Numbers in Bars Indicate Values in Excess of Scale Values + Max. Indicated Concentrations of HF and SO2are Shown in HC1 or HCN Columns Where the Latter were not Detected 1 oz/y3.39 x 10-3 8/cm FIGURE 6. Smoke and gas concentrations for individual materials-sheets. Numbers in Bars Indicate Values in Excess of Scale Values + Max. Indicated Concentration of S02 is Shown in HCN Column Where Latter was not Detected 1 oz/yd 3.39 x 101 8/cm FIGURE 7. Smoke and gas concentrations for individual materials-assemblies, laminates, films. "Maximum" indicated concentrations of gases are listed in appendix 3 along with the smoke data. These values are based on the average of two separate determinations, except that additional tests were made where large discrepancies (greater than a factor of 2) between duplicate values were obtained. Unlike the measurement of optical density of smoke, which is recorded continuously to obtain a maximum, the concentrations of selected components was measured periodically. Particularly for components which change rapidly, therefore, the indicated concentration values may not necessarily be the true maximum values. For the materials tested, the highest indicated concentrations were 2200 ppm CO, 2500 ppm HCl, and 90 ppm HCN. These concentrations refer to the same exposed area of specimen and chamber volume used, but to a wide range of specimen weights. Since the primary objective of this study was to ascertain approximate values, no extensive efforts were made to improve reproducibility. As a test of reproducibility for a PVC material (specimen No. 44), 5 separate smoldering exposure tests were conducted with the results shown in figure 8. This figure shows the five replicate smoke curves and a tabulation of indicated gas concentrations at spe cific times during each test. The measurement ranges were on the order of ±20 percent for CO and HCN and ±30 percent for HCl, and such variations may be considered typical of the maximum indicated concentration values under the test conditions. Because the plastic materials studied were from many manufacturers and generally contained plasticizers, fillers, and other additives, it is difficult to relate quantitatively gaseous product concentrations with polymer composition. In general, HC was produced by polyvinyl chloride and modacrylic materials, HF from polyvinyl fluoride, HCN from wool, urethane, ABS, and modacrylics, and SO2 from polysulfone and rubber materials. CO was produced by almost all the samples in varying amounts depending on the type of material. It has been shown [5] that the amount of a given gas produced during pyrolysis and its rate of generation are strongly temperature dependent. Thus, any materials or processes which affect the temperature profile across the specimen (e.g. fillers and plasticizers which produce surface crusting, intumescence, etc.), could readily influence the concentration of gaseous products. For certain materials, higher concentrations of some gases may be produced under conditions of insufficient air, e.g. 10 percent oxygen [6]. Sampling was performed sequentially, proceeding generally from HCl and HF to HCN to CO, and was initiated when optical density of the smoke approached its peak. This procedure was followed because of the fairly rapid decay in halogen acid concentration resulting from adsorption on (and reaction with) moisture, smoke particles, and chamber surfaces. To facilitate subsequent data comparison, sampling for HCl and HF was generally initiated at the beginning of the minute close to the maximum smoke level, and at 2-min intervals thereafter for other gases. Gas temperature at the sampling tube inlet generally ranged from 46 to 52 °C (115 to 126 °F), the higher temperatures occurring during flaming tests on heavier materials. Due to the cooling effect of the precleaning layers of the indicator tubes, the temperature of the gases passing the indicating layers were within the prescribed maximum temperature limits. The sampling rate was generally unaffected by either the elevated temperature of gases or by heavy smoke particle concentrations. Hydrogen chloride is generally released rapidly during combustion or pyrolysis of polyvinyl chloride, modified acrylics and other retardant-treated materials [7, 8]. Maximum levels were generally higher under flaming compared to smoldering exposure conditions presumably due to the higher temperature involved and the resultant greater rate of release. The HCl concentration changed rapidly as a result of its high reactivity, solubility in water, and adsorption on smoke particles and wall surfaces. The type of surface as well as the total area of the interior walls have a pronounced influence on the adsorption and settling (or decay) rate of HCl and smoke. To illustrate the decay of both HCl and CO, a suitable concentration of the pure component was metered into the bottom of the chamber under both smoke-free (D = 0) and smoke-filled conditions. Figure 9 shows the indicated concentrations of HCl and CO. In these tests involving smoldering specimens only, the gas concentration levels are obviously higher because a portion of the gas is introduced by combustion. The decay rates are also higher. 4. Discussion In the work described in this report, it was presumed that the test specimens were representative in thickness and density of the materials intended for actual use as interior finishes. For a few materials supplied in thicknesses greater than 1 in, the test specimen was trimmed to a thickness of 1 in to fit the size of the specimen holder. It should be evident that the density of smoke, the concentration of gaseous products, and the heat release characteristics are properties of the specimen as tested and will be different for other thicknesses and densities. 10 general illumination level, on the contrast threshold and the extent to which the observer's eyes have been dark-adapted, as well as on the irritating nature of the smoke. In figure 10, five lines are shown for transmission values ranging from 80 to 2.5 percent (optical density 0.1 to 1.6) corresponding to a wide range of visual limits [3]. Using this figure, sample computations have been made in Table 1 for 3 selected values of D.. If it is assumed that a lighted exit sign can be seen when the transmission is down to 40 percent (optical density 0.4), and an aircraft cabin has a volume of 10,000 ft within which smoke is uniformly dispersed, then Table 1 shows the estimated area A of material, the smoke from which may just begin to limit seeing the exit sign at various distances L. Up to this point, only geometrical factors have been considered, but time is certainly important, and the choice of a critical specific optical density for each material can presumably also be based on a prescribed time period which is sufficiently long to permit escape or defensive action. From appendix 3, it may be noted that the time periods to attain a critical specific optical density of 16 ranged from 0.2 to over 20 min. It was previously noted [3] that a specific optical density of 16 could represent a possible critical limit. Although the three factors, total smoke accumulation (Dm), maximum rate of smoke accumulation (Rm), and the time period to reach a "critical" optical density (t.), are directly related to the smoke obscuration hazard, their relative weighting is not entirely obvious. One suggestion for a single overall hazard index based on the results of this test was made in the appendix of reference 3. However, it should be emphasized that additional experimental verification would be desirable prior to establishing rigorous smoke hazard limits for interior materials. This study was concerned with the limited problem of measuring the optical density of smoke as it relates to the obscuration of human vision. No attempt was made to evaluate complications due to eye irritations, to respiratory effects from inhaled smoke particles, or to hysteria or associated physiological or psychological factors. The indicated concentrations of gaseous products listed in appendix 3 represent values measured at the sampling location and are associated with the prescribed exposure conditions on a specimen of given exposed area (26 in square) within a totally enclosed chamber of 18 ft3 volume. Specimens were tested in the thickness and weight supplied, which varied over a wide range. Concentration measurements were made periodically from the time when the optical density of the smoke approached its peak. Any realistic evaluation of the gas concentrations likely to be encountered in a real fire situation must take into account actual areas and thicknesses of the materials exposed and the volumes in which the gases are dispersed. Also of importance are the rate of fire growth, the effects of adsorption and reaction, the extent of ventilation, dilution, and/or application of extinguishing agents, and other factors outside the scope of this study. Where specimen area and chamber volume are the only variables and uniform mixing is assumed, an approximate relationship between the gas concentration measured in the smoke chamber and the projected concentration within a much larger chamber, such as an aircraft cabin, is given by |