An analysis of low ambient pressure on fire risks associated with rigid polyurethane building façade insulation under various width constructions. Issue 7 (August 2019)

• Record Type: Journal Article
• Authors: Ma, Xin; Tu, Ran; Zeng, Yi; Fang, Tingyong
• Is Part Of: Indoor + built environment. Volume 28:Issue 7(2019)
• Page Start: 905
• Page End: 913
• Publication Date: 2019-08
• Subjects: Rigid polyurethane foam -- Low air pressure -- Building façade -- Pool fire -- Heat transfer; Indoor air pollution -- Periodicals; Sick building syndrome -- Periodicals; Buildings -- Environmental engineering -- Periodicals; Environmental health -- Periodicals; 613.5
• DOI: 10.1177/1420326X18798003
• Languages: English
• ISSNs: 1420-326X

Abstract

The fire risk associated with the use of rigid polyurethane foam (PUR) as thermal insulation material in building façades was investigated. To assess the effect of ambient pressure, the combustion of PUR boards was examined at both Lhasa (with an ambient pressure of 66.5 kPa) and Hefei (99.8 kPa) in China, utilizing pool fire theory. The effects of pressure and sample width on important combustion parameters were analysed including flame spreading rate, flame height and mass loss rate. Effects of pressure on the combustion mechanism of the PUR were determined based on resulting data. The mass loss rate at lower pressure was found to be slower and unique, and relatively stable melting-flowing pool fire flame front was observed. In this respect, the combustion behaviour of PUR was different to that of other solid fuels commonly applied to building façades. Data obtained at the lower pressure using various widths showed that the average flame spreading rate was proportional to the pressure raised to an exponent ranging from 0.47 to 0.73. This result is similar to trends normally observed when the burning rate is largely driven by radiative heat transfer. The flame height was only minimally affected by pressure but increased with increases in the sample width. Finally, a linear relationship was determined between the pressure index values obtained from plots of flame height versus pressure and the sample width.