International Journal on Engineering Performance-Based Fire Codes, Volume 4, Number 3, p.60-67, 2002 60 NECESSITY OF TESTING FURNITURE MATERIALS WITH A CONE CALORIMETER H.W. Au Yeung and W.K. Chow Department of Building Services Engineering, The Hong Kong Polytechnic University, Hong Kong, China (Received 27 March 2002; Accepted 15 April 2002) ABSTRACT It is proposed and demonstrated in this paper that fire-safe furniture must be developed with appropriate tests. A cone calorimeter should be the minimum requirement, though some full-scale burning tests are necessary. The component materials of a typical furniture foam sofa commonly used in Hong Kong were selected for assessing the fire behaviour with a cone calorimeter. A high radiative heat flux of 50 kWm -2 was used to consume all materials in the tests. The sustained ignition time, peak heat release rate, average heat release rates in 60 s and 180 s after ignition, total heat release rate, mass loss percentage, total smoke release, carbon monoxide and carbon dioxide yields of the component materials in the arrangement, were measured. These data can be applied for assessing the possibility to flashover and making recommendations on what should be considered in selecting furniture. Further, the importance of protecting the foam by fabric and cotton covering; and treating the foam with fire retardant should be considered. Results are useful for the Authority to set up regulations on fire-safe furniture. 1. INTRODUCTION Big accidental fires due to burning furniture or cushion foam had raised public concern on using fire-safe furniture in Hong Kong since 1997 [e.g. 1]. This is important for public entertainment places like those karaokes with partition walls made of timber product [2,3]. The heat release rates of burning furniture and their contributions to a compartment fire at the preflashover stage should be clearly understood [e.g. 4,5]. As reviewed recently on furniture fires [e.g. 6], heat released from a burning foam might be strong enough to ignite adjacent items which are not easy to ignite accidentally by electric sparks or cigarettes. In fact, the government has set up some regulations to specify ‘fire-safe’ furniture [e.g. 7,8] by following overseas practices [9,10]. But without knowing the fire behaviour of common furniture and their constituting materials, it is difficult to convince the citizens that such regulations can really give fire- safe furniture. Now, regulations might not be passed by the Legislative Council so easily as before, especially there are cost implications. The recent Karaoke Establishments Bill [11] is an obvious example and there are numerous concerns [3]. Full-scale burning tests must be carried out to support the argument. An experimental rig, known as the PolyU/HEU Assembly Calorimeter [12], is now developed in China as a collaboration project between The Hong Kong Polytechnic University (PolyU) and Harbin Engineering University (HEU). There have been extensive studies on upholstered furniture over the past years [e.g. 13-19]. One of the recent biggest projects is perhaps the project on Combustion Behaviour of Upholstered Furniture (CBUF) in Europe [17-19]. Upholstered furniture is quite complicated with several fuel elements including cover fabric, seat cushion material and the padding. How these would burn depends [20] on the material composition, thickness and density. As reported in the literature, results measured from a cone calorimeter [e.g. 4,13,21,22] would be useful in understanding the fire behaviour, both heat and smoke aspect, of the materials including sandwich panel [23]. Results from cone calorimeter should be taken as a ‘yardstick’ for assessing whether the furniture is safe in a fire, in complementary with full-scale burning tests. Therefore, local furniture materials should be tested at least, by a cone calorimeter [e.g. 4,13,22,24]. A ranking system [21,25] should be worked out for ‘grading’ the materials concerned. Reporting how to get cone calorimeter results becomes the objective of this paper. Materials for a typical local furniture foam sofa arrangement were selected to demonstrate how a cone calorimeter can be used for assessing the fire behaviour. The heat release rate Q(t) (in kWm -2 ), smoke release rate S R (in s -1 ), carbon monoxide (CO) and carbon dioxide (CO 2 ) generated under an incident heat flux of 50 kWm -2 of the component materials and the furniture arrangement were measured. Using such a high heat flux could ensure that all the materials would be burnt out. Eleven parameters [e.g. 4,13,21,22] commonly deduced from cone calorimeter results were calculated. Finally, the smoke aspect [22] was also analysed.