<p>This paper describes a project which sets out to define a practical predictive test method for heat- induced explosive spalling of concrete. The method is based on a novel testing apparatus known as the Heat-Transfer Rate Inducing System (H-TRIS), previously developed at The University of Edinburgh. A series of 75 samples were tested using a large-scale standard fire testing furnace at CERIB (France) and using H-TRIS in Edinburgh. Within the scope of this project, a thorough examination of the thermal exposure was carried out in order to ensure repeatability of the thermal exposure imposed during testing. The H-TRIS method and apparatus was successful in accurately replicating the thermal exposures experienced by samples in the fire testing furnace when testing to both the ISO 834 and modified hydrocarbon (HCM) standard fire curves. The testing has also provided insights into the influence of sample size, polypropylene fibre content, and moisture content on the propensity for heat-induced concrete spalling.</p>

This paper evaluates the effect that a permeable mould, such as would be used to create fabric-formed concrete, may have on the heat-induced explosive spalling performance of cast concrete, using a novel experimental fire testing method and supported by scanning electron microscopy. Recent research suggests that a concrete cast using fabric formwork will gain durability enhancements at the cast surface that may negatively affect pore-pressure expulsion during severe heating. Six concrete samples were cast using high strength concrete including silica fume and tested using the University of Edinburgh's Heat-Transfer Rate Inducing System (H-TRIS), receiving thermal loading on one surface. Three samples were cast in permeable moulds, formed using a Huesker HaTe PES 70/70 single layer woven geotextile with a characteristic opening size (O 90) of 0.1x10-3 m. Three samples were cast in conventional impermeable timber moulds. The tests showed no conclusive evidence of differences in thermal profile or differential thermal deflections between the two casting methods; no occurrences of heat-induced explosive spalling were observed for either casting method. However, scanning electron microscopy undertaken on additional samples showed that the test face of samples cast in permeable moulds were over four times less porous compared to their impermeably cast equivalents. This could increase the risk of spalling of samples, particularly in cases where pore-pressure spalling dominates the material response. However, additional fire testing using H-TRIS is needed under a range of heating and loading conditions, before definitive conclusions on the spalling propensity of fabric-formed concrete can be made.

Summary A series of six exploratory experiments are presented investigating the behaviour of laminated glass when exposed to radiative heating of 25, 50 and 75 kW/m2. The laminated glass type used is typical of that which might be used in balustrades. Three experiments examined the potential for achieving flaming ignition of the inter‐layer should the glass break. At a radiant heat flux of 25 kW/m2, although the glass did not shatter, piloted ignition of the volatile gases released from the specimen edge was possible. At radiant heat fluxes of 50 and 75 kW/m2, the exposed glass layer shattered after 5 minutes 7 seconds and 2 minutes 24 seconds, respectively, and subsequently there was auto‐ignition of the volatile gases from the inter‐layer. Eventually flames were observed over the whole face of the specimen. Using an average measured mass loss and representative heat of combustion for the inter‐layer result in calculated heat release rates per unit area of 65 to 69 kW/m2. A further three experiments were carried out to examine the potential for the glass layers to separate when heated. Only one layer of glass was supported, and unsupported layer nearest the radiant heat source was observed to slip away from the other layer, exposing the inter‐layer, resulting in ignition and flaming.

Dear Editor,We have read the recent paper on the 'Qualitative assessment of fire hazard posed by laminated glass balcony balustrades on fire spread' published in Fire Technology by Faudzi et al. [1] with interest. We are also undertaking ongoing work on the topic of the fire performance of laminated glass [2] and part of that study has also been investigating some of the recent incidents in which fires have occurred on balconies with glass balustrades. We would like to offer an additional perspective on three of the incidents discussed by Faudzi et al., namely the West Hampstead square fire (incident #3) in the Orwell Building, London, UK in 2018, the Lacrosse Docklands fire (incident #1), Melbourne, Australia, in 2014, and The Lighthouse, Manchester in 2017 (incident #2). Our findings are based on the design of the existing building prior to the incident, information from the incident itself and any post-incident investigations.

No abstract