At a of acrylic carpets has a significant effect on ignition and flame spread. The other experimental parameter is the duration of preheating. Except at the highest radiant flux used in this study, 1.15 w/cm2, a preheating time longer than three minutes does not increase speed of flame spread significantly. radiant flux of 1.15 W/cm2, the longer preheating time produces smoking and partial charring of the sample surface before it is ignited by the pilot burner. After pilot ignition and a transient stage, very rapid flame spread is observed. It spreads so fast that accurate measurement of its speed by eye is very difficult. The order of speed is a few centimeters per second. With a still higher radiant flux, this speed would eventually reach the flame propagation speed for the mixture of decomposed fuel vapor and surrounding air, which is about an order of magnitude larger. This rapid flame spread phenomenon is also observed in the full scale experiments [1]. This phenomenon can spread fire and smoke throughout a building, causing loss of life and extensive property damage. Therefore, the relation between radiant flux and preheating time leading to this rapid flame spread is an important factor by which to judge the fire safety of a carpet. There are three important parameters: ignitability, minimum energy flux to sustain constant flame spread, and the domain of rapid flame spread controlled by radiant flux and preheating time. The general trends of these parameters with incident radiant flux and pilot burner duration time are shown in figure 4. It is considered here that the minimum radiant flux is an asymptotic value of external Figure 4. Qualitative relations between ignition and incident PREHEATING TIME radiant flux for sustained flame propagation. The plotted curves are relative ones and each carpet has its own characteristic curves. From the test standard point of view, ignitability and the domain of rapid flame spread should be measured by a pilot ignition study (because the preheating times and the external radiant fluxes necessary to cause rapid flame spread are well correlated with the pilot ignition delay time) and minimum radiant flux to sustain flame spread should be measured by a flame spread test under a radiation flux field which decreases in intensity along the sample surface. Thus, we have discussed flame spread and ignition under an external energy flux. However, in an actual fire, there is another important energy source which is heat release from the burning carpet. This energy will also contribute to the rate of flame spread. For this reason, the heat release rate from a carpet over which the flame spreads with constant speed was measured. First, the total heat release from the unburned carpet is measured by the oxygen-bomb [2] method. Then, heat release from the char remaining after the flame spread test is also measured with the same technique. The difference between these values, adjusted to represent equal areas, is the net heat release which is measured per unit surface area. The product of this net heat release and the constant flame spread speed is the net heat release rate per unit width. The results are shown in figure 5 for various radiant fluxes. The data indicate that the net heat release rate increases rapidly with increasing incident radiant flux. It is interesting to note that net heat release rates of low pile density carpets A-2 and A-5 are smaller than those of other carpets at low radiant fluxes although flame spread speeds of A-2 and A-5 are faster than those of other carpets. This means that flame spread over A-2 and A-5 involves only a shallow zone near the surface. However, the other carpets burn more deeply. This is confirmed by the measurement of weight loss of carpets after flame spread tests. Net heat release rate decreases considerably when the carpets are burned without a pad even for low pile density carpets A-2 and A-5. This behavior is quite different from the negligible effect of a pad on ignition and flame spread speed for A-2 and A-5. This is illustrated in figure 6, which compares the flame shape with a pad and without a pad under the same radiant flux. Without a pad, there is only a main flame. This means that the carpet burns longer and over a larger area with these smaller flames. This behavior is observed only at high radiant fluxes. At low radiant flux, the width of flame increases slightly. 1. 2. 3. 4. 4. SUMMARY For carpets which pass the pill test there exists a minimum radiant flux necessary to sustain flame spread over the carpet surface without extinguishment. Speed of flame spread increases rapidly with increasing radiant flux and approaches values of flame propagation speed characteristics of the gas phase. The construction of the pile layer of acrylic carpets significantly affects ignition, flame spread speed and net heat release rate. With nylon carpets the effect is smaller due to melting of the pile ahead of the flame front. The effect of a pad on ignition and flame spread speed is significant for carpets of heavy pile density, but small for acrylic carpets of light pile density. However, the effect on heat release rate is significant for all carpets tested. NATIONAL BUREAU OF STANDARDS SPECIAL PUBLICATION 411, Fire Safety Research, Proceedings of a Symposium Held at NBS, Gaithersburg, Md., August 22, 1973, (Issued November 1974) PHYSIOLOGICAL AND TOXICOLOGICAL EFFECTS OF THE PRODUCTS OF THERMAL DECOMPOSITION FROM POLYMERIC MATERIALS M. M. Birky1 National Bureau of Standards, Washington, D.C. I. N. Einhorn, M. L. Grunnett, S. C. Packham, J. H. Petajan, J. D. Seader A program that combines the capabilities of the College of Medicine and the College of Engineering of The University of Utah has been instituted to evaluate the physiological and toxicological effects of the products of thermal degradation and combustion of cellulose, a polyvinyl chloride, a flexible polyurethane, and wood (Douglas fir). The products produced from these materials are being identified and quantified with a gas chromatograph-mass spectrometer-computer system. In addition, a National Bureau of Standards smoke chamber has been modified with a weight loss transducer to correlate, on a continuous basis, the quantities of smoke produced with sample weight loss. Extensive studies on the effects of these degradation products on rats is in progress. The results of exposure of the rats to carbon monoxide are reported. All of the laboratory results are being correlated with full-scale fire studies at the National Bureau of Standards. Key words: Combustion; polymer; pyrolysis; smoke; specific optical density; toxic gases; toxicity. The flammability research program at The University of Utah is sponsored by the National Science Foundation's Research Applied to National Needs (RANN) Program. The program is directed to the physiological and toxicological aspects of humans during fire exposure and is divided into the following thirteen tasks: 1This paper was written during the 1973-1974 academic year while Dr. Birky was a visiting Professor at The University of Utah. [1] [2] Figure 6. The effect of pad on flame shape. Flame travels from right to left (0.86 W/cm2). Fung, F. C. W., Suchomel, M. R. and Oglesby, P. L., Fire Journal, Vol. 67, 41 (1973). Ambrosius, E. E. and Fellows, R. D., Mechanical Engineering Laboratory NATIONAL BUREAU OF STANDARDS SPECIAL PUBLICATION 411, Fire Safety Research, Proceedings of a Symposium Held at NBS, Gaithersburg, Md., August 22, 1973, (Issued November 1974) PHYSIOLOGICAL AND TOXICOLOGICAL EFFECTS OF THE PRODUCTS OF THERMAL DECOMPOSITION FROM POLYMERIC MATERIALS M. M. Birkyl National Bureau of Standards, Washington, D.C. I. N. Einhorn, M. L. Grunnett, S. C. Packham, J. H. Petajan, J. D. Seader A program that combines the capabilities of the College of Key words: Combustion; polymer; pyrolysis; smoke; specific optical The flammability research program at The University of Utah is sponsored by the National Science Foundation's Research Applied to National Needs (RANN) Program. The program is directed to the physiological and toxicological aspects of humans during fire exposure and is divided into the following thirteen tasks: 1This paper was written during the 1973-1974 academic year while Dr. Birky was a visiting Professor at The University of Utah. |