As shown earlier, steady-state permeation of hydro­gen through materials is normally governed only by solubility and diffusivity. It has been shown23 that at low pressures, permeation can also be limited by dissociation at the surface. Due to limited data in the literature on this effect (and questions about whether this condition ever really exists), we do not consider this effect in this chapter. It is also possible for permeation to be limited by the rate at which atoms can recombine back into molecules. With the exception of extremely high temperatures, this recom­bination is necessary for hydrogen to be released from a material. Wherever the release rate from a surface is limited by recombination, the boundary condition at that boundary is given by:

Jr = krc2 [19]

where kr is the recombination-rate constant and c is the concentration of hydrogen near the surface (for this discussion, we assume that there is no surface rough­ness). The units for k and c are m4 s-1 per mol of H2 and mol H2 m~3, respectively.

There are two specific types of conditions that can lead to the hydrogen release being rate limited by recombination. One of them occurs for plasma­facing materials in which the recombination-rate coefficient is relatively low, and the implantation rate is high. With this condition, the concentration of hydrogen in the very-near plasma-exposed surface will increase to the point at which Jr is effectively equal to the implantation rate. It is not exactly equal to the implantation rate because there is permeation away from that surface to the downstream surface. The other condition that can lead the hydrogen release being controlled by recombination is when the rate of ingress at the upstream boundary is very low. This condition can occur either when the upstream pressure is extremely low or a barrier is placed on the upstream surface, and the downstream surface has a relatively low recombination-rate con­stant. In the extreme case, the release rate at the downstream side is so slow that the hydrogen con­centration becomes uniform throughout the material. The release rate from the downstream surface will be krc2, where c now represents the uniform concentra­tion. From c = KpTT and eqn [19], it can easily be shown that the recombination-limited permeation is linearly dependent on pressure, rather than having the square root of pressure dependence of diffusion — limited permeation.

There are various derivations and definitions of the recombination-rate constant. In the case of intense plasma exposure in which extreme near-surface concentrations are generated, Baskes24 derived the recombination-rate constant with the assumption that the rate was controlled by the process of bulk atoms jumping to the surface, combining with surface atoms, and then desorbing. His expression for the recombination-rate constant is

s

-t Кexp where C is a constant, s is the sticking constant, which depends on the cleanliness of the material surface, and EX = DHs + Ed > 0, otherwise EX = 0. The sticking constant can be anywhere from 1 for clean surfaces to 10-4 or smaller for oxidized surfaces.

Pick and Sonnenberg25 solved the recombination — rate constant for the case where the near-surface con­centration of hydrogen is small. In the limit of low surface concentration, the rate of atom jump to the surface does not play an important role in the recom­bination rate, thus eliminating EX from the exponen­tial. The sticking constant in the Pick and Sonnenberg model is thermally activated: s = s0 exp(-2Ec/RT), where s0 is the sticking coefficient and EC is the activation energy for hydrogen adsorption.

Wampler26 also studied the case of low near­surface concentration to arrive at an expression for the recombination-rate constant. He assumed equi­librium between hydrogen atoms in surface chemi­sorption sites and atoms in solution, deriving the recombination-rate constant as

nsn 2D Hs

(bnL)2 exp RT" where ns is the area density of surface chemisorption sites, and n is the jump frequency.

These expressions differ, but also display many similarities. Unfortunately, the surface cleanliness dominates the rate of recombination and these theo­retical relationships are relevant only for sputter — cleaned surfaces and very low pressures. For example, Causey and Baskes27 showed that the Baskes24 model predicts fairly accurate results for plasma-driven per­meation of deuterium in nickel. Comparison with values in the literature for nickel showed other results to differ by as much as four orders of magnitude and to have significantly different activation energies.

4.16.2.3 Irradiation and Implantation

Irradiation and implantation can affect the transport of hydrogen isotopes in materials. Since these effects can be complex and depend on the conditions of the mate­rials and the environment, it can be difficult to draw broad conclusions from the literature. Nevertheless, changes in apparent transport properties are generally attributed to damage and the creation of hydrogen traps28-31 (see also Chapter 1.03, Radiation-Induced Effects on Microstructure). Therefore, the effects of
irradiation and implantation will depend sensitively on the characteristics of the traps that are created by these processes. The density of damage is an important con­sideration: for example, it has been shown that helium bubbles are not effective trapping sites for steels,32 likely because in these experiments, the density of helium bubbles was relatively low. The energy of the trap will determine the coverage as a function of tem­perature (eqn [9]): generally, the effect of trapping will be stronger at low temperatures, especially in materials with a low solubility (Figure 2), which can result in substantial increases in hydrogen isotope inventory compared to hydrogen content predicted from lattice solubility. Additionally, irradiation may increase ioni­zation of hydrogen isotopes, thus enhancing apparent permeation.29 Reactor environments can defeat per­meation barriers, for example, by damaging the integ­rity of oxide layers; this is discussed at the end of this chapter.

There are many design proposals for a He-cooled first wall and divertor concept for DEMO and ARIES-CS.37,160,221 Among these, the He-cooled modular design with jet cooling (HEMJ)102 is the most developed and qualified in terms of microstruc­tural response,103 having survived at reduced coolant temperatures of 450-550 °C at least 100 cycles at 11 MW m~2 without failure. In contrast to the results obtained for water-cooled components for ITER, no influence of grain orientation on the components performance was observed.102 This might be a result of the higher temperature, which was always above the DBTT.

Nevertheless, some difficulties in the design still have to be resolved. First, there are problems related to temperature with a desired coolant temperature of >600 °C; these include material recrystallization at the top surface and the necessary high temperature joining to the W-based heat sink material. Second,

issues related to the material’s mechanical properties must be solved, in particular for the ductility of W-based structural material and its neutron irradia­tion resistance (see Section 4.17.4.3). Finally, the manufacturing reproducibility has to be at a high level because of the large number of small units (1unit« 3 x 10~4m2) necessary for cladding the DEMO divertor.

PWIs have been recognized to be a key issue in the realization of practical fusion power since the beginning of magnetic fusion research. By the time of the first tokamaks in the 1960s in the USSR and subsequently elsewhere, means of reducing the level of carbon and oxygen were being employed.19,20 These included the use of stainless steel vacuum vessels and all-metal seals, vessel baking, and discharge cleaning. Ultimately, these improvements, along with improved plasma confinement, led to the first production of relatively hot and dense plasmas in the T3 tokamak (^1keVand ^3 x 1019m~3).21,22 These plasmas, while being cleaner and with low — Z elements fully stripped in the core, still had unacceptable levels of carbon, oxy­gen, and metallic impurities. The metallic contamina­tion inevitably consisted of wall and limiter materials.

Early in magnetic fusion research, it was recog­nized that localizing intense PWIs at some type of ‘sacrificial’ structure was desirable, if only to ensure that more fragile vacuum walls were not penetrated. This led to the birth of the ‘limiter,’ usually made to be very robust, from refractory material and posi­tioned to ensure at least several centimeters gap between the plasma edge and more delicate struc­tures like bellows, electrical breaks, vacuum walls, etc. Typical materials used for limiters in these early days included stainless steel in Adiabatic Toroidal Compressor (ATC)23 and ISX-A24 and many others, molybdenum in Alcator A25 and Torus Fontenay-aux-Roses (TFR),26 tungsten in symmetric tokamak (ST)27 and Princeton Large Torus (PLT),28 and titanium in poloidal divertor experiment (PDX).29

Poloidal divertors have been very successful at loca­lizing the interactions of plasma ions with the target plate material in a part of the machine geometrically distant from the main plasma where any impurities released are well screened from the main plasma and return to the target plate.30 By the early 1980s, it was also recognized that in addition to these functions, the divertor should make it easier to reduce the plasma temperature immediately adjacent to the ‘limiting’ sur­face, thus reducing the energies of incident ions and the physical sputtering rate. Complementary to this, high divertor plasma and neutral densities were found. The high plasma density has several beneficial effects in dispersing the incident power, while the high neutral density makes for efficient pumping. Pumping helps with plasma density control, divertor retention of impurities and, ultimately, in a reactor, helium exhaust.

By the late 1970s, various tokamaks were starting to employ auxiliary heating systems, primarily neutral beam injection (NBI). Experiments with NBI on PLT resulted in the first thermonuclear class temperatures to be achieved.28,31,32 PLT at the time used tungsten limiters, and at high powers and relatively low plasma densities, very high edge plasma temperatures and power fluxes were achieved. This resulted in tungsten sputtering and subsequent core radiation from partially stripped tungsten ions. For this reason, PLT switched limiter material to nuclear grade graphite. Graphite has the advantage that eroded carbon atoms are fully stripped in the plasma core, thus reducing core radia­tion. In addition, the surface does not melt if over­heated — it simply sublimes. This move to carbon by PLT turned out to be very successful, alleviating the central radiation problem. For these reasons, carbon has tended to be the favored limiter/divertor material in magnetic fusion research ever since.

By the mid-1980s, many tokamaks were operating with graphite limiters and/or divertor plates. In addition, extensive laboratory tests and simulations on graphite had begun, primarily aimed at under­standing the chemical reactivity of graphite with hydrogenic plasmas, that is, chemical erosion. Early laboratory results suggested that carbon would be eroded by hydrogenic ions with a chemical erosion yield of Y~ 0.1 C/D+, a yield several times higher than the maximum physical sputtering yield. Another process, radiation-enhanced sublimation (RES), was discovered at elevated temperatures, which further suggested high erosion rates for carbon. Carbon’s abil­ity to trap hydrogenic species in codeposited layers was recognized. These problems, along with graphite’s poor mechanical properties in a neutron environment (which had previously been known for many years from fission research33), led to the consideration of beryllium as a plasma-facing material. This was pri­marily promoted at JET.34 A description of the opera­tion experience to date with Be in tokamak devices is provided in Section 4.19.2.3.

At present, among divertor tokamaks, carbon is the dominant material only in DIII-D. Alcator C-Mod at Massachusetts Institute of Technology (MIT), USA35 uses molybdenum. ASDEX-Upgrade (Axially Symmetric Divertor Experiment) is fully clad with tungsten,36 and JET has completed in 2011 a large enhancement programme37 that includes the installa­tion of a beryllium wall and a tungsten divertor. New superconducting tokamaks, such as Korea Supercon­ducting Tokamak Advanced Reactor (KSTAR) in Korea38 and experimental advanced superconducting tokamak (EAST) in China,39 employ carbon as material for the in-vessel components, but with provisions to exchange the material later on in operation.

The current selection of plasma-facing materials in ITER has been made by compromising among a series of physics and operational requirements, (1) minimum effect of impurity contamination on plasma performance and operation, (2) maximum operational flexibility at the start of operation, and (3) minimum fuel retention for operation in the DT phase. This compromise is met by a choice of three plasma-facing materials at the beginning of opera­tions (Be, C, and W). It is planned to reduce the choices to two (Be and W) before DT operations in order to avoid long-term tritium retention in car­bon codeposits during the burning plasma phase. Beryllium has been chosen for the first-wall PFCs to minimize fuel dilution caused by impurities released from these surfaces, which are expected to have the largest contamination efficiency.40-44 More­over, the consequences of beryllium contamination on fusion performance and plasma operations are relatively mild. This has been demonstrated by experiments in tokamaks (see Section 4.19.2.3).

The main issues related to the use of beryllium in ITER are (1) the possible damage (melting) during transients such as ELMs, disruptions, and runaway electron impact, and its implications for operations and (2) the codeposition of tritium with beryllium which is eroded from the first wall and deposited at the divertor targets (and possibly also locally rede­posited into shadowed areas of the shaped ITER first wall). Both issues are part of ongoing research, the initial results of which are being taken into account in the ITER design so that the influence of these two factors on ITER operation and mission is minimized. This includes ELM control systems based on pellets and resonance magnetic perturbation (RMP) coils, disruption mitigation systems, and increased temper­ature baking of the divertor to release tritium from the beryllium codeposited layers. Carbon is selected for the high power flux area of the divertor strike points because of its compatibility with operation over a large range of plasma conditions and the absence of melting under transient loads. Both of these characteristics are considered to be essential during the initial phase of ITER exploitation in which plasma operational scenarios will require development and transient load control and mitiga­tion systems will need to be demonstrated.

The main differences between a future power reactor and ITER are the much longer operation time (e. g., >108s vs. >107s), high duty cycle, and the higher temperature of the fluid to cool the PFCs to maintain a high plant energy conversion efficiency. The higher surface temperature of the PFC will affect particle recycling, tritium uptake, chemical erosion, and material-mixing effects.

Erosion rates at the divertor target are very diffi­cult to predict in conventional fusion power plant concepts with solid high-Z targets because the net erosion or deposition is strongly dependent on plasma parameters. The fraction of ions arriving above the sputtering threshold is crucial, as is the efficiency of the prompt local redeposition. ELMs have not really been considered in this context but we can see from the analysis in Section 4.19.6.2.2.2 that ELMs in power plant systems will have to be extremely small — much smaller than will be allow­able in ITER, which still has a relatively low duty cycle. The ideal would in fact be a quiescent ELM free high density steady state edge plasma.

Calculations of the minimum erosion rate for the main wall are somewhat more robust as there has to be a hot plasma in the main chamber and the rate of leakage of neutrals into the main chamber from the divertor can be calculated using Monte-Carlo codes. In a recent study, Be, C, Fe, Cu, Mo, and W walls were compared,206 with the conclusion that in all cases the erosion rate was 1-2 t per year of continuous operation. A Be or CFC wall will erode too rapidly in a reactor and the large amount of eroded material might give rise to deleterious problems as far as control of the tritium and dust inventories are concerned. A medium-Z material, such as Fe, does not seem to be acceptable purely from the standpoint of erosion lifetime. As molybdenum is unfavorable for long-term activation problems, W is the best and only solution we have available for a reactor. Effects of plasma contamination from Mo and W at the wall of tokamaks are being addressed in Alcator C-Mod, ASDEX-Upgrade and in the near-future at JET.

There is considerable gross erosion by sputtering for all materials. The contributions ofions and neutrals from the plasma to this erosion are ofthe same order of magnitude. The integrated total erosion due to ions and the energetic neutrals for the different wall mate­rials (Figure 24) show that because of the larger sput­tering yields for the low-Z materials, the number of atoms eroded for these materials is a factor of 10-20 larger than for high-Z materials such as W. However, the total mass loss is similar for all materials, up to several kilograms per day or about 1 t per year.

The maximum wall thinning for the low-Z mate­rials is about 3.5 mmyear~ , while for high-Z materi­als, such as W, it is 0.22 mm year, that is, lower by about a factor of 15. These values are in reasonable agreement with erosion measurements at the JET vessel walls.223 With respect to wall thinning, W is favorable for the use at the vessel walls because it has the longest ‘erosion lifetime’ (Figure 24(b)). With respect to plasma contamination, the probability of the eroded atoms entering into the plasma core, their lifetime in the plasma core, and the tolerable concen­tration of these ions in a burning fusion plasma all have to be taken into account.206 The tolerable con­centration of W in the plasma is nearly three orders of magnitude lower than for low-Z atoms, such as Be and C. However, recent observations have shown that W can be effectively removed from the plasma center by central heating.224 As this central heating is natural for burning plasmas, W may be a possible plasma­facing material, even from the viewpoint of plasma contamination. The ion and neutral flux densities on the vessel walls are of the order of 102°m~2s~1, which may be critical with respect to the tritium implantation, accumulation in and permeation through the vessel walls.

Electrical resistance, more generally discussed in terms of the electrical conductivity (the inverse of the resistance), is an important basic parameter for numerous systems and components including the NBI (neutral beam injector) heating system, ICRH (ion cyclotron resonant heating) windows and sup­ports, magnetic coils, feedthroughs and standoffs, MI cables, and wire insulation. Any reduction in the electrical resistance of the insulator material in these components may give rise to problems such as increased Joule heating, signal loss, or impedance change. The main candidate material for these applications is Al2O3 and is also the one which has been most extensively studied, both in the polycrys­talline alumina form and as single crystal sapphire. To a lesser extent, MgO, BeO, MgAl2O4, AlN, and SiO2 have also been studied. At the present time, three types of electrical degradation in a radiation environment are recognized and have been investi­gated; these are radiation-induced conductivity (RIC), radiation-induced electrical degradation (RIED), and surface degradation.

Of these types of degradation, RIC was the first to be addressed in a fusion context, as this enhance­ment of the electrical conductivity is flux dependent and hence a possible cause for concern from the onset of operation of any fusion device. Fortunately, RIC had been studied for many years, and a sound theoretical understanding already existed.55-59 The ionizing component of the radiation field causes an increase in the electrical conductivity because of the excitation of electrons from the valence to the conduction band and their subsequent trapping in levels within the band gap near to the conduction band from where they are thermally excited once again into the conduction band. Figure 1 shows sche­matically RIC as a function of irradiation time and ionizing dose rate (flux). The increase in saturation depends not only on the dose rate as indicated, but also in a complex way on the temperature and sample impurity content, as may be seen in Figure 2 for MgO:Fe.60 Nevertheless, such behavior, including the initial step, is well predicted by theory.57 However, at the dose rates of interest for fusion applications, in the range of approximately 1 Gys-1 to >100 Gy s~ , saturation is reached within minutes to seconds, and it is this saturation level which is usually the value of interest. The RIC process can lead to increases in the electrical conductivity of many orders of magnitude. For example, a standard high-purity alumina has a room temperature conductivity of generally less than 10~16Sm~ which increases to approximately 10~ 0Sm-1 for an ionizing dose rate of only 1 Gys~161 The first experiments carried out within a fusion application context, that is, refractory oxide materials, high-dose rates, and temperatures, gave an insight into the effects of dose rate, tem­perature, and material impurity, and established the well-known relationship at saturation, between the total electrical conductivity measured during irradia­tion and the ionizing dose rate: stotal = s0 + KRR where s0 is the conductivity in the absence of radiation, R is the dose rate, and K and d are constants.59,61-63 Although d « 1, the detailed studies found tempera­ture, dose, and dose rate dependence in this parameter, with extreme values in certain cases ranging between 0.5 and 1.5, and in addition a temperature dependence was observed for K. At the present time, extensive RIC data are available for materials irradiated with X-rays, g-rays, electrons, protons, positive ions, and fission and

14MeV neutrons. Many of the additional results, although in some cases limited to one temperature, and/or one dose rate, add confirmation to the earlier extended studies, but more importantly show that RIC is essentially a function of the ionization, independent of the irradiating particle or source. With very few

exceptions, all the data taken together over a range of dose rates from <1 Gys-1 to about 104Gys-1 show d « 1, as may be seen in Figure 3, and lie within a narrow band with the spread in conductivity values at any given dose rate being about two orders of magnitude13; see also, for example, Noda et a/.,66

where 14 MeV neutron results are given together with a small selection of other RIC data. For all the RIC data available, because of the different experimental conditions, it is difficult to draw any conclusions as to the reason for the spread in RIC values at any given dose rate. However, data obtained from electron

irradiations of different aluminas and other materials under identical conditions of dose rate and temper­ature give an indication that the RIC is inversely proportional to the sample impurity content.19 From these results (Figure 4), two general conclusions/ indications may be drawn:

RIC (single crystal) > RIC (polycrystal) and

RIC (pure) > RIC (impure)

However, the indication on the impurity dependence needs to be qualified, as certain impurities intro­duce levels near to the conduction band, and increase the RIC.59,60 This would imply therefore that the vast majority of the impurities in the materials act as recombination centers for the electrons and holes, thereby reducing the free charge carrier life­times, and do not introduce electron levels near to the conduction band. The reduction of the electron lifetime in the conduction band has important con­sequences for the RIED effect in different materials, as discussed below.

From all the data available, at the present time one can safely say that RIC is sufficiently ‘well understood’ to allow this type of electrical degrada­tion to be accommodated by the design, and that materials exist which give rise to electrical conduc­tivities <10~6Sm_1 for ionizing dose rates of up to >103Gys~ . One only expects possible problems or influence near the first wall. Unfortunately, this is precisely the region where magnetic coil diagnostics that can tolerate only very low leakage conductivity will be employed. It is important to remember that RIC is a flux-dependent effect and will be present from the onset of operation of the next-step machines. Hence, devices which are sensitive to impedance changes, which will occur for example in MI cables,
must take RIC into account. Furthermore, as RIC is strongly affected by impurity content, the buildup of transmutation products will modify the RIC with irra­diation time (fluence), although this is not expected to be of serious concern for ITER.

In contrast to RIC, RIED is a more serious prob­lem because it has been observed under certain con­ditions to permanently increase, that is, degrade, the electrical conductivity with irradiation dose. Figure 5 shows a schematic RIED-type degradation. The ini­tial increase in the conductivity corresponds to the RIC as described above. Following a certain irradia­tion time, or accumulated dose, the conductivity again begins to increase as s0 degrades. In Al2O3 for which most work has been performed, RIED is observed as a permanent increase or degradation of the electrical conductivity (s0) when a small electric field («100 kV m j is applied during irradi­ation at moderate temperatures («450 °C). At con­siderably higher temperatures and voltages, but without an irradiation field,67 or for irradiations per­formed without an applied electric field,68 no degra­dation occurs. Even at the present time, this type of degradation is still not fully understood; nor is there general agreement as to whether RIED is a real degradation in the volume.

Following the first report of RIED effect in electron-irradiated sapphire (Al2O3) and MgO,8 numerous experiments were carried out to assess its possible relevance to fusion insulator applications. These addressed the effect of the applied electric

field, DC or AC/RF69 and voltage threshold,70 the irradiation temperature,71,72 and the ionizing dose rate,73 as well as observations that in addition to electrons, RIED occurred with protons (Figure 674), as,75 and neutrons,76-78 and the observation of RIED effects in other materials, for example, MgAl2O4.74 In addition, further experiments were performed in which RIED-like effects were also observed in sapphire that was electron irradiated in air,79 for thin Al2O3 films,80 and MgO insulated cable.81 In contrast, some experiments did not observe any RIED effect, with some reporting enhanced surface conductivity or even cracking of the material.82-88 This led to suggestions that the RIED degradation is not a real volume effect, but is caused by surface contamina — tion.82,86 Because of the potential importance of elec­trical degradation and the uncertainty, extensive discussions on RIED were held at several IEA Workshops,89,90 including the experimental techni­ques employed in the irradiations to separate and identify volume degradation from surface effects. It was pointed out at an early stage of the discussions that important factors such as dose rate, and in partic­ular material-type differences, and irradiation temper­ature, all of which could cause RIED not to be observed were not being taken into account.73 For example, under identical conditions RIED was observed in Vitox alumina but not in Wesgo AL995 alumina,75 strongly suggesting a material (possibly impurity and/or grain size) dependence, and further
observations showed that the low purity, large grain size Wesgo AL995 material was highly susceptible to surface degradation when irradiated in high vacuum.91 The in-reactor RIED experiment in HFIR at ORNL also threw light on the complex RIED problem.92,93 Initial results indicated no significant increase in elec­trical conductivity for 12 different samples. However, moderate to substantial electrical degradation was later reported for some ofthe sapphire and alumina samples, so material type is an important parameter.94 One of the major difficulties for in-reactor experiments is the determination of s0, the conductivity in the absence of radiation, and its temperature behavior. The use of nuclear heating and the residual reactor radiation level mean that changes in this parameter with temperature and its corresponding activation energy are not gener­ally measured, although these are the main indicators for the onset of degradation; hence, RIED only becomes measurable when the material conductivity in the absence of radiation is larger than the RIC; that is, s0 > KR. Furthermore, some experiments were performed at temperatures either near room tempera — ture85 or above 600 °C,95 considerably outside the expected effective temperature range for RIED of approximately 400-500 °C.

In an attempt to clarify the situation, work was performed to identify possible basic causes of RIED. These experiments detected specific volume effects in Al2O3 that are observed only for irradiations carried out with an applied electric field. A marked

enhancement of the well-characterized F+-center (oxygen vacancy with one trapped electron) was observed,71 and TEM identified large regions of g-alumina within the bulk of RIED degraded Al2O3.96 The increase in F+-center production gave rise to enhanced oxygen vacancy mobility, and led to vacancy aggregation and aluminum colloid for­mation, as may be seen in Figure 7 97 This clarified the observed close similarity between the RIED effect and colloid production in the alkali halides,68 and helped to explain the formation of g-alumina and associated bulk electrical and mechanical degrada- tion.96 The combined work led to a RIED model being formulated, which successfully explained the role of the electric field (both DC and AC/RF), the ionization, and the anion (oxygen) vacancies.98 The model predicted a threshold electric field for degradation depending on the impurity/defect con­centration which, as mentioned above in the discus­sion of RIC, reduces the free electron lifetime. This helps to explain the negative RIED results for Wesgo AL995 alumina where the applied experimental field was below the predicted value of >0.6 MV m-1.75,87 It also highlighted the importance of the ionization, in agreement with earlier conclusions.73,84 Additional support for the model, and RIED as a volume effect, came with the TEM identification of aluminum colloids, as well as previously observed g-alumina, in Al2O3 irradiated with an electric field applied.99 At that time, an alternative model based on charge buildup and breakdown was also developed, but
was not extended to explain many of the important observations.1

During the intense activities related to RIED during the 1990s, two important factors emerged, one concerned with surface electrical degradation, and the other related to the importance ofthe exper­imental irradiation environment. For insulating com­ponents in future fusion devices, surface electrical degradation may prove to be more serious than the RIC and RIED volume effects. At that time, two types of surface degradation were reported, a con­tamination caused by poor vacuum, sputtering, or evaporation,83,88 and a real surface degradation related to radiation-enhanced surface vacuum reduc­tion and possibly impurity segregation.101,102 Both forms are affected by the irradiation environment and ionizing radiation. However, the real surface degradation effect is strongly material dependent, and occurs in vacuum but not in air or helium.102 This stresses the extreme importance of a represen­tative irradiation environment for material testing. Most insulating materials required for fusion applica­tions in ITER and beyond must indeed operate in high vacuum, and in consequence accelerator experiments to study electrical conductivity have been performed in vacuum, whereas to date, with few exceptions, in-reactor experiments

for technical reasons have been performed in helium. Another significant aspect of in-reactor experiments performed in helium is the radiation-induced leakage current in the gas,53 which makes it difficult to

determine volume conductivity.81’104 One should also mention that severe electrical surface degradation has recently been observed when oxide insulator materials are bombarded with keV H and He ions.105 The mechanism giving rise to such surface degrada­tion is believed to be the loss of oxygen from the vacuum insulator surface region due to preferential radiolytic sputtering. Similarly’ in future fusion devices such as ITER ceramic insulators and win­dows may also degrade’ as they will be bombarded by energetic H isotope and He ions because of ionization of the residual gas by g radiation and acceleration by local electric fields.54 At the present time’ the role of the irradiation environment in electrical degradation clearly requires further study. Additional difficulties experienced in performing in-reactor experiments include temperature control and also component testing.104’106-108 It is also impor­tant to note that several in-reactor experiments have suffered from electrical breakdowns related to the difficulty of maintaining high voltages in a radiation field, precisely what is required for some H&CD and diagnostics systems in a next-step device. Whether or not these are due to RIED, temperature excursions, He gas breakdown, or problems with the MI cables, terminations, and feedthroughs remains unexplained.

As concrete ages, changes in its properties will occur as a result of continuing microstructural changes (i. e., slow hydration, crystallization of amorphous constitu­ents, and reactions between cement paste and aggre­gates), as well as environmental influences. These changes do not have to be detrimental to the point that concrete will not be able to meet its performance requirements. Concrete, however, can suffer unde­sirable changes with time because of improper spec­ifications, a violation of specifications, or adverse performance of its cement paste matrix or aggregate constituents under either physical or chemical attack.

Concrete durability and the relationship between durability and performance, a review of the historical perspective related to concrete and longevity, a description of the basic materials that comprise rein­forced concrete, and information on the environmental factors that can affect the performance of NPP con­crete structures have been provided. Primary aspects related to management of aging of NPP concrete structures have been noted: degradation mechanisms, damage models, and material performance; assessment and repair (i. e., component selection, in-service inspection, nondestructive examinations, and remedial actions); and estimation of performance at present or some future point in time (i.e., application ofstructural reliability theory to the design and optimization of in­service inspection/maintenance strategies, and deter­mination of the effects of degradation on plant risk).

Several areas have been identified where addi­tional research would be of benefit to aging manage­ment of NPP concrete structures: (1) compilation of material property data for long-term performance and trending, evaluation of environmental effects, and assessment and validation ofnondestructive eval­uation methods; (2) evaluation of long-term effects of elevated temperature and radiation on concrete behavior; (3) improved damage models and accep­tance criteria for use in assessments of the current condition as well as estimation of the future condition of the structures; (4) improved constitutive models and analytical methods for use in determination of nonlinear structural response (e. g., accident conditions); (5) nonintrusive methods for inspection of thick-walled, heavily reinforced concrete structures and basemats; (6) global inspection methods for metal­lic pressure boundary components (i. e., steel contain­ments and liners of concrete containments) including inaccessible areas and backside of liner; (7) data on application and performance (e. g., durability) of repair materials and techniques; (8) utilization of structural reliability theory incorporating uncertainties to address time-dependent changes to structures to ensure that minimum accepted performance require­ments are exceeded and to estimate on-going compo­nent degradation to estimate end of life; and (9) application of probabilistic modeling of component performance to provide risk-based criteria to evaluate how aging affects structural capacity.

The following definitions are used:

• ps = density of the pebble base material

• pp = pebble density

• Pp = total pebble porosity = average fraction of porosity within (macroscopic, single) pebble volume

• Ppb = pebble-bed density = ratio of pebble-bed mass mpb to pebble-bed volume Vpb

• g = packing factor = ratio of pebble volume to pebble-bed volume

The quantities are connected by

ppb mpb/Vpb (1 Pp)psg [1 ]

In literature, the term ‘fraction of theoretical density’ is also used. This term is identical to (1—Pp). The pebble-bed density ppb is often named ‘apparent’ or ‘tap’ density. The pebble-bed density ppb is an impor­tant quantity for nuclear calculations as these require the specific density of lithium and other constituents as inputs. In pebble-bed engineering, the packing factor g (generally given in percentage) is the charac­teristic quantity.

The packing factor is influenced by the following parameters: filling procedure, pebble shape, surface roughness diameter d, diameter distribution, and con­tainer (cavity) dimensions: height H, diameter D.

Fillingprocedure. The selection of an optimum pro­cedure is not trivial, especially for larger mock-ups: filling should be assisted by vibration in such a way that a particle flow within the cavity is prevented or, in case of too small vibration energies, friction forces between pebbles and walls can be overcome.

Pebble shape and surface roughness: ceramic breeder pebble shapes are almost spherical, partly with in­dentations (melting, spraying processes) or egg-like (extrusion, spheronization, sintering processes). The pebbles originating from melting have a smoother surface than the others. The packing factor is influ­enced by the surface roughness but negligibly by the pebble shape.

Diameter distribution: The diameter d is of influence with respect to the cavity dimensions, discussed later. Ideally, the packing factor increases with increasing diameter distribution; see, for example, McGeary.87

However, sedimentation during the filling is a critical issue.

For blanket pebble beds, two groups are of inter­est: (1) Pebble beds with a relatively small d variation: for example, Li2TiO3 pebbles with a nominal diame­ter of 1 mm ranging from 0.8 to 1.2 mm and Li4SiO4 pebbles ranging from 0.25 to 0.63 mm. Production costs generally favor a wider spread, which is not detrimental in respect to the packing factor, at least for the diameter ranges cited earlier. Packing factors in the range of 63-66% are reached; (2) Binary beds: here, large pebbles, dl, and small pebbles, ds, are used, with ds« 0.1dl. First, the large pebbles are vibrated in the cavity. Then the pebbles are fixed in their posi­tion, for example, by a sieve in the filling pipe, to avoid sedimentation during the following pouring in ofthe small pebbles, which fill the spaces between the large pebbles. Packing factors of about 82% were obtained for beryllium beds with dl« 2 mm, ds« 0.2 mm.94 Critical issues are (1) the stability of such pebble packing structure during thermal cycles in the blanket and (2) the buildup of large stresses due to irradiation-induced swelling causing pebble deg­radation, leading to fracture.

Diameter d, bed height H, diameter D: For finite cavities, there exist two characteristic ratios: H/d and D/d. The reason why g depends on H/d and D/d is the existence of two different characteris­tic pebble packing structures as demonstrated by

tomography experiments101,105,106: in the pebble

bulk, there is no preferential direction of contacts between pebbles, whereas at the wall, a zone with a thickness of about 4d exists, where regular packing structures are observed and contact zones are no longer homogeneously distributed.

Figure 17 shows tomography results106 for a cylindrical container (D = 49 mm, H= 50 mm) filled with spherical pebbles (d = 2.3 mm). Distinct pebble layers close to the walls can be clearly seen in Figure 17(a) where the positions of the pebble cen­ters are plotted in a horizontal plane (left) and a vertical plane (right).

Figure 17(a) shows axial void fraction distribu­tions. The degree of regularity is largest for the pebbles in contact with the wall and decreases with increasing wall distance. The regularity is most expressed for pebbles in contact with plane walls (D = 1); here, large isles with a dense hexagonal pebble arrangement are observed; see Reimann et al}05 The average packing factor of a plane wall zone is about equal to the bulk packing factor if the bed height is sufficiently large for the two wall zones

Figure 17 Visualization of pebble-wall layers by tomography: (a) Positions of sphere centers in region of interest (ROI) and b) Void fraction distributions. Reproduced from Pieritz, R. A.; Reimann, J.; Ferrero, C. Adv. Eng. Mater. 2011,13, Nr.3, 145-155.

to interfere. This is an important result because it demonstrates that the often cited statement that wall zones are characterized by reduced packing factors is not generally valid. However, with decreasing D, both the regularity of the packing and the packing factor decrease, as seen clearly in Figure 17(b). Corners in pebble-bed cavities and inserts, for example, thermocouples, generally decrease the packing factor. This is also a reason why in small experimental setups the achieved packing factor is often distinctively below the maximum value. Pebble beds in HCPB blanket concepts are typi­cally ‘thin’ pebble beds, that is, one dimension; the bed height H, is small compared with the other dimensions. For the ceramic breeder pebble beds, H« 10 mm; for beryllium, H« 30 mm. For a height of H« 10 mm and a pebble diameter of ~1 mm, the maximum packing factor is already difficult to achieve.

Best location for filling

X

Figure 18 Pebble-bed filling experiments relevant for ITER Test Blanket Modules. Reproduced from Abou-Sena, A.; Neuberger, H.; Ihli, T. Fusion Eng. Des. 2009, 84, 355-358.

A disadvantage of a large pebble diameter is also the smaller heat transfer coefficient, as outlined below.

The blanket modules will probably be filled with pebbles through small pipes, as schematically shown in Figure 18. For simple cubic cavities, a homogeneous filling was achieved only by carefully vibrating the container in different tilted positions.97,100

The filling of a TBM mock-up using glass balls for the beryllium pebble beds was investigated by Abou-Sena eta/.,107 by varying the number and posi­tions of the filling pipes. The best results were obtained by using a single filling pipe close to the box corner and, keeping the filling pipe at the highest position, tilting the box around two axes. Here, a packing factor of 63.6% was achieved, which is considered to be very close to the maximum obtain­able value.

In the HELICHETTA experiment (see Dell’Orco et a/.108,109) where elongated pebble-bed cavities (10 x 100 x 480 mm3) were filled through a pipe at the top (largest dimension orientated vertically), packing factors of 60% and 62% were achieved for Li2TiO3 pebbles with diameters between 0.8 and

1.2 mm, and 0.6 and 0.8 mm, respectively. For Li4SiO4 pebbles with d between 0.25 and 0.63, the packing factor was more than 64%.

The only report on the solubility of hydrogen in boron carbide is that by Shirasu eta/206 They exposed crystals of boron carbide to hydrogen gas at various temperatures and pressures for 20 h, and subse­quently outgassed them during anneals in which the temperature was linearly increased at the rate of 20 K min-1 up to 1273 K. The uptake was seen to increase with the square root of pressure, and to decrease with increasing temperature (exothermic). Schnarr and Munzel207,208 measured the diffusivity of tritium in both irradiated and unirradiated boron car­bide. While the actual expression for the diffusivity for each case was not given, it can be extracted from the figures. It was noted that the apparent diffusivity decreased with increasing radiation damage until the percentage of 10B exceeded 10%. Elleman eta/.209 used the 6Li-neutron reaction to generate tritium profiles in samples of boron carbide. Diffusivity was determined by examining the rate of release of the tritium during isothermal anneals at elevated temperatures.

The diffusivity and the solubility of hydrogen in silicon carbide (a material described in Chapter 2.12, Properties and Characteristics of SiC and SiC/ SiC Composites and Chapter 4.07, Radiation Effects in SiC and SiC-SiC) have been measured twice by Causey et a/.21 ’ 1 In the first set of experi­ments,210 various grades of silicon carbide were implanted with tritium, using the neutron reaction with 6Li on the sample surfaces. The diffusivity for each material was then determined by fitting the release curves determined during isothermal anneal to those predicted by the analytical solution to the diffusion equation. The results were seen to differ strongly depending on the type and purity of the silicon carbide. As an example, the measured diffusiv — ity in hot pressed and aluminum-doped a-silicon car­bide was approximately five orders of magnitude greater than that in vapor-deposited p-silicon carbide at 1273 K. The lowest diffusivities were reported for vapor-deposited p-silicon carbide and single-crystal a-silicon carbide. In all cases, the activation energy of the diffusivity was >200 kJ mol-1 (suggesting that chemical bonding plays a strong role in the diffusion). For the diffusion of tritium in vapor-deposited silicon carbide, the diffusivity was given as D = 1.58 x 10- exp(-37 000/T) m2 s — . Deuterium solubility was also determined for the vapor-deposited silicon car­bide. The values were determined by exposing samples at elevated temperatures to deuterium gas followed by outgassing to determine the amount ofuptake. Because equilibrium retention was not obtained in the experiments, inherent in the calculations was the assumption that the diffusivity values determined in the implantation experiments were valid in the gas­eous uptake experiments. The amount of uptake was assumed to be the product of diffusivity, the solubil­ity, the sample area, and the square root of pressure. The solubility was given as K = 1.1 x 10- exp (+18 500/T) mol H2 m-2 MPa-1/2. Again, the nega­tive value of the activation energy would suggest chemical bonding of the hydrogen to the host mate­rial. In the later work by Causey et a/.,2 vapor — deposited silicon carbide was again tested. In these experiments, the implantation of energetic particles into the silicon carbide was avoided. Samples were exposed to gas containing 99% deuterium and 1% tritium at a temperature of 1573 K for 1 h. The sam­ples were subsequently outgassed at temperatures from 1373 to 1773 K. The outgassing rates were then fitted to release curves predicted by the solution to the diffusion equation to determine the diffusivity. In this case, the diffusivity was given by the expres­sion D = 9.8 x 10-8 exp(-21 870/ T) m2 s-1, one to two orders of magnitude faster than the values deter­mined earlier with energetic particles.21 The solu­bility was also determined in this study. Samples were exposed to the deuterium/tritium gas at tem­peratures from 1273 to 1873K for sufficient duration to achieve equilibrium loading. The samples were then outgassed to determine this equilibrium amount. The expression for the solubility in this case was K= 2.2 x 10-2 exp(+7060/T) mol H2 m-3 MPa-1/2. This solubility is one to two orders of magnitude lower than the one determined in the earlier experi — ments.210 If one assumes the migration of hydrogen in silicon carbide to occur along active sites on the edges of the grains, it is not unexpected that radiation damage produced by the implantation of energetic particles would increase the apparent solubility and proportionately decrease the apparent diffusivity. If hydrogen can exist only on the grain bound­aries by being attached to trap sites, higher trapping means higher apparent solubility. Conversely, higher trapping means slower diffusion. It was the apparent higher solubility on small-grained samples that led Causey et a/.211 to propose the trap-controlled grain boundary diffusion model.

The permeation of hydrogen isotopes through sili­con carbide has been measured by several groups.212-21 Verghese eta/.213 measured the permeation of a hydro — gen/tritium mixture through a KT silicon carbide tube that was manufactured by wet extrusion and sin­tering. The permeability reported for the experiments
is given by Ф = 3.8 x 108 exp(—66 000/ T) mol H2 m-1 s-1 MPa-1/2. Sinharoy and Lange212 measured the permeation of hydrogen through a tungsten tube with a CVD coating of silicon carbide. The retarding effect of the tungsten was taken into consideration in the calculation. The recorded permeation for these experiments was Ф = 2 x 10-4 exp(-6830/T) mol H2 m-1 s-1 MPa-1/2. Yao eta/.214 performed perme­ation experiments on a steel sample that had been RF sputter-coated with silicon carbide. The thickness of the coating was estimated to be 1.3 pm and contained several percent oxygen and traces ofiron. The coating was seen to decrease the permeation rate of steel by about two orders of magnitude, but did not change the activation energy. In this case, the coating was clearly porous, and the reduction in permeation was simply due to a reduction in the effective permeation surface area. The plot of the permeation values for the vapor — deposited silicon carbide by Causey et a/.21 (calcu­lated as the product of diffusivity times solubility), KT silicon carbide by Verghese et a/.,213 and CVD silicon carbide by Sinharoy and Lange212 is shown in Figure 20. The differences in the absolute values of the permeability as well as the differences in the activation energy of the process are extreme. It is difficult to even imagine that the values are for the same material. In fact, the materials are not the same. As mentioned for the original study by Causey eta/.,2 differences in impurities play a significant role in determining the behavior of hydrogen in silicon car­bide. If hydrogen does migrate along the grain bound­aries, impurity metals along those grain boundaries reduce the fraction of migrating hydrogen chemically bound to the silicon carbide. Likewise, the apparent diffusivity would be much more rapid if hydrogen trapping at the grain boundaries is reduced. In the case of the permeability measured by Sinharoy and Lange,2 2 it is difficult to believe that the measured permeation is not really controlled by permeation through the underlying tungsten with the specific surface area limited by the porous silicon carbide coating. The activation energy for the permeation in the report by Verghese eta/.213 is difficult to under­stand. The value of 555 kJ mol-1 is even greater than the chemical bond of hydrogen to carbon.215 The permeation was seen to vary by as much as an order of magnitude at the same temperature. There is no apparent explanation for the rapid change in perme­ation with temperature.

Titanium carbide has also been tested as a perme­ation barrier. Due to adhesion problems with direct deposition on steel, titanium nitride was used as an intermediate layer between the steel and titanium carbide. Forcey et a/.202 measured deuterium perme­ation through 3-pm thick layers of TiC and TiN on steel, observing a PRF of ten. For the experiments performed over the temperature range of 550-740 K, extended defects were listed as the reason for the relatively small improvement over bare steel. Checchetto eta/.201 used ion-beam assisted deposition of TiN-TiC films on steel in their permeation experiments. When the film was deposited on the downstream side, little reduction in permeation was seen. Using the deposited film on the upstream side

did yield a PRF of ^50. Shan et at216 used a CVD process to deposit their 2.5-pm thick film on steel and noted a permeation reduction of five to six orders of magnitude. It is obvious from these three studies that the deposition oftheoretically dense thin films is very difficult. There is also the question of cracking of such thin films during thermal cycling. This is dis­cussed later in this chapter.

L. L. Snead

Oak Ridge National Laboratory, Oak Ridge, TN, USA

M. Ferraris

Politecnico di Torino, Italy © 2012 Elsevier Ltd. All rights reserved.

Introduction

Background

Plasma-Facing Materials Particle-Matter Interactions The Advantages of Carbon as a PFC Plasma Impurities and the Need for Graphite Materials Thermomechanical Loading of PFMs Transient Loading of PFMs

Irradiation Effects on Thermophysical Properties of Graphite and CFCs

Graphite Irradiation Damage Surface Effects

Properties and Property Evolution of Graphite Fiber Composite

Irradiation-induced dimensional changes in CFCs Irradiation-induced changes in strength and modulus Thermal conductivity degradation Plasma-Particle Interactions Chemical Erosion

Doping of Graphite to Suppress Erosion Physical Sputtering Radiation-Enhanced Sublimation Erosion of Graphite in Simulated Disruption Events Tritium Retention in Graphitic Materials HHF Component Technology Joining of CFC to Heat Sink Evaluation of HHF Joint Summary and Conclusions

4.18.1 Introduction

4.18.1.1 Background

Graphite-moderated, gas-cooled reactors led the way into the nuclear age starting with the Chicago Pile-1 reactor, where the first controlled and sus­tained critical nuclear reaction was initiated in December 1942. The first commercial nuclear power plant, Calder Hall in the United Kingdom, went critical in 1956. As the graphite moderator was literally at the core of these early reactors, graphite became one of the first and most exten­sively studied nuclear materials. As discussed in Chapter 4.10, Radiation Effects in Graphite, the fission-born neutron results in significant thermo­physical property changes in graphite. Moreover, depending on the type of fission reactor, other environmental factors such as graphite oxidation become extremely important. In addition to being the moderator of gas-cooled reactors, graphite has found a number of new nuclear power applications. As examples, pyrolytic graphite is a key functional element in TRi ISOtropic (TRISO) fuels, which con­tinue to be developed and utilized for gas-cooled reactors; carbon fiber composites (CFCs) are now under development for core application in high — temperature gas-cooled reactors1 and have been widely used as plasma-facing components (PFCs) in fusion reactors.2 The latter application began in 1978, when the Princeton Large Torus made a transition from tungsten to graphite ‘limiters.’ This enabled the first thermonuclear temperatures, beginning the widespread application of graphite materials in fusion systems, the subject of this chapter. As will be discussed, the primary motivation for the use of graphite in fusion systems is not (as in fission reac­tors) for neutron moderation, but for reasons related to its exceptional high temperature performance and its relatively innocuous interaction with the plasma. However, the fusion reactor environmental effects on graphite, including irradiation-induced property evolution, are very similar to those of their fission reactor analogs.

In contrast to the fission of heavy elements such as uranium or plutonium, which releases a large amount of energy in their fission fragments and a moderate amount of energy in the form of neutron kinetic energy (mean about 1 MeV), fusion can occur for a number of light elements, some of which have reac­tions that release very high-kinetic-energy neutrons. Several possible routes to fusion are shown below in

eqn [1]:

1H1 +1 H1 !1 D2 + positron = 1.4MeV

1H1 +1 D2 !2 He3 = 5.5 MeV

1H1 +1 T3 !2 He4 = 19.9 MeV

1D2 +1 D2 !2 He3 + neutron = 3.3 MeV

1D2 +1 D2 !1 T3 +1 H1 = 4.0MeV

1D2 +1 T3 !2 He4 + neutron = 17.6MeV

1D2 +2He3 !2 He4 +1H1 = 18.2 MeV I1]

For any of these reactions to take place, the ionized atoms must be brought together with sufficient force to overcome the coulombic barrier. In thermonuclear fusion, this is accomplished by heating the ‘plasma’ of these atoms to the point where the kinetic energy is sufficient to overcome that barrier. Currently, it is thought that D+T fusion is the most accessible route to fusion, though the gaseous temperature required for D+T reaction is more than 50 million Kelvin.

Control and containment of high-temperature, high-density fusion plasmas is the primary challenge and obstacle to fusion power. Many reactor concepts have been studied in the past and attention is now focused on the ‘tokamak’ system. This toroidal con­finement machine system was developed in the mid-1960s in Russia. In this design, a high-strength twisted helix of magnetic lines forms a magnetic bottle. Ions, which are trapped within a certain gyro — radius, travel along these lines circulating around the helix in opposition to the plasma electrons. For non — collisional plasmas, the ions can be heated by mag­netic induction or through various external means to the extreme temperature necessary for the fusion reac­tion to take place. This concept is the basis for the four largest present-day fusion machines (Table 1), and is the premise for the ITER machine currently under construction. To give an idea of scale, in all of the present-day machines listed in Table 1, the helical cavity is big enough in size for an adult to walk within, and the radius from the center of the machine to the middle of the helix is typically several meters. A depiction of the inside of the JET torus, complete with beryllium-coated CFC wall, is given in Figure 1.

4.19.4.1 General Considerations

A comprehensive, although not recent, review of the science and technology of beryllium can be found in Beryllium Science and Technology.120

Several reviews have been published recently related to use of beryllium in tokamaks and the status of the investigations of the Be properties for the fusion application.3’121-126 Various production and proces­sing methods of beryllium metal fabrication have been reviewed in Dombrowski.127 The majority of methods are based on powder metallurgy and include powder preparation from cast product by grinding (i. e., attrition milling, impact grinding, ball mill grinding); further powder consolidation (i. e., by cold pressing (CP), cold isostatic pressing (CIP), vacuum hot pressing (VHP), hot isostatic pressing (HIP)); and possible additional mechanical treatment (e. g., extru­sion, rolling, forging). Beryllium protective armor can also be produced by plasma spray (see Section 4.19.4.3) and vapor deposition.

Several proposals were made at the beginning of the ITER Research Programme during the ITER Engineering Design Phase to develop a fusion grade beryllium with high ductility, high resistance to heat flux, and high radiation resistance. However, it was recognized that this development would require sig­nificant efforts and could not be supported only by requests from the fusion community.

There are various beryllium grades, which have been developed for different applications. These grades differ by chemical composition (BeO content, impuri­ties), by method of powder preparation, by method of consolidation, etc. The nonexhaustive list of various beryllium grades from the US and the Russian Feder­ation is presented in ITER Materials Properties Handbook (MPH).12 Grades with similar composi­tion are under production in Kazakhstan and in China.

We briefly discuss below some of the most relevant physical and mechanical properties of beryllium, in relation to its application as armor for PFCs.