The first part of this study focuses on the &ect of cone calorimeter test variables on polyurethane flexible foam properties such as ignitability, heat release rate, effective heat of combustion and mass loss. Three of the main commercial foam types were used, i.e. conventional slabstock foams, high-resilience slabstock foams and all-MDI (methylene diphenyldiisocyanate) moulded foams. A decrease in heat flux (down to 40%) with increasing distance from the conical heater was measured. As a consequence, results were found to depend to a large extent on the thickness and the melting behaviour of the foam samples. To achieve a sufficiently constant and uniform heat flux exposure, sample thickness had to be limited to 25 mm. In addition, repeatability was found to be good under various conditions, with percentage standard deviations for effective heat of combustion, peak rate of heat release and mass loss below 10%. Levels of radiant flux above 25 kW m-' were found to be very severe to test flexible polyurethane foams. Under such conditions, foams that show large differences in combustion performance in small-scale flammability tests performed almost identically in the cone calorimeter. In the second part of this study the effects of foam variables, such as foam type, density and melamine content, are defined. These effects were clearly pronounced at radiant flux levels of 15 -25 kW m-'. Density was found to be the key variable in controlling ignition resistance. In addition, high-resilience slabstock foams and all-MDI moulded foams performed better than conventional slabstock foams of the same density. Melamine addition resulted in a delay of ignition for all three foam types and an incomplete combustion, decreased heat release and effective heat of combustion in HR-slabstock and all-MDI moulded foams However, melamine is not effective as a heat sink in conventional slabstock foams. The different performance of the foam types under study can be explained by a different melting behaviour.

The relative contributions of fabric and foam on important fire hazard parameters as measured in the cone calorimeter were determined for various mock-up combinations. Three commercial types of high-resilience PU flexible slabstock foams, representing a wide range of combustion performance, were combined with three different fabrics: polyester, polypropylene and combustion-modified cotton. Total heat release of the combinations correspond well with the value calculated from the individual components and is determined by the density and effective heat of combustion of the components, and the foam/fabric weight ratio. The foams show a smaller variation in effective heat of combustion than the fabrics. Covering of PU flexible foams with a fabric results in general in a delay of ignition and peak rate of heat release. Performance of the mock-ups is mainly determined by the fabric but is also influenced by the quality of the foam.

No abstract

A Horizontal Burn Test (HBT) has been successfully developed to evaluate the ignition resistance, flame propagation and heat release of a flexi ble polyurethane foam. The test was used to evaluate the combustion charac teristics of foams of different type and density, unfilled or filled with different types of combustion modification additives. The best differentiation between foam samples was found on the basis of relative heat release. On the basis of this measurement and the oxygen ignition index, a correlation between com bustion performance in the Horizontal Burn Test and in the BS 5852 Source 5 Test was established. In this way, the excellent BS 5852 Source 5 performance of melamine-filled high resilience slabstock foams was predicted on the basis of Horizontal Burn Test results. Furthermore, all-MDI foam formulations filled with melamine or other selected solid fillers were defined as alternatives to pass the BS 5852 Source 5 Test.