$site = "publications.iafss.org"; $fullsite = "publications.iafss.org"; $basePath = "/home2/firesag5/private/data/"; ?>
The problem considered here is â€œWhen does a window exposed to fire become a vent?â€ Both compartment fires and urban/wildland interface fires provide applications for this work, which was chosen for the 2002 Howard W. Emmons Lecture because he introduced that topic to fire research. Glass breaks when exposed to fire because the temperature difference between the exposed pane and its shaded perimeter produces a strain at the edge due to the excess thermal expansion of the central heated pane. When the stress induced by that strain exceeds the glass breaking stress, a brittle fracture crack is initiated at the edge which travels, usually on multiple paths, through the pane at ~ 1.5 km/s. Practical examples of real fires where glass breaking played a critical role are cited. The literature is reviewed. Multipane windows are discussed. Techniques are presented for calculating the glass breaking time as a boundary condition in field and zone models for fire safe designs. The primary difficulties in application are identifying the paneâ€™s proper glass properties and the fireâ€™s radiative and convective heating coefficients. Suggested properties at 50Â°C for the soda lime float glass common in windows include: density ? = 2500 kg/m3, specific heat capacity cp = 820 J/kgK, thermal conductivity k = 0.95 W/mK, Youngâ€™s Modulus E = 72 GPa, thermal coefficient of linear expansion ? = 9x10â€“6 K-1 and breaking stress ?b = 10 to 50 MPa. The breaking stress has a wide range because it is a function of the glass edge condition and history. Comparisons are made with experimental results from the Building Research Institute in Japan, the University of Maryland in the USA and the University of Ulsterâ€™s FireSERT Centre in the UK. The glass fall-out problem is formulated.