Performance of Unprotected Steel and Composite Steel Frames Exposed to Fire

Type of content
Reports
Publisher's DOI/URI
Thesis discipline
Degree name
Master of Engineering
Publisher
University of Canterbury. Civil Engineering
Journal Title
Journal ISSN
Volume Title
Language
Date
2002
Authors
Wastney, Clayton
Abstract

This report examines the performance of unprotected steel and steel composite beams as a component of steel framed buildings subjected to severe fire. The question of whether thermal protection is necessary for all structural steel is asked. The behaviour of structural steel and composite beams is analysed at elevated temperatures using the non-linear finite element computer software SAFIR (Franssen et al: 2001). SAFIR is used to first study the two-dimensional behaviour of beams with theoretical support conditions, and then with more realistic support conditions by the addition of columns. Steel and composite construction are both common and popular forms of construction used around the world. Structural steel commonly has thermal protection for fire resistance, which according to full-scale experimental fires (Clifton: 2001) may not be necessary. At this stage it is not well understood how the fire resistance mechanisms of unprotected steel works, nor how the changes in material properties influence the behaviour of the composite or steel beam. It is the intention of this report to provide some detail on single span two-dimensional beam behaviour in relation to material properties, support conditions and thermal exposure. A 610 UB 101 steel beam both with, and without a 120mm thick composite concrete floor slab is exposed to three-sided heating, simulating the effects of a compartment fire. It was found that the theoretically idealised beam with supports having axial and moment restraint performed poorly compared to beams with axial restraint only. It was also found that the beam without axial restraint at one support had a run-away failure mechanism. With the addition of columns the beam had varying degrees of axial and moment restraint at the supports, causing much lower midspan deflections during the early stages of the fire. This compares well with the findings of the real fire single beam test of the Cardington fires (Clifton: 2001). It was also found that when the EC3 Proportional and EC3 Yield Limit stresses were reached in the steel section, displacements, axial force and bending moments along the section were affected.

Description
Citation
Keywords
Ngā upoko tukutuku/Māori subject headings
ANZSRC fields of research
Fields of Research::40 - Engineering::4005 - Civil engineering::400510 - Structural engineering
Field of Research::09 - Engineering::0905 - Civil Engineering::090599 - Civil Engineering not elsewhere classified
Rights
Copyright Clayton Wastney