A computational fluid dynamics analysis of transient flow through a generic Chemical Biological Radiological and Nuclear respirator canister

•Pressure drops under transient and steady state flows are the same.•The residence time distribution differs between steady state and transient flow.•Lower minimum residence times will be seen under transient conditions.•Similar flow profiles will be seen at similar points in the inhalation.•The ratio of residence times between steady and transient flows remains constant. Unsteady Reynolds Averaged Navier–Stokes (RANS) simulations of a generic Chemical Biological Radiological and Nuclear (CBRN) respirator canister have been carried out in order to assess the performance of the canister under transient flow in a realistic breathing profile. The performance has been assessed with respect to pressure drop across the canister and residence time distribution within the carbon filter. It has been demonstrated that the pressure drop across the canister provides a good agreement with that predicted from a steady-state simulation for the corresponding equivalent continuous flow conditions. However, the mean residence time at the filter outlet has been shown to exhibit a clear “history” effect under transient flow conditions, which could not be predicted from the equivalent steady-state simulation. Three representative breathing rates (for adult males under light, moderate and heavy workloads) have been analysed. All three breathing rates exhibit a similar residence time distribution with respect to the fractional progress through the inhalation. For any chosen breathing profile and target residence time, a similar mean air age front could be seen at the moment of breakthrough. The minimum residence time of air leaving the carbon bed for each breathing profile has been found to be considerably smaller than that for the equivalent steady-state continuous flow. The minimum residence time for transient and steady continuous flow showed a consistent ratio of approximately 1:3 for all breathing patterns studied for the canister analysed here, suggesting there is some scope for predicting breakthrough times under real breathing patterns from steady-state data.