Identification of airborne pollutant sources within a slot ventilated porous enclosure depending on backward model and downwind scheme

Abstract

Identification of pollutant source, depending on the inverse computational fluid dynamics (CFD) methodology, within a slot-ventilated porous enclosure has been proposed in this paper. The whole fluid flow system is assumed to be two-dimensional, laminar and porous-saturated. As the inverse CFD modeling was ill-posed, a downwind scheme is employed to discretize the convection term to improve the accuracy and the corresponding time step range is derived to achieve numerical stability. Afterward, the downwind scheme simplifies the discretization process and its direct inversion algorithm. The effects of the Reynolds number (), the Darcy number (), the Schmidt number () and locations of pollutant source on the backward time CFD simulations are investigated, respectively. Results show that the temporal size/step is a critical factor of computational stability. In addition, the thickness of velocity or concentration boundary layer directly impacts the sensitivity of the inverse CFD simulation. The present research could aid in the safety of gas-supplying pipe network and relevant pollutant leakage accidents. Highlights Identification of prior-unknown sources was proposed regarding gas-supplying network. Original transport equations were maintained to enhance the accuracy and generality. The downwind convection scheme was implemented to enhance backward CFD procedures. Diverse influencing factors were numerically tested to recover pollutant sources. The proposed new algorithm was validated thoroughly by experiments and benchmark solutions.