In order to devise rational tests and performance requirements for combustible linings in various positions in buildings it is necessary to have some quantitative design criteria either from direct experience or from research. There are still many difficulties in presenting a clear description of the processes of fire growth and the contribution of the many factors involved. In parallel with an international study covering the factors in fire growth , certain fundamental features are being studied in detail, and this report is one of a series investigating the characteristics of flames beneath various kinds of ceiling.
At some stage in the growth of a fire in an enclosed space the flames from materials on the floor are deflected by the ceiling and extend horizontally. It was shown in part 1 that, even when the ceiling is incombustible, the horizontal flames dramatically increase the radiation to unburnt fuel away from the fire and this alone leads to fire spread irrespective of other influences.
This report shows that a combustible lining will cause an increase in the
length of the horizontal flames and a more rapid increase in the intensity of downward radiation than an incombustible one. The types of lining investigated were all cellulose-based building boards and with most of these (including untreated fiber insulating board) the lengths of horizontal flames depended on the size of the fire on the floor; when this was restricted fire would spread indefinitely beneath only one of the boards tested, a stove-enamelled hardboard. Apart from this material, differences between linings lay in the rate of increase of downward radiation much more than in the intensity finally attained.
The measurements of downward radiation were used to estimate the spread of fire over a narrow strip of combustible material on the floor, the assumption was made that the flames from a narrow strip would be small so that they would not affect the ceiling fire and it twas found that the initial rate of spread varied with performance of the material on the Fire Propagation Test. The final rate of spread with the exception of stove-enamelled hardboard did not depend on the type of lining.
The importance of the initial rate of spread would be greatly increased
where 'feedback' from a larger floor fire to the ceiling fire could occur because the spread would in effect then be an accumulation of successive initial rates of spread and would accelerate until other factors such as a shortage of oxygen intervened. It appears that under these circumstances the performance of a cellulose based board in the Fire Propagation Test is a good measure of its hazard when used as a ceiling lining.