Compartment fire experiments: Comparison with models

• Published in: Fire safety journal 1995-10, Vol.25 (3), p.187-227
• Main Authors: Dembsey, N.A., Pagni, P.J., Williamson, R.B.
• Format: Article
• Language: English
• Subjects: Applied sciences; Building technical equipments; Buildings; Buildings. Public works; Exact sciences and technology; Fire protection
• ISSN: 0379-7112
• DOI: 10.1016/0379-7112(96)00002-1

Twenty full-scale compartment fire experiments suitable for model comparison were conducted. Ceiling jet temperatures, surface heat fluxes and heat transfer coefficients which have not been previously reported are discussed. The ceiling jet temperatures 0·10 m below the ceiling show the effects of compartment ventilation, near-field entrainment conditions and burner location on the ceiling jet. Net and radiant incident heat fluxes to the upper and lower-walls and the floor are estimated. Combined (radiation and convection) interior heat transfer coefficients for the three surfaces are reported. As compartment fire models such as CFAST and FIRST continue to develop in sophistication, it is important that they be compared to experimental data. Data at three heat release rates: 330, 630 and 980 kW, are used to evaluate these comprehensive compartment fire models and two simpler models for the upper-layer gas temperature. CFAST predicts upper-layer gas temperatures 150–260°C hotter than the measured bulk outflow gas temperatures. The increased temperatures appear to be due to insufficient heat transfer through the compartment surfaces. FIRST predicts upper-layer gas temperatures that are slightly cooler (on average, 20C) than the measured bulk outflow gas temperatures. The two simpler models are within 40°C, on average, of the measured upper-layer gas temperatures