Dielectric window development for the ITER ICRF vacuum transmission line

The choice of dielectric material for the (double) dielectric window of the ITER ICRF Transmission Line is optimized with respect to nuclear, mechanical, and thermal properties in due consideration of material availability, fabrication issues and response to cyclic loads. The window is optimized to minimize electric field and thermal heating in the dielectric material with specific cooling conditions and with constraints posed by manufacturing. Electric field, temperature distribution, and thermal stresses are evaluated for beryllia and alumina dielectrics for unirradiated and irradiated material data using IVOFEM/IVOHEAT and ANSYS finite element code packages. The analysis is made for two annular ceramic septa joined to inner (Ø 110mm) and outer (Ø 196mm) coaxial water cooled conductors having maximum operating 50kV peak RF voltage. Neutron radiation at the window is evaluated from MCNP-4 calculations with account for the neutron streaming through the structure of the ICRF Array/Shield/VTL assembly. Placing beryllia windows at the vacuum vessel feedthrough or outside is acceptable, while alumina windows have to be positioned outside the bio-shield unless additional shielding (30-60 cm thick) behind the ICRF array exists. For unfavourable radiation conditions (5x10-2 dpa), alumina is found to be heated excessively near 1000 oC with unacceptable stresses. For beryllia (10-3 dpa) or for unirradiated alumina (97.5 % purity), the temperature is found to stay between the cooling temperature and 275 oC with maximum principal stress less than 125 MPa, provided niobium, titanium or materials with similar thermal expansion coefficients are used as a conductor. Stresses with steel, copper, or aluminium conductor are unacceptable. Beryllia (BeO) is chosen as a candidate for the window ceramic because of its better radiation resistance, weaker temperature growth, and because of its better heat conductivity helping the manufacturing process. The dependence of the stress distribution on the material data is presented including the additional impact from the residual thermal stresses. The characteristics for the ceramic/metal joints are estimated for candidate conductor and ceramic materials. In particular, the benefits from compatibility of thermal expansion in ceramics and metal are investigated. Vacuum brazing using active filler materials obviously provides sufficiently good conditions for heat conduction, and appears to be tight enough. The vacuum tightness, the strength against the relative movements of the outer and inner conductors, thermal expansion, and pressure shocks are investigated with a full-scale preprototype experiment where two alumina dielectrics (in an X-shaped geometry) are brazed to a titanium coaxial housing and the tests shall be performed for the system.