Development of high voltage power supply for the KSTAR 170 GHz ECH and CD system

A 3.6 MW (66 kV/55 A) DC power supply system was developed for the 170 GHz EC H&CD system in KSTAR. The power supply system consists of a cathode power supply (CPS), an anode power supply (APS) and a body power supply (BPS). The cathode power supply is capable of supplying a maximum voltage of ?66 kV and a current of 55 A to the cathode with respect to the collector using pulse step modulation (PSM). The high voltage switching system for the cathode is made by a fast MOS-FET solid-state switch which can turn off the high voltage to the cathode within 3 μs in the occurrence of gyrotron faults. The APS is a voltage divider system consisting of a fixed resistor and zener diode units with the capability of 60 kV stand-off voltage. The anode voltage with respect to the cathode is controlled in a range of 0–60 kV by turning the MOS-FET switches connected in parallel to each zener diode on and off. For high frequency current modulation of the gyrotron, the parallel discharge switch is introduced between the cathode and anode in order to clamp the charged voltage in the stray capacitance. The BPS is a DC power supply with the capability of 50 kV/160 mA. The nominal operation parameter of BPS was 23 kV and 10 mA, respectively, and the voltage output is regulated with a stability of 0.025% of the rated voltage. The series MOS-FET solid-state switch is used for on/off modulation in the body voltage sychronizing with anode voltage. The parallel discharge switch is also introduced between the body and collector for high frequency RF modulation. This paper describes the key features of the high voltage power supply system of the KSTAR 170 GHz gyrotron as well as the test results of the power supply.     

Preparation of ultrafine tungsten wire via electrochemical method in an ionic liquid

Ultrafine tungsten wire less than 10 μm in diameter is often used as wire array load applied in Inertial Confinement Fusion (ICF) physical experiments. In order to obtain a higher yield of X-ray, both initial radius and line quality of metal wire were required to be of high quality simultaneously. This paper has studied the electrochemical method to corrode tungsten wires uniformly in an ionic liquid electrolyte containing 1 wt% sodium hydroxide. A three electrode system composed of a tungsten anode electrode, a stainless steel cathode and a saturated calomel electrode as a reference electrode, was used in the electrochemical experiments. Liner sweep voltammetry (LSV) and Tafel experiments were used to investigate the electrochemical behaviors of tungsten wires in ionic liquid and aqueous solution. Based on scanning electron microscope (SEM) observation, the morphologies of tungsten wire surface with uniform corrosion under different applied voltages have been demonstrated. X-ray diffraction (XRD) methods were employed to track the evolution of the crystal structure before and after corrosions, and there is an obvious difference in peak intensities. The ultrafine tungsten wire with a uniform diameter of 8.5 μm was obtained under the optimized electric potential (2.5 V) applied for decreasing diameter at 30 °C.     

Residual stresses and tetragonal phase fraction characterisation of corrosion tested Zircaloy-4 using energy dispersive synchrotron X-ray diffraction

To improve the understanding of the oxidation mechanism in zirconium alloys for fuel clad applications, detailed residual stress and phase fraction analysis was carried out for the oxides formed on Zircaloy-4 after autoclave exposure at 360 °C for various times by means of synchrotron X-ray diffraction. In a post-transition sample (220 days), significant stress variation through the oxide thickness was found for the monoclinic phase in individual oxide layers, with maximum in-plane compressive stresses located towards the metal–oxide interface and a discontinuity in the residual stress profile. The depth of this discontinuity matched well with the depth at which electron microscopy analysis showed an interface between two distinct oxide layers. Analysis of the tetragonal phase with exposure time demonstrated changes of the total volume of tetragonal phase before and during transition. These observations are put into the context of residual stress evolution presented previously, to provide further insight into the importance of phase transformations and residual stresses in determining the corrosion kinetics of Zr alloys.     

Application of a boron doped diamond (BDD) electrode as an anode for the electrolytic reduction of UO2 in Li2O–LiCl–KCl molten salt

A boron doped diamond thin film electrode was employed as an inert anode to replace a platinum electrode in a conventional electrolytic reduction process for UO₂ reduction in Li₂O–LiCl molten salt at 650 °C. The molten salt was changed into Li₂O–LiCl–KCl to decrease the operation temperature to 550 °C at which the boron doped diamond was chemically stable. The potential for oxygen evolution on the boron doped diamond electrode was determined to be approximately 2.2 V vs. a Li–Pb reference electrode whereas that for Li deposition was around ?0.58 V. The density of the anodic current was low compared to that of the cathodic current. Thus the potential of the cathode might not reach the potential for Li deposition if the surface area of the cathode is too wide compared to that of the anode. Therefore, the ratio of the surface areas of the cathode and anode should be precisely controlled. Because the reduction of UO₂ is dependent on the reaction with Li, the deposition of Li is a prerequisite in the reduction process. In a consecutive reduction run, it was proved that the boron doped diamond could be employed as an inert anode.     

Fatigue assessment of the ITER TF coil case based on JJ1 fatigue tests

The material of the TF coil case in the ITER requires to withstand cyclic electromagnetic forces applied up to 3 × 10⁴ cycles at 4.2 K. A cryogenic stainless steel, JJ1, is used in high stress region of TF coil case. The fatigue characteristics (S–N curve) of JJ1 base metal and welded joint at 4.2 K has been measured. The fatigue strength of base metal and welded joint at 3 × 10⁴ cycles are measured as 1032 and 848 MPa, respectively. The design S–N curve is derived from the measured data taking account of the safety factor of 20 for cycle-to-failure and 2 for fatigue strength, and it indicates that an equivalent alternating stress of the case should be kept less than 516 MPa for the base metal and 424 MPa for the welded joint at 3 × 10⁴ cycles. It is demonstrated that the TF coil case has enough margins for the cyclic operation. It is also shown the welded joint should be located in low cyclic stress region because a residual stress affects the fatigue life.     

Synthesis of nanostructured multiphase Ti(C,N)/a-C films by a plasma focus device

Nanostructured multiphase Ti(C,N)/a-C films were deposited using a 3.3 kJ pulsed plasma focus device onto silicon (1 0 0) substrates at room temperature. The plasma focus device, fitted with solid titanium anode instead of usual hollow copper anode, was operated with nitrogen and Ar/CH₄ as the filling gas. Films were deposited with different number of shots, at 80 mm from top of the anode and at zero angular position with respect to anode axis. X-ray diffraction results show the diffraction peaks related to different compounds such as TiC₂, TiN, Ti₂CN, Ti and TiC_0.62 confirming the deposition of multiphase titanium carbo-nitride composite films on silicon. X-ray photoelectron spectroscopy confirms the formation of Ti–C, C–N, Ti–N, Ti–O and C–C bonds in the films. Scanning electron microscopy reveals that the nanostructure grains are agglomerates of smaller nanoparticles about 10–20 nm in size. Raman studies verify the formation of multiphase Ti(C,N) and also of amorphous graphite in the films. The maximum microhardness value of the composite film is 14.8 ± 1.3 GPa for 30 shots.     

Development of technology for recovering lithium from seawater by electrodialysis using ionic liquid membrane

Tritium fuel for fusion reactors is produced by reacting lithium-6 (⁶Li) with neutrons in tritium breeders. This study demonstrates a method for Li recovery from seawater, wherein Li does not permeate from the anode side to the cathode side through an ionic liquid N,N,N-trimethyl-N-propylammonium–bis(trifluoromethanesulfonyl) imide. Almost all Li ions remain on the anode side (seawater), whereas the other ions in the seawater permeate to the cathode side through the ionic liquid with an applied electric voltage of 2–3 V.     

Development and experimental study of beryllium window for ITER radial X-ray camera

Radial X-ray camera (RXC) is a diagnostic device planned to be installed in the ITER Equatorial Port #12. Beryllium window will be installed between the inner and outer camera of RXC, which severs as the transmission photocathode substrate and also the vacuum isolation component. In this paper the design and manufacture process of two types of beryllium windows were introduced. Although 50 μm thickness of beryllium foil is the best choice, the 80 μm one with X-ray threshold of 1.34 keV was selected for safety consideration. Using the intermediate layer (low purity of beryllium) between the beryllium foil and the stainless steel base flange is an effective strategy to limit the welding thermal deformation and thermal stress of the thin foil caused by bonding between different materials. By using ANSYS software, the feasibility of the aperture design was analyzed and validated. Metal sealing ring was applied in the mechanical clamped beryllium window for its good stability under high temperature and neutron radiation. Although both of the hollow metal sealing ring with 0.03 mm silver coating and the pure silver sealing ring can satisfy the sealing requirement, the later one was chosen to produce the final product. Two hours 240 °C high temperature baking test, two hours 3.3 Hz vibration test and fatigue test were performed on the two types of beryllium windows. Based on the tests results, the two types of beryllium windows could stand the high temperature baking during the wall conditioning phase of ITER tokamak and the vibration during transportation without causing large leakage. Both of the two types of beryllium windows could bear impact load (0.1 MPa pressure difference) for many times without failure.     

Effect of proton irradiation on deuterium diffusion in zirconium oxide layer

Deuterium diffusion in proton-irradiated oxide layer of zirconium alloy has been in situ examined at 573 K by using a combined technique of deuterium plasma exposure and nuclear reaction analysis of the D(³He,p)⁴He reaction. The zirconium alloy used was GNF-Ziron (a high iron Zircaloy-2 type alloy), which had been corroded in high temperature steam, producing an oxide layer of 1.6–1.7 μm thickness. Oxidized specimens were irradiated at ambient temperature with 350 keV H⁺ ions, and the total fluence was 1 × 10¹⁷ cm^?2. An outer non-protective oxide layer of 0.5–0.6 μm thickness, which was observed in the unirradiated oxide layer, evolved into the protective barrier oxide due to the proton irradiation. The evaluated diffusion coefficients in the barrier layer were almost identical for both the unirradiated and irradiated oxides. From X-ray diffraction measurements, lattice expansion and high compressive stress were found in the proton-irradiated oxide. The most probable mechanism for evolution of the diffusion property in the irradiated oxide was ascribed to the increase of the compressive stress induced by the constraint of the damage-accumulated oxide layer by the thick metal substrate.     

Irradiation creep relaxation of void swelling-driven stresses

Swelling-driven-creep test specimens are used to measure the compressive stresses that develop due to constraint of irradiation void swelling. These specimens use a previously non-irradiated 20% CW Type 316 stainless steel holder to axially restrain two Type 304 stainless steel tubular specimens that were previously irradiated in the US Experimental Breeder Reactor (EBR-II) at 490 °C. One specimen was previously irradiated to fluence levels in the void nucleation regime (9 dpa) and the other in the quasi-steady void growth regime (28 dpa). A lift-off compliance measurement technique was used post-irradiation to determine compressive stresses developed during reirradiation of the two specimen assemblies in Row 7 of EBR-II at temperatures of 547 °C and 504 °C, respectively, to additional damage levels each of about 5 dpa. Results obtained on the higher fluence swelling-driven-creep specimen show that compressive stress due to constraint of swelling retards void swelling to a degree that is consistent with active load uniaxial compression specimens that were irradiated as part of a previously reported multiaxial in-reactor creep experiment. Swelling results obtained on the lower fluence swelling-driven creep specimen show a much larger effect of compressive stress in reducing swelling, demonstrating that the larger effect of stress on swelling is on void nucleation as compared to void growth. Test results are analyzed using a recently proposed multiaxial creep-swelling model.     

Design and construction of Alborz tokamak vacuum vessel system

The Alborz tokamak is a D-shape cross section tokamak that is under construction in Amirkabir University of Technology. At the heart of the tokamak is the vacuum vessel and limiter which collectively are referred to as the vacuum vessel system. As one of the key components for the device, the vacuum vessel can provide ultra-high vacuum and clean environment for the plasma operation. The VV systems need upper and lower vertical ports, horizontal ports and oblique ports for diagnostics, vacuum pumping, gas puffing, and maintenance accesses. A limiter is a solid surface which defines the edge of the plasma and designed to protect the wall from the plasma, localizes the plasma–surface interaction and localizes the particle recycling. Basic structure analyses were confirmed by FEM model for dead weight, vacuum pressure and plasma disruptions loads. Stresses at general part of the VV body are lower than the structure material allowable stress (117 MPa) and this analysis show that the maximum stresses occur near the gravity support, and is about 98 MPa.     

Osmosis-induced water uptake by Eurobitum bituminized radioactive waste and pressure development in constant volume conditions

The chemo-hydro-mechanical (CHM) interaction between swelling Eurobitum radioactive bituminized waste (BW) and Boom Clay is investigated to assess the feasibility of geological disposal for the long-term management of this waste. These so-called compatibility studies include laboratory water uptake tests at the Belgian Nuclear Research Center SCK?CEN, and the development of a coupled CHM formulation for Eurobitum by the International Center for Numerical Methods and Engineering (CIMNE, Polytechnical University of Cataluña, Spain).In the water uptake tests, the osmosis-induced swelling, pressure increase and NaNO₃ leaching of small cylindrical BW samples (diameter 38 mm, height 10 mm) is studied under constant total stress conditions and nearly constant volume conditions; the actual geological disposal conditions should be intermediate between these extremes. Two nearly constant volume tests were stopped after 1036 and 1555 days to characterize the morphology of the hydrated BW samples and to visualize the hydrated part with microfocus X-ray Computer Tomography (μCT) and Environmental Scanning Electron Microscopy (ESEM). In parallel, a coupled CHM formulation is developed that describes chemically and hydraulically coupled flow processes in porous materials with salt crystals, and that incorporates a porosity dependent membrane efficiency, permeability and diffusivity.When Eurobitum BW is hydrated in (nearly) constant volume conditions, the osmosis-induced water uptake results in an increasing pressure to values that can be (in theory) as high as 42.8 MPa, being the osmotic pressure of a saturated NaNO₃ solution. After about four years of hydration in nearly constant volume water uptake tests, pressures up to 20 MPa are measured. During this hydration period only the outer layers with a thickness of 1–2 mm were hydrated (as derived from μCT and ESEM analyses), and only about 10–20% of the initial NaNO₃ content was released by the samples. In the studied test conditions, the rates of water uptake and NaNO₃ leaching are low because of the low porosity, and thus low permeability, of the hydrated BW samples in combination with a highly efficient semi-permeable bitumen membrane. In contrast to the hydration in free swelling conditions, the increase in porosity is limited by the high pressures in the nearly constant volume tests. Furthermore, at the interface with the stainless steel filters, a low permeable re-compressed bitumen layer is formed, as observed on the ESEM images.The experimental results of pressure increase and NaNO₃ leaching, as well as observations on μCT and ESEM images (e.g. compression of leached layers, high dissolved NaNO₃ concentration in hydrated BW after about four years), were reproduced rather successfully by the coupled CHM formulation for Eurobitum BW. A long-term model prediction of the evolution of the osmosis-induced pressure in the nearly constant volume tests shows that the pressure would reach a maximal value of about 20 MPa after about 5.5 years, after which the pressure would start to decrease. After 10,000 days (～27 years) the pressure would have decreased to a value of ～2 MPa.     

Joining C/C composite to copper using active Cu–3.5Si braze

A simple technique was developed to join C/C composite to Cu using active Cu–3.5Si braze for nuclear thermal applications. The brazing alloy exhibited good wettability on C/C substrate due to the reaction layer formed at the interface. A strong interfacial bond of the brazing alloy on C/C with the formation of TiC + SiC + Ti₅Si₃ reaction layer was obtained. The produced CC/Cu/CuCrZr joint exhibited shear strength as high as 79 MPa and excellent thermal resistance during the thermal shock tests.     

Compatibility of CLAM steel weldments with static LiPb alloy at 550 °C

CLAM steel is considered as a structural material to be used in the Test Blanket Module as a barrier or blanket adjacent to liquid LiPb in fusion reactors. In this paper, CLAM steel is welded by tungsten inert gas (TIG) welding, and the compatibility of the weldment with liquid LiPb is tested. Specimens were corroded in static liquid LiPb, with corrosion times of 500 h and 1000 h, at 550 °C, and the corresponding weight losses are 0.272 mg/cm² and 0.403 mg/cm² respectively. Also the corrosion rate decreases with increased corrosion time. In the as-welded condition, corrosion resistance of the weld zone is higher than that of the HAZ (Heat Affected Zone). Likely, thick martensite lath and large residual stresses at the welding zone result in higher corrosion rates. The compatibility of CLAM steel weld joints with high temperature liquid LiPb can be improved to some extent through a post-weld tempering process. The surface of the as-welded CLAM steel is uniformly corroded and the concentration of Cr on the surface decreases by about 50% after corrosion. Penetration of LiPb into the matrix is observed for neither the as-welded nor the as-tempered conditions. Influenced by thick martensite lath and large residual stresses, the welded area, especially the weld zone, is easily corroded, therefore it is of primary importance to protect the welded area in the solid blanket of the fusion reactor.     

Vacuum plasma-sprayed tungsten on EUROFER and 316L: Results of characterisation and thermal loading tests

Vacuum plasma-spraying (VPS) can be used for the industrial deposition of thick W coatings on actively water-cooled components made of low activation steel or stainless steel. Mock-ups made of martensitic steels, EUROFER and F82H, as well as steel 316L, were coated with 2 mm thick W-VPS layers. The coated materials are candidates for first wall components (ITER and DEMO) receiving moderate heat load of up to 1 MW/m². Mixed tungsten/steel interlayers were applied to reduce the residual and thermal stresses at the substrate–coating interface and to improve the adhesion of the coating. The advantage of this mixed interlayer is that no further (high activation) materials have to be introduced to improve coating adhesion.The characterisation of the W-VPS layers includes the evaluation of the coating microstructure, the measurement of physical and mechanical properties and the metallographical examination before and after heat load tests.Heat load tests with steady state operation up to 2.5 MW/m² and cycling heat loads of 2 MW/m², were successfully completed. They confirm the thermomechanical suitability of industrially manufactured W-VPS coatings for plasma facing first wall components made of steel.     

Fatigue cracking of a bare steel first wall in an inertial confinement fusion chamber

Inertial confinement fusion power plants will deposit high energy X-rays onto the outer surfaces of the first wall many times a second for the lifetime of the plant. These X-rays create brief temperature spikes in the first few microns of the wall, which cause an associated highly compressive stress response on the surface of the material. The periodicity of this stress pulse is a concern due to the possibility of fatigue cracking of the wall. We have used finite element analyses to simulate the conditions present on the first wall in order to evaluate the driving force of crack propagation on fusion-facing surface cracks.Analysis results indicate that the X-ray induced plastic compressive stress creates a region of residual tension on the surface between pulses. This tension film will likely result in surface cracking upon repeated cycling. Additionally, the compressive pulse may induce plasticity ahead of the crack tip, leaving residual tension in its wake. However, the stress amplitude decreases dramatically for depths greater than 80–100 μm into the fusion-facing surface. Crack propagation models as well as stress-life estimates agree that even though small cracks may form on the surface of the wall, they are unlikely to propagate further than 100 μm without assistance from creep or grain erosion phenomena.     

Design and fabrication of the KSTAR in-vessel cryo-pump

In-vessel cryo-pump (IVCP) of the Korea Superconducting Tokamak Advanced Research (KSTAR) has been designed, fabricated, and installed in the vacuum vessel for effective particle control by pumping through a divertor gap. For the final engineering design of the IVCP supports to withstand all external forces, a structure analyses were performed for two cases. The first is the thermal stress due to cool-down from room temperature to operating temperature (cryo-panel: 4.4 K, thermal shield: 77 K), and the other is the electro-magnetic stress due to the induced eddy currents during plasma disruptions. When the plasma disrupts, the maximum stress and displacement on the supports were estimated to be 849 MPa and 5.36 mm, respectively. These results were taken into account in the support design. The IVCP system was fabricated in two half-sectors and a pre-assembling test was successfully completed in the factory. Final installation of the IVCP in the vacuum vessel was fulfilled in parallel with a pressurization test (thermal shield: 30 bar, cryo-panel: 10 bar), a helium leak test, and a thermal shock test using liquid nitrogen. As a result, the IVCP system was successfully installed in the vacuum vessel.     

Creep-to-rupture of the steel P92 at 650 °C in oxygen-controlled stagnant lead in comparison to air

Creep-to-rupture experiments were performed on 9%-Cr ferritic–martensitic steel P92 in the CRISLA facility. The specimens of P92 were examined at 650 °C and static tensile stress between 75 and 325 MPa in both stagnant lead with 10^?6 mass% dissolved oxygen and air. The steel showed an insignificant difference in time-to-rupture, t_R, and ductile fracture in both environments at >100 MPa, corresponding to t_R < 3,442 h. At 75 MPa in Pb (t_R = 13,090 h), the steel, however, featured purely brittle fracture pointing to liquid metal embrittlement. Structural changes in the steel and surface oxidation in the different environments were studied using metallographic techniques. The Laves phase that forms during thermal aging at 650 °C was found along prior austenite grain boundaries and martensite laths already after relatively short testing time, along with chromium carbides that are already present in the as-received condition of the steel.     

Phase transformation induced by ion implantation in cubic stabilized zirconia

Cubic yttria-stabilized zirconia possesses a high stability against radiation. No amorphization of this material has been observed, even at high ion fluences leading to the production of a large amount of defects. Nevertheless irradiation with energetic particles may induce microstructural evolutions and phase transformations. In the present paper we demonstrate that a cubic-to-rhombohedral phase transformation occurs in yttria-stabilized zirconia implanted with He ions. This transformation consists in a rhombohedral deformation of the cubic cell along the 〈1 1 1〉 directions due to residual stresses induced by implantation.     

Model validation of GAMMA code with heat transfer experiment for KO TBM in ITER

By considering the requirements for a DEMO-relevant blanket concept, Korea (KO) has proposed a He cooled molten lithium (HCML) test blanket module (TBM) for testing in ITER. A performance analysis for the thermal–hydraulics and a safety analysis for the KO TBM have been carried out using a commercial CFD code, ANSYS-CFX, and a system code, GAMMA (GAs multicomponent mixture analysis), which was developed by the gas cooled reactor in Korea. To verify the codes, a preliminary study was performed by Lee using a single TBM first wall (FW) mock-up made from the same material as the KO TBM, ferritic martensitic steel, using a 6 MPa nitrogen gas loop. The test was performed at pressures of 1.1, 1.9 and 2.9 MPa, and under various ranges of flow rate from 0.0105 to 0.0407 kg/s with a constant wall temperature condition. In the present study, a thermal–hydraulic test was performed with the newly constructed helium supplying system, in which the design pressure and temperature were 9 MPa and 500 °C, respectively. In the experiment, the same mock-up was used, and the test was performed under the conditions of 3 MPa pressure, 30 °C inlet temperature and 70 m/s helium velocity, which are almost same conditions of the KO TBM FW. One side of the mock-up was heated with a constant heat flux of 0.3–0.5 MW/m² using a graphite heating system, KoHLT-2 (Korea heat load test facility-2). Because the comparison result between CFX 11 and GAMMA showed a difference tendency, the modification of heat transfer correlation included in GAMMA was performed. And the modified GAMMA showed the strong parity with CFX 11 calculation results.