A correct description of the fire is essential for getting a correct evaluation of the effects on the people and on the structures and, ultimately, to obtain a sufficient and reliable level of safety in case of fire.
Current prescriptive design and verification methods for fire safety often rely on quite simplified description on the fire action, typically based on free burn test under a cone calorimeter or standard fire test in an oven or a small test room. Nevertheless, it is well known that the enclosure and boundary conditions affect the fire significantly.
In particular, the occurrence and the onset of flashover are strongly influenced by thermal feedback of the enclosure, which depends on the size and geometry of the compartment, on the ventilation conditions, and also by type of linings the enclosure is made of. This affect the reliability of simplified pre- and post-flashover models, used to design for local and structural fires respectively.
1. Pre-flashover fires
Full scale fire testing complemented by computer modeling has provided significant knowhow about the risk, prevention and suppression of fire in terrestrial systems (cars, ships, planes, buildings, mines, and tunnels). In comparison, no such testing has been
With respect to the pre-flashover fire, tests and projects carried out by the group have indicated that, in case of enclosure with very low thermal inertia, the onset of flashover can be significantly overestimated by the traditional flashover criterion based on the temperature of the smoke layer. This reduces the time for the evacuation and results in a non- conservative design of the safety of the occupants. Furthermore, the difference observed in tests for room burn conditions and free burn conditions indicate that models based on free burn results cannot be safely used for describing pre-flashover fires in room.
2. Post-flashover fires
With respect to post-flashover fires, special attention should be paid to the following aspects:
Low-energy requirements for new buildings translate in lower ventilation conditions during fires. This is due to the employment of thicker double glass layer windows, which are expected not to break during a post-flashover fire, as normally assumed in calculations for traditional compartments. Due to the lower ventilation, the fire will be milder but longer and the temperature reached by the elements during the fire, which is strongly influenced by the duration of the fire, will be higher. This means that the current resistance classes, calibrated to model the structural effect of fires in traditional compartments, can’t ensure sufficient safety in new buildings and that a longer time of resistance should be used when designing with standard fire curve.
Since the size of the compartment influences the thermal feedback of the enclosure and the occurrence of flashover, post-flashover models based on tests in small compartments cannot be applied to rooms with a floor area larger than 500 m2, according to current prescriptive regulations. Nevertheless, large compartments are necessarily present in industrial hall and public buildings and, more and more often, open spaces represent a desired feature for office premises as well. This leaves a dangerous hole in the design methods of actual buildings and highlights a safety problem in case of fire in large compartments.
Tests and studies carried out by the group have shown that when the combustible material can be assumed to be reasonably uniform and the large compartment has a regular geometry (such as, for example, long rectangular office spaces), the assumption of flashover is acceptable, as it yields the same or higher safety level than in small compartments.
Nevertheless, recent studies published in literature have shown that the assumption of flashover may not be conservative, when a traveling fire develops, as the long duration of the mild fire, which slowly moves from one combustible material to another, may yield higher temperature in the elements. On the contrary, if the distribution of the combustible material is highly non-uniform (such as in case of staked storage materials in industrial halls), an very intense local fire may have unaccounted effects on the structural elements above the flame, which suffers also from a higher thermal restrain offered by the surrounding elements, which remain much colder.
PhD Project
Fire Models and design fires, A. Poulsen, 2012
Publications
Poulsen A., Jomaas G. (2011):
Experimental Study on the Burning Behaviour of Pool Fires in Rooms with Different Wall Linings,
Journal of Fire Technology, vol: 48, pages: 419-439, 2011
Poulsen A., Jomaas G.:
Experimental Study on the Influence of Thermal Feedback on the Burning Behavior of Flexible Polyurethane
Poulsen A., Jomaas G., Bwalya, A. (2012):
Evaluation of the Onset of Flashover in Room Fire Experiments,
Journal of Fire Technology, vol: 49, issue: 4, 2012
Hertz K.D. (2012):
Parametric Fires for Structural Design
Fire Technology, vol: 48, issue: 4, pages: 807-823, 2012
Giuliani L., Gentili F., Bontempi F.:
Effects of combustible stacking in large compartments
Journal of Structural Fire Engineering, vol: 4, issue: 3, pages: 187-205, 2013
Contact
Kristian Hertz
Luisa Giuliani