Particle shapes ranged from "stringer" type smokes containing discrete spherical units forming chains for rigid PVC and ABS to the elliptical shaped liquid residue for red oak. Through a knowledge of the magnification of the scanning electron microscope photographs, typical particle sizes were estimated as follows: No attempt was made at this time to estimate particle size distributions. It is interesting to note, however, that there appears to be a noticeable difference in shape between rigid and flexible PVC smoke particles, the former containing relatively long aggregated chains while the latter has a greater number of smaller aggregates more spherical in nature. 5. FUTURE STUDIES The research described in this paper has recently been started and represents only a part of the overall program. The program will be limited to the six materials described earlier and will consist of three phases: 1. The determination of the effects of heat flux, oxygen depletion, 2. 3. The determination of the relationship between optical density and The determination of smoke particulate shape and size distribution (if possible) for each material under flaming and smoldering exposures. The study is designed to provide information relating to the following questions: 1. 2. 3. 4. What are the relative smoke and carbon monoxide generating characteristics of these six materials and do they change significantly with higher temperatures and oxygen deficiencies? From a knowledge of weight loss, heat flux, and oxygen availability can the maximum smoke densities and CO generation rates be estimated on samples of similar composition to those studied? After experimentally determining the relationship between optical density and smoke mass density over a narrow range can these data be extrapolated to include higher optical densities so that one might predict total airborne smoke mass? How does smoke particle shape and size vary with material composition and burning conditions? Based on the limited study to date, carbon monoxide accompanies smoke development in the flaming combustion process, but there does not appear to be a unique relationship between the two products for materials of different composition. In the case of red oak, it is possible to produce considerable amounts of CO without much smoke if the burning process remains in the flaming mode. The rate of carbon monoxide evolution for the plastic materials studied under nonflaming conditions was observed in all cases to be relatively low. The complex interrelationship between smoke mass, optical density, and particulate size and shape appears to be strongly dependent upon material composition. [1] Lee, T. G., Interlaboratory Evaluation of Smoke Density Chamber, Nat. Bur. Stand. (U.S.), Tech. Note 708 (Dec. 1971). [2] Gross, D., Loftus, J. J. and Robertson, A. F., Method for Measuring Smoke 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) A FIELD STUDY OF NON FIRE-RESISTIVE MULTIPLE DWELLING FIRES Francis L. Brannigan1 Montgomery College, Rockville, Maryland A field study was made of structural and building design factors contributing to the spread of fire in more than 40 non fire-resistive, multiple occupancy dwellings, typically "Garden Apartments". Most deficiencies could be corrected by preserving the integrity of a gypsum board sheath serving as a fire barrier. Examples are given of penetrations and openings in fire barriers which permitted substantial fire spread. Key words: Apartments; building codes; fire; fire walls; garden apartments; gypsum board; insurance; livability; multiple dwellings. 1. INTRODUCTION From January 1971 till the present the author has been under contract to the National Bureau of Standards to investigate serious multiple dwelling fires to determine what structural elements or practices contributed to the extension of fire beyond the area of origin. Several local fire department radio channels are routinely monitored by the author, and he responds to fires of interest. In addition, he keeps in close touch with fire officials in the local area, and is notified of fires of interest. All investigations were made either while the fire department was still in possession of the premises, or if later, accompanied by a fire official and with the permission of and access provided by the owners' representative. Wherever possible firefighters early on the scene were interviewed.2 A few observations are based on examination of buildings under construction. No attempt was made to determine the extent to which any particular building conformed to the code in effect at the time the particular building was built. Obviously, not all the defects cited exist in all jurisdictions, or in all structures examined. Each report provides a capsule description of the building, the fire, the manner in which the fire spread from the point of origin, appropriate 35mm slides, and any recommendations the investigator might have on how modifications of present practices might improve the fire containment design of similar buildings. The reports are circulated to the appropriate staff members and provide a body of information which aids the staff in developing and supporting improved building design. Through other channels the investigator shares his findings and fire tactical recommendations with firefighting forces. If a fire department is well aware of the defects of a building or types of buildings it may be possible to "plug the gap" tactically and reduce the consequences of the designers' errors. 1Mr. Brannigan serves as a consultant to the Program for Fire Control Construction, Programmatic Center for Fire Research, National Bureau of Standards, Washington, D.C. 2 Appreciation is expressed to the Montgomery County Department of Fire and Rescue Services and the Fire Departments of Bethesda, Kensington, Silver Spring, Rockville, Gaithersburg, Prince Georges County, Anne Arundel County, Clarkesville and Annapolis in Maryland, Arlington and Fairfax Counties in Virginia and the District of Columbia. 2. PRIVATE HOMES VS. MULTIPLE DWELLINGS "A man's home is his castle" is a long established principle of our common law. This principle is carried forward into statute law such as building codes, in that fire protection aspects of building codes bear very lightly on single family dwellings. Regulations calling for masonry exterior walls and spacing between buildings limit one's "right" to burn down his neighbor, though recent relaxation of rules against wooden shingles have restored this "right" in some areas. Generally, inspection laws don't give the Fire Inspector access to private dwellings. The principal impact of the statutes is in the regulation of the initial installation of electricity and heating appliances, but there is generally no warrant for reinspection to determine that installations have not been substantially altered after initial approval. Even when requirements are upgraded, usually no attempt is made to apply them retroactively to existing dwellings. For example, current efforts to legislate installation of combustion products detectors in one-family homes are limited to new homes; little thought is given to ordering installation in existing homes. It is impossible to house each family in the United States in its own individual "castle". Multiple dwelling units are necessary. In the multiple dwelling unit there is an interaction between a fire in any unit, and the safety of the occupants in all the other units. In addition, multiple dwelling units contain common areas accessible to all occupants, and to varying degrees, to any outsider. A fire in the common area threatens all the occupants. Historically, multiple dwellings in many cities were created by the subdivision of large one-family homes into smaller units as the former owners moved away and neighborhoods deteriorated. In cities like New York, the supply of deteriorated housing was insufficient to meet the needs of hordes of immigrants, so tenements were built to house several family units, structurally indistinguishable from the multi-story one-family house of the well-to-do (the typical New York "Brownstone"). The first official recognition of any special fire safety problem was the requirement that outside fire escapes (vertical ladders, with a platform at each floor level) be provided. Terrible losses of life occurred in these tenements (now "old-law" tenements in New York parlance).1 About 1903, as a result of sociological pressures, sweeping tenement house legislation was passed in New York giving rise to the "New Law" tenement. Though not so stated, the principle which underlay the requirements was that no person should lose his life due to a fire which started outside his apartment. The entire thrust was towards easy escape of the occupants, and for many years there was no loss of life in such a building due to extension of fire. While the fire protection emphasis was almost exclusively directed towards safe escape, requirements for life safety also tended to limit extension of fire. Fire resistive stairways, self-closing fire resistive apartment doors and similar requirements for dumbwaiters together with the universal complete plastering of all interior surfaces, served to contain almost all fires to the apartment of origin, for many years. The provisions for easy escape of occupants lightened the burden of rescue operations so that firefighters were available to swiftly check for, and suppress, any extension of fire. Exterior fire escapes provided direct secondary access to all apartments. For a further discussion of the circumstances of the New York City Tenement House Act see Brannigan, F. L., "Building Construction for the Fire Service", National Fire Protection Association (1971), pp. 146-150. 2.4. Heat Flux Smoke and carbon monoxide concentrations from burning rigid PVC samples were studied as a function of heat flux (flow rate of energy across or through a surface). Radiant heat flux levels were varied from 1.87 to 3.07 W/cm2 under flaming type burning conditions while the gas and air mixture to the ignition burners was constant. As shown in figure 5 as the heat flux increased the maximum specific optical densities were increased, the maximum rates of smoke and carbon monoxide generation were increased, and the times to reach Dm were decreased. The marked two-fold increase in maximum smoke density resulting from increasing the radiant heat from 1.87 to 3.07 W/cm2 can be attributed in part to a greater total weight loss (ca. 70 percent and 84 percent respectively). Other factors which could contribute might be that less time for particulate decomposition on the chamber surface is available and there are possibilities of different reaction pathways which favor more smoke production at higher tempera tures. 3. SMOKE MASS Smoke mass density has been related to optical density measurements in the smoke chamber by using frequency monitored electrostatic precipitation. The change in frequency during a given time interval is directly proportional to the total particulate mass. In this study smoke was sampled at a point half way up the chamber wall under steady state conditions after reaching the maximum smoke density. A plot of smoke optical density per meter versus smoke mass density (milligrams per cubic meter) for red oak (flaming and smoldering exposure) and ABS, and rigid and plasticized PVC (smoldering exposure) is shown in figure 6. Figure 5. Effect of heat flux on smoke and carbon monoxide CARBON MONOXIDE- PPM |