ASSESSMENT OF EROSION AND SURFACE TRITIUM INVENTORY ISSUES FOR THE ITER DIVERTOR

We analyzed sputtering erosion and tritium codeposition for the ITER v ertical target divertor design using erosion and plasma codes (WBC/RED EP/DEGAS+) coupled to available materials data. Computations were made for a beryllium, carbon, and tungsten coated divertor plate, and for three edge plasma regimes. New data on tritium codeposition in berylli um was obtained with the tritium plasma experiment (TPE) facility. Thi s shows codeposited H/Be ratios of the order of 10% for surface temper atures less than or equal to 300 degrees C, beryllium thereby being si milar to carbon in this respect. Hydrocarbon transport calculations sh ow significant loss (10-20%) of chemically sputtered carbon for detach ed conditions (T-e approximate to 1 eV at the divertor), compared to e ssentially no loss (100% redeposition) for higher temperature plasmas. Calculations also show a high, non-thermal, D-T molecular flux for de tached conditions. Tritium codeposition rates for carbon are very high for detached conditions (similar to 20 g T/1000 s discharge), due to buildup of chemically sputtered carbon on relatively cold surfaces of the divertor cassette. Codeposition is lower (similar to 10X) for high er edge temperatures (similar to 8-30 eV) and is primarily due to dive rtor plate buildup of physically sputtered carbon. Peak net erosion ra tes for carbon are of the order of 30 cm/burn yr. Erosion and codeposi tion rates for beryllium are much lower than for carbon at detached co nditions, but are similar to carbon fdr the higher temperatures. Both erosion and tritium codeposition are essentially nil for tungsten for the regimes studied.