Metals are characterized by a very low electrical resistance, which increases with temperature and about doubles as a consequence ofmelting. In general, the electrical resistivity of LM increases, when impu­rities enter the melt. However, in the case of a liquid

Figure 11 Dynamic viscosity of liquid Na, Pb-Bi(e), and Pb at normal atmospheric pressure.

300 500 700 900 1100 1300 1500 1700 1900 2100

Temperature (K)

Figure 12 Kinematic viscosity of liquid Na, Pb-Bi(e), and Pb at normal atmospheric pressure.

alloy system that is composed of polyvalent compo­nents’ the resistivity sometimes shows a negative deviation from the additivity of the components’ resis­tivities. The electrical resistivity of LM with rare exceptions (such as Na) increases almost linearly with temperature below the boiling temperature; at high temperatures close to the boiling point, it can increase more rapidly. In many cases, the parabolic function can be used:

r(T;p) = ru + Ar(T — TM) + Br (T — Tm)2 [20]

In 1985, the data on the electric resistivity of liq­uid Na were estimated by Ho and James.82 Later, Bretonnet83 reviewed different sources to obtain cor­relations for the electric resistivity of many pure LM and found that a linear function could be used in most cases. However, for Na, the use of a parabolic function was proposed, which describes its electric resistivity with an uncertainty <4% in the temperature range from normal melting to normal boiling point.34

The electric resistivity of liquid lead at normal atmospheric pressure was measured rather well from the melting point up to about 1300 K. In this temper­ature range, it can be described with a linear function with an uncertainty <2%.34 The precision of data obtained at high temperatures is lower, and they indi­cate a more rapid increase of the Pb electric resistivity with temperature.84 Only a few reliable data sources exist on the electric resistivity of Pb-Bi(e), which are limited to a temperature of 1073 K.25,70,85 The differ­ence between these sources is about 7-8%.

The coefficients of correlation [20] for liquid Na, Pb, and Pb-Bi(e), at normal atmospheric pressure, taken from Sobolev,34 are presented in Table 12;

Figure 13 illustrates the temperature dependence of the calculated electrical resistivities.

Any burnable poison material needs to meet a num­ber of important materials property requirements if it is to be used in nuclear fuel. Such materials must be physically stable under high-temperature conditions, they must not be susceptible to corrosion in the harsh chemical environment, and they must be compatible with the other materials present in nuclear fuel, such as UO2, Zircaloy, stainless steel, and the coolant materials water, steam, or gas as appropriate to the reactor type. The burnable poison materials must also be well behaved under the intense neutron irra­diation and gaseous releases must be accommodated. This latter consideration is particularly important for boron, because the neutron absorption leads to helium production through the reaction

10B + n!7 Li +4 He

If used in discrete poison rods, there must be suffi­cient free volume to accommodate the helium release without overpressurization. With IFBA, the helium released accumulates in the plenum volume provided to accommodate the gaseous fission products and adds to the end-of-life rod internal pressure.

Additionally, burnable poisons must be designed so that there is no possibility that the poison material could melt or slump out of the active core in an accident scenario or in any way interfere with the insertion of control rods. The migration of absorber material away from the active core would add reac­tivity during the accident scenario that could poten­tially worsen the consequences of an accident.

The zirconium diboride coating thickness used in IFBA fuel is thin enough that there is only a very small temperature drop across it and virtually no impact on fuel temperatures and fuel behavior, apart from the higher internal gas pressure due to helium generation noted earlier. In contrast, the
presence of a second ceramic phase in the fuel pellets, as is the case for gadolinia, erbia, and dysprosia, does have an impact on fuel pellet behavior, as discussed in the next section.

Many of the properties that make graphite attractive for a particular application have been discussed above. However, the following characteristics have been ascribed to synthetic, polygranular graphite6 and are those properties that make graphite suitable for its many applications: chemical stability; corro­sion resistance (in a nonoxidizing atmosphere); non­reactive with many molten metals and salts; nontoxic; high electrical and thermal conductivity; small ther­mal expansion coefficient and consequently high thermal shock resistance; light weight (low bulk den­sity); high strength at high temperature; high lubric­ity; easily dissolved in iron, and highly reductive; biocompatible; low neutron absorption cross-section and high neutron-moderating efficiency; resistance to radiation damage. The latter properties are what make graphite an attractive choice for a solid moder­ator in nuclear reactor applications.

Nuclear applications, both fission and fusion (of keen interest the reader), are described in detail in Chapter 4.10, Radiation Effects in Graphite, and Chapter 4.18, Carbon as a Fusion Plasma­Facing Material. Accounts of nuclear applications have also been published elsewhere.9, — 0 Graphite is used in fission reactors as a nuclear moderator because of its low neutron absorption cross-section and high neutron moderating efficiency, its resistance to radia­tion damage, and high-temperature properties. In fusion reactors, where it has been used as plasma facing components, advantage is taken of its low atomic number and excellent thermal shock characteristics.

The largest applications of nuclear graphite involve its use as a moderator and in the fuel forms of many thermal reactor designs. These have included the early, air-cooled experimental and weapons materials producing reactors; water-cooled graphite-moderated reactors of the former Soviet Union; the CO2-cooled reactors built predominantly in the United Kingdom, but also in Italy and Japan; and helium-cooled
high-temperature reactors, built by many nations, which are still being operated in Japan and designed and constructed in China and the United States. All of the high-temperature reactor designs utilize the ceramic Tri-isotropic (TRISO) type fuel (see Chapter 3.07, TRISO-Coated Particle Fuel Performance), which incorporates two pyrolytic graphite layers in its form. Graphite-moderated reactors that were molten-salt cooled have also been operated.

Electrical resistivity of ZrCx is plotted as a function of temperature in Figure 16. Room temperature resistivity ranges from 60 to 200 pQ cm, depending on C/Zr ratio and microstructure. In an intermediate temperature range from approximately 100 to 2000 K, resistivity increases linearly with temperature.45,46,56,57 Modine et a/58 measured resistivity of single crystal ZrCx (x = 0.89, 0.93, and 0.98) between 4 and 1000 K. The authors deemed the data well represented by the Bloch-Gruneisen model for temperature depen­dence of resistivity of metals, with resistivity varying as T5 at low temperatures (4-100 K) and linearly at intermediate temperatures. At a high enough tem­perature (1000-2000 K), resistivity deviates from linear behavior and tends to saturate at a constant value which decreases with C/Zr ratio. The higher — temperature measurements on single-crystal ZrC0.93 of Hinrichs et a/.59 are consistent with the trend established for single crystal ZrC0.93 at lower tem­peratures by Modine et a/. (Figure 16).

E

о

a

З

>.

>

Temperature (K)

Figure 16 Electrical resistivity of ZrCx as a function of temperature.

The resistivity of single crystals exceeds that of polycrystals up to 2200-2500 K where the former begins to saturate; resistivity of polycrystalline ZrCx saturates only near the melting temperature, although few measurements have been made in this tempera­ture range. The effects of free carbon and oxygen/ nitrogen impurities on resistivity have not been explored. Measurements on pyrolytic ZrCx53 lie in the same range as those of other polycrystalline speci­mens, but a detailed study of the effects of grain size, texture, porosity, and other microstructural factors on electrical resistivity is needed.

Room temperature electrical resistivity as a function of C/Zr ratio is plotted in Figure 17. Resis­tivity is lowest for near-stoichiometric compositions and increases with deviation from stoichiometry. A decrease in C/Zr ratio increases the concentration ofcarbon vacancies, which scatter conduction electrons. Storms and Wagner35 fit the available experimental data to the formula 1 1

0.00382 +

55 + where p is electrical resistivity (p. Q cm) and x is C/Zr ratio, which is plotted in Figure 17.

The flow sheet for UO2 pellet production is shown in Figure 10. The UO2 pellet fabrication process con­sists of mixing the UO2 powder with additives such as binder, lubricant and pore former materials, gran­ulating to form free-flowing particles, compaction in an automatic press, heating to remove the additives, sintering in a controlled atmosphere, and grinding to a final diameter. The process varies slightly according to the nature of the starting UO2 powder.

2.15.4.2.1 Powder preparation

In the pelletizing process, UO2 powder must be filled easily and consistently into dies. UO2 powder from the AUC process is free-flowing and can be pressed

without granulation. Usually it is mixed with a small amount of U3O8 to control the density and pore distribution of the pellets. The fine particle size of the integrated dry route (IDR) powders prevents them from being free-flowing when produced; these powders are therefore prepressed into briquettes, fractured, sieved to produce granules, and a dry

lubricant added. ADU powder is slurried with a sol­vent and a volatile binder such as polyethylene glycol or polyvinyl alcohol, spray dried and sieved to size. The obtained material flows freely and will con­sistently fill pellet dies but an extra operation is required to remove the binder. Additives known as pore formers are often included to give uniform final density: 95-97% TD for LWR UO2 and MOX, and 85-95% TD for FBR MOX fuel pellets. The pore former will decompose in the dewaxing process to leave closed pores that are stable in-reactor.

The fresh fuel plutonium concentration and distribu­tion (homogeneous or heterogeneous microstructure) and the induced variations in the local stoichiome­try are supplementary parameters for MOX fuels. The plutonium concentration decreases with burnup whereas it increases in UO2. Therefore, the differ­ences observed in fresh fuels linked to the plutonium effect can be expected to decrease with burnup, at least for homogeneous MOX fuels. In hetero­geneous MOX, the microstructure may have an enhanced impact because of the presence of Pu-rich agglomerates in which the burnup is much higher than in the matrix. However, recent studies showed that this effect is in fact negligible.27,28

2.17.2.3.2 Microstructure

The microstructure of the fuel is linked to the pres­ence of additives and porosity, voids, and fission gas bubbles, their concentrations changing with the radial position. Other parameters concern the fuel matrix itself: the densification at the beginning of life and the grain size evolution. The effect of poros­ity and voids can be approximated by applying a factor to the matrix conductivity, determined for instance by the formula of Maxwell giving the effec­tive thermal conductivity 1eq of a medium consti­tuted by a matrix of conductivity 1m containing a small volume fraction vf of inclusions having a con­ductivity 1f (eqn [3]). This approximation is obtained only when 1m ^ If, and is therefore not usable for ceramic or metallic precipitates. Approximate formu­lae, validated for porosity in nuclear fuels, were pro­posed by Brandt and Neuer.29

l if + 2lm + 2vf(if — im) , 2(1 — vf) 3

1eq “ 1m if + 21m + Vf (1m — if) ifiC^if 1m^ + VT ^

The macroscopic cracking and fracture of the pellet are heterogeneities that cannot be taken into account for the definition of an equivalent thermal conduc­tivity because their size is not small compared to the size of the pellet. Cracking can be separated into radial and tangential, the latter decreasing the appar­ent thermal conductivity.

Extensive microstructure changes take place with the formation of the HBS starting from the periphery of the pellet for discharge burnups higher than about 40 MWd kg HM—1 or local burnups higher than about 60 MWd kg HM—1. This structure is characterized by a reduced grain size, an increase in porosity, and a depletion of fission gas from the UO2 matrix.30,31

2.17.2.3.3 Burnup

The fuel burnup reflects the proportion of fissioned atoms and is the most commonly used parameter for the interpretation of thermal conductivity degrada­tion. In principle, this parameter integrates the phe­nomena that do not depend on the irradiation conditions, that is, the concentration of nonvolatile fission products. In practice, it is often the only param­eter of the correlations and is used to account for (almost) all irradiation effects. Some particular effects exist at low burnup: the fast buildup of irradiation damage, and at high burnup: the formation ofthe HBS.

Silicon carbide has a myriad polytypes depending on the varied stacking of closed atomic planes.23 Only CVD SiC material is inherently highly crystalline, pure, and stoichiometric, which is critical to irradia­tion stability. Much emphasis is placed on CVD SiC in this chapter, as it corresponds very closely to the matrix of CVI SiC/SiC. The reader will find further details on the SiC structure-property relationships in the excellent comprehensive review by Snead and colleagues.23 Here the main data from Snead’s paper are summarized.

Only the 3C-SiC crystal, known as p-SiC, has the sequence showing cubic symmetry out of the infinite number of variations. All the other polytypes which show noncubic symmetry are classified as a-SiC. a-SiC is formed above 2373 K and p-SiC at 1273-1873 K.

Various fabrication techniques, such as sintering, direct conversion, gas-phase reaction, and polymer pyrolysis, are currently used for the synthesis ofSiC. The CVD technique is one ofthe most familiar gas — phase reaction methods for the synthesis of highly crystalline, stoichiometric, high-purity p-SiC.

As this review has illustrated, ZrC possesses a combi­nation of thermodynamic, thermal, and mechanical properties that are promising for nuclear fuel appli­cations requiring high-temperature resistance and structural integrity. However, it is also clear that more data are needed. The body of mechanical prop­erty data is limited. The degree of scatter in experi­mental data indicates that methods for fabricating dense, pure, homogeneous, stoichiometric ZrC are not mature. Properties of ZrC are known to be affected by oxygen and nitrogen impurities, but the thermodynamics of the Zr-C-O-N system has not been elaborated. The effects of irradiation on prop­erties and performance must be more thoroughly characterized. ZrC is stable over a large range of nonstoichiometry, which is promising for irradiation damage tolerance, but carbon vacancies are shown to cause a decrease in bond strength, reflected in decreased heat capacity and hardness and increased CTE. The introduction of lattice defects also reduces thermal and electrical conductivity. With the renewed interest in ZrC for advanced composite nuclear fuels, much work lies ahead in building the required knowl­edge base in both processing and performance that will enable the nuclear community to take full advan­tage of ZrC as a high-temperature structural ceramic. Chapter 1.02, Fundamental Point Defect Proper­ties in Ceramics; Chapter 1.05, Radiation-Induced Effects on Material Properties of Ceramics (Mechanical and Dimensional); and Chapter 3.08, Advanced Concepts in TRISO Fuel.

2.14.1 Introduction

Liquid metals (LM), such as sodium (Na), lead (Pb), and lead-bismuth (Pb-Bi) eutectic (e), are considered as potential coolants for the fast spectrum nuclear reactors of the next generation.1 In the period 1960­1980, a lot of studies were performed for the creation of adequate databases of thermophysical properties of Na in the frameworks of development, construc­tion, and operation of liquid metal (cooled) fast breeder reactors LMFBR in the United States, European Union, and in the former USSR.2 Most of these results were collected later and published in review reports and handbooks.3- 1 Since that time, the interest for fast reactors with Na coolant increased significantly worldwide, especially after the launch of the Generation IV International Forum (GIF) initia­tive,1 where the sodium fast reactor (SFR) is considered as the main candidate for future nuclear power plants, which can be used for both electricity production and transuranium elements (TRU) incineration — a way for closing the nuclear fuel cycle. In the GIF documents, the lead-cooled fast reactor (LFR — cooled by Pb or Pb-Bi) is considered as the second candidate.1 An interest to use Pb and Pb-Bi(e) in the civil power reactors appeared mainly after communications in open literature about the preliminary design studies on BREST12 and SVBR13 reactors in the Russian

Federation. LFR systems are now considered in few other countries: PBWFR, 14 SLPLFR,15 and CAN­DLE16 in Japan, PEACER17 and BORIS18 in Korea, SSTAR19 in the United States, and ELSY20,21 in the European Union. At present, the available data on thermophysical properties of Pb and Pb-Bi(e) in the temperature range of interest are still incomplete and often contradictory. This complicates the design calcu­lations and the prediction of the normal and abnormal behavior of nuclear installations where they will be used. Intensive studies have been performed in differ­ent countries aiming at better understanding of their properties needed for design and safety analysis of the nuclear installations. Recently, a review of the Na properties was performed in Argonne National Labo — ratory22; compilations with the recommendations for the properties ofPb and Pb-Bi(e) were prepared by the WPFC (OECD) Expert Group on Heavy Liquid Metals Technology23,24 and for all three LM ofinterest by expert groups of the IAEA.25,26

This chapter gives a brief review of the compila­tions and recommendations developed for the main thermophysical properties of Na, Pb, and Pb-Bi(e), which include melting temperature, boiling temper­ature, critical point parameters, saturated vapor pres­sure, melting and boiling enthalpies, surface tension, density, thermal expansion, adiabatic and isothermal compressibility, speed of sound, heat capacity at con­stant pressure and at constant volume, enthalpy, elec­trical resistivity, viscosity, and thermal conductivity. The properties of these coolants were measured in many laboratories but mainly at normal atmospheric pressure and at relatively low temperatures (except for Na). In general, the reliability of the data is satisfactory; however, a large uncertainty still exists in some properties of Pb and Pb-Bi(e). A set of correlations for the estimation of the main properties of Na, Pb, and Pb-Bi(e), as a function of tempera­ture and pressure, is proposed based on the previous reviews and new results that appeared in open litera­ture. For the prediction of the missing properties at high temperatures and pressures, relevant equations of state (EOS), based on the proven physical models and available experimental data, were indicated. Taking into account that the critical parameters for the considered metals are not yet well defined with adequate precision, the EOS validation at high tem­peratures and pressures is still a problem.

In Section 2.14.2, general properties of liquid Na, Pb, and Pb-Bi(e) at a temperature of 400 °C, which is within the typical temperature range of coolant in SFR and LFR, are compared.

Characteristic temperatures (melting, boiling, and critical), and the enthalpies of melting and boiling are given in Section 2.14.3. Thermodynamic proper­ties are reviewed in five subsections of Section 2.14.4. Section 2.14.5 is devoted to transport prop­erties. Conclusions are formulated in Section 2.14.6. Compatibility of structure materials with liquid sodium and lead alloys is considered in Chapter 5.13, Material Performance in Sodium and Chapter 5.09, Material Performance in Lead and Lead-bismuth Alloy, respectively.

2.14.2 General Properties

Almost all main thermophysical properties of liquid Na, Pb, and Pb-Bi(e) (such as density, thermal expan­sion, compressibility, heat capacity, surface tension, sound velocity, and compressibility) are measured with satisfactory precision in the region close to normal melting temperature. An exception is the saturated vapor pressure, which is rather small at these temperatures to be measured with high preci­sion. The aforementioned parameters of liquid Na, Pb, and Pb-Bi(e), calculated with the recommended correlations given below for normal atmospheric pressure and at the typical mean temperature 400 °C (673 K) of the normal operation of the considered LM coolants in a Gen IVreactor, are compared in Table 1.

The main advantages of Na coolant in comparison with Pb and Pb-Bi(e) are the lower density and viscosity and the higher other transport coefficients (electrical conductivity and thermal conductivity); its disadvantages are higher compressibility, high satura­tion vapor pressure, and lower surface tension. More detailed information about the thermophysical prop­erties of liquid Na, Pb, and Pb-Bi(e) is given in the following sections.

The pressure diapason of the normal operation of the LM coolants in nuclear installations usually ranges from 0.1 to 1-1.5 MPa and the temperature diapason is between 300 °C (573 K) and 600 °C (873 K). Under accidental conditions, the coolant temperature can locally increase up to the fuel melting temperature, and the local coolant pressure can increase up to the cladding failure limit. Therefore, coolant properties have to be known within larger temperature and pressure ranges. For the development of an EOS of the LM coolants needed for the correct extension of the properties’ recommendations to higher tempera­tures and pressures, their critical parameters (temper­ature, pressure, and density) should be known.

2.14.3 Characteristic Temperatures, Pressures, and Heats

A temperature range of the normal operation of a liquid metal coolant is usually determined by its melting and boiling temperatures. Under accidental conditions, it can even be close to the critical point region. The melting temperature increases very weakly with pressure, but the boiling temperature increases rapidly, so the temperature range is larger at the higher pressures.

The melting temperatures of the chemically pure Na, Pb-Bi(e), and Pb were measured with high pre­cision at normal atmospheric pressure.3-30 However, the difference between the values coming from dif­ferent sources sometimes reaches a few tenths of degree of Celsius (a few tenths of Kelvin) for Na and Pb, and a few degrees for Pb-Bi(e). This disper­sion is mainly explained by the presence ofimpurities in the samples.30 Moreover, for Pb-Bi(e), supplemen­tary uncertainties exist due to a possible deviation from the eutectic composition (currently, most of the

researches fix it at 45.5 wt% Pb + 55.5 wt% Bi24’31; see Figure 1) and due to a possible presence of metastable phases.24’32 For the technically pure metals, the best estimated values of the melting temperature (TM 0) for the considered metals at the standard atmospheric pressure are 371.0 ± 0.1 K for Na, 600.6 ± 0.1 K for Pb, and 398 ± 1K for Pb — Bi(e).9’24-34

In 1999, Stolen and Gronvold35 performed the critical assessment of the available data on the enthalpy of melting (AHM0) of pure metals, used as enthalpy standards, at the standard atmospheric pres­sure. The values recommended by them for Na and Pb are 2.60 ± 0.03 and 4.78 ± 0.03 kJ mol-1, respec­tively. For the Pb-Bi(e) melting enthalpy, the analysis of the available data was performed in Sobolev and Benamati24 and Sobolev,34 where the mean value of

8.4 ± 0.06 kJ mol-1 of three more reliable sources Bogoslovskaia et a/.,25 Kirillov et a/.,26 Cevolani36 was recommended, which is adopted in this work.

The boiling temperatures of these LM were measured with a lower precision than their melting temperatures. The uncertainty ranges from 10 to 20 K for Pb and Pb-Bi(e)24 and about 1-2 K for Na.22,27,30 The selected ‘best estimate’ values of the normal boiling temperatures (TB0) of Na, Pb, and Pb-Bi(e) are 1155 ± 2 K, 2021 ± 3 K, and 1927 ± 16 K, respectively.

The available data on the enthalpy of boiling of Na were analyzed in Fink and Leibowitz22 and Kirillov et a/.,26 and the selected recommended value
of AHB 0(Na) = 97.4 ± 0.1 kJ mol-1 is recommended for the standard atmospheric pressure. The literature values on the boiling latent heat of Pb at normal atmospheric pressure were reviewed in Sobolev and Benamati.24 The difference <3% was found be­tween the maximum and minimum reported values: AHB 0(Pb) = 177.9 ± 0.4 kJ mol-1. In the same compi­lation,24 the mean value of 178 ± 1 kJ mol-1 of only three available sources25,26,36 was recommended for the latent heat of boiling of LBE at normal pressure.

The most probable values of the melting and boiling temperatures, together with the latent enthal­pies of the melting and boiling of Na, Pb-Bi(e), and Pb at normal atmospheric pressure, recommended earlier, and the operation temperature ranges of these liquid metal coolants, are presented in Table 2.

In order to improve the economical aspects of MOX pellet fabrication and to extend the fabrication pro­cess to MOX pellets containing MAs, various R&D programs have been started especially in France, Germany, Japan, and Russia.

In France, several coconversion processes have been developed and combined with the development of reprocessing processes. One typical coconversion process, called the CO-EXtraction (COEX) process,

Figure 23 Flow sheet for the short process.

has been developed at the ATALANTE (Atelier Alpha et Laboratoires d’Analyses des Transuraniens et d’Etudes de Retraitement)61,69 In this, a mixture of uranyl and plutonium nitrate solutions containing MAs is coconverted to MOX powder following the oxalate precipitation method. According to the results of COEX pellet fabrication tests in the MELOX test chain, MOX pellets produced by the COEX pro­cess have mean grain size larger than 6 pm. These are compatible with current MOX manufacturing values obtained in the MELOX.61 In parallel with the above development, fuel fabrication processes have also been developed in the ATALANTE and LEFCA (Laboratoire d’Etudes et de Fabrications Experimen — tales de Combustibles Nucleaires Avances).

In Germany, basic R&D concerning fabrication processes for MOX fuel bearing MAs have been carried out at the Institute for Transuranium Ele­ments (ITU).70 One of the fuel irradiation test pro­grams carried out by ITU was the SUPERFACT experiment. In this experiment, SUPERFACT fuels bearing Np or Am were fabricated by the sol-gel method and they were irradiated in various fast

70,71

reactors.

In Japan, a simplified MOX pellet fabrication pro­cess, the short process, has been developed on the
basis of the MH method, for the above purposes. The flow sheet for this process is shown in Figure 23. A 300 g scale laboratory test of the short process has been successfully completed.72

In the short process, three different solutions, ura — nyl nitrate, plutonium nitrate, and a nitrate solution in which rejected MOX pellets are dissolved, are mixed to obtain the desired plutonium content in the final mixed solution. Then, the mixed solution is converted to the MH-MOX powder with desired plutonium content by the MH method. This con­verted MH-MOX powder is tumbling-granulated after adding an adequate amount of water as a binder to improve its flowability. The tumbling-granulated MH-MOX powder is calcined at 750 °C in air and reduced to MH-MOX powder at 750 °C under an atmosphere of N2 + 5% Ar mixed gas. The MH-MOX powder so obtained is directly pressed into green annular pellets using a die-wall lubrication method. These are then sintered without heat treat­ment in the dewaxing furnace because the amount of organic compounds contained in the green pellets is controlled at a lower value than that in pellets prepared by the conventional MOX fuel fabrication process. Sintered MOX pellets are ground by a cen­terless grinder, and subsequently, the geometrical

density and appearance of each pellet are inspected. The MOX pellets rejected at the inspections are dis­solved in nitric acid and used as part of the final blending solution as shown in Figure 23. Figure 24 shows a photograph of annular MOX pellets prepared by the short process. A ceramograph of a pellet prepared by the short process is shown in Figure 25.

The MOX pellets manufactured by the short pro­cess have a larger mean grain size than those manufac­tured by the other processes such as the SBR, MIMAS, JAEA, and COEX processes. The development of a series of small scale (kg scale) test devices was started in 2007.73 In parallel with this work, JAEA has an irradiation test program for MOX pellets bearing MAs, to understand their irradiation behavior. In this program, MOX pellets bearing the MAs, Am, and Np were prepared by the JAEA process and irradiated in the JOYO. These irradiated pellets were subjected to postirradiation examinations and the results obtained have been reported in Maeda et a/.74,75

In Russia, RIAR (Research Institute of Atomic Reactors) has proposed the demonstration program concept DOVITA (Dry reprocessing, Oxide fuel, Vibropac, Integral, Transmutation of actinides) and many R&D activities related to them have been car­ried out. From this program, vibro-packing technol­ogy has been applied to load MOX granules into a cladding tube.76,77

Other chapters in this volume present more details on the fundamental nature and aspects ofthe primary damage state in irradiated metals and alloys, and on the detailed effects of irradiation on mechanical properties’ behavior. This chapter simply highlights some changes in microstructure caused by fission or fusion reactor neutron irradiation, and the changes in properties that they cause in 300 series austenitic stainless steels, to facilitate easy comparison to the unirradiated behavior properties described above. Various other sections of this volume deal in far more detail with the effects of irradiation in various kinds of alloys.

In LWRs at 20-250 °C, the interstitials migrate freely to sinks, while the vacancies or their clusters are relatively immobile, so this has been termed the ‘low-temperature regime’ of microstructural evolu­tion in austenitic stainless steels.14 In sodium-cooled FBRs, temperatures are not lower than the sodium coolant, so they are typically 300-350 °C or above, which is termed the ‘intermediate-temperature regime,’ and both vacancy and interstitial defects can migrate to sinks. Transmutation-produced helium atoms are another form of primary radiation damage that varies with reactor environment (high in LWR and magnetic fusion reactor (MFR) systems, low in FBR systems). Thermal neutrons produce helium in austenitic stainless steels from boron atoms directly, and by a two-step reaction with nickel atoms.17 He/dpa ratios for LWR systems can be very high, over 100appm He/dpa, while in mixed-spectrum fission reactors used for radiation-effects studies on materials, the ratios vary from 1 to 70 appm He/dpa. MFRs, with 14MeV neutrons from D-T fusion reactions, have linear He/dpa ratios of about 14 appm dpa-1 in a stainless steel first-wall component. The FBR reactors with mainly fast fission neutron spectra produce very low He/dpa ratios of 0.1-0.5 appm He/dpa in austenitic stainless steels. Irradiated mate­rials properties data discussed in the remainder of this section is mainly from LWR or mixed-spectrum fission reactor facilities used to study irradiation effects for MFR applications, so they have relatively high He/dpa ratios as well as a wide range of irradia­tion temperatures.

The major effects of irradiation in mixed — spectrum fission reactors, such as Oak Ridge Research Reactor (ORR) or High Flux Isotope Reactor (HFIR), on mechanical properties in the low-temperature regime are dramatic hardening (increased YS) and reduced ductility in SA and 316 and Ti-modified 316 stainless steels, and more modest hardening and ductility reduction in 20-25% CW steels. The increased YS for irradiated SA steels are illustrated in Figure 10.1 The SA stainless steels have 250-300MPa YS in the unirradiation condition, and 50% or more total elongation at room temperature and up to 250-300 °C, but irradiation increases the YS to 600-800 MPa or more, and reduces ductility to 10% or less. However, the fracture mode in this irradiation temperature regime still remains ductile.8,9,19 After irradiation, 20-25% CW steels have YS of 800-1000 MPa, and less ductility, but still retain ductile fracture. This is an important feature to note, and despite transmutation-produced helium levels of 1000-2000 appm, they do not embrittle, because helium and vacancy complexes are immobile in this temperature regime. However, most tensile test­ing results are in vacuum or air, and radiation-induced sensitization in water is not found after irradiation at 20-200 °C, but does become an embrittling factor to consider for irradiation above 300 °C.2

Irradiation-induced hardening of austenitic stain­less steels at room temperature to <250 °C is caused by the microstructural changes produced by irradia­tion in this low-temperature regime. Effects of alloy composition are small in this regime, but the effects of processing condition prior to irradiation (SA or 20-25% CW) are very large. Both SA and 25% CW steels, like 316 or Ti-modified 316, have very dense dispersions of ‘black-spot’ interstitial loops (2-4 nm diameter) uniformly within the grains,2,1 ,21 as illu­strated for 25% CW Ti-modified steel in Figure 11. However, the SA steels also have larger (10-50 nm) diameter Frank (faulted) interstitial loops and no network dislocations, whereas the 25% CW steels

have a recovered dislocation network and virtually no large Frank loops (Figure 12). These microstructural effects directly reflect the fact that interstitial defects are main point defects migrating freely to sinks in this temperature regime. Large Frank loops cannot

nucleate and grow until the concentration of network dislocations is below some critical concentration. This also affects mechanical behavior, because the ‘black-dot’ and larger Frank loops are sessile until they unfault, whereas the network dislocations can
climb and glide in response to stress or as they absorb point defects.

Radiation-induced microstructural changes are definitely different at 300 °C and above. In the dislo­cation structure, the ‘black-dot’ loop damage clearly observed at 200-250 °C is absent at 300-330 °C, and the dislocation structure consists of larger Frank loops and networks that add up to a fairly high total dislocation density.14,21,22 There is now also a cavity component of the microstructure, with nanoscale helium bubbles visible at 300-330°C, and larger voids and helium bubble visible at 400 °C, after irra­diation at high He/dpa ratios in mixed-spectrum reactors (ORR, HFIR), or just voids in FBR irradia­tions at 350-400 °C.12,14,22 The appearance of cavities is a clear indication that vacancy or vacancy clusters and complexes (and helium atoms) are migrating in this temperature regime.

Tensile properties of austenitic stainless steels irradiated at 300 °C and above reflect the micro­structural changes, particularly the dislocation compo­nent of the microstructure. This higher temperature regime in austenitic stainless steels is marked by stronger and more complex temperature and dose dependencies of all the microstructural components, including precipitation and micro/nano-compositional changes.14,22 The YS declines from the 800 MPa values at 300 °C to values of about 400 MPa at 500 °C (Figure 10), which approach the YS of un­irradiated steels, because all components of the radiation-induced microstructure coarsen, and dislo­cation density falls by several orders of magnitude. Ductility can vary significantly, but is generally higher (>10%) at 400-500 °C, but not as high as that of unir­radiated materials. However, the effects also depend on He/dpa ratio. For FBR irradiations (low He/dpa ratio, <50dpa), total elongation can be good even at 600­650 °C, but for irradiations in mixed-spectrum reactors such as ORR or HFIR (high He/dpa ratio, >20dpa), ductility becomes very low above 500 °C, with almost no ductility and brittle grain-boundary fracture at 600 °C due to severe grain-boundary helium embrit­tlement (>500-1000 appm He). For more detailed information on tensile properties after irradiation, see Chapter 1.04, Effect of Radiation on Strength and Ductility of Metals and Alloys.

Microstructural changes produced by irradiation at temperatures of 400 °C and above manifest the intense effects of radiation-induced solute segrega­tion (RIS), which drive nonequilibrium flows and buildups of solute-atoms to sinks (bubbles, voids, dis­location loops and networks, and grain boundaries), because they are coupled to the point defect flows. Such changes are important to note because prolonged aging at <550 °C produces little or no change to the as-fabricated microstructure. Undersized atoms, such as Ni and Si, strongly couple to interstitial defects, and migrate with them to all sinks. Inverse — Kirkendall effects cause fast-diffusing elements such as Cr and Mo to migrate away from sinks with vacancy fluxes diffusing toward them, whereas slow — diffusing Ni atoms build up at such vacancy sinks. The original austenitic solid-solution alloy phase then unmixes after prolonged irradiation into differ­ent kinds of micro/nano-alloys (Figure 7). Regions around the point-defect sinks (voids, loops, and grain boundaries) become enriched in Ni and Si, while the remaining alloy left behind in between such sinks is rich in Cr, and poor in Si and Ni.12 The different micro/nano-alloy regions become unstable as dose increases, and then transform into various precipitate phases, most of which are radiation-induced or mod­ified relative to the natural thermal precipitation that would form in austenitic alloy during aging at higher temperatures (550-650 °C).12,14,15 The most obvious Ni — and Si-rich radiation-induced phase is Ni3Si g0, which forms abundantly in reactor-irradiated SA 316, as shown in Figure 13, but would not form at all in thermally aged SA 316 (Figure 12).15 Another extreme effect of such RIS, found in some FBR — irradiated steels, is the actual decomposition of the austenite parent phase into austenite shells around voids and other sinks, and ferrite regions in between.15 These effects tend to maximize at about 450-550 °C, and then all diminish with increasing irradiation tem­perature. At 650-700 °C, RIS effects are nearly gone and are replaced by basically thermal-aging effects with slightly enhanced kinetics due to radiation — enhanced diffusion.

In addition to the irradiation-produced mechani­cal properties described above, irradiation in this higher temperature regime also causes void/cavity swelling to occur. Void swelling is caused by the biased (or preferred) flows of interstitial and vacancy defects to different sinks, with more vacancies flow­ing to cavities (helium bubbles and voids) and more interstitials flowing to Frank loops and/or radiation — induced precipitates. An example of precipitation — enhanced void swelling in a SA 316 + Ti steel irradiated in ORR at 500 °C to 11 dpa (200 appm He) is shown in Figure 14; clearly the larger voids are directly associated with RIS-induced G-phase (Mn6Ni16Si7) silicide particles. Formation of such voids is the direct cause of volumetric swelling in

G-phase

50 nm

Figure 14 Transmission electron microscopy of radiation — induced voids in SA PCA steel irradiated in ORR at 500 °C to 11 dpa. Largest voids have G-phase particles attached. Reproduced from Maziasz, P. J. J. Nucl. Mater. 1989, 169, 95-115.

reactor-irradiated steel, with an example of swelling of SA 316 steel as a function of dose for FBR irradia­tion at 420 °C shown in Figure 152 For more detailed information on swelling, see Chapter 4.02, Radiation Damage in Austenitic Steels. Such void swelling is generally observed in various FBR or mixed-spectrum reactor environments at 400-650 °C. If very high concentrations of helium bubbles, disloca­tions, or precipitates become the dominant sinks for

point defects, then all the radiation-induced point defects recombine at those sinks (critical radius for void growth becomes very large), and both void swelling and RIS are suppressed.12,14,15,22,24 Such delayed void swelling is seen for dense dispersions of Ti-rich MC carbide particles and dislocation networks
in CW Ti-modified 316 steel (D9 or prime candidate alloy, PCA), as also shown in Figure 15. Very high concentrations of helium bubbles suppress void swelling at 300 °C in mixed-spectrum reactors such as ORR or HFIR (high He/dpa), but as those bubbles coarsen with increased temperature, void swelling is observed, particularly at 500-600 °C. For FBR irra­diations (low He/dpa), void swelling will abate at 650-700 °C, with only tiny helium bubbles being visi­ble at grain boundaries at high doses. However, in HFIR (high He/dpa) cavity swelling due to very large helium bubbles can still be 8-10% or more above 650 °C, with large grain-boundary cavities.12

One aspect of RIS effects on mechanical proper­ties worth noting and highlighting is RIS causing grain-boundary sensitization. For specimens of 25% CW Ti-modified 316 irradiated in ORR at 330 and 400 °C to 6-7 dpa, electrochemical testing to detect grain-boundary sensitization revealed grain­boundary grooving only at 400 °C, suggesting RIS — induced sensitization due to lower Cr at the grain boundaries.20 Similar ORR irradiation of SA 316 at 400 °C shows ductile fracture when tested in vacuum, but very brittle intergranular fracture when tested in oxygenated water at 300 °C (Figure 16), suggesting severe RIS-induced sensitization. In LWR systems, concerns about irradiation-assisted stress-corrosion­cracking (IASCC) at about 300 °C and <10 dpa are important for extended service,25 and these data sup­port such concerns and their connection to RIS. For more detailed information on IASCC, see Chapter 5.12, Material Performance in Supercritical Water.