Application of high-fidelity modelling approach to predict smoke and fire propagation in a nuclear fire scenario

Abstract

Fire and smoke propagation from an electrical cabinet fire in a multi-compartment geometry was simulated using computational fluid dynamics (CFD) in this study. The URANS turbulence modeling approach, and variants of two-equation models available within the CFD code STAR-CCM + were tested for non-vented and vented conditions. Specifically, the ability of the turbulence models to predict the smoke and hot gas spread was analysed qualitatively. The ceiling jet and plume spilling phenomenon reported in literature was predicted in the progression of smoke propagation. It was observed that the ventilation created better mixing, enhanced the spreading of the stratified smoke layer and resulted in delayed progression of smoke propagation compared to the non-vented cases. A qualitative assessment of CFD predictions for temperature distribution was also undertaken using a slightly different geometry. It was found that the qualitative predictions obtained by the application of two-equation turbulence models for the temperature and smoke distribution were realistic and in-line with earlier studies using a similar building configuration. The findings from this study are expected to be beneficial in the assessment of existing two-equation turbulence models to predict smoke propagation in multi-compartment fires.