Assessment of Buoyancy-Corrected Turbulence Models for Thermal Plumes

Abstract A computational investigation of thermal plume is important because such flows are encountered in various industrial applications. The investigation is also important because thermal plume can be considered as a test case for modeling of fire which can help designers and safety engineers to develop preventive measures and fire safety systems. The present study primarily focuses on the investigation of thermal buoyant plume in the self-similar region. In the present study, an assessment of three buoyancy-corrected turbulence models, namely k-ω, the standard k-ε model and the RNG k-ε model, has been conducted for a thermal buoyant plume. Modifications to the turbulence models have been made to account for the effect of buoyancy on the production and dissipation of turbulent kinetic energy and these modifications are based on the simple gradient-diffusion hypothesis and generalized gradient-diffusion hypothesis. The model based on the simple gradient diffusion hypothesis is shown to under-predict contribution to the generation of turbulence kinetic energy due to buoyancy. A comparison with the experimental measurements reported in the literature shows that the generalized gradient-diffusion hypothesis along with both turbulence models correctly predicts the mean flow field, temperature field and spread rates. The results of the present simulations using the RNG k-ε model with the generalized gradient-diffusion hypothesis are shown to be in good agreement with the corresponding experimental results reported in the literature for thermal plumes.