Active Metal Casting Research Articles

Status of High Heat Flux Components at W7-X

The actively water-cooled in-vessel components (IVCs) of the stellarator Wendelstein 7-X consist of the divertor, the first wall protection components, the port liners, each designed for different loading conditions, and the associated pipework, the control coils, the cryo-pump system, the Glow discharge electrodes, and a set of diagnostics. The divertor, designed for high heat fluxes (HHFs), is a set of 10 target and baffle units arranged along the plasma surface. The design and production of these HHF components is a challenging task. The divertor target elements, which are based on flat carbon-carbon fiber composite tiles bonded via active metal casting onto CuCrZr cooling structures required intensive development and testing to reach a reliable performance; removing, under stationary conditions, 10 MW/m2. Industrially manufactured high quality target elements have been delivered and assessed, and the process of incorporating them into assembly units, so-called modules, has begun. The time scale for the completion of the HHF divertor has been held for the last four years and the final delivery of the HHF divertor is still planned in 2017. In parallel to the realization of the divertor, most of the remaining IVCs have been defined, developed, designed, and fabricated and the installation of many of these components has begun. Some of these components can also be expected, for a short period of time, to receive high heat loads approaching those of the divertor. These components will be described, in detail, from conception to realization.

The development of joining doped graphite to copper for first wall application in HT-7 tokamak

Two joining methods have been developed for joining carbon based plasma facing material to copper based heat sink material for the potential application in HT-7 and EAST tokamak. The first joining method is based on brazing technique by using a rapidly solidified foil-type Ti–Zr based amorphous filler with a melting temperature of 850 °C. The other joining method is direct active metal casting–casting the premixed powders of copper and active transition metals on the mechanical machined carbon surface directly. SEM observations demonstrate high quality of joining surface for both joints. The brazing technique is more promising for fabrication joint with larger size compared with the direct active alloy casting method. High heat flux test using an e-beam device was performed on the actively cooled C/Cu joint fabricated by brazing method. There has no damage occurred on the joint after heat loading at 6 MW/m 2.

Results of the examinations of the W7-X pre-series target elements

The target elements of the WENDELSTEIN 7-X (W7-X) divertor are designed to sustain a stationary heat flux of 10 MW/m 2 and to remove a maximum power of 100 kW. CFC Sepcarb ® NB31 tiles are bonded to a water-cooled CuCrZr heat sink in two steps: active metal casting (AMC ®) of an AMC ®-copper interlayer to CFC tiles, electron beam welding (EBW) or hot isostatic pressing (HIP) of the AMC ®-NB31 tiles to CuCrZr. The fabrication of the whole amount of CFC NB31 has been completed. The key target of the pre-series phase is the qualification of this bond based on a series of examinations. The introduction of silicon during the AMC ® process significantly improved the strength of the joint between CFC and AMC ®-copper. The strength of the bond is preserved after either EBW or HIP processes. High heat flux testing carried out in the ion beam facility GLADIS exhibited a too high percentage of defective tiles. Pre-series activities have been extended to reduce the stress concentration at the interface between tiles and heat sink by optimizing the design.

Examination of C/C flat tile mock-ups with hypervapotron cooling after high heat flux testing

Two C/C flat tile mock-ups with a hypervapotron cooling concept, have been successfully tested beyond ITER specification (3000 cycles at 15 MW/m 2, 300 cycles at 20 MW/m 2 and 800–1000 cycles at 25 MW/m 2) in two electron beam testing facilities [F. Escourbiac, et al., Experimental simulation of cascade failure effect on tungsten and CFC flat tile armoured HHF components, Fusion Eng. Des., submitted for publication; F. Escourbiac, et al., A mature industrial solution for ITER divertor plasma facing components: hypervapotron cooling concept adapted to Tore Supra flat tile technology, Fusion Eng. Des. 75–79 (2005) 387–390]. Both mock-ups provide a SNECMA SEPCARB ® NS31 armour, which has been joined onto the CuCrZr heat sink by active metal casting (AMC) and electron beam welding (EBW). No tile detachment or sudden loss of single tiles has been observed; a cascade-like failure of flat tile armours was impossible to generate. At the maximum cyclic heat flux load of 25 MW/m 2 all tested tiles performed well except one, which revealed already a clear indication in the thermographic examination at the end of the manufacture. Visual examination and analysis of metallographic cuts of the remaining tiles demonstrated that the interface has not been altered. In addition, the shear strength of the C/C to copper joints measured after the high heat flux (HHF) test has been found to be still above the interlamellar shear strength of the used C/C material. The high resistance of the interface is explained by a modification of the C/C to copper joint interface due to silicon originating from the used C/C material.

Technologies for ITER divertor vertical target plasma facing components

The ITER divertor vertical target has to sustain heat fluxes up to 20 MW m−2. The concept developed for this plasma facing component working at steady state is based on carbon fibre composite armour for the lower straight part and tungsten for the curved upper part. The main challenges involved in the use of such components include the removal of the high heat fluxes deposited and mechanically and thermally joining the armour to the metallic heat sink, despite the mismatch in the thermal expansions. Two solutions based on the use of a CuCrZr hardened copper alloy and an active metal casting (AMC®) process were investigated during the ITER EDA phase: the first one called ‘flat tile geometry’ was mainly developed for the Tore Supra pumped limiter, the second one called ‘monoblock geometry’ was developed by the EU Participating Team for the ITER project. This paper presents a review of these two solutions and analyses their assets and drawbacks: pressure drop, critical heat flux, surface temperature and expected behaviour during operation, risks during the manufacture, control of the armour defects during the manufacture and at the reception, and the possibility of repairing defective tiles.

Experience feedback from high heat flux component manufacturing for Tore Supra

Tore Supra is involved in flat tile carbon armoured plasma facing components (PFCs) since 1985. In 1997, a third generation of components, based on the original concept developed with Plansee Company, called active metal casting (AMC®), has been launched. Since 1998, 660 elementary components for the toroidal pump limiter (TPL) are in production. The route of the manufacture is rather complex and many controls were requested all along the fabrication to insure a high reliability of the elements. One of the main controls is the final infrared (IR) test allowing to determine the quality of the bonding between the carbon fibre composite (CFC) tiles and the heat sink made of copper–chromium–zirconium alloy (CuCrZr). Although results for the first batch of elements were as expected (less than 5% rejected at the final test), unexpected defects appeared with the followings batches. Investigations on the fabrication processes underlined the importance of having a better heat treatment of the pieces in copper alloy (CuCrZr), however this was not sufficient to completely explain the observed defects.

First plasma experiments in Tore Supra with a new generation of high heat flux limiters for RF antennas

During the 1997 and 1998 Tore Supra shutdown, a first set of new antenna guard limiters was installed on one of the three ion cycloton resonance heating (ICRH) antennas of Tore Supra. This limiter, which was one of the main technological studies of the 1998 campaign, was widely experimented in real plasma conditions, thus allowing the validation in situ, for the first time, of the technology of active metal casting (AMC) for plasma facing components. The huge improvement in the thermal response of the new limiter generation, compared to the old one, is shown on plasma pulses made identical in terms of antenna position and injected RF power profile. By using the infrared cameras installed inside Tore Supra and viewing the antennas front, the power density fluxes received by the carbon fibre composite (CFC) surface of the limiter were evaluated by correlation with the heat load tests made on the electrons beam facility of CEA/Framatome.

Evidence of damage in carbon fibre composite tiles joined to a metallic heat sink under high heat flux fatigue

The two years experience with Active Metal Casting (AMC) flat bonds shows that this technology is suitable for the heat fluxes expected in Tore Supra (10 MW/m 2). Tests were pursued up to 3330 cycles, with elements still functional. At higher heat fluxes, fatigue damage is observed, but the bond resists remarkably well with no tile detachment. Examination of such deliberately damaged bonds showed distributed cracking, proving the absence of any weak link. The limitations to those higher heat fluxes are more related to the design and the base materials than to the bond itself.

Advanced solutions for beryllium and tungsten plasma-facing components

Beryllium and tungsten are candidate plasma-facing armour materials for the International Thermonuclear Experimental Reactor (ITER). These armours are proposed for areas with low heat flux (⩽5 MW m −2); however, in the divertor, surface melting during abnormal events may occur. This paper reports the progress made in developing novel approaches to solving the difficulties posed in designing with these armours. A Be monoblock brazed to an oxygen free high conductivity (OFHC) 10 mm ID Cu tube using InCuSil `ABA' braze alloy has survived 130 cycles of 10–11 MW m −2 for 6 s, with surface temperatures of 1250°C. No visible surface cracking occurred. The same monoblock was then exposed to several cycles of 20–22 MW m −2 for 8 s, creating a 2 mm deep molten layer. High cycle fatigue was then performed. The test results are detailed in this paper. Comparison between experimental and theoretical results are made. W and Cu have a large mismatch in their thermal expansion coefficients and two designs are proposed that minimise the interface stresses. These are: a `brush'-like structure with rectangular fibres set in a Cu substrate using the `active metal casting' (AMC) technique; and thin monoblocks (or lamellae) brazed or active metal cast onto a Cu tube. Analyses of the lamellae concept for steady-state heat loads of 5 MW m −2 are presented. Fatigue analyses show that both solutions are theoretically viable (∼10 4 cycles). A `brush' mock-up has been manufactured and progress on its testing is reported. Results of all tests and their relevance to the ITER design are discussed.

Thermal fatigue tests with actively cooled divertor mock-ups for ITER

Mock-ups for high heat flux components with beryllium and CFC armour materials have been tested by means of the electron beam facility JUDITH. The experiments concerned screening tests to evaluate heat removal efficiency and thermal fatigue tests. CFC monoblocks attached to DS-Cu (Glidcop Al25) and CuCrZr tubes by active metal casting and Ti brazing showed the best thermal fatigue behaviour. They survived more than 1000 cycles at heat loads up to 25 MW m −2 without any indication of failure. Operational limits are given only by the surface temperature on the CFC tiles. Most of the beryllium mock-ups were of the flat tile type. Joining techniques were brazing, hot isostatic pressing (HIP) and diffusion bonding. HIPed and diffusion bonded Be/Cu modules have not yet reached the standards for application in high heat flux components. The limit of this production method is reached for heat loads of approximately 5 MW m −2. Brazing with and without silver seems to be a more robust solution. A flat tile mock-up with CuMnSnCe braze was loaded at 5.4 MW m −2 for 1000 cycles without damage. The first test with a beryllium monoblock joined to a CuCrZr tube by means of Incusil brazing shows promising results; it survived 1000 cycles at 4.5 MW m −2 without failure.

Design, fabrication and testing of an improved high heat flux element, experience feedback on steady state plasma facing components in Tore Supra

Actively cooled plasma facing components (PFC) have been developed and used in Tore Supra since 1985. One of the main technological problem is due to the expansion mismatch between graphite armour and metallic heat sink material. A first technology used graphite tiles with or without a reinforcement and a compliant layer, brazed with titanium copper–silver (TiCuAg) alloy. The next technology used carbon fiber material (CFC) tiles with a 2 mm pure copper compliant layer, since the good mechanical strength of the CFC allowed the reinforcement layer to be suppressed. No destructive inspection during the manufacturing procedure was found to be essential to insure a good reliability of the elements. A recent technology was developed for the new actively cooled toroidal pump limiter of Tore Supra (designed to remove up to 15 MW during 1000 s with incident heat flux up to 10 MW m −2). This new technology uses an active metal casting (AMC) of copper onto CFC tiles whose surface is prepared by a laser treatment. Copper compliant layer is then electron-beam welded to the heat sink. This technology is silver free and could be ITER relevant. The first bonding is X-ray controlled and the second one ultrasonic (US) tested. Finally all the elements are controlled by an infrared imaging technique during a hot water test.