Evaluating uncertainties in computational fluid dynamic simulations of human exposure to paint-spray aerosols.

Preliminary numerical simulations of human exposure to paint-spray aerosols demonstrate the ability of computational fluid dynamic (CFD) software to discriminate between two different orientations of spraying a flat plate in a cross-flow ventilated spray booth. To conduct exposure-scenario simulations using CPD, a conceptual model of reality must be created that is compatible with the computer code. If this conceptual model is not a sufficient representation of reality with regard to the desired outcome, then no matter how accurate the simulation, the results will be of limited value. Conversely, good conceptual models will be inadequate, if there is insufficient numerical resolution. A balance between these two components is essential to achieve meaningful results given finite resources. This work examines some of the uncertainties involved in the conduct of such simulations with an eye toward developing efficient modeling approaches for optimizing control decisions based on exposure reduction.