A· r a ş t ı r m a The thermal properties of the enclosure versus the real fire building were alsa consiclered. it was cleterminecl that 0.5 in, (1.25 cm) gypsumed wallboarcl was sufficiently therınally similar to the brick and plasterecl construction. As previously clescribed,1 2 the effects of the upward projected makeup air jet woulcl cause a significant ınass iınpact at the upper terıninus of the atrium. The correlation usecl to approximate this phenome nan was that reported in ASHRAE Transactions12 as: While this mass flow would have the beneficial result of ınoving colcl cheınical smokes off the floor ancl up to the exhaust inlets (as previously cliscussed), it would significantly interfere with the effective smoke exhaust capabilities during a real fire. The series of tests conductecl variecl the fire size in scale from an equivalent of approxiamtely 1 MW to 5 MW. The air flows at the jet were varied from no farced flow to a scaled equivalent of 25,000 cfın (11 800 L/s). The exhaust rates were variecl from 38,000 cfın (17 900 L/s) to 76,000 cfm (35 900 L/s) scalecl equivalents. CONCLUSION The results of these tests (as shown in Figure 2) were in goocl agreement with the mathematical predictions, the eyewitness accounts ancl subsequent investigations of the real fire inciclent. The pri ncipal issues of significance were: -The 6 adı of exhaust rate was totally incapable of provicl ing sufficient removal of smoke in the real inciclent and of fire products in the scale model; and - The induction jet at the base of the atriuın perfarmed exactly as predictecl i n ali tests of the scale model, resulting in clistribution of fi re proclucts throughout the model as was evidencecl in the real fire. There are two unmistakable conclusions to be clrawn from this investigation. First, hydraul ic scale ınodeling is an effective engineeri ng tool to examine flow regimes in complex geometries far the purpose of determining smoke removal during real fires in actual buildings, whether in design or under investigation. Seconcl, current US code practices are probably most often ineffective in cleal ing with sınoke removal in such spaces ancl probably contribute greatly to i ncreased smoke daınage ancl certainly to increased sınoke clistribution within theın. Ali of this could have been preclictecl scientifically using methocls set farth in the literature far soıne time. References 1 - Morgan, H. 1979. Sınoke Control, Methods in Enclosecl Shopping Coınplexes of üne or More Storeys: A Design Suınmary. Borehaınwood, Hertfardshire, United Kingdoın: Fire Research Station. 2- Hinkley,P. 1971. Soıne Notes on the Control of Sınoke in Enclosed Shopping Centers. Fi re Research Note No. 875. Borehaınwoocl, Hertfordshi re Unitecl Kingcloın: Fire Research Station. 3-Butcher, E., Parnell, A. 1979. Sınoke Control in Fire Safety Design. Landon, United Kingdoın: E. & F.N. Spon. 4- Klote, J., Milke, J. 1 992. Design of Smoke Manageınent Systems. Atlanta, Georgia: ASHRAE. 5- ICBO, 1988. Unifarın Building Code. Whittier, California: lnternational Conference of Building Officials. 6- BOCA, 1988. National Building Code. Country Club Hills, lllinois: Builcli ng Officials and Code Adıninistrators lnternational. 7- SBCCI, 1988. Standard Building Cocle. Birmingham, Alabama: Southern Building Code Congress 1 nternational. 8-NFPA, 1988. Life Safety Code. Quincy, Massachusetts: National Fire Protection Association. 9- Quintiere, J. 1989. "Scaling applications in fire research". Fire Safety Journal. Essex. England: Elsevier Science Ltd. 1 0- Turner, J. 1 973. Bu oyan ey Effects ancl Fluicls. Cambridge, United Kingdom: Caınbridge at the University Press. 11-Taınura, G., Shaw, C. 1976. "Studies on exterior wall ai r tightness and ai r infiltration of tali buildi ngs." ASHRAE Transactions. Atlanta, Georgia: ASHRAE. Vol. 82, Pt. 1. 1 2- Dil lon, M. 1983. "The design of a smokemanagement system fara hotel atri um." ASHRAE Transactions. Atlanta, Georgia: ASHRAE. Vol. 89. Pt. 2B. Editor's note: This article is based on a paper first presented by the author at the Clima 2000 Conference in November 1993 in Landon, England. Far further infarmation, contact CIBSE in Landon, fax 081-675-5449. 111
RkJQdWJsaXNoZXIy MTcyMTY=