Turbulent three-dimensional air flow and trace gas distribution in an inhalation test chamber

Steady incompressible turbulent air flow and transient carbon monoxide tran sport in an empty Rochester-style human exposure chamber have been numerica lly simulated and compared with experimental data sets. The system consiste d of an inlet duct with a continuous carbon monoxide point source, 45- and 90-degree bends, a round diffuser, a round-to-square transition, a rectangu lar diffuser, the test chamber, a perforated floor, and again transition pi eces from the chamber to an outlet duct. Such a configuration induced highl y nonuniform vortical flow patterns in the chamber test area where a pollut ant concentration is required to be constant at breathing level for safe an d accurate inhalation studies. Presented are validated momentum and mass tr ansfer results for this large-scale system with the main goals of determini ng the development of tracer gas (CO) distributions in the chamber and anal yzing the contributions to CO-mixing. Numerical simulations were conducted employing a k-epsilon model and the latest available RNG k-epsilon model fo r air and CO-mixing. Both models predict similar velocity fields and are in good agreement with measured steady and transient CO-concentrations. It wa s found that secondary flows in the inlet section and strong vortical flow in the chamber with perforated flooring contributed to effective mixing of the trace gas at breathing levels. Specifically, in the height range of 1.4 m<h<2.0 m above the chamber floor predicted CO-concentrations rapidly reac hed a near constant value which agrees well with experimental results. This work can be extended to analyze trace gas mixing as well as aerosol disper sion in occupied test chambers with or without flow redirection devices ins talled in the upstream section. A complementary application is particle tra nsport and deposition in clean rooms of the electronic, pharmaceutical, and health care industries. [S0098-2202(00)01702-8].