Connection behaviour and the robustness of steel-framed structures in fire

The full-scale fire tests at Cardington in the 1990s, and the collapse of at least one of the WTC buildings in 2001, illustrated that connections are potentially the most vulnerable parts of a structure in fire. Fracture of connections causes structural discontinuities and reduces the robustness pro...

Full description

Bibliographic Details
Main Author: Burgess Ian
Format: Article
Language:English
Published: EDP Sciences 2018-01-01
Series:MATEC Web of Conferences
Online Access:https://doi.org/10.1051/matecconf/201814901008
id doaj-8418f1ed7cc9435f92f0389b83c6bc57
record_format Article
spelling doaj-8418f1ed7cc9435f92f0389b83c6bc572021-02-02T06:41:51ZengEDP SciencesMATEC Web of Conferences2261-236X2018-01-011490100810.1051/matecconf/201814901008matecconf_cmss2018_01008Connection behaviour and the robustness of steel-framed structures in fireBurgess IanThe full-scale fire tests at Cardington in the 1990s, and the collapse of at least one of the WTC buildings in 2001, illustrated that connections are potentially the most vulnerable parts of a structure in fire. Fracture of connections causes structural discontinuities and reduces the robustness provided by alternative load paths. An understanding of connection performance is essential to the assessment of structural robustness, and so to structural design against progressive collapse. The forces and deformations to which connectionscan be subjected during a fire differ significantly from those assumed in general design. The internal forces i generally start with moment and shear at ambient temperature, then superposing compression in the initial stages of a fire, which finally changes to catenary tension at high temperatures. If a connection does not have sufficient resistance or ductility to accommodate simultaneous large rotations and normal forces, then connections may fracture, leading to extensive damage or progressive collapse of the structure. Practical assessment of the robustness of steel connections in fire will inevitably rely largely on numerical modelling, but this is unlikely to include general-purpose finite element modelling, because of the complexity of such models. The most promising alternative is the component method, a practical approach which can be included within global three-dimensional frame analysis. The connection is represented by an assembly of individual components with known mechanical properties. Component characterization must include high-deflection elevated-temperature behaviour, and represent it up to fracture.In reality a connection may either be able to regain its stability after the initial fracture of one (or a few) components, or the first failure may trigger a cascade of failures of other components, leading to complete detachment of the supported member. Numerical modelling must be capable of predicting the sequence of failures of components, rather than considering the first loss of stability as signifying building failure. It is necessary to use a dynamic analysis, so that loss of stability and re-stabilization can be tracked, includingthe movements of disengaging members and the loadsharing mechanisms which maintain integrity and stability within the remaining structure, until total collapse occurs.https://doi.org/10.1051/matecconf/201814901008
collection DOAJ
language English
format Article
sources DOAJ
author Burgess Ian
spellingShingle Burgess Ian
Connection behaviour and the robustness of steel-framed structures in fire
MATEC Web of Conferences
author_facet Burgess Ian
author_sort Burgess Ian
title Connection behaviour and the robustness of steel-framed structures in fire
title_short Connection behaviour and the robustness of steel-framed structures in fire
title_full Connection behaviour and the robustness of steel-framed structures in fire
title_fullStr Connection behaviour and the robustness of steel-framed structures in fire
title_full_unstemmed Connection behaviour and the robustness of steel-framed structures in fire
title_sort connection behaviour and the robustness of steel-framed structures in fire
publisher EDP Sciences
series MATEC Web of Conferences
issn 2261-236X
publishDate 2018-01-01
description The full-scale fire tests at Cardington in the 1990s, and the collapse of at least one of the WTC buildings in 2001, illustrated that connections are potentially the most vulnerable parts of a structure in fire. Fracture of connections causes structural discontinuities and reduces the robustness provided by alternative load paths. An understanding of connection performance is essential to the assessment of structural robustness, and so to structural design against progressive collapse. The forces and deformations to which connectionscan be subjected during a fire differ significantly from those assumed in general design. The internal forces i generally start with moment and shear at ambient temperature, then superposing compression in the initial stages of a fire, which finally changes to catenary tension at high temperatures. If a connection does not have sufficient resistance or ductility to accommodate simultaneous large rotations and normal forces, then connections may fracture, leading to extensive damage or progressive collapse of the structure. Practical assessment of the robustness of steel connections in fire will inevitably rely largely on numerical modelling, but this is unlikely to include general-purpose finite element modelling, because of the complexity of such models. The most promising alternative is the component method, a practical approach which can be included within global three-dimensional frame analysis. The connection is represented by an assembly of individual components with known mechanical properties. Component characterization must include high-deflection elevated-temperature behaviour, and represent it up to fracture.In reality a connection may either be able to regain its stability after the initial fracture of one (or a few) components, or the first failure may trigger a cascade of failures of other components, leading to complete detachment of the supported member. Numerical modelling must be capable of predicting the sequence of failures of components, rather than considering the first loss of stability as signifying building failure. It is necessary to use a dynamic analysis, so that loss of stability and re-stabilization can be tracked, includingthe movements of disengaging members and the loadsharing mechanisms which maintain integrity and stability within the remaining structure, until total collapse occurs.
url https://doi.org/10.1051/matecconf/201814901008
work_keys_str_mv AT burgessian connectionbehaviourandtherobustnessofsteelframedstructuresinfire
_version_ 1724300776434565120