1967-2137 (review special issue), with permission from IAEA.

substrate at its interface with a tungsten, beryllium or carbon tile of 10 mm thickness during a typical VDE releasing about 60 MJ m~2 to the surface in 300 ms.7 Tungsten and carbon armors of similar thickness usually result in a similar and higher cop­per surface temperature than that of beryllium armor of the same thickness. This is because most of the incident plasma energy is removed by the beryllium’s higher surface vaporization rate, which leaves little energy to be conducted through the structural material. In order to reduce the tempera­ture at the copper interface, thicker tiles would be required. Only beryllium tiles of reasonable thickness (<5-10 mm) or very thick carbon or W tiles (>20mm) can withstand the acceptable temperature rise in the copper structure for the conditions shown. The coolant flux and, conse­quently, the Be/Cu interface temperature increase with decreasing Be thickness. The evaporated and melting thickness and temperature at the Be/Cu alloy interface during each VDE is shown in Table 5 for Be tiles (5 and 10 mm thick); for two values of the VDE energy density (30 and MJ m~ ) and for two VDE durations (10 and 100 ms).

The study of intense radiation effects in metals has been closely associated with the development of nuclear fission reactors, and as a result at the begin­ning of the 1980s when the urgent need to consider radiation damage aspects of materials to be employed in future fusion reactors was fully realized, a consid­erable amount of knowledge and expertise already existed for metallic materials.29 This was not the case for the insulating materials, mainly because of the fact that the required use of insulators in fission — type reactors is in general limited to low radiation regions, well protected from the reactor core. How­ever, despite the late start and the reduced number of specialists working in related fields at the time, together with the complexity of the mechanisms involved in radiation damage processes in insulators, considerable progress has been made not only in assessing the possible problem areas, but also in finding viable solutions. Several general reviews give a good introduction to the specific problem of radiation damage in insulators.30-36

The materials employed in the next-step fusion machine will be subjected to fluxes of neutrons and gammas originating in the ignited plasma. The radiation intensity will depend not only on the dis­tance from the plasma, but also in a complex way on the actual position within the machine because of the radiation streaming along the numerous pene­trations required for cooling systems, blanket struc­tures, heating systems, and diagnostic and inspection channels, as well as the radiation coming from the water in the outgoing cooling channels due to the 16O(n, p)16N nuclear reaction. However one-, two-, and even three-dimensional models are now available, which enable the neutron and gamma fluxes to be calculated with confidence at most, if not all, machine positions.37-40

Radiation damage is generally divided into two components: displacement damage and ionization effects. In a fusion environment, displacement dam­age, which affects both metals and insulators, will result from the direct knock-on of atoms/ions from their lattice sites by the neutrons, giving rise to vacancies and interstitials. Those primary knock-on atoms (PKAs) with sufficient energy may go on to produce further displacements, so-called cascades. The numerous point defects thus produced may either recombine, in which case no net damage results, or they may stabilize and even aggregate producing more stable extended defects. These sec­ondary processes which determine the fate of the vacancies and interstitials are governed by their mobilities. These mobilities are highly temperature dependent, and in the case of insulators even depend on the ionizing radiation level (radiation-enhanced diffusion). Displacement damage is measured in ‘dpa’ (displacements per atom) where 1 dpa is equivalent to displacing all the atoms once from their lattice sites. At the first wall of ITER, the primary displacement dose rate will be of the order of 10~6dpas~

In contrast, ionizing radiation although absorbed by both metals and insulators, in general, only produces heating in metals. However, certain aspects of radia­tion damage in metals, such as radiation-enhanced corrosion and grain boundary modification are related to ionization. The effects of ionization on insulators are in comparison quite marked because of the exci­tation of electrons from the valence to the conduction band giving rise to charge transfer effects. Ionizing radiation is measured in absorbed dose Gy (Gray) where 1 Gy = 1J kg-1. At the first wall of ITER, the dose rate will be of the order of 104Gy s-1.

The response of insulators to both displacement and ionizing radiation is far more complex than in the case of metals. Apart from a few specific cases (diamond for example), insulating materials are polyatomic in nature. This leads to the following:

(i) We have in general two or more sublattices which may not tolerate mixing.

(ii) This gives rise to more types of defects than can exist in metals.

(iii) Because of the electrically insulating nature, the defects may have different charge states, and hence different mobilities.

(iv) The displacement rates and thresholds, as well as the mobilities, may be different on each sublattice.

(v) We may have interaction between the defects on different sublattices.

(vi) Defects can be produced in some cases by purely electronic processes (radiolysis); however, in the insulating materials of interest for fusion, this is generally not the case.

As a consequence of these factors, while radiation damage affects all materials, the insulators are far more sensitive to radiation damage than metals. While stainless steel, for example, can withstand sev­eral dpa and GGy with no problem, some properties of insulating materials can be noticeably modified by as little as 10-5 dpa or a few kGy. Because of this, the present ongoing programs of radiation testing for diagnostics are concentrating mainly on the insulat­ing components of the systems. The results of these radiation damage processes are flux — and fluence — dependent changes in the physical and mechanical properties ofthe materials, which may be particularly severe for the insulators. The properties of concern which suffer modification are the electrical and thermal conductivity, dielectric loss and permittivity, optical properties, and to a lesser extent the mechan­ical strength and volume. The effects of such changes are that the insulators may suffer Joule heating because of the increased electrical conductivity or lower thermal conductivity, and absorption in windows and fibers can increase from the microwave to the optical region and they emit strong lumines­cence (radioluminescence, RL); in addition, the materials may become more brittle and may suffer swelling. Clearly, some materials are more radiation resistant than others. The organic insulators, which are widely used in multiple applications in general, degrade under purely ionizing radiation and are not suitable for use at temperatures above about 200 °C; as a result their use will be limited to superconducting magnet insulation and remote handling applications during reactor shutdown. Inorganic insulators of the alkali halide class have been widely studied and are used as optical windows; however, they are suscepti­ble to radiolysis (displacement damage induced by electronic excitation) and in general become opaque at low radiation fluences. Of the numerous insulating materials, it is the refractory oxides and nitrides, which in general show the highest radia­tion resistance, and of these the ones which have received specific attention within the fusion program include MgO, Al2O3, MgAl2O4, BeO, AlN, and Si3N4. In addition, different forms of SiO2 and materials such as diamond and silicon have been examined for various window and optical transmis­sion applications.

One other aspect of radiation damage that should be mentioned is nuclear transmutation. The high — energy neutrons will produce nuclear reactions in all the materials giving rise to transmutation pro — ducts.1 These will build up with time and represent impurities in the materials, which may modify their properties. The physical properties of insulators are particularly sensitive to impurities. Furthermore, some of these transmutation products may be radio­active and give rise to the need for remote handling and hot cell manipulation in the case of component removal, repair, or replacement. For the structural materials, in the present concepts mainly steel alloys, considerable work has been carried out on the devel­opment of so-called low or reduced activation mate­rials (LAM, RAFM — reduced activation ferritic/ martensitic) for possible use in DEMO and future commercial fusion reactors.41-45 This work with the aim of reducing the amount of nuclear waste has studied not only the substitution ofradiological prob­lem alloying elements such as Mo and Nb in steels, but also the viability of other materials such as vana­dium and SiC/SiC composites. In the case of the insulating materials, no equivalent study or development has been carried out, in part because of the small fraction of the total material volume repre­sented by the insulators, and also because the impor­tant physical properties of these materials are expected to be degraded before the transmutation products become of concern. Certainly, for a next-step machine such as ITER, transmutation products, with the possi­ble exception of hydrogen and helium, are not expected to present a serious problem.

Deterioration of reinforced concrete generally will result in cracking, spalling, or delamination of the cover concrete. Whenever damage is detected, cor­rective actions are taken to identify and eliminate the source of the problem, thereby halting the degrada­tion process. The first step in any repair activity is a thorough assessment of the damaged structure or component including evaluation of (1) cause of dete­rioration, (2) extent of deterioration, and (3) effect of deterioration on the functional and performance requirements of the structure or component. From this information, a remedial measures strategy is for­mulated based on the consequence of damage (e. g., effect of degradation on structural safety), time requirements for implementation (e. g., shutdown requirements, immediate or future safety concern), economic aspects (e. g., partial or complete repair), and residual service life requirements (e. g., desired residual service life will influence action taken).105 Basic remedial measures options include (1) no active intervention; (2) more frequent inspections or con­ducting specific studies; (3) if safety margins are presently acceptable, taking action to prevent deteri­oration from getting worse; (4) carrying out repairs to restore deteriorated or damaged part of the structure to a satisfactory condition; and (5) demolishing and rebuilding all or part of the structure. Basic guidance on the repair of degraded structures is available,27,106 and a workshop has been held addressing repair of NPP concrete structures.46 Results of the workshop indicate that improved guidance is required on the assessment of defects (e. g., cracks), and information is desired on the performance and effectiveness of subsequent repairs to concrete structures in NPPs (e. g., durability of repair materials). Information on past performance and current practices for repair materials and systems for general civil engineering structures is being assembled (http://projects. bre. co. uk/conrepnet/pages/contents. htm).