|Structural Fire Design: Travelling Fires Versus Traditional Design Fires|
Structural Fire Design: Travelling Fires Versus Traditional Design Fires
By Egle Rackauskaite & Guillermo Rein
Structural fire engineering design tools available in current standards (e.g. standard and parametric fires) are based on small scale fire tests (<100 m2) and assume uniform fire conditions in the compartment. That is, they assume the involvement of full compartment in a fire and homogeneous temperature distributions. While this assumption may be suitable for small enclosures, in large open-plan compartments fires tend to spread along the compartment burning over a limited area at any one time [1,2]. These fires are referred to as travelling fires. Accidental events where fires were observed to travel include those at the World Trade Centre Towers 1, 2 & 7 (2001) and Windsor Tower fire in Madrid (2006). Such fires are not considered in current standards.
To account for travelling fires Dr G. Rein while at the University of Edinburgh during 2007 - 2011 led the team who developed the new design concept of the Travelling Fires Methodology (TFM) [1,2]. The most recent version of the TFM now developed at Imperial College London is referred to as the Improved Travelling Fires Methodology (iTFM) . In the latter version TFM has been advanced to account for more realistic fire dynamics and range of fire sizes within the limits of the available experimental and accidental fire data.
TFM considers design fires to be composed of two moving regions: the near-field (flames) and the far-field (smoke) (see Fig. 1). The near-field represents the flames directly impinging on the ceiling and assumes the peak flame temperatures. The far-field model represents smoke temperatures which decrease with distance away from the fire due to mixing with air. Any structural element will experience cooler far-field temperatures which correspond to preheating and/or cooling for much longer duration than the short hotter near-field.
In recent studies on the thermal response it has been found that structural members are likely to reach higher temperatures (up to 200°C for steel rebar) when subjected to travelling fires in comparison to uniform fires. This challenges the inherent assumption behind the traditional design fires that uniform fires represent the worst case scenario. Illustration of a typical temperature distribution along the fire path with time in steel beams is shown in Fig. 2. Travelling fire sizes in the range of 10 – 25 % of the floor area generally result in the highest peak temperatures that tend to occur towards the end of the fire path.
In the terms of the structural response results from the extensive 2D multi-story steel frame analysis  indicate that travelling fires are likely to trigger structural mechanisms which might have been previously unnoticed by assuming a uniform fire. In this study 80 fire scenarios have been considered by varying fire exposure and fire location. Different fire exposures included travelling fires, Eurocode parametric curves, ISO-834 standard fire and a constant compartment temperature curve. Uniform fire scenarios were found to result in higher compressive axial forces in beams compared to travelling fires. However, for travelling fires results show irregular cyclic development of stresses, which are not observed for uniform fires. Peak beam bending moments were found to be similar for both travelling fires and uniform fires.
Another important and widely applied structural performance criteria is the deflection of the members. The results from this study showed that the rate and magnitude of the highest beam mid-span deflections depend mainly on the fire duration and not the fire type (i.e. TFM or uniform fire) for the steel frame. Short and hot fires result in faster development of deflections while long and cool fires result in larger peak deflections. On the other hand, the locations where these peak deflections occur and deflection patterns are different for TFM and uniform fire scenarios (see Fig.3).
The overall results indicate that, depending on the structural metric examined, both travelling fires and uniform fires can result in the most severe scenario. Thus, in order to ensure a safe building fire resistance design, a range of different fires including both travelling fires and uniform fires need to be considered. A fire event is a highly stochastic process by nature and these results indicate that there is no single worst case fire scenario under which a structure could be designed and deemed to be safe. Arup’s approach  notes the limitations of only using uniform fires for the design as well. Use of travelling fires in addition to uniform fires in a building design in the probabilistic approach allows a better understanding of the overall building performance subject to a range of conditions.
TFM offers a paradigm shift in the structural fire engineering of modern buildings and changes the way modern buildings are designed for fire safety. The concept has already been applied by Arup, BuroHappold, AECOM, and Trenton Fire on dozens of iconic buildings in the UK.
Egle Rackauskaite and Guillermo Rein are with the Department of Mechanical Engineering, Imperial College London, UK