Comprehensive nuclear materials 4 12 vanadium for nuclear systems

Comprehensive nuclear materials 4 12 vanadium for nuclear systems Comprehensive nuclear materials 4 12 vanadium for nuclear systems Comprehensive nuclear materials 4 12 vanadium for nuclear systems Comprehensive nuclear materials 4 12 vanadium for nuclear systems Comprehensive nuclear materials 4 12 vanadium for nuclear systems Comprehensive nuclear materials 4 12 vanadium for nuclear systems Comprehensive nuclear materials 4 12 vanadium for nuclear systems

4.12 Vanadium for Nuclear Systems T Muroga National Institute for Fusion Science, Oroshi, Toki, Gifu, Japan ß 2012 Elsevier Ltd All rights reserved 4.12.1 Introduction 391 4.12.2 4.12.3 4.12.4 4.12.5 4.12.6 4.12.7 4.12.8 4.12.9 4.12.10 4.12.11 4.12.12 4.12.13 References Vanadium Alloys for Fusion Reactors Compositional Optimization Fabrication Technology Fundamental Study on Impurity Effects Thermal Creep Corrosion, Compatibility, and Hydrogen Effects Radiation Effects Tritium-Related Issues Development of Advanced Alloys Critical Issues Vanadium Alloy Development for Fusion Blankets Summary 391 392 393 396 396 398 400 401 402 403 403 404 405 Abbreviations DBTT dpa flibe GTA HFIR HIP IFMIF Ductile–brittle transition temperature Displacement per atom Molten LiF-BeF2 salt mixture Gas tungsten arc High Flux Isotope Reactor Hot isostatic pressing International Fusion Materials Irradiation Facility IP Imaging plate ITER International Thermonuclear Experimental Reactor LMFBR Liquid Metal Fast Breeder Reactor MA Mechanical alloying PWHT Postweld heat treatment RAFM Reduced activation ferritic/martensitic REDOX Reduction–oxidation reaction TBM Test Blanket Module TBR Tritium breeding ratio TEM Transmission electron microscope 4.12.1 Introduction Vanadium alloys were candidates for cladding materials of Liquid Metal Fast Breeder Reactors (LMFBR) in the 1970s.1 However, the development was suspended mainly because of an unresolved issue of corrosion with liquid sodium Vanadium alloys attracted attention in the 1980s again for use in fusion reactors because of their ‘low activation’ properties At present, vanadium alloys are considered as one of the three promising candidate low activation structural materials for fusion reactors with reduced activation ferritic/martensitic (RAFM) steels and SiC/SiC composites Overviews of vanadium alloys for fusion reactor applications are available in the recent proceedings papers of ICFRM (International Conference on Fusion Reactor Materials).2–6 This chapter highlights the recent progress in the development of vanadium alloys mainly for application in fusion nuclear systems 4.12.2 Vanadium Alloys for Fusion Reactors Various tritium breeding fusion blanket concepts have been studied with different combinations of structural materials, tritium breeding materials, and cooling materials Vanadium alloys have been used in most cases with liquid lithium as the breeding and cooling materials (self-cooled V/Li blankets) for advanced concepts of DEMO (fusion demonstration power plant) and commercial fusion reactors.7,8 Because of high atomic density of Li atoms in liquid Li relative to Li-ceramics, Li–Pb, and molten-salt 391 392 Vanadium for Nuclear Systems Flibe, V/Li systems can obtain high tritium breeding ratio (TBR) without using the neutron multiplier Be A neutronics calculation showed that ‘tritium self sufficiency’ can be satisfied without Be both in Tokamak and Helical reactor systems.9 Without the necessity of using beryllium as a neutron multiplier, the replacement frequency of the blanket will be reduced because the blanket system is free from the periodic replacement due to the lifetime of Be, which can lead to enhanced plant efficiency V/Li blankets can be designed with a simple structure as schematically shown in Figure The blanket is composed of Li cooling channels made of vanadium alloys, reflectors, and a shielding area, which is in contrast to more complex solid breeder blankets that need a solid breeder zone, a neutron multiplier beryllium zone, cooling channels using gas or water, and tritium recovery gas flow channels in addition to reflectors and shielding A self-cooled Li blanket using neutron multiplier beryllium was also designed in the Russian program.10 This concept can downsize the blanket area because of efficient tritium generation per zone However, the blanket structure must be more complex than V/Li and new issues need to be solved such as Li/Be compatibility General requirements for structural materials of fusion blankets include dimensional stability, compatibility with breeder and coolants, high-temperature strength and low-temperature ductility during irradiation For vanadium alloys, issues concerning industrial maturity such as developing large-scale manufacturing technology need to be resolved Vanadium alloys could be a candidate structural material for molten-salt Flibe (LiF–BeF2) blankets For this application, a concept was proposed to dissolve WF6 or MoF6 into Flibe for corrosion protection of the wall surfaces by precipitation of W or Mo and reduction of the tritium inventory in vanadium alloys by enhancing reaction from T2 to TF, which is more highly soluble in Flibe than T2.11 The TBR of Flibe/V blankets may be marginal, but the neutron shielding capability for the superconductor magnet systems may be superior relative to V/Li according to neutronics investigation.12 In this system, precipitates of Wor Mo formed as a result of reaction from T2 to TF needs to be recovered from the flowing Flibe Table summarizes the blanket concepts using vanadium alloys with the advantages and critical issues Flowing liquid lithium 4.12.3 Compositional Optimization Superconducting magnet Neutron D-T plasma Shield Coating with W, Be, or C Reflector Vanadium alloy structures Blanket Figure Illustration of self-cooled Li blanket with V–4Cr–4Ti structural material Table Vanadium alloys potentially have low-induced activation characteristics, high-temperature strength, and high thermal stress factors For the optimization of the composition, both major alloying elements and minor impurities need to be controlled For maintaining the low activation properties, use of Nb and Mo, which used to be the candidate alloying elements for application to LMFBR, need to be avoided Cr was known to increase the strength of vanadium at high temperature and Ti was known to enhance ductility of vanadium by absorbing interstitial impurities, mostly oxygen However, excess Cr or Ti can Breeding blanket concepts using vanadium alloys Concept V/Li V/Be/Li V/Flibe Breeder and coolant materials Use of neutron multiplier Be Advantages Critical issues Liquid Li Liquid Li Molten-salt Flibe No Simple structure MHD coating, T recovery from Li Yes High TBR MHD coating, Li/Be compatibility, T recovery from Li No Small MHD pressure drop REDOX control, recovery of W or Mo, increase in TBR Vanadium for Nuclear Systems lead to loss of ductility Hence, optimization of Cr and Ti levels for V–xCr–yTi has been investigated It was known that with x þ y > 10%, the alloys became brittle6 as shown in Figure With systematic efforts, V–4Cr–4Ti has been regarded as the leading candidate For low activation purposes, the level of Nb, Mo, Ag, and Al needs to be strictly controlled Large and medium heats of V–4Cr–4Ti have been made in the United States, Japan, and Russia 100 1150 ЊC 50 ±20 ЊC DBTT (ЊC) 1100 ЊC 950 ЊC –50 1000 ЊC –100 Annealing temperature 950 ЊC 1000 ЊC 1050 ЊC 1100 ЊC 1150 ЊC –150 –200 –250 10 15 Cr + Ti (Wt %) 20 25 Figure DBTT as a function of Cr ỵ Ti (wt%) of V–Cr–Ti alloy for various annealing temperatures Reproduced from Zinkle, S J.; Matsui, H.; Smith, D L.; Rowcliffe, A L.; van Osch, E.; Abe, K.; Kazakov, V A J Nucl Mater 1998, 258–263, 205–214, with permission from Elsevier 393 An especially high-purity V–4Cr–4Ti ingot produced by the National Institute for Fusion Science (NIFS) in collaboration with Japanese Universities (NIFS-HEAT-1 and 2) showed superior properties in manufacturing due to their reduced level of oxygen impurities.4 Figure compares the contact dose rate after use in the first wall of a fusion commercial reactor for four reference alloys The full-remote and full-hands-on recycle limits are shown to indicate the guideline for recycling and reuse.13 SS316LN-IG (the reference ITER structural material) will not reach the remoterecycling limit after cooling and hence the recycling is not feasible F82H (reference RAFM steel) and NIFS-HEAT-2 behave similarly, but NIFS-HEAT-2 shows significantly lower dose rate before the 100year cooling The dose rate of F82H and NIFSHEAT-2 reached a level almost two orders lower than the remote-recycle limit by cooling for 100 and 50 years, respectively The dose rate of SiC/SiC composites (assumed to be free from impurities because of lack of reference composition) is much lower at 100 year cooling relative to F82H and NIFS-HEAT-2 4.12.4 Fabrication Technology Figure summarizes the microstructural evolution during the breakdown process of NIFS-HEAT-2 105 104 Contact dose rate (Sv h–1) 103 Reduced activation ferritics (F82H) 102 101 100 10–1 V–4Cr–4Ti (NIFS-HEAT) FFHR Li blanket first wall neutron 1.5 MW m–2 operation SS316 for ITER (SS316LN-IG) Pure SiC/SiC Full-remote recycling 10–2 10–3 Full-hands-on recycling 10–4 10–5 10–2 10–1 100 101 102 103 Cooling time after shutdown (years) 104 Figure Contact dose after use in first wall of a fusion commercial reactor for four reference alloys SS316LN-IG: the reference ITER structural material F82H: reference reduced activation ferritic/martensitic steel NIFS-HEAT-2: reference V–4Cr–4Ti alloy SiC/SiC: assumed to be impurity-free 394 Vanadium for Nuclear Systems Ingot Hot forging 1423 K Formation Heat treatment Hot/cold roll 1373 K/RT 973 K 1273 K 1373 K 1573 K Ti-rich blocky precipitates (with N, O, C) Elongation, band structure Dissolution Ti–O–C thin precipitates Formation Coarsening Dissolution V–C on GB 50 mm 50 mm 25 mm mm mm mm 50 mm Figure Microstructural evolution during the breakdown process of V–4Cr–4Ti ingots Reproduced from Muroga, T.; Nagasaka, T.; Abe, K.; Chernov, V M.; Matsui, H.; Smith, D L.; Xu, Z Y.; Zinkle, S J J Nucl Mater 2002, 307–311, 547–554 260 240 Vickers hardness (Hv) ingots.4 Bands of small grains aligned along the rolling direction were observed at the annealing temperature below 1223 K The grains became homogeneous at $1223 K The examination showed that optimization of size and distribution of Ti-CON precipitates are crucial for good mechanical properties of the V–4Cr– 4Ti products Two types of precipitates were observed, that is, the blocky and the thin precipitates The blocky precipitates formed during the initial fabrication process The precipitates aligned along the working direction during the forging and the rolling processes forming band structures, and were stable to 1373 K Since clustered structures of the precipitates result in low impact properties, rolling to high reduction ratio is necessary for making a thin band structure or homogenized distribution of the precipitates The thin precipitates were formed at $973 K and disappeared at 1273–1373 K At 1373 K, new precipitates, which were composed of V and C, were observed at grain boundaries They seem to be formed as a result of redistribution of C induced by the dissolution of the thin precipitates The impact of the inhomogeneous microstructure can influence the fracture properties.14 Figure shows the hardness as a function of final heat treatment temperature for three V–4Cr–4Ti materials: NIFS-HEAT-1, NIFS-HEAT-2, and USDOE-832665 (US reference alloy).15 The hardness has a minimum at 1073–1273 K, which corresponds to the temperature range where formation of the thin precipitates is maximized With the heat treatment higher than this temperature range, the hardness increases and the ductility decreases because the NIFS-HEAT-1 NIFS-HEAT-2 US-DOE 832665 220 200 180 160 140 V–4Cr–4Ti 120 200 400 600 800 1000 1200 1400 1600 Annealing temperature (K) Figure Vickers hardness as a function of annealing temperature for NIFS-HEAT-1, NIFS-HEAT-2, and US-DOE 832665 Reproduced from Heo, N J.; Nagasaka, T.; Muroga, T J Nucl Mater 2004, 325, 53–60 precipitates dissolve enhancing the level of C, N, and O in the matrix Based on the evaluation of various properties in addition to the hardness as a function of heat treatment conditions, the optimum heat treatment temperature of 1173–1273 K was suggested Plates, sheets, rods, and wires were fabricated minimizing the impurity pickup and maintaining grain and precipitate sizes in Japanese, US, and Russian programs Thin pipes, including those of pressurized creep tube specimens, were also successfully fabricated Vanadium for Nuclear Systems in Japan maintaining the impurity level, fine grain size, and straight band precipitate distribution by maintaining a constant reduction ratio between the intermediate heat treatments.16 The fine-scale electron beam welding technology was enhanced as a result of the efforts for fabricating the creep tubes, including plugging of end caps.17 In the United States, optimum vacuum level was found for eliminating the oxygen pick-up during intermediate annealing to fabricate thin-walled tubing of V–4Cr–4Ti.18 In Russia, fabrication technology is in progress for construction of a Test Blanket Module (TBM) for ITER (International Thermonuclear Experimental Reactor).19 Joining of V–4Cr–4Ti by gas tungsten arc (GTA) and laser welding methods was demonstrated GTA 395 is a suitable technique for joining large structural components GTA welding technology for vanadium alloys provided a significant progress by improving the atmospheric control The results are summarized in Figure Oxygen level in the weld metal was controlled by combined use of plates of NIFSHEAT-1 (181 wppm O) or US-8332665 (310 wppm O) and filler wire of NIFS-HEAT-1, US-8332665, or a high-purity model alloy (36 wppm O) As demonstrated in Figure 6, ductile–brittle transition temperature (DBTT) of the joint and the oxygen level in the weld metal had a clear positive relation This motivated further purification of the alloys for improvement of the weld properties.20 Only limited data on irradiation effects on the weld joint are available at present 15 Absorbed energy (J) EU = 13 J 10 Plate/filler NH1/HP 128 K US/HP 183 K NH1/NH1 US/US 188 K 320 K 50 100 150 200 250 Test temperature (K) 350 400 US/US 300 DBTT (K) 300 US/HP 200 NH1/NH1 100 NH1/HP DBTT = +60 K/100 wppm O 0 50 100 150 200 250 300 350 400 Oxygen in weld metal (wppm) Figure Upper: Absorbed energy of Charpy impact tests of V–4Cr–4Ti weld joints as a function of test temperature for various combinations of plates and fillers Lower: DBTT of V–4Cr–4Ti weld joints as a function of oxygen level in the weld metal NH1, NIFS-HEAT-2 (O: 181 wppm); US, US-DOE 832665 (O: 310 wppm); HP, high-purity model V–4Cr–4Ti alloy (O: 36 wppm) Reproduced from Nagasaka, T.; Grossbeck, M L.; Muroga T.; King, J F Fusion Technol 2001, 39, 664–668 396 Vanadium for Nuclear Systems The welding results in complete dissolution of TiCON precipitates and thus results in significant increase in the level of C, O, and N in the matrix In such conditions, radiation could cause embrittlement Some TEM observations showed enhanced defect cluster density at the weld metals However, the overall evaluation of the radiation effects remains to be performed Especially, elimination of radiation-induced degradation byapplying appropriate conditions of postweld heat treatment (PWHT) is the key issue For the use of vanadium alloys as the blanket of fusion reactors, the plasma-facing surfaces need to be protected by armor materials such as W layers Limited efforts are, however, available for developing the coating technology A low pressure plasma-spraying method was used for coating W on V–4Cr–4Ti for use at the plasma-facing surfaces The major issue for the fabrication is the degradation of the vanadium alloy substrates by oxidation during the coating processes Figure shows the result of bending tests of the coated samples The crack was initiated within the W layer propagating parallel to the interface and followed by cracking across the interface Thus, in this case, the quality of W coating layer is the issue rather than the property of the V–4Cr–4Ti substrate or the interface Hardening of substrate V–4Cr–4Ti by the coating occurred but was shown to be in acceptable range.21 Figure is a collection of the products from NIFS-HEAT-2 Research with model V–4Cr–4Ti alloys doped with O and N provided information on the partitioning of O and N into the precipitates and matrix The density of the blocky precipitates and thin precipitates increased with N and O levels, respectively Figure shows hardness as a function of N and O levels in V–4Cr–4Ti after melting and annealing at 1373 K for h.22 Hardness after annealing at 1373 K, where only the blocky precipitates were observed in the matrix, increased to a certain extent with O level ($4.5 Hv/100 wppm O), but only very weakly with N level ($0.9 Hv/100 wppm N) These data suggest that, after the annealing, most of the N is included in the blocky precipitates and stable to $1373 K On the other hand, O exists in the matrix, the blocky and the thin precipitates, and the partitioning changes with the heat treatment Thus, for the purpose of the property control of V–4Cr–4Ti, the level of N before the heat treatment is not so important but that of O is crucial It is to be noted, however, that N contamination during the operation can influence the properties of vanadium alloys seriously Fundamental information on the impurity distribution and interaction with solutes and dislocations is obtained by serrated flow in tensile deformation as shown in Figure 10 Temperature and stain rate dependence of the flow showed that the serrated flow above 673 K is related to C and O and above 773 K to N Small serration height at 673 K for NIFS-HEAT-1 (1–3 MPa) relative to that of US-832665 ($9 MPa) was observed and attributed to the difference in O level.23 4.12.5 Fundamental Study on Impurity Effects 4.12.6 Thermal Creep Effects of C, O, and N on the property of vanadium are a long-standing research subject However, research into the effects of C, O, and N on V–4Cr–4Ti is limited Thermal creep is a potential factor which can determine the upper operation limit of vanadium alloys Crack W V–4Cr–4Ti Intergranular fracture 500 µm 50 µm 10 µm Figure Cross-section of W coating on V–4Cr–4Ti after bending tests Fracture started in the W coating layer Vanadium for Nuclear Systems 397 (mm) f 4.57 ϫ 25 t ϫ 400 mm 6.6 t 0.5 t 26 t 1.9 t 1.0 t 4.0 t 0.25 t 2d 8d f 10 ϫ t ϫ 100 m m Plates, sheets, wires, and rods Thin pipes 20 mm Creep tubes W coating NIFS-HEAT-2 (V–4Cr–4Ti) 0.5 mm W coating by plasma spraying Laser weld joint mm Figure Collection of the V–4Cr–4Ti products manufactured by the Japanese program V–4Cr–4Ti, as-melted V–4Cr–4Ti, 1373 K Pure V, as-melted Pure V, 1373 K Vickers hardness (Hv) 300 250 200 150 100 50 200 400 600 800 1000 1200 Oxygen level (wppm) 200 400 600 800 1000 1200 Nitrogen level (wppm) Figure Vicker’s hardness as a function of O and N levels for V–4Cr–4Ti after melting and annealing at 1373 K for h Reproduced from Heo, N J.; Nagasaka, T.; Muroga, T.; Matsui, H J Nucl Mater 2002, 307–311, 620–624 Previously, uniaxial tensile creep tests and biaxial pressurized creep tube tests were carried out in vacuum for evaluation of the creep deformation characteristics Figure 11 shows summary of the creep deformation rate as a function of applied stress.3 In this plot, the creep data were described by a powerlaw equation24: de=dt ẳ ADGb=kT ịs=Gịn where de/dt is the creep rate, s is the applied stress, D is the lattice diffusion coefficient, G is the shear modulus, b is the Burgers vector, k is the Boltzmann constant, T is the absolute temperature, and A is a constant The exponent of the function (n) changed from 10 with the increase in the stress A new apparatus for biaxial creep testing in Li provided opportunities for examining creep 398 Vanadium for Nuclear Systems Stress (MPa) 200 10 30 20 Strain (%) Figure 10 Tensile deformation curves of V–4Cr–4Ti at various temperatures 10-5 Uniaxial tests 310 wppm O 10-6 (de/dt)kT/DGb n = 3.7 10-9 10-10 n = 4.3 n = 0.84 10-11 10-12 10-7 10-8 In vacuum 50 MPa 70 MPa 90 Mpa 10-9 10-10 Creep strain (%) In lithium 10 12 Figure 12 Creep strain rate as a function of creep strain for the same batch of NIFS-HEAT-2 creep tubes in vacuum and Li environments Modified from Li, M.; Nagasaka, T.; Hoelzer, D T.; et al J Nucl Mater 2007, 367–370, 788–793; Fukumoto, K.; Nagasaka, T.; Muroga, T.; Nita, N.; Matsui, H J Nucl Mater 2007, 367–370, 834–838 n = 13 10-7 10-8 Creep strain rate (1 s–1) 10-6 1073 K 973 K 873 K 773 K 673 K RT surface hardening during exposure to Li Further investigation is necessary for understanding the environmental effects on impurity redistribution and creep performance Microstructural observations of the creep tube specimens tested at 1123 K showed free dislocations and dislocation cell at 100 and 150 MPa, respectively This change of dislocation structure can cause the change in power-law creep behavior.27 Biaxial tests 699 wppm O 10-3 10-2 4.12.7 Corrosion, Compatibility, and Hydrogen Effects s/G Figure 11 Thermal creep deformation rate of V–4Cr–4Ti as a function of applied stress for uniaxial and biaxial tests The definition of the terms and the function from which n is extracted are indicated in the text Reproduced from Kurtz, R J.; Abe, K.; Chernov, V M.; Hoelzer, D T.; Matsui, H.; Muroga, T.; Odette, G R J Nucl Mater 2004, 329–333, 47–55 deformation in Li with that in vacuum.25 However, the correlation of creep data is subject to the alloy heat and manufacturing processes as well as test methods and environments Figure 12 shows the comparison of the NIFS-HEAT-2 creep strain rate versus creep strain data for tests in vacuum and Li environments at 1073 K, for the same batch of NIFSHEAT-2 creep tubes.25,26 The figure clearly shows reduced strain rate in Li environments A possible factor could be N pick-up from Li and the resulting In a Li/V blanket, it is believed that the interior of the wall needs to be coated with insulator ceramics for mitigating the pressure drop caused by magnetohydrodynamic effects (see also Chapter 4.21, Ceramic Coatings as Electrical Insulators in Fusion Blankets) Corrosion of vanadium alloys in liquid Li might not be a concern if the entire inner wall is covered with the insulating ceramic coating However, since the idea to cover the insulator ceramic coating again with a thin vanadium or vanadium alloy layer was presented for the purpose of preventing liquid lithium from intruding into the cracks in the ceramics coating, the corrosion of vanadium alloys in liquid lithium again attracted attention It is known that the corrosion of vanadium alloys in liquid lithium is highly dependent on the alloy composition and lithium chemistry Especially, the N level influences the corrosion in complex manners.28,29 Figure 13 shows a summary of the weight Vanadium for Nuclear Systems 0.1 Ti 50 I 30 +11.9 20 +11 Ti:Cr = 2:1 +2.5 +6.7 +5.8 Weight gain (mg mm–2) 0.08 40 +1.2 –19.0 –22.0 V–4Cr–4Ti V–4Cr–4Ti–0.5Si V–4Cr–4Ti–0.5Al V–4Cr–4Ti–0.5Y 0.06 0.04 II +0.4 10 –8.2 –2.1 –19.7 –21.0 20 –47.4 –52.5 –26.4 30 Cr 40 50 Figure 13 The compositions of V–Ti–Cr alloys (wt%) with increase (area I) and decrease (area II) of mass (g cmÀ2) after holding of samples in Li at 973 K, 500 h Reproduced from Eliseeva, O I.; Fedirko, V N.; Chernov, V M.; Zavialsky, L P J Nucl Mater 2000, 283–287, 1282–1286 gain and loss in V–xCr–yTi systems in Li.30 High Ti alloys showed a weight increase by forming a TiN layer and high Cr alloys exhibited a weight loss as a result of the dissolution of Cr–N complexes As the boundary of the two contradictory changes, Ti:Cr$2:1 was observed Recently, a corrosion test using monometallic thermal convection Li loop made of V–4Cr–4Ti was conducted at 973 K for 2355 h Because of the temperature gradient, weight loss and weight gain of V–4Cr–4Ti samples occurred at the hot leg and cold leg, respectively However, the loss rate corresponded to only