Large-eddy simulation of plume dispersion under various thermally stratified boundary layers

Abstract

Abstract. Contaminant gas dispersion in atmospheric boundary layer is of great concern to public health. For the accurate prediction of the dispersion problem, the present study numerically investigates the behavior of plume dispersion by taking into account the atmospheric stability which is classified into three types; neutral, stable, and convective boundary layers. We first proposed an efficient method to generate spatially-developing, thermally-stratified boundary layers and examined the usefulness of our approach by comparing to wind tunnel experimental data for various thermal boundary layers. The spreads of plume in the spanwise direction are quantitatively underestimated especially at large downwind distances from the point source, owing to the underestimation of turbulence intensities for the spanwise component; however, the dependence of the spanwise spreads to atmospheric stability is well represented in a qualitative sense. It was shown that the large-eddy simulation (LES) model provides physically reasonable results.

On the highly unsteady dynamics of multiple thermal buoyant jets in cross flows

“…As an example, though the selection of the proper value to be used for the turbulent Prandtl number in order to capture precisely experimental results is a still debated question, in line with the assumptions of Bastiaans et al, 94 and Li and Ma, 71 for the simulation of thermal plumes, and those by Murataa and Mochizuki 73 for the companion problem relating to turbulent heat transfer in rectangular ducts with transverse rib turbulators we fix Pr T to 0.5. Moreover, following the analyses by Nakayama and co-workers (Nakayama and Nagai 102 ; Nakayama et al, [103][104][105][106][107] ), who performed LESs of turbulent flows in the atmospheric boundary layer with thermal plumes for a variety of conditions (0.12Ri0.45, Re=O(10 5 ) obtaining good agreement with dedicated experiments, in the present paper we assume Cs=0.1 (notably, Li and Ma 71 used the same value of Cs for pure thermogravitational convection, i.e. Re=0 while Ciofalo and Collins 72 , and Lohász et al, 75 yet assumed Cs=0.1 for rib-roughened channels in the opposite condition for which there was no thermal buoyancy, i.e.…”

Numerical Models of Wind Effects on Temperature Loaded Object

“…The described plot corresponds to the flow with Reynolds number Re = 6•10 3 -2•10 4 . It concerns of flowing with a transition from the low turbulence at the beginning to the fully developed turbulence after the flown around obstacle [14,15,16]. The flown around temperature burdened object contributes to a change of momentum and turbulent flow properties [17] and there is a significant change in the momentum of the flow field in the transition area of the turbulent flow at low Re numbers.…”

Numerical modelling of flow around thermally loaded object

“…It is a flowing with the transition from the low turbulence at the beginning to the fully developed turbulence behind the obstacle [6][7][8]. A flow around and temperature loaded object contributes to a change of momentum and turbulent flow characteristics [9] that leads to a significant change in the momentum in the transition area of turbulent flow at low Re numbers.…”