Response Surface Methodology for Analysis of an Air Curtain Used as Emergency Ventilation System in a Tunnel Fire

Abstract

This paper presents results obtained by a multi-variable method combined with CFD to study the performance of an air curtain in a fire scenario inside a tunnel. An air curtain is a plane stream of air blown across an opening to create a singularity that hampers the free air movement through this opening. An air curtain is used to produce a barrier effect while permitting traffic of people, vehicles, materials or objects between the areas the air curtain separates. They are widely used at building entrances to ensure a constant inside-outside temperature difference. Air curtains are also often mounted in the front of refrigerated food counters and open-shelves for customers to easily see products without having to open a door. In many circumstances, an air curtain can be seen as an impinging plane jet. The main flow is then composed of three or four basic regions depending on the jet height H to width e ratio: a free jet (just downstream of the discharge nozzle exit), a development region (absent if H/e is small), an impingement region including a stagnation zone, and a wall-jet region. The air curtain investigated in this study is part of an emergency ventilation system designed to stop smoke propagation during fire in a tunnel. The whole system is made of 2 air curtains forming a confinement cell within which smoke should be retained. A traditional CFD study would take a lot of time and computational resources. Therefore, CFD was combined with Response Surface Methodology (RSM), a useful tool to produce results that take into consideration many variables and ranges in an optimal way. The virtual set-up investigated in this study describes a 2D tunnel of length 100 m with one confinement cell and a fire source designed from real fire simulations for testing ventilation systems. The Reynolds number based on the flow velocity and width at the discharge duct of the air curtain was varied from 4·105 to 1·106. The height-to-width (H/e) ratio was varied from 2 to 30. The efficiency of the air curtain was assessed through the temperature and smoke concentration difference between the tunnel exit and the confined section. The RSM method proved to be an efficient way to study multiple ventilation conditions for a given fire scenario from a reduced number of traditional CFD simulations. Air curtains are shown to be effective barriers able to confine heat, reduce the temperature in the tunnel outside of the confinement cell, and thus facilitate evacuation. Smoke leakage occurs through the curtain due to two main reasons: the first one is based on the development condition of the impinging jet, and the second is related to mass conservation within the confinement section. The results show that the best air barriers are those with a high Reynolds number and a small H/e ratio.