No abstract

The result of 14 experiments in which a single glazing assembly was exposed to fires of increasing severity in the centre of an enclosure are presented. The glazing assembly was exposed to a two zone fire enclosure environment with consistency and repeatability within each set of three experiments conducted at each fire severity level evident from the enclosure gas temperature profiles. The results presented include heat release rates, enclosure and local gas temperature, heat flux distribution, glass surface temperatures, shaded glass temperatures, thermally induced strains, crack bifurcation patterns and loss of integrity of the glazing assembly. Limits regarding the maximum incident heat fluxes required to cause catastrophic failure of the glazing assembly are suggested.

The results of 19 experiments in which a single pane glazing assembly was exposed to enclosure corner 5res of increasing severity are presented. The glazing assembly was exposed to a two zone 5re enclosure environment, with consistency and repeatability within each set of three experiments conducted at each 5re severity level, evident from the enclosure gas temperature pro5les. The results presented include heat release rates, enclosure and local gas temperatures, heat 6ux distributions, glass surface temperatures, shaded glass temperatures thermally induced stains, crack bifurcation patterns and loss of integrity of the glazing assembly. The results and observations obtained are presented in the form of a toolkit for the estimation of the occurrence of cracking and failure of single glazing in enclosure 5res. Figure 5. Enclosure local gas temperature distribution across panes, Ure pan size 500 mm;500 mm.Figure 6. Enclosure local gas temperature distribution across panes, Ure pan size 900 mm;900 mm.126 T. J. SHIELDS E¹ A¸.Figure 11. Total heat Vux incident on window area, pan Ure size 0.5 m;0.5 m.Figure 12. Total heat Vux incident on window area, pan Ure size 0.6 m;0.6 m.Figure 15. Total heat Vux incident on window area, pan Ure size 0.9 m;0.9 m.Figure 16. Incident total Vux proUle on exposed glass surface by time to Urst crack with location, time and subsequent crack and Unal crack pattern, 0.5 m;0.5 m pan Ure size. GLAZING ASSEMBLY AND CORNER FIRES 131 Figure 17. Incident total Vux proUle on exposed glass surface by time to Urst crack with location, time and subsequent crack and Unal crack pattern, 0.6 m;0.6 m pan Ure size.Figure 18. Incident total Vux proUle on exposed glass surface by time to Urst crack with location, time and subsequent crack and Unal crack pattern, 0.7 m;0.7 m pan Ure size.Figure 19. Incident total Vux proUle on exposed glass surface by time to Urst crack with location, time and subsequent crack and Unal crack pattern, 0.8 m;0.8 m pan Ure size. GLAZING ASSEMBLY AND CORNER FIRES 137

The effect of melting behaviour on upward flame spread of thermoplastic materials when subjected to small ignition sources and considered to suffer no external flux was studied using large‐scale tests. For moderate fire conditions the cone calorimeter was utilized, with the sample set in a vertical orientation to study the melting behaviour of the specimens. Under these conditions the results indicate that the melting behaviour significantly affects upward flame spread behaviour. A pool of the melt which formed at the base of the vertically oriented sample tested creates a pool fire which then controls the fire growth and flame spread. In contrast, it was found that some thermoplastic materials which have higher glass transition temperatures or undergo a special pyrolysis process such as depolymerization, intumescing or charring do not experience significant melting behaviour when exposed to the same thermal insult. As a result, they behave very differently in terms of upward flame spread. The study also indicates that the melting behaviour of thermoplastic materials is an important characteristic in fires which should be taken into account in the development of modelling, in particular for upward flame spread models. © 1997 by John Wiley & Sons, Ltd.

Three timber-based materials were investigated by subjecting them to a constant uniform heat flux in the range 2k70 kW m-z using a Cone Calorimeter and the I S 0 Ignitability Apparatus. The specimens were examined in the vertical and horizontal orientations in the Cone Calorimeter using gas flame pilot, spark pilot and spontaneous methods of ignition. They were also studied using the IS0 Ignitability Apparatus in the normal and inverted horizontal orientations using gas flame pilot and spontaneous methods of ignition. The results obtained are compared by specimen, orientation and mode of ignition.