Citation

Wu Y, Bakar MZA. Fire Safety J. 2000; 35(4): 363-390.

Copyright

(Copyright © 2000, Elsevier Publishing)

DOI

unavailable

PMID

unavailable

Abstract

The "critical velocity" is the minimum air velocity required to suppress the smoke spreading against the longitudinal ventilation flow during tunnel fire situations. The current techniques for prediction of the values of the critical velocity for various tunnels were mainly based on semi-empirical equations obtained from the Froude number preservation combining with some experimental data. There are a few uncertainties in the current methods of prediction of the critical ventilation velocity. The first is the influence of the fire power on the critical ventilation velocity. The second is the effect of the tunnel geometry on the critical velocity. Both problems lead to the issues of the scaling techniques in tunnel fires. This study addressed these problems by carrying out a series of experimental tests in five model tunnels having the same height but different cross-sectional geometry. Detailed temperature and velocity distributions in the tunnels have been carried out. The experimental results showed that the critical velocity did vary with the tunnel cross-sectional geometry. It was also shown clearly that there are two regimes of variation of critical velocity against fire heat release rate. At low rates of heat release the critical velocity varies as the one-third power of the heat release rate, however at higher rates of heat release, the critical velocity becomes independent of fire heat release rate. Analysis of the distribution of temperature within the fire plumes showed that there were two fire plume distributions at the critical ventilation conditions. The change of the fire plume distribution coincided with the change of the regime in the curves of the critical velocity against fire heat release rate. The study used dimensionless velocity and dimensionless heat release rate with the tunnel hydraulic height (tunnel mean hydraulic diameter) as the characteristic length in the experimental data analysis. It was shown that the experimental data for the five tunnels can be correlated into simple formulae which can be used for scaling. The new scaling techniques are examined by applying the scaling techniques to the present experimental results and three large-scale experimental results available in the public literature. A good agreement has been obtained. This suggests that the scaling techniques can be used with confidence to predict the critical ventilation velocity for larger-scale tunnels in any cross-sectional geometry. Comprehensive CFD simulations have been carried out to examine the flow behaviour inside the tunnels. Validation against the experimental results showed that the CFD gave slightly lower but satisfactory prediction of the flow velocity. However the temperature prediction in the fire region was too high. The findings from the CFD simulations supported the ones from experimental tests.