Geometrical and environmental effect on fire behaviour of polyurethane foam slabs

Abstract

This study presents the results of a set of large-scale fire tests performed on horizontal polyurethane foam slabs. Slab thickness, ignition location, and environment (free burn or corner wall) were varied over 23 individual tests. In a continuation of a previous study by the authors, four different slab thicknesses, five different ignition locations, and three environmental modification combinations were tested to examine the potential effects these parameters may have on the test outcomes. Wall configurations were composed of either gypsum plaster board or particle board wall linings and enclosed two sides of the foam slabs. The results obtained were heat release rate, peak HRR, total heat release, and effective heat of combustion, smoke production rate, total smoke production, specific extinction area and soot yield along with flame spread rate calculations obtained using a specifically designed sample tray. These were then analysed to quantify the effects of the test input parameter changes. Results highlight the influence of test input parameter changes on almost all the chosen fire metrics analysed. Slab thickness showed relatively straight forward linear changes to global metrics such as total heat release and total smoke production, while effective heat of combustion showed no change. The smoke production parameters of specific extinction area and soot yield showed minor trends towards higher values with increased slab thickness. Flame spread rates also showed an increased velocity with thicker slabs. Changes to the environment surrounding the foam slabs, from a free-burning scenario to a corner wall configuration that obstructed two sides of the foam slab generally resulted in faster growth of HRR and higher peak values when compared to the free-burn tests. This outcome was most obvious for the wall tests using combustible materials, where growth rates were comparative to an ultra-fast t^2 fire, and peak HRR value rose to over triple that of the free-burn test results. Flame spread rates were highly impacted by the addition of a corner wall, spread rates increased, and the overall pattern of spread rates across the face of the slabs changed entirely. The influence on smoke production values was minor, and for other smoke metrics, the scenario changes showed little influence. However, the values obtained for soot yields were approximately an order of magnitude lower than those given in standard reference books, such as the SFPE handbook for flexible polyurethane foam. This was an interesting additional finding as soot yield values are often directly used in CFD models to simulate smoke spread and determine life safety criteria in fire safety engineering (FSE).