Journal of Constructional Steel Research 62 (2006) 303–315 www.elsevier.com/locate/jcsr Effect of fire on composite long span truss floor systems Graeme Flint a , Asif Usmani a, , Susan Lamont b , Jose Torero a , Barbara Lane b a School of Engineering and Electronics, University of Edinburgh, Alexander Graham Bell Building, The King’s Buildings, Edinburgh, EH9 3JL, UK b Arup Fire, Ove Arup and Partners, 13 Fitzroy Street, London, W1T 4BQ, UK Received 3 May 2005; accepted 16 August 2005 Abstract Much research has been and is currently being done to try and accurately recreate the effects of fire on structures including the effects of redistribution within the structure and the associated changes in load carrying mechanisms. This work has mostly focussed upon the most common and simplest structural forms currently used. As such most of the research to date tends to involve Universal Beam and Universal Column sections in a grid formation with spans of 6–9 m. This paper reports on an investigation of the effects of heating on a long span truss floor system. The ABAQUS finite element package is used to model the structure including fully non-linear behaviour and thermal expansion effects. Different boundary conditions and heating regimes are investigated to understand the response of the truss members to fire. The effects of heating on the lateral restraint available from the slender floor systems to a column have also been studied. The results and analysis indicate that composite truss flooring systems may not fail suddenly. Individual member buckling seems to be a much more gradual occurrence linked to material failure and expansion based geometry change rather than sudden “failure”. c 2005 Elsevier Ltd. All rights reserved. Keywords: Tall buildings; Structures in fire; Composite construction; Long span floor systems 1. Introduction Much research has been and is currently being done to try and accurately recreate the effects of fire on structures including the effects of redistribution within the structure and the associated changes in load carrying mechanisms [1,2,14]. This work has mostly focussed upon the most common and simplest structural forms currently used. As such most of the research to date tends to involve Universal Beam (UB) and Universal Column (UC) sections in standard grid formations of 6–9 m spans. In the aftermath of the events of 11th September 2001 it has become more apparent that other structural forms need to be investigated under fire conditions. Modern tall office buildings, especially, are using structural designs that are distinctly different to the more common forms of 9 m spans of UBs on a standard column grid. Such buildings often incorporate long span, slender floor systems to maximize the lettable area without columns interfering with the space. Typical systems Corresponding author. Tel.: +44 131 650 5789; fax: +44 131 650 6781. E-mail address: asif.usmani@ed.ac.uk (A. Usmani). include cellular steel beams or trusses composite with the floor slab. This paper reports on an investigation of the local effects of heating on a long span truss floor system. Different boundary conditions and heating regimes are investigated to understand the response of the truss members to fire. The effects of heating on the lateral restraint available from the slender floor systems to a column have also been studied. The ABAQUS finite element package is used to model the structure including fully non-linear behaviour and thermal expansion effects. 2. Model description The results in this paper are based on two different structural models created and analysed using the ABAQUS Finite Element Analysis software. The first is a 3D model of a single 18.5 m long truss acting compositely with 2 m width of concrete slab above. A pinned connection into a column is used as support at one end of the truss, the other end is free to rotate but restrained from rotation (assumed to be connected to a rigid core structure). The second structural model is a 2D representation of a substantial portion of a long span composite 0143-974X/$ - see front matter c 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.jcsr.2005.08.002