A wall model developed for the analysis of Tore Supra wall loading exp
eriments has been applied to an experiment on DIII-D which demonstrate
d a substantial capacity for retention of deuterium gas in an all-grap
hite environment, and which showed the efficacy of the pumped divertor
to deplete a gas-loaded wall. The Tore Supra model has been extended
and applied to evaluate the particle exchange mechanisms between the c
ore, divertor, and wall. Data-constrained plasma modeling is done for
the discharges of the load/unload sequence. The poloidal distribution
of the charge exchange flux profile to the divertor and outer wall is
determined from the Eirene neutral transport code, to estimate the eff
ective working areas for particle exchange and saturation. The deposit
ion and saturation of the hydrogenic efflux in the aC:H layer and grap
hite is modeled with the 1-D WDIFFUSE code, applied to the regions fue
lled by charge exchange flux to predict the instantaneous local wall r
ecycling coefficient. A mechanism is proposed to explain the previous
paradoxical result that exhaust quickly (similar to 3 s) balances the
only particle input, due to beam fueling, whereas a long term net wall
depletion is observed over similar to 10 discharges. The saturation a
nd depletion of wall layers fuelled by energetic charge exchange parti
cles provides such a mechanism.