Validation and Use
of Digital Image Correlation for High Temperature Deformation
Measurement in Prestressing Steel John
Gales (proposed speaker) and Luke Bisby University of Edinburgh
A potentially
important issue in structural fire engineering arises when modern
fire safety engineering calculations and design rely on experimental
results from antiquated materials testing. The mechanical properties
of prestressing steel tendons at elevated temperatures are an
example of such an issue. Contemporary prestressing steel’s
composition is different from that used several decades ago, and
contains additional alloying compounds that were not present when
historical high temperature tests were conducted on these types
of steels. With added complexity in an alloy, high temperature
creep – plastic flow – resistance may change. For
prestressing steel, this can lead to errors for predicting failure
when creep deformation (i.e. in-service tendon stress relaxation)
is overestimated based on high temperature material properties
from older steels. Unfortunately, accurately deriving high temperature
material properties for prestressing steels (up to failure at
elevated temperature) involves instrumentation limitations, failure
or even destruction. Recently, a novel strain and displacement
measurement method, termed Digital Image Correlation (DIC), has
been assessed in an attempt to alleviate these experimental pitfalls.
This technology is being used at the University of Edinburgh to
accurately quantify the high temperature mechanical behaviour
of modern prestressing steels. Good accuracy has been shown when
preliminary results have been used for predictive modelling of
prestressing steel for both steady state and transient testing.
Consequences for modelling high strength steel at elevated temperature
will be discussed.
The COMPFIRE project
covers the behaviour and robustness of practical connections
to composite columns in fire. It involves 5 European universities
and two industrial organisations. A series of short presentations
by the partners will summarise the current position at about
the real half-way stage.
Overview
of COMPFIRE - Objectives and Progress
Luis Simoes da Silva Universidade de
Coimbra, Portugal
Temperature
distribution in reverse channel connections in fire Yong Wang University of Manchester
Detailed Numerical Modelling
Milan Veljkovic, Tim Heistermann and Naveed Iqbal
Luleå University, Sweden
Preparation of a Large-Scale Demonstration
Fire Test Frantisek Wald CVUT Prague, Czech Republic
A summary of the preparation
of the demonstration fire test of project COMPFIRE, which
will take place in September 2011. Prediction of the behaviour,
and the design of instrumentation including the selection
of thermocouples, strain gauges, deflectometers, and radiometers
will be described.
Munich
Fire Tests on Membrane Action Martin Mensinger and Martin Stadler Technische Universität München, Germany
In Great Britain and
Switzerland membrane action is already used to design composite
beam-slab systems in fire. To ensure that it also can be used
in Germany further investigations are required. Available
design methods need to be adapted to German design rules,
and some remaining issues must be clarified. For this reason
a research project was initiated by the Technische Universiät
München and the Leibniz Universität Hannover.
Main objective of the
project is the behaviour of the intermediate beams between
two adjacent slab panels. Large rotations may lead to huge
cracks in the concrete chord above the edge beams which can
reduce the load-bearing capacity of these beams. Two large-scale
fire tests have been performed in Munich in 2010 to analyse
this issue, to calibrate numerical models and to validate
analytical assumptions.
Developing
OpenSees for Modelling Structures in Fire Asif Usmani University of Edinburgh
OpenSees is an open-source
object-oriented software framework developed for modelling
framed structures subjected to seismic loading. Work is progressing
at the University of Edinburgh on adding a "structures
in fire" modelling capability to the OpenSees framework.
This includes:
a fire
module;
a complete
2D/3D heat transfer module to determine temperatures in
structural members which takes boundary heat flux information
from the fire module (or directly provided by the user);
a
thermo-mechanical module for calculating structural response
to heating.
The current status of
the work will be reported to StiFF with a view to obtaining
useful feedback in further developing this work.
Comparison
of Resulting Steel Temperatures from Travelling Fires and
Traditional Design Methods - Case Study Guillermo
Rein & Anna M Jonsdottir University of Edinburgh Jamie Stern-Gottfried Arup Fire, London
This paper calculates
the steel temperatures resulting from a novel methodology
based on travelling fires to determine the design fire of
large enclosures that are outside the range of applicability
of current design methods. The methodology is applied to The
Informatics Forum, a modern building in the city of Edinburgh
finished in 2008 and which is seemed representative of complex,
modern architectural trends and with large enclosures. Analysis
of three different steel beams, protected and unprotected,
and the results compared to those from traditional design
methods (standard and parametric fire curves). The results
indicate that more severe conditions are predicted for small
travelling fires of size 5 to 10% of the floor plate than
for uniform fires assumed in traditional methods. Compared
to the parametric fire curve, travelling fires lead up to
95% higher steel temperatures for unprotected steel, and up
to 110% for protected with 39mm-gypsum. Traditional methods
are more conservative only when compared to travelling fires
larger than 85% of the floor plate.
(Due to problems with flights, only 3 of the scheduled
6 presentations were given)
The New Zealand
Slab Panel Method for tensile membrane action design of composite
slabs for fire
John Dowling Corus Ltd
(on behalf of Charles Clifton,
University of Auckland, New Zealand)
Download presentation slides
and notes (PDF
(2Mb))
Characterisation
of intumescent coating performance for performance-based design
Yong Zhang University of Manchester
Behaviour
of Fin-Plate Connections to CFT Columns at Ambient and Elevated
Temperatures Under Shear and Bending Load
Mark Jones University of Manchester
An
Approach for Modelling Membrane Action and Fire Tests on Composite Slabs, Focusing on the Behaviour
of Edge Beams
Martin Mensinger TU München, Germany
Membrane
Action of Concrete Floor Slabs at Ambient and Elevated Temperature Bok Man Chan University of Manchester
The concept of membrane action has been adopted to explain the
ultimate load carrying capacity of floor slabs above the traditional
yield-line load at ambient and elevated temperatures. However,
the detailed distribution of membrane forces within the concrete
and reinforcement components of the slab has not been fully investigated.
This research project extends the investigation of membrane action
in concrete slabs to fully understand the development of the internal
forces. A finite element model has been developed to simulate
horizontally unrestrained, isotropic reinforced, concrete slabs.
Both shell elements and solid elements were used at ambient temperature
and solid elements were used at elevated temperature to accurately
represent the highly non-linear temperature distribution through
the depth of the slab. The models have taken into account the
smear crack concept with thermal, geometric and material nonlinearity.
Two types of smooth bars, mild steel reinforcement and stainless-steel
reinforcement, with different diameters and spacing of bars in
rectangular and square slabs were simulated. The effect of the
membrane action within the slabs due to different reinforcement
ductility, reinforcement content and aspect ratio of the slab,
were also investigated. Numerical results were validated against
a number of small-scale slab tests carried out at the University
of Manchester and were shown to provide very good predictions.
It has been concluded that the distribution of the in-plane stress
and through-depth stress of the slabs, subjected to large displacement
at ambient temperature is similar to the slabs at elevated temperature.
The magnitude is different due to the different applied load and
weakening of the material due to elevated temperatures. Numerical
studies show clearly the need to model the tension softening curve
of the concrete with an accurate estimate of the fracture energy
and maximum tensile strength. It is expected that the conclusion
of this research project will be used for further development
of design codes and guidance.
Assessing
the Safety of Tall Buildings in Multiple Floor Fires David Lange, Charlotte Roben and Asif
Usmani University of Edinburgh With increasing acceptance and use of performance based design
methodologies, the question of the safety of tall buildings in
a multiple floor fire will sooner or later need to be addressed.
Particularly so in cases where the occurrence of such a scenario
is deemed to have an unacceptably high probability. A simple stability
assessment method for tall buildings subjected to multiple floor
fires is proposed. This method would allow structural engineers
to obtain an indication of the upper bound collapse mechanism
for tall buildings in multiple floor fires without the need to
use complex and labour-intensive computer models. It can also
be used to provide bounds on the results of a more comprehensive
computational analysis and aid the interpretation of the output,
thus helping to avoid errors. This method is strictly only applicable
to tall buildings which can be idealised as two dimensional frames.
However as many tall buildings tend to be rather regular in plan,
the method could be used quite generally, at least for preliminary
assessment.
Modelling
Intumescent Coating Behaviour in Fire
Jifeng Yuan University of Manchester
This research aims to provide a robust scientific footing to model
intumescent coating performance under various fire exposure conditions,
to help develop fire engineering design methods suitable for application
under different fire conditions. A mathematical model has been
established to investigate the fire protection performance of
intumescent coatings. In the model, once some basic chemical characteristics
and physical properties of the intumescent coating has been quantified,
the fire performance of this kind of intumescent coating exposed
to different external heat conditions (Cone-calorimeter, standard
fire, parametric fire, etc) can be determined. In the mathematical
model, the governing equations are based on energy and mass conservation,
and the model is solved by using the Finite Difference Method
(FDM). Heat transfer, heat generation (from chemical reaction),
and mass transfer (production of volatile gases and their movement)
models form the basis of the model. Arrhenius equation is used
to describe chemical kinetics of coating decomposition (typically
including three parts: melting, blowing, and charring), which
in turn is responsible for expansion. An important aspect of this
research is to develop and examine a thermal conductivity model
for the expanding porous intumescent coating foam, which should
not only include pure conduction in air, but also radiation through
the voids of the porous foam. The main objective of this research
is to identify the main fundamental parameters that are responsible
for describing the intumescence performance with sufficient accuracy
under different fire conditions.
The
Suitability of Fire-Field Modelling for Enclosure Fires involving
Complex Solid Fuel Loads Stuart
Winter University of Manchester Research
on the use of Computational Fluid Dynamics to model closed car-park
and compartment fires is presented. The Fire-Field model FDS (Fire
Dynamics Simulator) is used to simulate closed car-park fires
and successful comparisons with simulations performed as part
of the European Commission report on such fires are made. The
car fire assumptions used in these simulations, and consequently
the conclusions drawn from their output, are critically assessed.
An alternative modelling approach for car fires in fire-field
simulations is outlined by the author. The simulation of full-scale
compartment fire tests using FDS is also considered and a model
representing the important processes of wooden crib burning is
presented. The practical application of such models is discussed.
3D
Interpolation Method for the Determination of the Contribution
of Intumescent Coatings to the Fire Resistance Performance of
Structural Steelwork Protected with Intumescent Coatings Hans van de Weijgert International Fire Consultants
Technical
Update on the Torre Windsor Fire, Madrid Roger Pope BCSA
(The
slides contain currently restricted material, and cannot be downloaded
yet.)
Connections
of Unprotected Steel Members to Fire Walls Alexander Heise Arup Fire It is common practice to design single storey warehouse or
industrial structures for fire, with very limited or no fire protection.
Structural failure of these forms of single storey buildings in
a fire situation is tolerated, traditionally based on the low
risk to the life safety of building occupants these structures
are deemed to create. This originates from a typically limited
occupancy with multiple means of escape, and as such a rapid evacuation
time. However traditionally the risk to the life safety of fire
fighters has not been considered in the design, and this is resulting
increasingly in external and therefore limited fire fighting in
these forms of structure. Therefore the concept of property protection
and business continuity through protection of contents has become
an increasingly important concept in design. It is then essential
to maintain compartmentation within these large volume spaces
in order to limit fire spread and resulting damage to the overall
contents of the building and the extent of structure.
What this means is that
the large fire walls provided to sub-divide these space, cannot
be damaged when the unprotected steel structure fails. As a consequence,
the unprotected steel structure is usually erected independently
from the fire wall, preventing the use of the load bearing capacity
of the fire wall in the structural design concept. This results
in the need for additional columns and bracing. This paper presents
an alternative concept to this standard approach allowing connections
of unprotected steelwork to fire walls. Two different design concepts
for these connections are presented, which prevent damage to the
fire wall when unprotected steel members connected to the fire
wall fail. These two design concepts can be combined enabling
the designer to find the most cost effective solution for the
connection.
The first option is a design
concept based on maximum load bearing capacity of bolts. The maximum
reaction forces limited by the load bearing capacity of bolts
are calculated using design values for the upper limit of the
load bearing capacity of the bolts. A complete design concept
will be presented including combination values for additional
loads and design values to determine the resistance of the fire
wall in the fire situation. The statistical parameters of the
bolts determined on the basis of 3000 tests collated from literature
and bolt producers will be explained. The statistical parameters
for additional loads and resistance will be described in detail.
The design values for the maximum load bearing capacity of the
bolts and the design resistance for the fire wall is determined
using the FORM-Method. The determination of combination values
considering additional loads in the fire situation using the Borghes-Castanheta
method, will also be discussed.
Finite
Element Analysis of Concrete Filled Steel Columns in Fire Jun Ding University of Manchester Concrete filled steel columns are widely used all over the
world, due to their significant advantages in construction speed
and high fire resistance. This paper presents the results of a
finite element study to simulate the fire performance of concrete
filled steel columns exposed to the standard fire condition. The
current structural fire resistant design standards such as EC
4 Part 1.2 and BS 5950 Part 8 provide simplified and approximate
methods to calculate the fire resistance of these members, but
their range of application is limited and there is still a lot
of uncertainty over their accuracy. Finite element analysis offers
the possibility of accurately simulating the behaviour of this
type of columns, including non-uniform temperature distributions
and geometric and material non-linearity in fire. The objectives
of this study are to validate the finite element simulations for
both heat transfer and structural analyses, both being conducted
using a general finite element analysis package ANSYS. The temperature
distribution within a composite column cross section is analysed
by using a 2- D model, while a 3-D model is used to simulate the
structural behaviour of the column in fire. The results from the
finite element analyses were compared with a number of fire test
results and were found to agree well with the available test results.
A numerical parametric study was carried out to investigate the
sensitivity of simulation results to different assumptions with
regard to the effect of initial displacement, meshing of the cross-section
and concrete stress-strain relationship at elevated temperatures.
Quantitative
comparison of Fire Dynamics Simulator and parametric fire curves
with compartment fire test data
Nicholas Pope University of Manchester
This presentation gives a comparison of two parametric fire modelling
techniques (Eurocode 1, and the BFD curve method) and one field
model (Fire Dynamics Simulator) against large scale post-flashover
test data. Using a method of the product moment correlation coefficient,
it is shown that the BFD curve predictions are most closely representative
of reality. For the computational test data, two grid resolutions
are adopted in the FDS field model, the finer of which having
comparable results in terms of regression analysis to the BFD
method. Both the field model and the BFD curve method were found
to give better predictions compared to the Eurocode method over
the duration of the test. However, a direct comparison of the
maximum gas temperatures shows the field model to be poorer in
its predictive capability than the parametric methods, under-predicting
the maximum gas temperatures. In addition, a more in-depth analysis
of the FDS predictions indicates that by considering simply average
compartment temperatures the more inaccurate spatially-specific
temperature predictions were disguised. This study provides useful
quantitative data on the three techniques presented and discusses
more general issues concerning fire modelling.
Analysis
of reinforced concrete buildings In fire
Zhaohui Huang University of Sheffield A non-linear finite element, developed for three-dimensional
modelling of beam-column elements of general cross-sections in
fire conditions, will be described. Because of the changes in
material properties and the large deflections experienced in fire,
both geometric and material non-linearities are taken into account
in this formulation. The cross-section of the beam-column is divided
into a matrix of segments, and each segment may have different
material, temperature and mechanical properties. The more complicated
aspects of structural behaviour in fire conditions, such as thermal
expansion, cracking or crushing of concrete, and progressive change
of the constitutive properties of materials with temperature,
are modelled. Since it is possible to offset their nodes by pre-determined
distances the elements can easily be combined with shell or plate
elements to model concrete floor systems in fire. A high-deflection
numerical example using linear elastic material is presented to
demonstrate the accuracy at high deflections of the elements.
A full-scale standard fire test on a composite slim-floor beam
has been modelled to show the capabilities of the element. Finally
detailed 3D analyses of a reinforced concrete structure subject
to standard fire conditions are carried out.
Global
effects of fire on long span, truss based floor systems used in
multi-storey buildings Graeme Flint University of Edinburgh
Fire
safety on board ships: Lessons from the past - systems for the
future
Hans van de Weijgert International Fire Consultants Shipowners and shipyards, electrical engineers, naval architects
and classification societies are all confronted with fire safety
aspects of cable and pipe transits. Trouble free maintenance,
low cost installation, lower overall costs, protection of human
life and international regulations often turn out to be conflicting
as it comes to fire safety. And what about fire safety during
the construction process itself? The recent fire on board the
Diamond Princess at a yard in Nagasaki shows that fire safety
deserves attention throughout the lifetime of the ship –
from the moment the ship's design is on the drawing-board right
through till its final voyage.
This lecture focuses on
the principles behind type approval certificates for fire safety
devices such as cable and pipe transits. Based on these regulations
and on the lessons we have learned from daily practice, new technologies
and systems have been developed that fulfil the demands of all
parties.
An
Experimental Study of the Compressive Performance of Structural
Panels with Cold-Formed Thin-Walled Perforated Steel Channels Bashar Salhab University of Manchester
Design
of composite columns under high temperatures with special consideration
of imperfections
Anja Urbach TU Darmstadt Composite columns use the insulation properties of the concrete
to reach high fire resistance without additional fire proofing.
The design of these columns for a fire is influenced by the imperfections
due to uneven distribution of thermal stresses and geometric inaccuracies.
The first
problem is to find the temperature distribution over the cross
section, assuming that this distribution is constant over the
columns length. This can be calculated with normal FEM software.
Based on
the exact material laws for steel and concrete under temperatures
up to 1000 degrees Celsius the second step is made. By variation
of the curvature and the strain of the centre line, incorporating
the material laws and adding the thermally induced stresses the
interaction between moment, curvature and normal force is calculated.
The axial
buckling load is defined as the intersection point of the strain
dependent Euler buckling load calculated with the tangent modulus
of the stress-strain relationship and the course of the resistance
to axial compression of the total cross section against the strain.
The calculation
of the moment-curvature relationship gives also the stiffness
distribution over the column length which will allow calculating
the column’s collapse load by a second-order analysis including
geometrical imperfections.
The calculations
are based on the advanced calculation models of the Eurocode 4
Part 1-2
Tensile
membrane action of thin, lightly reinforced concrete floor slabs
under large deflection and at elevated temperatures Samantha Foster University of Sheffield
Numerical
simulation of composite floor slabs under fire condition Can Tesar
ETH Zurich In-plane forces in reinforced concrete slabs are caused by
large deflections. Under fire condition this membrane action increases
the load-carrying capacity of composite floor slabs dramatically.
A tailor-made finite element program \textit{SlabFem} is being
developed to
investigate the structural behavior and the load-carrying-mechanism
of heated composite floor slabs. The basic concepts of the program
- including the material laws and the elements models - are shown.
The numerical modeling of the shear connection between beam and
slab elements is presented in detail.
The
seventh Cardington full-scale fire test on the composite frame Tom Lennon
BRE
Performance-based
structural fire engineering design - a sceptic's view David Fletcher
NHBC
Fire
Resistant Design Criteria of Tall Building Structures
Y C Wang Manchester Centre for Civil and Construction Engineering
This paper argues that the structural fire resistant design criteria
of tall building structures should be considered differently from
those embedded in the current design philosophy where only stability
of the structure in fire is considered. The potential repairability
issue should be addressed at the design stage by conducting a
potential risk assessment. This risk assessment may determine
the acceptable extent of fire damage to a tall building that the
fire resistant design team should aim to achieve.
Analysis
of Catenary Action in Steel Beams Under Fire Cocnditions Yingzhi Yin Manchester Centre for Civil and Construction Engineering
This presentation investigates the large deflection behaviour
of a steel beam under fire conditions taking into consideration
the effect of the catenary action provided by the surrounding
structures. After validating the capability of ABAQUS against
available experimental results of fire tests, a numerical parametric
study is conducted with parameters investigated including beam
span, load ratio, uniform and non-uniform temperature distributions,
different levels of axial and rotational restraint at the beam
ends and the effect of lateral torsional buckling on axially restrained
beams. It is concluded that the existence of horizontal axial
restraint can significantly affect a beam's survival temperature
in fire. The development of catenary action in a steel beam at
large deflection can help the beam to hang to the surrounding
structure and prevent the runaway deflection. Clearly, fire engineering
design will need to consider the effect of axial forces in the
beam on the adjacent structure. A simplified hand calculation
method is then developed to model the beam behaviour at elevated
temperatures, which may be adopted as a design calculation method.
The proposed method can be used to predict the mid-span deflection
and the catenary force of a heated steel beam, which are the main
concerns of steel structure fire engineering design.
CFD
Modelling of Large Compartment Fires Nicholas Pope Manchester Centre for Civil and Construction Engineering
The modelling of compartment fires has, over the last twenty years,
evolved from simple hand calculations to zone models, and now
to field models. The basis of field models is the area of physics
called computational fluid dynamics (CFD). Much of the validation
work done on such complex models has often been confined to the
analysis of plumes or small scale fires. The major reason for
this is in testing on the small scale the input parameters and
output measurements are able to be measured relatively accurately.
Using the large scale test facilities at Cardington, large scale
compartment fire data was analysed using the Fire Dynamics Simulator
- a Large Eddy Simulation CFD package developed at NIST in USA.
The results of this large scale testing were compared directly
with the equivalent results predicted by the CFD model. These
predictions need to be understood in terms of the limitations
and approximations applied not only by the user but by the underlying
physics of the model itself.
Fire
Resistance: Implications for Regulations and Standards of the
September 11 Terrorist Attacks on the World Trade Center Tom Lennon BRE Following the initial shock, horror and outrage at
the nature and scale of the attack on the twin towers on the 11th
September 2001, thoughts have turned to ways in which such an
event can be prevented in the future, or, if this is not possible,
on ways in which the consequences of such an event can be minimised.
For engineers the complete collapse of the twin towers and the
manner of the collapse has led to much reflection on the nature
and purpose of the design and regulatory process. The balance
between risk and consequence is one that all designers consider
either explicitly through a risk assessment process or implicitly
through interpretation of the regulations and design codes and
standards. The nature of the terrorist attack where the building
was subjected to firstly a large impact force followed, almost
immediately, by the combustion of many tons of highly flammable
aviation fuel is not a normal design situation. However, the consequences
of failure of the twin towers or of any tall building accommodating
large numbers of occupants in a crowded urban environment are
such that complete collapse should be avoided under all conceivable
design scenarios. If this is not possible then the building should
survive for a period sufficient to evacuate all building occupants
and those in the immediate vicinity. This paper will consider
the concept of fire performance and fire resistance in terms of
traditional prescriptive fire design and in relation to performance
based fire engineering design. Results from a series of full scale
fire tests carried out on real buildings at BRE's whole building
test facility will be used to look at the effectiveness of fire
engineering design methods.
Risk-based
Approaches in Fire Safety Engineering Craig English WSP Group
a) Absolute risk in fire
(comparison of failure probability in fire with that for accidental
loads at ambient temperature).
b) Comparative risk. Comparison
of performance of design alternatives with code compliant solutions.
c) Correlating evacuation
times with structural failure times.
A
New Approach to Specifying Fire Resistance Requirements for Regulatory
Purposes
Brian Kirby Corus Group
Application
of a new limit state design method to the Cardington tests
Neil Cameron University of Edinburgh
A new method for the design of concrete slabs has recently been
developed at the University of Edinburgh and presented at a previous
STIFF meeting. The method has been used to analyse four of the
six Cardington fire tests. In comparison with the experimental
results the theoretical results agree well. They show that in
three of the four tests there was considerable spare load carrying
capacity.
Attendance
about 40. There were 4 oral presentations, plus a General Discussion
session.
24
September 2002
Attendance
29. There were 5 oral presentations, plus a General Discussion
session.
Utilising
academic research in consultancy: a case study Barbara Lane
& Susan Lamont Arup Fire The presentation will address the collaboration between academia
(Edinburgh University) and consultancy (Arup) and the resulting
benefits to the construction industry. Brief details of the collaboration
will be provided. A detailed summary of recent project work carried
out at Arup relying on this collaboration will be developed. The
non-linear analysis of composite-steel frame construction for
specific projects will be addressed. The assumptions made when
developing the building model, plus some results and consultation
will also be presented for discussion. A way forward for progressing
the integration of such complex analysis with practical design
but particularly in the context of the approvals process will
be suggested. The intent being to initiate some debate as to a
sensible means of developing quality design solutions based on
complex analysis but in a way that a quality review by the local
authorities can take place. The main aim being to enhance the
reputation of the fire engineering profession.
Behaviour
of Restrained Columns in Fire Yong Wang Manchester Centre for Civil and Construction Engineering
University of Manchester A restrained column, forming part of a complete structure,
can have many types of structural interaction with the adjacent
structure. As a result, the loading and boundary conditions of
the restrained column in fire will be different from those at
ambient temperature. At present, only Eurocodes offer a limited
amount of design guidance on the effective length of continuous
columns in fire. The objectives of this research, supported by
the EPSRC, are to provide experimental information to evaluate
how bending moments and effective lengths of restrained steel
and concrete filled columns change under fire conditions. In total,
34 fire tests, including 18 tests on restrained steel columns
and 16 tests on restrained concrete filled columns, have been
carried out. Parameters investigated include type of beam to column
connection, level of column axial load and amount of initial column
bending moment. The results of this investigation suggest that
for fire limit state design, the design bending moment of a restrained
column may be taken as that caused by the reactions of the connected
beams acting eccentrically. The results of this study also validates
the design recommendation on steel column effective length in
Eurocode 3 Part 1.2. For concrete filled composite steel and concrete
columns, the design recommendation on column effective length
in Eurocode 4 Part 1.2 is acceptable if there is no local buckling
in the steel tube. However, if local buckling occurs, the position
of local buckling should be considered as a pin. Further research
is necessary to determine when and where local buckling occurs.
Behaviour
of Pre-cast Hollow-Core Slabs in Fire Tom Lennon BRE Download presentation slides (PDF
0.7Mb)
A
New Method to Estimate the Ultimate Load Capacity of Composite
Floors in Fire Neil Cameron University of Edinburgh A new method for determining the ultimate load capacity of
composite floors systems in multi-storey building fires is presented.
Collapses of composite frame buildings in fire are an extremely
rare event therefore the manner of structural failure is not well
understood. One of the main contributions to the robust performance
of such structures is due to the tensile membrane mechanism in
the composite deck slab. The research group at Edinburgh has discovered
that the development of tensile membrane mechanism in fire is
much more reliable relative to ambient conditions. This is because
a large amount of thermal strain allows composite floor systems
to assume highly deflected shapes while limiting the magnitude
of damaging tensile mechanical strains, thereby retaining the
ability to carry loads for much longer. If the load capacity available
through the tensile membrane mechanism could be reliably quantified,
then it may not be necessary to apply protection to all beams
leading to significant cost savings.
Component-Based
Steel Connection Model Ian Burgess University of Sheffield
This presentation reports on the experimental furnace testing
and development of simple analytical models intended to initiate
the development of a Component Method for modelling of steel beam-to-column
connections in fire conditions. The main objective was to investigate
the behaviour of tension and compression zones of end-plate connections
at elevated temperatures. A series of experiments was carried
out. Simplified analytical models of the component behaviour have
been developed, and validated against the tests and detailed finite
element simulations. The simplified models seem very reliable
for this very common type of joint. The component models developed
have produced moment-rotation curves which correlate well with
the results of previous furnace tests on complete connection behaviour
in fire. The principles of the Component Method can be used directly
in either simplified or finite element modelling, without attempting
to predict of the overall joint behaviour in fire. This will enable
semi-rigid behaviour to be taken into account in the analytical
fire engineering design of steel-framed buildings, for which it
is inadequate simply to consider the degradation of the ambient-temperature
moment-rotation characteristics of a joint without taking account
of the high axial forces which also occur.
Attendance
22. There was one oral presentation, plus a General Discussion
session.
On
the application field of OZone V2 Jean-François Cadorin Université de Liège The computer code OZone V2 has been developed to help engineers
in designing structural elements submitted to compartment fires.
The code is based on several recent developments, in compartment
fire modelling on one hand and on the effect of localised fires
on structures on the other hand. It includes a single compartment
fire model that combines a two-zone model and a one zone model.
It takes into account the localised effect of a fire on a structural
element with the help of the Hasemi's model. It is thus a pre-
and post-flashover model. It calculates the temperature of a steel
section submitted to that compartment fire and, finally, evaluates
the fire resistance of simple steel elements, according to EC3.
The methodology implemented
in the tool OZone V2 is first briefly presented. Some important
sub-models are also described (combustion model, wall model).
Some limitations of the
zone model approach are first quoted and some improvements made
in OZone are explained.
Comparisons of the code
with full-scale fire tests are then presented. These comparisons
enable to find some limits to the application field of the code.
A particular emphasis will be given on large openings and on burning
rates.
General
Discussion : Collapse of the World Trade Center Towers
The following abstract was
sent by Ali Nadjai of The University of Ulster, who was unfortunately
unable to be present on the day:
Ali
Nadjai University of Ulster
The structure integrity of the World
Trade Center depends on the closely spaced columns around the
perimeter. Lightweight steel trusses span between the central
elevator core and the perimeter columns on each floor. These trusses
support the concrete slab of each floor and tie the perimeter
columns to the core, preventing the columns from buckling outwards.
The buildings
survived the impact of a plane, although a significant number
of perimeter columns on several floors of the building were completely
destroyed, but failed under fire. The high temperature fires from
burning aviation fuel and plane debris softened the steel core
of the structure. Eventually, the loss of strength and stiffness
of the materials resulting from the fire combined with the initial
impact have caused the collapse of the towers.
Codes and
their supervision need urgent investigation and updating, for
the safety of all those who live and work in tall buildings, both
here and around the world.
Attendance
21. There were 3 oral presentations, plus one extra contribution
during the Discussion session.
Update
on Codes and Standards Roger Pope BCSA A review of the current situation on development of codes
and standards relevant to steel construction in fire.
A
Comparison of Structural Behaviour in Response to a Short-Hot
Fire and a Long-Cooler Fire: Part II Susan Lamont University of Edinburgh At the last STIFF meeting we presented the differences in
structural behaviour of a small generic composite steel frame
in response to two fires. This highlighted a new phenomenon in
the primary beam of the frame, which could not be explained adequately
at the time. The initial behaviour was as expected with axial
compressions developing in all beams as the heating regime progressed
and beams expanded against supports. However in both fire scenarios,
but at two different temperatures, there was a sudden sharp reduction
in compression in the primary beams accompanied by a sharp increase
in deflection of the composite floor. This phenomenon was not
noted in the modelling of any of the Cardington tests. This presentation
will describe this behaviour further and attempt to explain the
reasons behind this apparent `instability' and its effect on the
surrounding structure.
A
Simplified Model of Heated Steel Beams at Large Deflection Ahmed Allam Arup Fire This presentation introduces a simplified model for a heated
steel beam at large deflection taking into consideration the effect
of the catenary action provided by the surrounding cold structure.
This model can be used to predict the mid-span deflection, and
the tensile axial force induced by catenary action, of a heated
steel beam. This tensile axial force at large deflection can lead
to integrity failure, and consequent fire spread, if sufficient
strength and ductility are not designed into key elements such
as beams and connections. However, provided that this can be done
in particular cases, then it may be possible to prove much greater
real fire resistance than is estimated by the simple calculation
methods currently used. The study highlights the effect of the
axial horizontal restraint which can be provided by the surrounding
cold steel frame on the behaviour of the heated beam at large
deflection. However for heated steel beams in steel construction
catenary action, which may help the beam to hang to the surrounding
structure, can prevent the runaway deflection when the tensile
axial force of the beam has been overcome. The prime objective
of this study is to study computationally and analytically how
different levels of restraint from surrounding structure, via
catenary action in beams, can affect the survival of steel framed
structures in fire.
Mechanical
Properties of Grade 8.8 Bolts at Elevated Temperatures Buick Davison University of Sheffield
Rotationally
Restrained Steel Columns in Fire Faris Ali
University of Ulster The presentation discusses the outcomes of the experimental
investigation on the performance of rotationally restrained steel
columns in fire. Half scale steel columns (within steel frames)
where tested in fire under two values of rotational restraint
0.18 and 0.93. Each case of rotational restraint was tested under
four loading levels 0.2, 0.4, 0.6 and 0.8. The presentation discusses
the test methodology and the main results and compares the results
with a case of axially only restrained columns in fire. A method
of calculating the effective length of fixed end columns is also
discussed.
Download presentation text
and figures (PDF
0.18Mb)
Behaviour
of Steel Beam-to-Column Joints: Test Results of the Compression
Zone at Elevated Temperatures Spyros Spyrou
University of Sheffield This presentation is an update on a test program being carried
out at the University of Sheffield, which investigates the behaviour
of steel beam-to-column joints at elevated temperatures and seeks
to use the component approach to model the overall structural
response. The behaviour of the compression zone is under investigation
and a comparison of the experimental results with a simplified
mathematical model is presented.
Theoretical
Approach for Concrete Masonry Walls Subjected to Fire Ali Nadjai
University of Ulster Use of theoretically based methods has gathered momentum in
the last two decades, especially with the advances in computer
technology. Computing power enables detailed evaluation of a dynamic
transient process. The aim is to generate the fire process and
develop a representative computer model. With sound knowledge
of material properties and the effects of temperature, then computer
models are capable of recreating the effects of material degradation
and subsequent thermo-structural response. Analysis can be very
complex and specialised, and designers may be unable to understand
the intricacies of tools such as finite element. Unless thermal
effects on material properties are thoroughly understood then
computer models will be flawed. For example before Anderberg [1]
no one understood the significant influence of transient strain
on concrete response, theoretical models prior to this time therefore
did not truly represent total strain response. Even today there
is still very limited data available on material property response
with temperature. One must also understand the effects of interaction
of various parameters. The simultaneous combined effects of load
and temperature acting in concrete specimens produce significantly
different response than the additive effects applied separately.
This paper describes the
development of one such approach, a mathematical model to predict
the deflection response of axially loaded concrete masonry walls
in fire situations. To validate the theoretical model a comprehensive
suite of experimental tests was conducted on walls of varying
slenderness ratio and different load. The methodology used for
the experimental testing was unique in that ½ scale walls were
used, requiring an enhanced fire curve to generate thermal similitude
in model and prototype. The theoretical model considers internal
equilibrium of axial forces and moments and generates an equivalent
linear strain and associated strain gradient (curvature) from
total strains through the wall cross section. The effect of load
and load eccentricity are incorporated by computing linear stress
distribution though the section using conventional structural
mechanics. The combined effects of temperature and stress are
considered at discrete points through the wall and are integrated
through the section to compute the resultant wall curvature and
subsequent deflection. The model under pins the significant influence
of Load Induced Thermal Strain (LITS) on deflection response.
A
Comparison of Structural Behaviour in Response to a Well-Ventilated
and an Under-Ventilated Fire Susan Lamont
University of Edinburgh During fires in highly redundant composite steel framed buildings
the structural behaviour is dominated by the thermal regime. This
has been highlighted in numerous publications by researchers in
recent years (Huang et al 1999, Rotter et al 1999, Usmani et al,
2001).
This presentation describes
the impact of a well-ventilated fire and an under-ventilated fire
on a generic composite steel frame building. Well-ventilated fires
will reach high temperatures but the duration of the post-flashover
phase will be short. In contrast under-ventilated fires will achieve
much lower maximum temperatures but the duration of the post-flashover
phase will be notably longer.
When exposed to these fires
the structure will respond in two distinct ways both thermally
and structurally. During the well-ventilated fire the unprotected
steel reaches temperatures similar to the fire atmosphere. The
lightweight concrete because of its much lower thermal conductivity
and the short fire exposure reaches temperatures only marginally
higher than at ambient. Thus a high gradient exists over the depth
of the composite section. In contrast the under-ventilated fire
(extended post-flashover duration) allows the concrete slab much
longer to respond to the heating regime. In this second scenario
both the steel and the concrete achieve considerable temperatures
and the mean overall temperature of the composite section is high.
Therefore, we have two distinct situations, one in which there
is a high gradient leading to structural performance influenced
by thermal bowing and a second scenario where the mean temperature
of the composite section is high and the behaviour of the structure
is dominated by thermal expansion effects.
This presentation looks
at a composite generic steel frame designed to comply with EC4
Part 1.1. Petterson’s (1976) temperature-time curves are adopted
to describe the two fires. Heat transfer to the concrete and protected
edge beams is calculated using the adaptive finite element heat
transfer code HADAPT (Usmani and Huang 1994, Lamont et al 2000)
and unprotected steel temperatures are calculated using the appropriate
equations in EC3 Part 1.2.
The behaviour of the frame
is analysed during a well ventilated and an under ventilated fire
and the key features of response are highlighted. Some preliminary
conclusions are made about the most detrimental heating regime.
Modelling
of the Cardington Demonstration Fire Test using ABAQUS/Explicit David O'Callaghan
Corus STC This presentation covers some of the modelling issues and
incorporates some of the understanding gained from previous tests
to carry out a simulation of the office fire demonstration test.
This test differed somewhat from previous tests and included such
differences as heated edge beams, a two-bay fire compartment and
different fire loading. Results are presented from a number of
simulations with various modelling assumptions, with each showing
differences in behaviour.
Testing
of Restrained Columns in Fire Beilei Hu
University of Manchester The presentation will address an experimental programme for
restrained columns in fire, which is funded by EPSRC. It involves
the types and dimensions of tested specimens, physical modelling
of boundary conditions and the details of the furnace. It will
also discuss the possible numerical modelling which will be carried
out at a later stage of the project based on the experimental
studies.
Natural
Fire Safety Concept: Full Scale Tests, Implementation in the Eurocodes,
and Development of a User-friendly Design Tool Tom Lennon
BRE, Garston The purpose of this presentation is to summarise BRE's role
in an ongoing research programme to develop a new concept for
design based on natural fire behaviour. The object of the research
is to move away from a reliance on the standard fire curve to
a more rational approach taking into account all the influences
on fire growth and development. The project is due to finish shortly.
This is the second phase of a European collaborative project funded
by the ECSC. The global approach to fire safety was formulated
in the first phase (NFSC1) and a one-zone model (OZONE)
was developed to predict fire behaviour and heat transfer to structural
members. However, no tests were carried out during this first
phase to validate the design approach. BRE's role was to carry
out a series of large-scale tests to look at post-flashover behaviour.
Complementary experimental programmes were carried out by other
partners to investigate localised fire behaviour and the reliability
of active measures (sprinklers). The output from the research
will be a validated model for predicting compartment fire behaviour,
and the implementation of the results into the Eurocodes.
Modelling
Studies of the Cardington BS Corner Test Martin Gillie
University of Edinburgh
The
Influence of Surrounding Cool Structure on the Behaviour of Composite
Buildings in Fire Zhaohui Huang
University of Sheffield Of the six large-scale fire tests carried out on the full-scale
composite building at BRE Cardington, the Restrained Beam Test
was subject to very high restraint from the surrounding cool structure,
while the BRE Large Compartment Test was subject to no restraint
at all. Comparing these two tests it can be found that the influence
of surrounding cool structure on the structural behaviour of the
fire compartment is important and should be investigated in more
detail. Because the cost of such tests is very high it is becoming
increasingly important to have analytical methods which can predict
the behaviour of structures when subjected to fire conditions.
In this paper a typical composite steel framed building is set
up using VULCAN and a series of analyses are carried out based
on several locations of fire compartment, in order to get a better
understanding of the interactions between cool and hot structure.
A number of different fire scenarios, for example the Standard
Fire and typical natural fires, are applied. The results are compared
with Eurocode 4 limits, and some conclusions on current fire engineering
design codes for composite structure are highlighted.
Investigation
of Tensile Membrane Action in Composite Slabs in Fire Using aGrillage Analogy Ahmed Allam
University of Sheffield It has become apparent recently that the ultimate survival
of composite framed buildings in fire is to a large extent controlled
by the ability of the concrete slabs covering the fire compartment
to sustain their loading when acting as tensile membranes at very
high deflections. These occur when temperatures are very high,
so that exposed steelwork has lost almost all of its strength.
The membrane action is a geometrically non-linear behaviour whose
exact nature depends on the supporting conditions which can be
maintained around the boundaries of the fire compartment, and
on the in-plane restraint imposed by surrounding structure. In
the development of design principles based on the ultimate maintenance
of compartmentation against fire spread it is important that this
form of structural action, and its modelling, should be fully
understood, so that it can be designed-in to buildings as part
of an integrated fire engineering design approach. The simplest
level is to model the slab as a beam grillage using non-linear
beam elements. Initially, simply supported and fixed-edge slabs
are studied. This is extended to heated slabs with different levels
of horizontal restraint stiffness from adjacent structure, and
to multi-bay slab systems. Finally a comparison is made of these
simplified grillage systems and high-deflection slab behaviour.
Behaviour
of Steel Beam-to-Column Joints: Test Results of the Tension Zone
at Elevated Temperatures Spyros Spyrou
University of Sheffield This presentation is an update on a test program being carried
out at the University of Sheffield, which investigates the behaviour
of steel beam-to-column joints at elevated temperatures. Using
the image acquisition and processing technique a total of twenty-five
tests, within the tension zone, and a comparison with a simplified
mathematical model will be presented.
3 January, 2012
