Since 1995, DIII-D has performed correction of magnetic field imperfections using a set of six external picture frame coils located on the vessel mid-plane. In 2000, these coils also demonstrated benefits when used for feedback of the resistive wall mode, an instability that limits the plasma performance at high beta. Modeling has shown that substantial performance improvements could be achieved by installing new coils inside the vessel and expanding the poloidal coverage above and below the mid-plane. Two prototype internal coils were installed in 2001 and were power tested successfully after several bakes to 350/spl deg/C. A full set of twelve internal coils and related magnetic sensors are now operational in the DIII-D tokamak. The design requirements for the new coil system was to maximize the magnetic field at the plasma edge, operate with a frequency range of dc to 1000 Hz, and fit behind the existing graphite wall tiles. The coil design adopted and installed is a water-cooled hollow copper conductor insulated with polyamide and housed inside a stainless steel tube that forms a vacuum boundary. The coil is rigidly mounted to the inside of the vacuum vessel. The primary challenge in the design of these coils was in joining of both the copper conductor and the stainless tube without overheating the polyamide insulator. Elastic-plastic analysis was used to demonstrate acceptable thermal stresses during baking conditions. Analysis determined the optimum water cooling channel diameter. The coils were tested in high toroidal field to the limit of the power supply of 4.5 kA DC with inductance-limited current for frequencies between 300 Hz and 1000 Hz. Recent results are presented.

The lower divertor of the DIII-D tokamak is being modified to provide improved density control of the tokamak plasma during operation in a high triangularity double null configuration. This divertor replaces the low triangularity advanced divertor installed in 1990. The design and analysis of the lower divertor is complete and hardware is being fabricated. Installation of the new divertor is scheduled to be completed by the end of 2005. The primary component of the lower divertor is a toroidally continuous flat plate. The plate is water cooled for heat removal. Three rows of graphite tiles are mechanically attached to the plate to shield it from plasma impingement. Owing to a concern over excessive erosion caused by plasma impingement, the through tile-face bolt holes have been eliminated from graphite in areas of high heat flux. The plate is water cooled for heat removal between shots and heated to 350degC with hot air and inductive current during vessel baking. The divertor plate is supported 100 mm from the vacuum vessel floor by two rows of 24 supports that must react the vertical loads due to halo currents. These supports are radially flexible to allow for differential radial thermal expansion between the divertor ring and the floor. The space below the plate forms a pumping plenum connecting the floor strike point to the lower cryopump. Upgraded floor tiles inboard of the plate will be installed to improve the target for the plasma strike point for outer leg pumping. The divertor plate is to be fabricated in four 90 deg sectors from type 316 stainless steel. Each sector consists of two plate halves with three machined coolant channels and is joined together by spot welds and perimeter seam TIG welds. The vacuum tight 90 deg plate sectors are welded together inside the vessel to form a toroidally continuous ring. The water cooling/air bake-out lines connecting the 4 sectors into two 180 deg cooling circuits will be welded in situ. Several plasma diagnostics wil- l require some modification or relocation for integration into the divertor system

—In September 2003, a water leak emanating from a magnet pole shield in one of the 80 keV neutral beam lines halted DIII-D physics experiments. Investigation showed that impingement of the diverted charged particle beam on the copper shield plate initiated melting and a through-wall crack that propagated through a brazed stainless steel cooling line. A subsequent survey of all pole shields in the 4 neutral beam lines found that similar damage was present in 9 of the other 15 shields but to a lesser extent. The decision was made to arrest the crack propagation in damaged shields and to begin fabrication of new pole shields, which will be installed during the long torus opening of 2005-

In September 2003, a water leak emanating from a magnet pole shield in one of the 80 keV neutral beam lines halted DIII-D physics experiments. Investigation showed that impingement of the diverted charged particle beam on the copper shield plate initiated melting and a through-wall crack that propagated through a brazed stainless steel cooling line. A subsequent survey of all pole shields in the 4 neutral beam lines found that similar damage was present in 9 of the other 15 shields but to a lesser extent. The decision was made to arrest the crack propagation in damaged shields and to begin fabrication of new pole shields, which will be installed during the long torus opening of 2005-2006. A summary is presented of the physical damage that was found in the shields and the assessments that were conducted to determine the cause of the damage. The custom tooling that was developed and then utilized inside the beam lines to arrest crack propagation in the shields is also presented. Modifications to the design of the shields were made to minimize stress concentrations and were implemented in the new pole shields. A novel technique was devised to fabricate the serpentine cooling tubes that are brazed into matching grooves in the shields. This presentation includes descriptions of the shield design modifications and the procedures that were employed to successfully complete the fabrication of the new pole shields

Since 1995, DIII-D has performed correction of magnetic field imperfections using a set of six external picture frame coils located on the vessel mid-plane. Recently, these coils have also demonstrated significant benefits when used for feedback of the resistive wall mode, an instability that limits the plasma performance at high beta. Modelling has shown that substantial performance improvements can be achieved by installing new coils inside the vessel and expanding the poloidal coverage above and below the mid-plane. Two prototype internal coils were installed in 2001 and have been tested successfully. Installation of a set of twelve internal coils and magnetic sensors in the DIII-D tokamak is to be completed in December 2002. The design requirement for the new coil system was to maximize the magnetic field at the plasma edge, operate with a frequency range of dc to 1000 Hz, and fit behind the existing graphite wall tiles. The coil design adopted and installed is a water-cooled hollow copper conductor insulated with polyamide and housed inside a stainless steel tube that forms a vacuum boundary. The coil is rigidly mounted to the inside of the vacuum vessel. The primary challenge in the design of these coils was in joining of both the copper conductor and the stainless tube without overheating the polyamide insulator.

The fusion engineering research experiment (FIRE) device is designed for high power density and advanced physics operating modes. Due to the short distance of the divertor from the X-point, the connection lengths are short and the scrape off layer thickness is small. A relatively high peak heat flux of 25 MW/m2 is expected on the divertor. The FIRE divertor engineering design is based on the design approaches developed for international thermonuclear experimental reactor (ITER). The geometry of the FIRE divertor consists of water cooled copper fingers and a tungsten brush armor as plasma facing material. The divertor assembly consists of modular units for remote handling. A 316 stainless steel back plate is used for support and manifolding. The backing plate is joined to the copper fingers by pins. The coolant channel diameter is 8 mm at a pitch of 14 mm. The total power flow to the outer divertor is 35 MW. Water at an inlet temperature of 30 °C, 1.5 MPa and a flow velocity of 10 m/s is used with two channels in series. A margin of ≈1.6 is obtained on the critical heat flux. A three dimensional thermal stress finite element (FE) analysis of this geometry was performed. Thermal hydraulic correlations derived for ITER were used to perform the thermal analysis. Design changes were implemented to reduce the stresses and temperatures to acceptable levels.

Recent calibration shots in which full toroidal field (TF) coil current interacted with the maximum poloidal field coils have produced increased lateral loads on the outer sections of the TF-coil. The increased lateral loads have resulted in deflections that have been sufficient to cause the TF-coil to contact adjacent equipment and produce a transient short to ground within the coil. The six outer turns of each TF-coil bundle are clamped together by insulated preloaded studs to provide increased bending stiffness. These sections of the outer bundles depend on friction to react the lateral loads as a bundle rather than six individual turns. A major concern is that the increased loads will produce slip between turns resulting in excessive lateral deflections and possible damage to the insulating sleeve on the preloaded studs. A finite element structural model of the TF-coil was developed for the calculation of deflections and the shear load distribution throughout the coil for the applied lateral loads from a full current calibration shot. The purpose of the updated structural model is to correlate the applied lateral loads to the total shear force between the unbonded sections of the outer turns. An allowable integrated lateral load applied to the outer turns is established based on the maximum shear force that can be reacted by friction. A program that calculates the magnetic fields and integrated lateral load along the outer turns can be incorporated into the plasma control system. The integrated load can then be compared to the calculated allowable value prior to execution of calibration shots. Calibration shots with a calculated total lateral load greater than the allowable value will be prevented.

Each GA ECH launcher mirrors is used to transmit 800 kW of power to the plasma. Until 2000, the pulse length for use of these mirrors was limited to 2 s due to temperatures of the mirrors resulting from 1) a high ratcheted bulk temperature and 2) a large increase in temperature of the mirror during the pulse. A new design was proposed and implemented which has extended the capability of the mirror to 10 s with passive cooling. The important features of the new design are 1) increase in the passive heat transfer rate during cooling and 2) a modified shape of the mirror. The analysis shows that, the new mirrors can be used for 10 s pulses. The new mirrors have been installed in DIII-D, they have been used for up to 2 s pulses.

The DIII–D program has completed a series of density control and plasma core confinement experiments this past year. These experiments were designed to investigate the performance of baffled and open divertors with single-null plasmas and particle control in double-null plasmas. The experiments utilized all three of the DIII–D divertor assemblies located in the lower outer corner, the upper outer corner, and the upper inner corner of the vessel, which were installed last year. Each divertor consists of a liquid helium cryopump, a shielded protective ring, and a gas puff system. The divertors were designed to optimize pumping performance and to withstand the electromagnetic loads from both halo and toroidal, induced currents. With theoretical pumping speeds varying from 15,000 to 32,000 l/s, the cryopumps, combined with the baffle structures, collect particles and prevent them from recirculating back into the plasma core. The intent of the gas puff systems is to inject neutral gases in and around the divertors to minimize the heat flux on the divertors, minimizing the impurities generated by the excessive heating of the divertor graphite tiles. This hardware permits either single- or double-null plasma experiments and enables continued research of well confined high beta divertor plasmas with noninductive current drive, which is one of the primary research goals of DIII–D.

The design of a double-null divertor for use in JT-60SU is presented. The free-boundary equilibrium code EFIT is used to establish a symmetric highly triangular double-null plasma shape. The baffle shapes are highly contoured to match the equilibrium, with the plasma-facing surfaces intersecting the flux surfaces at steep angles in the regions of high heat flux. These contoured surfaces also provide a tightly baffled design with small-aperture pumping gaps near both the inner and outer divertor strike points. The gaps provide adequate throughput of D-2 gas for active control of impurity entrainment at reasonable pressures. The structural design is shown to be consistent with both forces from disruptions and thermal stress during vacuum vessel bakeout.

General Atomics recently completed a divertor design study for JAERI for the JT-60 Super Upgrade (JT-60SU) tokamak. JT-60SU is being designed as a superconducting device for an integrated R&D investigation of steady-state operation in a tokamak. A divertor design was developed to accommodate double-null operation for a 1000 s discharge duration at 8 MA of plasma current and 80 MW of auxiliary heating. The work reported here is an extension of a previous design study.(1,2) The thermal requirements are a peak heat flux of 9 MW/m(2), a maximum surface temperature of 1600 degreesC, and a poloidal cooling flow configuration for the plasma facing components. The structural requirements are determined from both the predicted stresses due to halo currents as well as the stresses due to differential thermal expansion encountered during bakeout. The halo current loads are based on a nominal halo current of 0.19 Ip with a 2.0 toroidal peaking factor. Analysis has determined that the halo current load per centimeter of circumference is P = 2856 (1+cos theta) N/cm, where 8 is the toroidal angle. The loads due to differential thermal expansion are a result of an expected 100 degreesC temperature difference between the vacuum vessel and divertor during bakeout.Based on the aforementioned criteria, a divertor design was developed for all three areas of the JT-60SU divertor: the inner baffle, the private flux baffle, and the outer baffle. In order to have highly reliable divertor components, flexible supports sized to accommodate the structural loads are utilized in the design rather than insulators or sliding interfaces. The plasma facing components are mounted on a structural mounting plate to form a removable and remotely-maintainable segment which is in turn mounted on the supports. For outer and private flux baffles, these structural mounting plates are joined together using a double shear joint design to form a structurally continuous ring to react the halo current loads. The plasma facing components are broken into 80 segmentation; however, the outer and private flux baffles have an alternating 8 degrees and 16 degrees structural segmentation which forms the double shear toroidal structural joint. The inner baffle takes advantage of its relatively short poloidal length and its proximity to the vacuum vessel to provide structural integrity. The thermal design consists of a plasma facing material of flat CFC tiles that are brazed onto a poloidally cooled copper heat sink. Adequate gaps between the baffles and wall are provided for pumping of recycled gas.

Under the APEX program the He-cooled system design task is to evaluate and recommend high power density refractory alloy first wall and blanket designs and to recommend and initiate tests to address critical issues. We completed the preliminary design of a helium-cooled, W–5Re alloy, lithium breeder design and the results are reported in this paper. Many areas of the design were assessed, including material selection, helium impurity control, and mechanical, nuclear and thermal hydraulics design, and waste disposal, tritium and safety design. Systems study results show that at a closed cycle gas turbine (CCGT) gross thermal efficiency of 57.5%, a superconducting coil tokamak reactor, with an aspect ratio of 4, and an output power of 2 GWe, can be projected to have a cost of electricity at 54.6 mill/kW h. Critical issues were identified and we plan to continue the design on some of the critical issues during the next phase of the APEX design study.

An aggressive low aspect ratio scoping fusion reactor design (C.P.C. Wong, R. Cerbone, E.T. Cheng, R.L. Miller, R.D. Stambaugh, Proc. of 17th IEEE/NPSS Symp. on Fusion Engineering, pp. 1053) indicated that a 2 GW(e) reactor can have a major radius as small as 2.9 m resulting in a device with competitive cost of electricity at 49 mill/kWh. One of the technology requirements of this design is a high performance high power density first wall and blanket system. A 15 MPa helium-cooled, V-alloy and stagnant LiPb breeder first wall and blanket design was utilized. Due to the low solubility of tritium in LiPb, there is the concern of tritium migration and the formation of V-hydride. To address these issues, a lithium breeder system with high solubility of tritium has been evaluated. Due to the reduction of blanket energy multiplication to 1.2, to maintain a plant Q of >4, the major radius of the reactor has to be increased to 3.05 m. The inlet helium coolant temperature is raised to 430°C in order to meet the minimum V-alloy temperature limit everywhere in the first wall and blanket system. To enhance the first wall heat transfer, a swirl tape coolant channel design is used. The corresponding increase in friction factor is also taken into consideration. To reduce the coolant system pressure drop, the helium pressure is increased from 15 to 18 MPa. Thermal structural analysis is performed for a simple tube design. With an inside tube diameter of 1 cm and a wall thickness of 1.5 mm, the lithium breeder can remove an average heat flux and neutron wall loading of 2 and 8 MW/m2, respectively. This reference design can meet all the temperature and material structural design limits, as well as the coolant velocity limits. Maintaining an outlet coolant temperature of 650°C, one can expect a gross closed cycle gas turbine thermal efficiency of 45%. This study further supports the use of helium coolant for high power density reactor design. When used with the low aspect ratio reactor concept a competitive fusion reactor can be projected at 51.9 mill/kWh.

A cryocondensation pump for the upper inboard divertor on DIII-D is to be installed in the vacuum vessel in the fall of 1999. The cryopump removes neutral gas particles from the divertor and prevents recycling to the plasma. This pump is designed for a pumping speed of 18,000 l/s at 0.4 mTorr. The cryopump is toroidally continuous to minimize inductive voltages and avoid electrical breakdown during disruptions. The cryopump consists of a 25 mm Inconel tube cooled by liquid helium and is surrounded by nitrogen cooled shields. A segmented ambient temperature radiation/particle shield protects the nitrogen shields. The pump is subjected to a steady state heat load of less than 10 W due to conduction and radiation heat transfer

Beginning in 1992, the first of the DIII–D divertor baffles and cryocondensation pumps was installed. This open divertor configuration, located on the outermost floor of the DIII–D vessel, includes a cryopump with a predicted pumping speed of 50,000 l/s excluding obstructions such as support hardware. Taking the pump structural and support characteristics into consideration, the corrected pumping speed for D2 is 30,000 l/s [1]. In 1996, the second divertor baffle and cryopump were installed. This closed divertor structure, located on the outermost ceiling of the DIII–D vessel, has a cryopump with a predicted pumping speed of 32,000 l/s. In the fall of 1999, the third divertor baffle and cryopump will be installed. This divertor structure will be located on the 45° angled corner on the innermost ceiling of the DIII–D vessel, known as the private flux region of the plasma configuration. With hardware supports factored into the pumping speed calculation, the private flux cryopump is expected to have a pumping speed of 15,000 l/s. There was question regarding the effectiveness of the private flux cryopump due to the close proximity of the private flux baffle. This led to a conductance calculation study of the impact of rotating the cryopump aperture by 180° to allow for greater particle and gas exhaust into the cryopump’s helium panel. This study concluded that the cost and schedule impact of changing the private flux cryopump orientation and design did not warrant the possible 20% (3,000 l/s) increase in pumping ability gained by rotating the cryopump aperture 180°. The comparison of pumping speed of the first two cryocondensation pumps with the measured results will be presented as well as the calculation of the pumping speed for the private flux cryopump now being installed.

Beginning in 1992, the first of the DIII-D divertor baffles and cryocondensation pumps was installed. This open divertor configuration, located on the outermost floor of the DIII-D vessel, includes a cryopump with a predicted pumping speed of 50,000 l/s excluding obstructions such as support hardware. Taking the pump structural and support characteristics into consideration, the corrected pumping speed for D/sub 2/ is 30,000 l/s. In 1996, the second divertor baffle and cryopump were installed. This closed divertor structure, located on the outermost ceiling of the DIII-D vessel, has a cryopump with a predicted pumping speed of 32,000 l/s. In the fall of 1999, the third divertor baffle and cryopump will be installed. This divertor structure will be located on the 45/spl deg/ angled corner on the innermost ceiling of the DIII-D vessel, known as the private flux region of the plasma configuration. With hardware supports factored into the pumping speed calculation, the private flux cryopump is expected to have a pumping speed of 15,000 l/s.

The radiative divertor program of DIII-D is in its final phase with the installation of the cryopump and baffle structure (Phase 1B Divertor) in the upper inner radius of the DIII-D vacuum vessel at the end of this calendar year. This divertor, in conjunction with the Advanced Divertor and the Phase 1A Divertor, located in the lower and upper outer radius of the DIII-D vacuum vessel respectively, provides pumping for density control of the plasma while minimizing the effects on the core confinement. Each divertor consists of a cryo/helium cooling ring and a shielded protective structure. The cryo/helium-cooled pumps of all three diverters exhaust helium from the plasma. The protective shielded structure or baffle structure, in the case of the diverters located at the top of the vacuum vessel, provides baffling of neutral charged particles and minimize the flow of impurities back into the core of the plasma