Abstract
Stochastic geometries in Monte-Carlo simulations enable to simulate complex configurations such as the repartition of possible radioactive dust in a glove box. This paper compares several dust models that represent more or less explicitly the heterogeneous repartition of dust speckles in space. Indeed, assessing the contribution of dust to the dose received by the hands of an operator is a key problem for glove boxes. Results show that homogeneous models generally overestimate the dose, which is correct for radioprotection studies, but that dust aggregates produce doses that are much smaller than those obtained by homogenising dust. These heterogeneous models can also help estimating deposited dust quantities from dose measurements inside the glove box, whereas an homogenous model would grossly underestimate dust quantity.