108 A r a ş t ı r m a groups yet shares certain common elements is the mod ern exhibit hali. The current trend is toward halis in excess of 1,000,000 ft2 (92 900 m2), with �eilings at 40 to 60 ft (12 to 1 8 m) high. The diversity of goods dis played, the uncountable ignition sources, and the mazelike arrangement of aisles provide the potential for a major disaster. The problems that underlie all large enclosed spaces are; type of fuel, quantity of fuel, arrangement of the fuel package, effectiveness of the fire sprinkler system, and ability to provide safe and timcly egress. So long as such buildings are built and usecl, limitations on the fuel problem beyond very broacl and general restrictions are essentially impractical ancl virtually unenforce able. Until the next major breakthrough on reliable fixecl extinguishing systems occurs, the physical limitations of geometry (in particular, height) are going to continue to cast cloubt on sprinkler effectiveness. Because size and height have already complicatecl the egress issue, the problem left to examine is the smoke generated by the probable fire. Much has been written about the smoke problem. Per haps the most encyclopedic treatment of the problem in large enclosed spaces in that found in the ASHRAE publication Design of Smoke Management Systems by Klote and Milke.4 Examining issues of toxicity, irritation, obscuration and temperature is important, but of little practical use in designing for safe egress. As a rule, people exhibit a strong negative response to being immersed in any smoke regardless of its "tolerable" properties. This negative response may not result in appropriatc life-preserving actions. Therefore, the design goal for exit paths should transcend "tenable environmend" and approach, as nearly as possible, "smoke free". Herein lies the problem with smoke control design as generally practiced in the United States, Economic pressures generally limit the design to that of minimum code compliance. This consists of an exhaust by a fixed number of air changes, specified methocls of makeup air introduction, and demonstration of compliance capricious by an arbitrary and capricious test. Because the latter portion of this article is focused solely on atriums and because the problem is generic even though not so treated by US cocles, this is a good time to begin limiting the discussion to atriums. DISCUSSION The typical exhaust requirements in US codes for atrium smoke control are 6 air changes per hour (ach) for smaller atriums and 4 ach for larger ones,5·6.7,8 This, of course, is patently unscientific, and numerous authors have noted this repeatediy over the years. While there yet remains some argument over which empirical correlation provides the simplest and most reliable approximation of smoke generation, they ali recognize that fire size and space geometry are key parameters that must be employed to determine the appropriate exhaust rate. The typical makeup air requirements and test re quirements in US codes are intimately related. This mythical "Hydra" springs from two basically sound, but misapplied concepts: ♦First, that nature abhors a vacuum and, therefore, some percentage of air must be positively supplied to allow exhaust to occur. ♦Second, that "cold" smoke does not rise and this must be helped. Typically, the requirement for replacement air is set at 50 °'0 of the exhaust rate. This is myth number one. lf nature abhors a vacuum, it is no less dis pleased by a partial vacuum. lf the exhaust rate is to be achieved, then the supply rate as the sum of makeup air and combustion producst must be equal. As noted by Klote and Milke,4 the loca! air velocity approaching the fire plume can be unbeneficially clis ruptive if not controlled and maintained below about 200 fpm ( 1 m/s). This cannot necessarily be assured if only 50 % of the exhaust rate is being controlled. Typically, the test requirement is for some form of vis ibility in the space at the enci of a predetermined time period before which some form of chemical smoke was introduced into the space, usually 1 O to 1 5 minutes after turning the exhaust system on. This visibility
RkJQdWJsaXNoZXIy MTcyMTY=