Corus Tubes

Structural & Conveyance Business

Design guide for SHS concrete filled columns

This page intentionally

blank

Design Guide for Concrete Filled Columns

Original text by :-

S J Hicks BEng, PhD (Cantab.)

G M Newman BSc(Eng), CEng, MIStructE, MIFireE

Additional text by :-

M Edwards BSc (Lond)

A Orton BA (Cantab), CEng, MICE, MIStructE

A

ii

Publication 2002 Corus Tubes Care has been taken to ensure that the contents of this publication are accurate, but Corus UK Limited and its subsidiary companies do not accept responsibility for errors or for information which is found to be misleading. Suggestions for or descriptions of the end use or application of products or methods of working are for information only and Corus UK Limited and its subsidiaries accept no liability in respect thereof. Before using products supplied or manufactured by Corus UK Limited customers should satisfy themselves of their suitability.

Main text 2002 The Steel Construction Institute

Apart from any fair dealing for the purposes of research or private study or criticism or review, as permitted under the Copyright Designs and Patents Act, 1988, this publication may not be reproduced, stored or transmitted, in any form or by any means, without the prior permission in writing of the publishers, or in the case of reprographic reproduction only in accordance with the terms of the licences issued by the UK Copyright Licensing Agency, or in accordance with the terms of licences issued by the appropriate Reproduction Rights Organisation outside the UK.

Enquiries concerning reproduction outside the terms stated here should be sent to the publishers, The Steel Construction Institute, at the address given on the title page.

Although care has been taken to ensure, to the best of our knowledge, that all data and information contained herein are accurate to the extent that they relate to either matters of fact or accepted practice or matters of opinion at the time of publication, The Steel Construction Institute, the authors and the reviewers assume no responsibility for any errors in or misinterpretations of such data and/or information or any loss or damage arising from or related to their use.

iii

FOREWORD

This publication was collated and edited at Corus Tubes, Corby from original contributions by Dr Stephen Hicks and Mr Gerald Newman of SCI and supplementary text from Mr Mike Edwards and Mr Andrew Orton of Corus. It brings together the latest available information on the design and construction of buildings using concrete filled structural hollow sections and supersedes other publications by Corus and SCI.

The Steel Construction Institute develops and promotes the effective use of steel in construction. It is an independent, membership based organisation.

SCI's research and development activities cover many aspects of steel construction including multi-storey construction, industrial buildings, light steel framing systems and modular construction, development of design guidance on the use of stainless steel, fire engineering, bridge and civil engineering, offshore engineering, environmental studies, value engineering, and development of structural analysis systems and information technology.

Membership is open to all organisations and individuals that are concerned with the use of steel in construction. Members include designers, contractors, suppliers, fabricators, academics and government departments in the United Kingdom, elsewhere in Europe and in countries around the world. The SCI is financed by subscriptions from its members, revenue from research contracts and consultancy services, publication sales and course fees.

The benefits of corporate membership include access to an independent specialist advisory service and free issue of SCI publications as soon as they are produced. A Membership Information Pack is available on request from the Membership Manager.

The Steel Construction Institute, Silwood Park, Ascot, Berkshire, SL5 7QN. Telephone: +44 (0) 1344 623345 Fax: +44 (0) 1344 622944 Email: membership@steel-sci.com For information on publications, telephone direct: +44 (0) 1344 872775 or Email: publications@steel-sci.com For information on courses, telephone direct: +44 (0) 1344 872776 or Email: education@steel-sci.com World Wide Web site: http://www.steel-sci.org

iv

This page intentionally

blank

v

Contents Page No.

FOREWORD iii

SUMMARY vi

1 INTRODUCTION 1 1.2 Applications of concrete filled columns 4

2 NORMAL DESIGN 9 2.1 General 9 2.2 Material properties 9 2.3 Partial safety factors 11 2.4 Basis of design method 11 2.5 Restrictions on the simplified design method 13 2.6 Properties of cross-section 13 2.7 Column buckling resistance 17 2.8 Analysis of bending moments due to second-order effects 20 2.9 Combined compression and bending 21 2.10 Longitudinal and transverse shear 31 2.11 Load introduction 32

3 FIRE DESIGN 36 3.1 General 36 3.2 Protected columns 38 3.3 Unprotected columns 39 3.4 Partial safety factors 40 3.5 Properties of the cross-section 41 3.6 Column buckling resistance 42 3.7 Combined compression and bending 43 3.8 Material properties 45

4 FABRICATION AND CONNECTIONS 47 4.1 General 47 4.2 Column Splices and Flange plate connections 47 4.3 Beam-to-column connections 49 4.4 Shearhead connections to slabs 52 4.5 Base plate connections 53

5 PRACTICAL CONSIDERATIONS 54 5.1 Concrete Filling 54

6 SOFTWARE 58

7 REFERENCES 60

vi

SUMMARY

A concrete filled structural hollow section provides architects and engineers with a robust and inherently fire resistance column. This publication contains design information for these columns for both the normal and fire conditions. The information is based on Eurocode 4. Also included are case studies illustrating the use of concrete filled columns and practical guidance on concrete filling and connection design.

Design software, ConcFill 2, is described which will analyse sections for the normal and fire conditions.

1

1 INTRODUCTION

Structural Hollow Sections (SHS) are the most efficient of all structural steel sections in resisting compression. Their availability in the high yield material Celsius 355 gives them a high strength to weight ratio and produces slender attractive lines that make them a natural choice for building structures. In addition, SHS can achieve a constant external dimension for all weights of a given size, which enables them to achieve standardisation of architectural and structural details throughout the full height of the building.

Celsius is the brand name for Corus Tubes hot-finished structural hollow sections, Celsius 355 being produced to the European Standard, EN 10210 S355J2H. Celsius sections are produced by the electric weld process and the J2 denotation signifies that they have a Charpy impact minimum average energy value of 27J at –20oC, making them suitable both for internal and external applications. Celsius sections are produced to the technical delivery requirements of EN 10210-1:1994[1] with dimensions and tolerances to EN 10210-2:1997[2]. However, for Celsius sections, there is an improved corner profile of 2T maximum.

By filling hollow sections with concrete a composite section is produced (see Figure 1.1), which will increase the section’s room temperature load carrying capacity, whilst retaining all the advantageous features of the basic unfilled section. Alternatively, for the same original load capacity, it permits smaller composite sections to be used. The reduction in section size also gives advantages in subsequent construction processes, including a reduced surface area for painting or fire protection. In the fire condition the presence of the concrete filling acts as a heat sink.

Concrete or grout filled hollow sections can be divided into those that are externally protected against fire by fire-rated boards, lightweight sprayed protection or intumescent coatings, and those that have no such protection. A further division can be made, by differentiating between those that are filled with plain concrete mixes and those that contain steel reinforcement within the mix.

Figure 1.1 Concrete filled square hollow section

2

Externally protected composite sections are designed compositely at room temperature and external fire protection is applied to achieve the required fire rating of the column. The composite action is maintained in the fire limit state, the external protection serving to limit the rise in steel temperature such that the column capacity is always in excess of the fire limit state design load over the required fire resistance period. In general, externally protected sections will not need to contain reinforcement in the mix in order to achieve the desired fire rating - reinforcement is usually added to such columns so to enhance axial capacity while minimising or maintaining column SHS size. Reductions in the thickness of the external protection are possible because of the heat sink effect, which effectively reduces the section factor of the column; these reductions have been shown to be substantial in the case of filled hollow sections with intumescent coatings. In cases where smaller columns are used, particularly if speed of construction is an issue, consideration should be given to sizing the SHS such as to use a plain fill with external protection, so that reinforcement can be dispensed with. Note that because of the tensile capacity of the steel in the composite column, in almost all cases, grout may be used interchangeably with concrete as the filling material.

Filled hollow sections that are not externally protected against fire are designed using the concrete core alone to meet the fire limit state load requirements but the capacity of the composite section is checked for the room temperature design case. In general, such sections will need to contain reinforcement in the mix in order to minimise column dimensions and to sustain the required fire limit state design loads for such practical fire resistance periods of 60 minutes or more.

Cost comparisons

Cost comparisons have shown that to establish the cost competitiveness of a column it is necessary to take into account both the cost of the supplied and erected steel section together with the cost of its fire protection. Such comparisons have shown that hollow sections can offer a competitive first choice solution to structural columns in multi-storey construction.

Construction

Concrete or grout filling of structural hollow sections requires no special equipment and the filling operation may be integrated into other concreting operations. The enhancement in the overall efficiency obtained by filling a steel structural hollow section with concrete allows the designer a wider choice of sections. Filled hollow section columns combine the advantages of economy in the use of materials with the construction advantages of the use of steelwork. Columns, whether externally fire-protected or not, will usually arrive to site as fully finished elements with make-ups only at column splice joints, if any. Concrete filling of the hollow section columns can take place on or off site. If filling takes place on site, then the steel column and its connections are designed to carry all construction loads so that the operation of filling the columns can be taken off the critical path. In larger buildings, the best economy is obtained by planning for the simultaneous working of different trades at different levels or plan positions. The hollow section columns may be filled from the top with a self-compacting or other type of mix; alternatively, they can be filled from the bottom, through a gate-valve, with a pumpable concrete or grout mix. Grout is often used as an alternative to concrete and this has substantial advantages at the construction stage, such as easier filling and pumping and the avoidance of unfilled voids.

3

Advantages of composite hollow sections

In the initial construction period The steel section dispenses with the need for formwork. Concrete placement and compaction in many cases are unhindered by internal reinforcement.

Erection schedule is not dependent on concrete curing time.

During finishing The concrete filling is protected against mechanical damage. Additional external fire protection is not always necessary Slender columns reduce the application time and cost of applied finishes.

Completed building • Greater useable floor area.

• Higher visibility.

• Reduced maintenance.

• Aesthetically pleasing.

Figure 1.2 Detail at the facade line of a plain filled concrete tubular column structure

that uses an intumescent coating as external fire protection (Wellcome Trust Headquarters, London; structural engineer WSP Group)

4

Many of these advantages were recognised early in the history of the use of iron and steel hollow sections in construction, indeed the first known Patent relating to the concrete filling of circular hollow sections dates from 1898. The wider use of the composite concrete hollow section did not really begin until the mid 20th Century following the results of structural investigation and the availability of a comprehensive range of structural hollow sections.

A number of institutes in Britain and overseas have now undertaken considerable research into the structural and fire resistance performance of concrete hollow sections aimed at developing suitable design procedures for this form of construction. The results of this recent research are now incorporated into prEN 1994-1-1: 2002 (EC4-1-1), the European code for composite construction[18].

The material included in this manual makes the design of structural hollow section columns filled with concrete simple and rapid when standard sections are used in conjunction with concrete of common grades.

1.2 Applications of concrete filled columns 1.2.1 FLEET PLACE HOUSE, LONDON The site of Fleet Place, off Holborn Viaduct in the City of London was comprehensively redeveloped at the same time as the adjacent Thameslink station. This eight-storey high office block on the site was built using concrete filled external CHS columns on each longitudinal face of the building and has clear spans on the inside. So that they could be supported off existing pilecaps, the CHS columns were kinked at first floor level, except at the entrance leading to the Thameslink station behind the building, where the columns were kinked at second and third floor levels.

Figure 1.3 Tubular column with reinforced concrete infill without external fire protection

(Cheung Kong Center, Hong Kong; structural engineer Ove Arup & Partners)

5

Technical details Construction details The columns have a constant external diameter of 323.9 mm but, within this serial size, vary from 323.9 × 30 CHS at the first floor, 323.9 × 16 CHS between second and fifth floors and 323.9 × 12.5 CHS between sixth and seven floors. The CHS column material was Celsius hot-finished hollow section to EN10210 grade S355J2H.

The fire rating for internal elements was generally two hours but the requirement for the external columns was only 35 minutes in view of their position outside the cladding line. No fire protection was given to the columns, a 45 minute rating being achieved by concrete filling alone. The concrete infill used was to Grade 40 or Grade 60 depending on the vertical load.

Project data Client: Heron Property Corporation

Architect: Skidmore, Owings & Merrill

Structural Engineer: Waterman Partnership

1.2.2 MONTEVETRO APARTMENT BLOCK, LONDON Steel composite tubular columns have been used on the facade of this high specification apartment block, which looks out across the River Thames. The composite columns were chosen because of their small plan area and their slender shape which minimised the obstruction to the views out from the apartments.

Figure 1.4 Fleet Place House, London

6

Technical details Construction details CHS columns are used and, in the typical case, are only 244.5 mm in diameter. Within this serial size the columns vary from 244.5 × 16 CHS to 244.5 × 20 CHS; for the most heavily loaded areas, the columns were increased to a maximum size of 355.6 × 16 CHS.

The steel CHS columns are used to support concrete flat slab floors. At the highest point the apartment block has twenty storeys and there are heavy axial loads in the columns in these areas. The CHS column material was Celsius hot-finished hollow section to EN10210 grade S355J2H.

The fire rating required for the columns varied between one and two hours and was achieved by the combined effect of the concrete infill and an intumescent coating. No reinforcement was used inside the CHS columns.

Project data Client: Taylor Woodrow Capital Developments

Architect: Richard Rogers Partnership

Structural Engineer: Waterman Partnership

Figure 1.5 Montevetro Apartment Block, London

7

1.2.3 PECKHAM LIBRARY, LONDON This library building is of striking appearance. It is supported at the front by concrete-filled CHS columns, angled to form an irregular facade and enclosing a covered space, which is an extension of a new public square. There are 7 supporting columns, 18 metres long, which are angled out of the vertical, the columns supporting steel tubular trusses at the fourth floor and a steel roof above. At ground floor level, the columns have a height of 18 metres, giving a height to width ratio of 37. The angling of the columns helps to provide additional stiffness against lateral loads and limits the development of bending moment in the columns.

Technical details Construction details At ground floor level, the column line consists of a 323.9 × 20 CHS column at each end with 323.9 × 16 CHS columns in-between. At first floor level, these section sizes reduce to 323.9 × 6.3 CHS members. All columns were designed without external protection, utilising an infilled concrete core within the tubes to satisfy the requirement for a one-hour fire rating. The concrete infill is reinforced with 8 T12 longitudinal bars with T6 links at 175 centres. The columns were fabricated as full-height elements and were supplied in Celsius hot-finished hollow section to EN10210 grade S355J2H.

Project data Client: London Borough of Southwark

Architect: Alsop Architects

Structural Engineer: Adams, Kara, Taylor

1.2.4 QUEENSBERRY HOUSE, LONDON The steel columns used to support the floors of this six-storey office and commercial buildings are CHS tubular columns. The columns use a tube-in-tube system in which one CHS section is placed inside a larger one with all the voids grouted after erection of the floor structure.

Figure 1.6 Peckham Library, London

8

Technical details Construction details There is an atrium in the middle of the building, which is 6 metres wide, with 12 metre clear spans each side of the atrium in the transverse direction. In a typical case the column consists of a 457 × 10 CHS outer tube and a 323.9 × 6.3 CHS inner tube and supports two floor beams on each side. All columns were delivered to site in two three-storey lengths and were joined by means of an in situ concrete joint in the inner tube and by bolting and welding on the outer tube. After erection, the columns only needed to be made good at the joint and given a final finish coat of paint.

No external fire protection was necessary, the internal column and grout infill having sufficient load capacity by itself in the fire limit state. At room temperatures, the full section capacity of the tube-in-tube column is utilised. The CHS column material was Celsius hot-finished hollow section to EN10210 grade S355J2H.

Project data Client: General Accident and Capital & City

Architect: RHWL

Structural Engineer: Buro Happold

Figure 1.7 Queensberry House, London

9

2 NORMAL DESIGN

2.1 General In general, a composite column must be designed for the ultimate limit state. For structural adequacy, the internal forces and moments resulting from the most unfavourable load combination should not exceed the design resistances of the composite cross-sections. While local buckling of the steel sections may be eliminated, the reduction in the compression resistance of the composite column due to overall buckling should be allowed for, together with the effects of residual stresses and initial imperfections. Moreover, the second order effects in slender columns, as well as the effect of creep and shrinkage of concrete under long-term loading, must be considered if they are significant. The reduction of flexural stiffness due to cracking of the concrete in the tension area should also be considered. These are provided for either explicitly, or empirically, in prEN 1994-1-1: 1994[18] (EC4-1-1).

2.2 Material properties 2.2.1 Hot rolled structural steel Nominal values of the yield stress yf , and the ultimate tensile stress uf , for structural steel are presented in Table 2.1 below.

Table 2.1 Mechanical properties of Celsius Sections

Nominal thickness of element t (mm) t ≤ 40 mm 40 mm < t ≤ 100 mm

Nominal steel grade to

BS EN 100210-1 fy (N/mm²) fu (N/mm²) fy (N/mm²) fu (N/mm²)

S 275 275 430 255 410 S 355 355 510 335 490

Design values of other coefficients for the steel sections are given as follows:

aE = Modulus of elasticity = 210 000 N/mm²

aG = Shear modulus = ( )a

a

12 ν+E

aν = Poisson’s ratio = 0.3

aρ = Density = 7850 kg/m³

2.2.2 Structural concrete Concrete strengths are based on the characteristic cylinder strengths ckf measured at 28 days in accordance with Clause 3.1.2.2, of DD ENV 1992-1-1: 1992[16] (EC2-1-1). The different strength classes, and the associated cube strengths, given by this Eurocode are presented in Table 2.2 below. Classification grades of concrete, such as C20/25, refer to the cylinder/cube strength at the specified age.

For normal weight concrete, the mean tensile strength ctmf and the secant modulus of elasticity

cmE , for short-term loading are also given in Table 2.2. The effect of creep and shrinkage of concrete may be significant under long-term loading in some cases. As will be discussed in

10

Section 2.6.2, provision is given within EC4-1-1[18] to reduce the secant modulus of elasticity, depending on the proportion of permanent load acting on the column.

The density of structural concrete is assumed to be 2400 kg/m³ for plain, unreinforced, concrete and 2500 kg/m³ for reinforced concrete.

Table 2.2 Characteristic compressive strength fck (cylinders), mean tensile strength fctm and secant modulus of elasticity Ecm for structural concrete

Strength class of concrete

C20/25 C25/30 C30/37 C35/45 C40/50 C45/55 C50/60

fck (N/mm²) 20 25 30 35 40 45 50 fctm (N/mm²) 2.2 2.6 2.9 3.2 3.5 3.8 4.1 Ecm (N/mm²) 29000 30500 32000 33500 35000 36000 37000

2.2.3 Steel reinforcement bars In the UK, steel bars for the reinforcement of concrete should conform to BS 4449:1997[19]. DD ENV 10080: 1996[20], which is currently at the draft for development stage, will eventually replace this British Standard. However, the 1997 edition of BS 4449 has been revised considerably compared to its earlier versions, to bring it into line with the requirements of EC2-1-1[16]. The properties most frequently required in design calculations are referred to in Clause 3.2 of EC2-1-1; types of reinforcement steel are classified as follows:

• High (class H) or normal (class N), according to ductility characteristics.

• Plain smooth or, ribbed bars, according to surface characteristics.

Steel grades that should be used in construction are given in Table 2.3.

Table 2.3 Characteristic yield strength fsk, ductility requirements and modulus of elasticity Es of reinforcing steel

Reinforcing steel standard

BS 4449: 1997[19] DD ENV 10080: 1996[20]

Name 460A (class N) 460B (class H) B500A (class N) B500B (class H) fsk (N/mm²) 460 500 Total elongation at maximum force (%)

2.5 5.0 2.5 5.0

Elongation at fracture (%) 12 14 - -

Es (N/mm²) 210000† 210000† † According to EC4-1-1

As can be seen from Table 2.3, apart from the obvious difference in the characteristic yield strength skf , between reinforcing steels complying with BS 4449:1997[19] and DD ENV 10080: 1996[20], BS 4449 also specifies a minimum elongation at fracture: thereby guaranteeing the length of the plastic deformation plateau. Furthermore, in the 1997 edition of BS 4449, the minimum elongation at fracture of 14%, for 460B steel, is higher than the 12% requirement for this yield strength, given in earlier versions of this code of practice. A graphical representation of the difference in the elongation requirements for these two standards is shown in the stress-strain curve in Figure 2.1.

11

It should be noted, however, that although the ductility of reinforcing bars has a significant effect on the behaviour of continuous composite beams[21], this property is of little significance with respect to the design of composite columns at ambient temperature. Concrete filled hollow sections may be used without any reinforcement, except for reasons of fire resistance (see Section 3).

2.3 Partial safety factors National authorities are free to select appropriate values for partial safety factors for loads and materials, and substitute them for ‘boxed’ values in the Eurocodes. The boxed values and the UK National Application Document (NAD) values are:

Loads: EC4-1-1‘boxed’ values UK NAD

Imposed (variable) load, γQ 1.50 1.50

Dead (permanent) load, γG 1.35 1.35

Materials:

Steel, aγ 1.10 1.05

Concrete, cγ 1.50 1.50

Reinforcement, sγ 1.15 1.15

2.4 Basis of design method In EC4-1-1, isolated columns are defined as compression members that are integral parts of a braced or non-sway frame but which are considered to be isolated for design purposes.

Definitions of non-sway structures are given in EC2-1-1[16] as follows:

• Structures or structural elements, with or without bracing elements, for which the influence of displacements of the connections upon the design moments and forces may be neglected, are classified as non-sway. Otherwise, they are classified as sway.

Minimum elongat ion at f racture

Total elongat ion atmaximum force εuk

Stre

ss

St rain

Minimum elongat ion at f racture

Total elongat ion atmaximum force εuk

Stre

ss

St rain

Figure 2.1 Elongation limits for steel reinforcement bars

12

• Braced building structures, where substantial shear walls or core structures provide the bracing, may be assumed to be non-sway.

• Frames may be classified as non-sway if the first order displacements of the connections do not increase the effects of actions calculated without considering these displacements by more than 10%. In general, it is sufficient to consider only the relevant bending moments due to these second-order effects.

A similar definition of a non-sway frame in DD ENV 1993-1-1: 1992[28] (EC3-1-1) is also given for reference:

• A frame may be classified as non-sway if its response to in-plane horizontal forces is sufficiently stiff for it to be acceptably accurate to neglect any additional internal forces or moments arising from horizontal displacements of its nodes.

Two methods of design for isolated composite columns in braced or, non-sway frames are given within EC4-1-1:

2.4.1 General design method This comprehensive method is used for composite columns with non-symmetrical or non-uniform cross-section over the column length. It is also used for composite columns of doubly symmetrical, and uniform cross-section over the column height, when the limits of applicability for the simplified design method are not satisfied (see Section 2.5). In these circumstances, some of the important design issues which should be considered using the general method, are as follows:

• geometrical and material non-linearity;

• second order effects (on slender columns);

• creep and shrinkage of the concrete under long-term loading;

• contribution of the tensile strength of the concrete between cracks;

• imperfections for the calculation of internal forces and moments about both axes;

• distribution of internal forces and moments between the steel section and the concrete by means of a clearly defined load path;

• transfer of longitudinal shear stress at the interface between the steel section and the concrete under large transverse shear; and

• chemical bond and friction together with mechanical shear connection if necessary.

In order to allow for these design considerations, it is necessary to use sophisticated computer software, which operate with both geometrical and material non-linearity. In general, the design effort is considerable. Thus, this method is not preferred for use in practical design, and is outside the scope of this publication.

2.4.2 Simplified design method This method is used for composite columns of doubly symmetrical and uniform cross-section over the column height. It is based on certain assumptions relating to the geometrical configurations of the composite cross-sections. Moreover, it also adopts the European buckling curves for steel columns as the basis of column buckling design. The limits of applicability of this method given in EC4-1-1 are also listed in Section 2.5; when the limits are not satisfied, the above general design method should be used.

13

It should be noted that this method is formulated in such a way that only hand calculation is required in practical design. The simplified design method is presented in detail within this publication. The calculation procedure is in six parts, as follows:

(i) Check that the limits of the simplified design method are satisfied.

(ii) Calculate the properties of the cross-section.

(iii) Calculate the buckling resistance of the column.

(iv) Check whether second-order effects should be considered

(v) Calculate the effect of interaction between axial load and bending.

(vi) Calculate the longitudinal and transverse shear.

2.5 Restrictions on the simplified design method The application of the simplified design method is subject to various restrictions, as follows:

(a) The column is doubly-symmetrical and is of uniform cross-section over the height of the column.

(b) The steel contribution ratio δ must satisfy the following conditions:

9.02.0 ≤δ≤

If δ is less than 0.2, the column may be designed according to EC2-1-1[16]. If δ is larger than 0.9, the concrete is ignored in the calculations, and the column is designed as a bare steel section.

(c) The maximum non-dimensional slenderness ratio of the composite column λ is limited to 2.0.

(d) The maximum amount of longitudinal reinforcement that can be considered in the analysis is 6% of the concrete area. However, if design for fire resistance is not needed, according to prEN 1994-1-1: 2001[17], no minimum amount of reinforcement is normally necessary within a filled SHS column; in other words:

0% ≤ c

s

AA

≤ 6.0%.

2.6 Properties of cross-section 2.6.1 Squash (plastic) resistance, Npl,Rd The plastic resistance, to compression, of a composite cross-section represents the maximum load that can be applied to a short composite column. It is important to recognize that concrete filled circular hollow sections exhibit enhanced resistance due to the triaxial containment effects. Concrete filled rectangular sections (RHS) do not achieve such enhancement.

Local buckling of steel hollow sections Before the plastic resistance of the concrete filled hollow section is calculated, it should be insured that local buckling of the steel does not occur. To prevent premature local buckling, the width to thickness ratio of the steel section in compression must satisfy the following limits:

For concrete filled rectangular hollow sections (RHS) ε52≤th

14

For concrete filled circular hollow sections (CHS) 290ε≤td

where:

t is the wall thickness of the steel hollow section in mm.

h is the larger outer dimension of the rectangular hollow section in mm.

d is the outer diameter of the circular hollow section in mm.

ε y

235f

=

yf is the yield strength of the steel section in N/mm².

Local buckling in some rectangular hollow sections with large h/t ratios may be critical. No specific design recommendation is given within EC4-1-1, and design using sections which exceed the local buckling limits should be verified by tests.

Concrete filled rectangular hollow sections (RHS) The plastic resistance of a concrete filled rectangular hollow section (i.e., the so-called “squash load”) is given by the sum of the resistances of the components as follows:

c

ckc

s

sks

a

yaRdpl, γγγ

fAfAfAN ++=

where:

aA is the area of the steel section.

sA is the area of the reinforcement.

cA is the area of the concrete.

yf is the yield strength of the steel section.

skf is the characteristic yield strength of the steel reinforcement bars.

kcf is the characteristic compressive strength (cylinder) of the concrete.

For ease of expression, a

y

γf

, s

sk

γf

and c

ck

γf

are presented as design strengths of the respective

materials in the remainder of Section 2 as: ydf , sdf and cdf respectively. As a result of this simplification, the above equation for the plastic resistance of the composite column, can be rewritten in the following compact form:

cdcsdsydaRdpl, fAfAfAN ++=

Concrete filled circular hollow sections (CHS) For composite columns with concrete filled circular hollow sections, the increased resistance of concrete due to the confining effect of the circular hollow section may be included. This restraint to transverse strain in a three dimensional confinement results in increased concrete

15

resistance. At the same time, circular tensile stresses in the circular hollow section also arise, which reduce its axial resistance.

In general, the resistance of a concrete filled circular hollow section to compression may increase by up to 15% under simple axial loads when the effect of tri-axial confinement is considered. However, this effect on the resistance enhancement of concrete depends also on the slenderness of the composite columns and is significant only in stocky columns. For composite columns with a non-dimensional slenderness of λ > 0.5 (where λ is defined in Section 2.7), this effect should be neglected and the plastic resistance assessed as for rectangular hollow sections.

In addition, further linear interpolation is necessary to take account of any effective load eccentricities. However, the eccentricity, e of the applied load may not exceed the value d/10, where d is the outer diameter of the circular hollow section.

The eccentricity, e is defined as follows:

Sd

Sd

NMe =

where:

SdM is the maximum design moment (second order effects are ignored).

SdN is the design applied load.

The plastic resistance of a concrete filled circular hollow section may be obtained as follows:

+++=

ck

y1cdcsdsyd2aRdpl, 1

ff

dtfAfAfAN ηη

where:

t is the wall thickness of the steel hollow section in mm.

( )

η−+η=η

−η=η

de

de

101

101

20202

101

for 0< e ≤ d/10

=η=η

0.10

2

1 for e > d/10

The basic values 10η and 20η depend on the non-dimensional slenderness ratio λ , and are defined as follows:

210 175.189.4 λλη +−= but 010 ≥η

( )λη 2325.020 += but 0.120 ≤η

16

If the eccentricity e exceeds the value d/10, or if the non-dimensional slenderness ratio λ exceeds the value 0.5, then 010 =η and 0.120 =η . Table 2.4 gives the basic values 10η and

20η for different values of λ .

Table 2.4 Basic values of η10 and η20 to allow for the effect of triaxial confinement in concrete filled circular hollow sections

Non-dimensional slenderness ratio λ

0 0.1 0.2 0.3 0.4 ≥ 0.5

10η 4.90 3.22 1.88 0.88 0.22 0

20η 0.75 0.80 0.85 0.90 0.95 1.00

2.6.2 Effective flexural stiffness Short-term loading The effective flexural stiffness of the composite column ( )eEI is obtained from adding the up the flexural stiffnesses of the individual components of the cross-section:

( ) c cm ss a a e 6 . 0 I E I E I E EI + + =

where:

aI , sI and cI are the second moment of area, about the appropriate axis of bending, for the steel section, the reinforcement and the concrete (assumed uncracked) respectively.

aE and sE are the elastic moduli for the structural steel and the reinforcement respectively.

c cm 6 . 0 I E is the effective stiffness of the concrete component (the 0.6 factor is an

empirical multiplier, which has been determined from a calibration exercise, to give good agreement with test results).

cmE is the secant modulus of elasticity for structural concrete; see Table 2.2.

Long-term loading For composite columns under long-term loading, the creep and shrinkage of concrete will cause a reduction in the effective elastic stiffness of the composite column, thereby reducing the buckling resistance. However, this effect is only significant for slender columns; as a simple rule, the effect of long-term loading should be considered if the buckling length to depth ratio of a composite column exceeds 15.

If the eccentricity of loading (see Section 2.6.1) is more than twice the cross-section dimension, the effect on the applied bending moment distribution caused by increased deflections, due to creep and shrinkage of the concrete, will be very small. Consequently, it may be neglected and no provision for long-term loading is necessary. Moreover, no provision is necessary if the non-dimensional slenderness λ of the composite column is less than the limiting values given within Table 2.5 below.

17

Table 2.5 Limiting values of λ for long-term loading

Frame type Braced non-sway frame

Sway frames and/or unbraced frames

Concrete filled hollow sections ( )δ−18.0

( )δ−1

5.0

The steel contribution factor δ, given in Table 2.5 above, is defined as follows:

Rdpl,

yda

NfA

=δ

If the eccentricity of loading is less than twice the cross-section dimension and the non-dimensional slenderness λ of the composite column is less than the limiting values given within Table 2.5, the effect of creep and shrinkage of concrete should be allowed for by reducing the effective elastic modulus of the concrete to the value:

−=∞

Sd

SdG,cdc

5.01

NN

EE

where:

SdN is the design applied load.

SdG,N is the part of the design load permanently acting on the column.

Table 2.5 also allows the effect of long-term loading to be ignored for concrete filled hollow sections with 0.2≤λ , provided that δ is greater than 0.6 for braced (or non-sway) columns, and 0.75 for unbraced (and/or sway) columns.

2.7 Column buckling resistance The plastic resistance to compression of a composite cross-section Rdpl,N represents the maximum load that can be applied to a short column. However, for slender columns, with low elastic critical load, overall buckling considerations may be more significant.

In Figure 2.2(b), the buckling resistance of a column is expressed as a proportion χ of the plastic resistance to compression NPl,R , thereby non-dimensionialising the vertical axis compared to Figure 2.2(a). The horizontal axis may be non-dimensionalised similarly by use of the Euler buckling load crN as is also shown in Figure 2.2(b).

By incorporating the effects of both residual stresses and geometric imperfections, the European buckling curves may be drawn on this basis as shown in Figure 2.2(c). These curves form the basis of column buckling design for both steel and composite columns.

18

The buckling resistance is calculated from the plastic resistance and the Euler (elastic) critical load using the EC3-1-1[28] buckling curve ‘a’ (N.B. at the fire limit state, curve ‘c’ is used due to its close agreement with the results from fire tests; see Section 0). The Euler buckling load is given by:

( )2

e2

cr lEI

Nπ

=

where:

( )eEI is the effective elastic flexural stiffness of the composite column (see Section 2.6.2).

l is the buckling length of the column.

EC4-1-1 suggests that the buckling length l of an isolated non-sway composite column may conservatively be taken as equal to its system length L. Alternatively, the buckling length may be determined using Annex E of EC3-1-1.

The non-dimensional slenderness ratio is given by:

cr

Rpl,

NN

=λ

ab

c1.00

2.01.000

0.2

λ = pl,Rcr

χ = Rdpl,Rd

N N

NN

(c)

l/r Slenderness

pl

cr

N

N

(a)

χ =pl

plcr

NN

NN(b)

N

Figure 2.2 (a) Idealised column buckling curve, (b) Non-dimensionalised column buckling curve, (c) European buckling curves according to EC3-1-1

19

where:

Rpl,N is the plastic resistance of the composite cross-section to compression, according to

Section 2.6.1, with 0.1csa === γγγ .

The resistance of a composite column in axial compression (buckling load) is obtained from:

Rd pl,Rd .N N χ =

where:

χ is the reduction coefficient for buckling obtained from curve ‘a’ of EC3-1-1, and is dependant on the non-dimensional slenderness ratio λ .

The reduction factor may be determined from:

[ ] 5.022

1

λ−φ+φ=χ but 0.1≤χ

where:

φ ( )[ ]22.015.0 λλα +−+=

α is an imperfection parameter depending on the buckling curve considered.

Relevent Buckling Curves and Imperfection Factors According to prEN 1994-1-1, circular or rectangular hollow section columns filled with plain concrete or containing up to 3% reinforcement can be designed using buckling curve ‘a’ with an imperfection factor, α, = 0.21. However, concrete filled sections containing between 3% to 4% reinforcement must be designed using buckling curve ‘b’ with an imperfection factor, α, = 0.34 (see Figure 2.3 (a) and (b) below).

In addition, concrete filled circular hollow section columns as shown in Figure 2.3(c) containing an additional open Section used as primary steel can also be designed as a composite section using buckling curve ‘b’ with an imperfection factor, α, = 0.34.

Figure 2.3 Typical column cross-sections

20

Although not explicitly stated, Clause 4.8.3.2 of EC4-1-1, while defining the partial safety factors implies that isolated non-sway composite columns need not be checked for buckling, if any of the following conditions is satisfied:

(i) the axial force in the column is less than cr1.0 N ; or

(ii) the non-dimensional slenderness ratio λ is less than 0.2.

2.8 Analysis of bending moments due to second-order effects

Under the action of the design axial load SdN on a column with an initial imperfection 0e , as shown in Figure 2.4, there will be a maximum internal moment of 0SdeN . It is important to note that this ‘second order moment’, or ‘imperfection moment’, does not need to be considered separately, as its effect on the buckling resistance of the composite column is already accounted for in the European buckling curves as shown in Figure 2.2(c).

However, in addition to axial forces, a composite column may be also subject to end moments as a consequence of transverse loads acting on it or, because the composite column is a part of a frame. The moments and displacements obtained initially are referred to as ‘first order’ values. For slender columns, the ‘first order’ displacements may be significant and additional, or ‘second order’, bending moments may be induced under the actions of the applied loads. As a simple rule, the second order effects should be considered if the buckling length to depth ratio of a composite column exceeds 15.

The second order effects on bending moments for isolated non-sway columns should be considered if both of the following conditions are satisfied:

1) 1.0cr

Sd >NN

where:

SdN is the design applied load.

crN is the Euler buckling load.

2) ( )r−>λ 22.0

where:

λ is the non-dimensional slenderness ratio

r is the ratio of the smaller to the larger end moment (see Figure 2.5). If there is any transverse loading, r should be taken as 1.0.

NSd NSd

eo

Figure 2.4 Initially imperfect column under axial compression

21

The second order effects in an isolated non-sway column may be allowed for by modifying the maximum first-order bending moment max,SdM , with a correction factor k, which is defined as follows:

0. 1

1 cr,eff

Sd ≥

−

β =

N N

k

where:

SdN is the design applied load

cr,eff N is the elastic critical load of the composite column based on the system length, L, and a reduced design value of effective stiffness (EI)e,II

where:

(EI)e,II = 0.9(EaIa + EsIs + 0.5EcmIc)

β is the equivalent moment factor.

For columns with transverse loading within the column length, the value for β should be taken as 1.0. For pure end moments, β can be determined as follows:

r44.066.0 +=β but 44.0≥β

2.9 Combined compression and bending The design for a composite column subjected to combined compression and bending is carried out in stages as follows:

• The composite column is isolated from the framework, and the end moments, which result from the analysis of the system as a whole, are taken to act on the column under consideration. Internal moments, and forces within the column length, are determined from the structural consideration of end moments, axial and transverse loads.

• For each axis of symmetry, the buckling resistance to compression should be checked with the relevant non-dimensional slenderness of the composite column.

• In the presence of applied moment about one particular axis e.g., the y-y axis, the moment resistance of the composite cross-section should be checked with the relevant non-dimensional slenderness of the composite column i.e., yλ , instead of zλ , although zλ may

be larger, and thus more critical, than yλ .

• For slender columns (see Table 2.5 and Section 2.8), both the effect of long-term loading and the second-order effects are included.

SdSd

-1 r +1≤ ≤rMM

Figure 2.5 Ratio r of the end moments

22

It should be noted that, by adopting the EC4-1-1[18] simplified method, imperfections within the column length need not be considered as they are taken account of in the relevant buckling curve when determining the buckling resistance of the column (see Section 2.7).

2.9.1 Combined compression and uni-axial bending In EC4-1-1[18], the resistance of a composite column subjected to combined compression and bending is determined from an interaction curve. For a bare steel section, the interaction curve is characterised by a continuous reduction of the moment resistance with a corresponding increase in axial load.

However, as shown from the interaction curve in Figure 2.6: a short composite column may exhibit an increase in moment resistance under axial load. The reason for this increase is that, under some favourable conditions, the compressive axial load prevents concrete cracking, and therefore makes the cross-section of a short composite column more effective in resisting moments.

An interaction curve between compressive axial load and moment can be obtained for a short composite column by considering several possible positions of the neutral axis within the cross-section, and determining the internal forces and moments from the resulting plastic stress blocks. For the simplified method given within EC4-1-1, sufficient accuracy in estimating the effects of combined compression and bending may be found by constructing the interaction curve, shown in Figure 2.7, from 4 or 5 points.

1.0

1.0

λ ≈0

Rd

MN

Rd

pl,RdN

N

MM pl,Rd

Figure 2.6 Interaction curve for a composite column subjected to compression and uni-axial bending

23

For composite columns, which are doubly symmetrical and of a uniform cross-section over their height, the following approach given in EC4-1-1[18], and the UK NAD, may be adopted.

Figure 2.8 shows the plastic stress distributions within the cross-section of a concrete filled RHS at point A, B, C, D and E of the interaction curve given in Figure 2.7. The significance of each of these points are as follows:

• Point A indicates the plastic resistance of the cross-section to compression, in the absence of an applied bending moment:

0A

Rdpl,A

=

=

M

NN

• Point B corresponds to the plastic moment resistance of the cross-section, in the absence of an applied axial load:

Rdpl,B

B 0MM

N==

• At point C, the axial compression and moment resistance of the composite column are given as:

Rdpl,C

cdcRdc, Rdpm,C ) (or

MM

fANNN

=

==

N

A E

C

D

B

M

pl,Rd

pm,Rd

pm,Rd

N

N

½N

pl,Rd max,RdM M

Figure 2.7 Interaction curve with linear approximation

24

The expressions may be obtained by combining the stress distributions of the cross-section at points B and C; the compression area of the concrete at point B is equal to the tension area of the concrete at point C. The moment resistance at point C is equal to that at point B, since the stress resultants from the additionally compressed parts cancel one another out in the central region of the cross-section. However, these additionally compressed regions create an internal axial force, which is equal to the plastic resistance to compression of the concrete alone i.e., Rdpm,N or Rdc,N .

• At point D, the plastic neutral axis coincides with the centroid of the cross-section, and the resulting axial force is half of the value at point C, i.e.:

Rdmax,D

Rdpm,D 2/MMNN

=

=

Point A cd yd sd

pl.Rd– ––

–

cd yd sd

–

cd yd sd

– ––

cd yd sd–

cd yd sd

– ––

–

Point B

–n

–

+ +

B

Point C

n+ +

CC

+

Point D

Point E

– –

+

D

E

E

E

f f f

f f f

f f f

f f f

f f f

h

h

No momentN

M = Mzero axial force

pl,.Rd

pl,Rdc,Rd

M = MN = N

2

max,Rdc,Rd

M

Nn

h

M = M N

DN =

h /2

h/4 (N.B., the moment resistance Mmax,Rd, at point D, is not allowed in the UK NAD)

Figure 2.8 Stress distributions for the points on the interaction curve for concrete filled hollow sections, according to EC4-1-1[18]

25

Generally, point D is less than point C in design.

• Point E is mid-way between A and C, and is often required for highly non-linear interaction curves, in order to achieve a better a better approximation. For concrete filled structural hollow sections, the use of point E will yield a more economical design; however, much more calculation effort is required. Thus, to retain simplicity, point E tends not to be used.

According to the UK NAD, the additional moment resistance of the composite cross-section (indicated by point D within Figure 2.7), should not be taken account of in design. Therefore, in the UK, an interaction curve consisting of A-C-B or A-E-C-B may only be considered.

The plastic moment resistance of a concrete filled hollow section may be evaluated as follows:

)()(5.0)( psnpssdcnppccdpanpaydRdpl, WWfWWfWWfM −+−+−=

where:

ydf , sdf , cdf are a

y

γf

, s

sk

γf

and c

ck

γf

respectively

paW , pcW , psW are the plastic section moduli for the steel section, the concrete of the composite cross-section (assumed to be uncracked) and the reinforcement respectively.

panW , pcnW , psnW are the plastic section moduli of the corresponding components within the region of n2h from the centre-line of the composite cross-section.

The values of the relevant parameters in the above equation for concrete filled hollow sections are:

Rectangular hollow sections

( )( ) ( ) ps23

2

pc 24

32

422 WrthrrthtbW −

−−π−−−

−−=

where: r is the internal radius of the corners to the hollow section

Circular hollow sections ( )

ps 3

pc 6 2 W t d W −

− =

In general, for both types of section:

( )( )cdydcd

cdsdsncdcn 242

2fftbfffAfA

h−+−−

=

where: snA is the area of reinforcing bars within the region of n2h from the centre–line of the composite cross-section.

For rectangular hollow sections, it can be explicitly stated that:

Wpan = 2t.hn2

26

( ) psn2

npcn 2 WhtbW −−=

Similar simple explicit equations cannot be written for circular sections. However, the above equations can be conservatively applied to circular sections with a high accuracy by substituting diameter d for breadth b

For the calculation of the resistances at the additional point E, RdE,N and RdE,M (see above), the neutral axis is located between nh and the border of the section, so that

hhh 25.05.0 nE += . Using rectangular hollow sections, the axial resistance of the column for this case is:

( ) ( )( ) ( ) Rdpm,cdsdsEcdydnEcdnERdE, 222 NffAffhhtfhhbN +−+−−+−=

where sEA is the sum of the areas of reinforcement lying in the additional compression region between Eh and nh .

The magnitude of RdE,M is calculated from the above equations but with Eh substituted for nh in the values of panW and pcnW .

Again, the above equations can be applied to circular sections by substituting diameter d for breadth b but may become highly over-conservative. In such circumstances it may be preferable to simply apply a linear interpolation between points A and C

The principal for checking the composite cross-section under combined compression and uni-axial bending, in accordance with EC4-1-1[18], is illustrated graphically in Figure 2.9. Firstly, the resistance of the composite column under axial load is determined in the absence of bending, which is given by Rdpl,Nχ (see Section 2.7). The moment resistance of the composite column should then be checked with the relevant non-dimensional slenderness, which is in the same plane of the applied moment. As mentioned earlier, the initial imperfections of columns have been incorporated within the appropriate buckling curve, and no additional consideration of geometrical imperfections is necessary in the evaluation of bending moments within the column height.

Consider the interaction curves for combined compression and uni-axial bending shown in Figure 2.9. Under an applied force SdN equal to Rdpl,Nχ , the horizontal coordinate Rdpl,k Mµ represents the second order moment due to imperfections within the column, otherwise known as the ‘imperfection moment’. It is important to recognise that the moment resistance of the column has been fully utilised in the presence of the imperfection moment; the column, therefore, cannot resist any additional applied moment. Moreover, the influence of the imperfections decreases when the axial load ratio is less than χ, and it is assumed to vary linearly between nχ and χ. For an axial load ratio less than nχ , the effect of imperfections is neglected.

It is important to note that the value nχ accounts for the fact that the influence of the imperfections and that of the bending moment do not always act together unfavourably. For columns with only end moments, nχ may be determined as follows:

( ) χχ4

1n

r−=

27

where r is the ratio of the small to the large end moment (see Figure 2.5).

If transverse loads occur within the column height, then r must be taken as unity and nχ is thus equal to zero (i.e., it coincides with the origin of the interaction curve shown in Figure 2.9).

With a design axial load of SdN , the design axial load ratio dχ is defined as follows:

Rdpl,Sdd / NN=χ

By reading off the horizontal distance from the interaction curve (see Figure 2.9), the moment resistance ratio µ may be obtained, and the moment resistance of the composite column under combined compression and uni-axial bending may be evaluated. Details of the UK NAD method for calculating µ, and its limitations, are discussed below.

EC4-1-1[18] considers that the design is adequate when the following condition is satisfied:

Rdpl,Sd 9.0 MM µ≤

where:

SdM is the design bending moment, which may be factored to allow for second-order effects, if necessary (see Section 2.8).

µ is the moment resistance ratio obtained from the interaction curve.

Rdpl,M is the plastic moment resistance of the composite cross section.

The interaction curve shown in Figure 2.9 has been determined without considering the strain limitations within the concrete. Hence the moments, including second-order effects (if necessary), are calculated using the effective elastic flexural stiffness ( )eEI , and taking into

Rd Rd

χ

χn

χd

1.0

µ

µdkµ

Rd

(a) (b)

χ

χd

1.0Rd

χn

µdkµ0 0

µ

Cross-sectioninteraction curve A

C

B

χpm

1.0 1.0

NN

NN

MM

MM

pl,Rd pl,Rd

pl,Rd pl,Rd

Figure 2.9 Interaction curve for compression and uni-axial bending using (a) the EC4-

1-1 method; and (b) the simplified method in the UK NAD

28

account the entire concrete area of the cross-section (i.e., the concrete is uncracked). Consequently, in order to allow for these simplifications, the 0.9 constant, shown in the above equation, is applied to the moment resistance.

For concrete filled hollow sections, the interaction curve of A-E-C-B (shown in Figure 2.7) may be preferred to A-C-B (shown in Figure 2.9(b)), as it will give a more economical design: especially for columns with high axial load and low end moments (although much more calculation effort is required). For a better approximation, the position of point E may be chosen to be closer to point A rather than being mid-way between points A and C. For further information, refer to EC4-1-1[18].

Requirements of the UK National Application Document (NAD) The following additional requirements are specified in the UK NAD:

• For columns under combined compression and bending, the ratio nχ should be determined[24,25] as follows:

(i) Concrete filled rectangular hollow sections

Providing the non-dimensional slenderness λ does not excess 1.0, the ratio pmd χχ ≥

may be determined from the equation shown above. For values of λ in the range of 1.0 to 2.0, 0n =χ .

(ii) Concrete filled circular or square hollow sections

The ratio nχ may be determined from the equation shown above, with no limits place on the non-dimensional slenderness λ .

• The design moment resistance, in combined compression and uni-axial bending, of a composite column should not exceed the design plastic moment resistance Rdpl,M , irrespective of the applied load N (i.e., point D on the interaction curve shown in Figure 2.7 is not allowed).

Combined compression and uni-axial bending to the UK National Application Document In order to comply with the UK NAD, the moment resistance ratio µ, for a composite column under combined compression and uni-axial bending should be evaluated as follows:

µ ( )( )( )( )npm

nd

11

χχχχχχ

−−−−

= when pmd χχ ≥

( )( )( )( )npm

nd

11

1χχχχχχ

−−−−

−= when pmd χχ <

where:

pmχ is the axial resistance ratio due to the concrete, Rdpl,Rdpm, / NN

dχ is the axial resistance ratio, Rdpl,Sd / NN

29

χ is the reduction factor due to column buckling (see Section 2.7).

For concrete filled rectangular hollow sections

nχ ( ) χ

41 r−

= for 0.1<λ

0= for 0.20.1 ≤≤ λ

For concrete filled circular or square hollow sections

nχ ( ) χ

41 r−

= for 0.2≤λ

The expression, for the moment resistance ratio µ, is greatly simplified by taking 0n =χ as follows:

µ ( )pm

d

1 χχχχ

−−

= when pmd χχ ≥

( )

( )pm

d

11

1χχ

χχ−

−−= when pmd χχ <

The above expressions are obtained from a general consideration of the geometry of the UK NAD interaction curve shown in Figure 2.9(b). The simplified expression for the moment resistance ratio µ, is always conservative; since, by taking 0n =χ it is implied that r = 1.0 (i.e., the end moments are equal, and constant over the column length, see Figure 2.5).

2.9.2 Combined compression and bi-axial bending For the design of a composite column under combined compression and bi-axial bending, the axial resistance of the column in the presence of bending moment for each axis, has to be evaluated separately. In general, it will be obvious which of the axes is more likely to fail and the imperfections need to be considered for this direction only: as shown in Figure 2.10. If it is not obvious which plane is more critical, checks should be made on both planes.

30

After the evaluation of the moment resistance ratios yµ and zµ for both axes, as described in the previous section, the interaction of the moments must also be checked using the linear curve shown in Figure 2.10 (c). This linear interaction curve is cut off at y9.0 µ and z9.0 µ . The

design moments, Sdy,M and Sdz,M , related to the respective plastic moment resistances, must lie within the moment interaction curve.

EC4-1-1[18] considers the check is adequate when all the following conditions are satisfied:

9.0Rdy,pl,y

Sdy, ≤MM

µ

9.0Rdz,pl,z

Sdz, ≤MM

µ

and 0.1Rdz,pl,z

Sdz,

Rdy,pl,y

Sdy, ≤+MM

MM

µµ

As it is only necessary to consider the effect of geometric imperfections in the critical plane of the column buckling, the moment resistance ratio µ in the other plane may be evaluated without the consideration of imperfections, which is presented as follows:

1.0χ

χd

χn

kµ µd1.0

µz

a 1.0

χd

1.0

Rd Rd

b

µz

Rd Rd

(a) Plane expected to fail, withconsideration of imperfections

(b) Plane without considerationof imperfections

(c) Moment interaction curvefor bi-axial bending

c

µyµy0.9

µyµy0.9

z

y

NN

NN

MM

MM

MM

MM

pl,Rd pl,Rd

pl,Rd pl,Rd

pl,y,Rdy,Rd

z,Rdpl,z,Rd

Figure 2.10 Verification for combined compression and bi-axial bending

31

µ pm

d

11

χχ

−−

= when pmd χχ >

0.1= when pmd χχ ≤

These expressions are based on the simplified interaction curve, given in the UK NAD.

2.10 Longitudinal and transverse shear In general, the applied internal forces and moments from a member connected to the ends of a composite column are distributed between the steel section and the concrete. EC4-1-1[18] requires that adequate provision should be made for the distribution of these internal forces and moments.

For structural hollow sections, the shear resistance between the steel section and the concrete is achieved by both chemical bond and friction at the interface. In these circumstances, the design shear resistance, developing at the interface between the concrete and the inner wall of the steel section, is limited to[17]: 0.40 N/mm² for a square or rectangular hollow section (RHS); and 0.55 N/mm² for a circular hollow section (CHS).

For axially loaded columns, it is usually found that this interface shear is sufficient to utilise the combined strengths of both materials at the critical cross-section (mid-column height). For columns with significant end moments, a horizontal shear force is required, which demands the development of longitudinal shear forces between the concrete and the steel.

Similarly, the design transverse shear forces may be assumed: to act on the steel section alone; or to be shared between the steel section and the concrete. For the latter case, the shear force to be resisted by the concrete must be assessed in accordance with EC2-1-1, whereas the shear force to be resisted by the steel section may be checked according to von Mises yield criterion. However, it is simpler in design to assume that the whole of the transverse shear force acts on the steel alone. Figure 2.11 indicates the reduction in the design strength of the shear area (web) that will occur within a steel section subjected to transverse shear stress.

For design purposes, any reduction in the design steel strength in the shear area of the steel section may be transformed into a reduction in steel thickness. For a steel section under major axis bending, the effective wall thickness of the ‘web’ w,dt in the presence of transverse shear may be evaluated as follows:

cd yd–f f

– red.fyd

+

sd –

+

f

–

τ

Figure 2.11 Reduction of design strength of steel within shear area in the presence of

transverse shear stress

32

−−=

2

Rdpl,a,

a,Sdww,d 1

21

VV

tt

where:

a,SdV is the design shear force to be resisted by the cross-section

Rdpl,a,V is the plastic resistance of the steel cross-section in shear = 3

ydv

fA

vA is the shear area of the steel section.

For rectangular hollow sections of uniform thickness:

Load parallel to depth, h, )/(v hbAhA +=

Load parallel to breadth, b, )/(v hbAbA +=

For circular hollow sections and tubes of uniform thickness, π= /2v AA

However, no reduction in the web thickness is necessary when:

Rdpl,a,a,Sd 5.0 VV <

Using the effective wall thickness of the ‘web’ w,dt of the steel hollow section, the moment resistance of the composite cross-section may be evaluated using the same set of expressions given within 2.9.1: without any modification.

For simplicity, the division of the shear force between the hollow section and the concrete may be neglected, and the design shear force is assumed to be resisted by the steel section alone.

2.11 Load introduction Where a load is applied to a composite column, it must be ensured that the load is distributed between the individual components of the cross-section in proportion to their design resistances within a specified introduction length. For composite columns using SHS, this may be achieved as follows:

(i) No shear connection needs to be provided for load introduction through a cap plate, at the top of a column, if the full interface between the concrete section and endplate is permanently in compression: after due consideration of the effects of creep and shrinkage. Otherwise, the load introduction has to be verified according to (v). For concrete filled circular hollow sections, the effect caused by the confinement may be taken into account for load introduction according to Section 2.6.1, but using the values 10η and 20η for λ = 0.

(ii) If the cross-section of a cap-plate is only partially loaded (see Figure 2.12), loads may be distributed with a ratio of 1:2.5, over the thickness of the end plate. The concrete stresses should be limited then in the area of the effective load introduction area for concrete filled hollow sections according to Figure 2.12, Figure 2.13 and (vi) below.

33

(iii) In absence of a more accurate method, when loads are introduced at an intermediate position of an SHS length, the introduction length should be assumed not to exceed 2.5d, where d is the minimum transverse dimension in the case of concrete filled rectangular hollow sections or the outside diameter of the column for circular hollow sections.

(iv) Shear connectors should be provided in the load introduction area, and in areas with change of cross-section, if a design shear strength at the interface between the steel and concrete exceeds the values given in Section 2.10 viz.: 0.40 N/mm² for RHS; and 0.55 N/mm² for CHS. The shear forces should be determined from the change of sectional forces of the steel or reinforced concrete section within the introduction length, where the sectional forces should be determined by plastic theory. If the loads are introduced only into the concrete cross section, the values resulting from an elastic analysis considering creep and shrinkage should be taken into account. At a beam connection position, it is necessary to check that:

For an RHS column: sSd AV /)1( δ− < 0.40 N/mm² with dbAs 5.2=

For a CHS column: sSd AV /)1( δ− < 0.55 N/mm² with 4/5.2 2dAs π=

where:

SdV is the design shear load to be transferred to the column by a beam connection.

δ is the steel contribution ratio.

sA is the usable shear area/connection at the concrete interface.

b is the breadth of RHS face at a shear connector.

d is the minimum dimension of an RHS or diameter of a CHS.

If load introduction would give rise to excessive interface shear stresses, then additional shear stud connectors, or a through gusset plate (Figure 2.13), should be provided in the load introduction area, to enable the additional load to be introduced into the concrete core.

(v) Shear studs may be designed using the usual method given in EC4-1-1[18], based on the following assessment, namely that the design shear strength of a stud should be determined as the lower of:

F

Steel

Concrete1:1

1:2.5 1:2.5

1:1

Figure 2.12 Load dispersion through a locally loaded cap plate

34

( ) vuRd dfP γπ= /4/8.0 2

or

( ) vcmckRd EfdP γα= /29.0 2

with ( )[ ]1/2.0 −=α dh for 3 ≤ h/d ≤ 4

0.1=α for h/d > 4

where:

uf is the specified ultimate strength of the shear stud material (but not greater that 500 N/mm²)

ckf is the characteristic cylinder strength of the concrete.

cmE is the secant modulus of the concrete as given in Table 2.2.

d is the diameter of the shank of shear stud.

h is the length of the shear stud within the concrete core

vγ is a partial safety factor of 1.25.

(vi) When a concrete filled circular or rectangular square hollow section is only partially loaded by plate stiffeners at a cap column divider plate position (column section type A-A in Figure 2.13), or from a gusset plate through the profile at an intermediate column length position (section type B-B in Figure 2.13), the local design resistance strength of concrete

N Sd

eg

t s

t cσ c,Rd

A A

Section A - A Section B - B

B B

σ c,Rd

SdMN Sd

≤ f yd

te

e

A1

cst + 5t

Figure 2.13 Load introduction into a concrete core through a gusset plate

35

Rdc,σ under the gusset plate or stiffener, resulting from the sectional forces of the concrete section, should be determined by:

l

cdc

l

c

ck

yclcddRc, 1

AfA

AA

ff

taf ≤

+= ησ

where:

fcd and fck are the design strength of the steel and the characteristic strength of the concrete respectively.

t is the wall thickness of the steel tube.

a is the diameter of the tube or the width of the rectangular section.

Ac is the cross-sectional area of the concrete.

Al is the loaded area under the gusset plate according to Figure 2.13.

ηcL is 4.9 for circular steel tubes; and 3.5 for rectangular sections.

The ratio Ac / Al in the equation above should not exceed 20. Welds between the gusset plate and the steel hollow sections should be designed according to Section 3 of prEN 1993-1-8: 2002[29].

(vii) For concrete filled circular hollow sections, longitudinal reinforcement may be fully taken into account when assessing cross-sectional design parameters, even where the reinforcement is not welded to the end plates or in direct contact with the endplates, provided that the gap eg between the reinforcement and the end plate does not exceed 30 mm (see also column section type A-A in Figure 2.13).

Alternatively, proprietary nailed connectors can used to effect the required shear transfer capacity. These must be shot fired through the tube wall from the outside in a defined pattern before concrete filling. Typically, they can have a design shear capacity of 12 kN/connector and are placed at a spacing of 50 mm between connectors[34] .For further information see the cited reference.

36

3 FIRE DESIGN

3.1 General The presence of load bearing concrete within a hollow steel column has a beneficial effect on the fire resistance of the steel section. Columns may be fire protected in the conventional way using externally applied protection but, in many cases, significant periods of fire resistance can be obtained without the need for external protection. Guidance on both methods of achieving fire resistance is given in this section; however, the emphasis is on the use of unprotected sections. Nevertheless, in most practical cases, applying external protection has the practical advantage of removing the need to use reinforcment bars to obtain longer periods of fire resistance.

Note that in any particular case, reducing the applied loads will increase the fire resistance of a column.

Extensive experimental and theoretical investigations on the fire performance of concrete filled columns, without applied protection, have been carried out in Europe and the UK with the support of: the European Coal and Steel Community (ECSC); the International Committee for the Development and Study of Tubular Structures (CIDECT); and various national governments. These studies have led to the development of the design rules that are now included in the design codes of many countries in addition to DD ENV 1994-1-2: 2000[3] (EC4-1-2) and BS5950-8: 1990[22].

The design guidance presented in this publication is based on the Eurocode methodology rather than BS5950-8. SCI and Corus are of the opinion that the Eurocode methods are more advanced than BS5950-8 and are therefore preferred.

Software to design concrete filled hollow sections for both normal conditions and fire has been written by SCI and is available on a Corus CD for general use. The software only covers unprotected sections. Details of the software are given in Section 6.

3.1.1 Behaviour in fire Under ambient conditions the steel and concrete material in a concrete filled hollow section move together, and hence longitudinal steel and concrete strains are equal. Accordingly, the stress in each material is proportional to the ratio of the elastic moduli of the two materials.

On heating, the steel will try to expand more rapidly than the concrete, and will therefore begin to resist a greater proportion of the applied load. However, at the same time, the steel yield stress and modulus of elasticity will begin to reduce and, eventually, the steel will begin to shed load into the concrete.

Heat from the steel shell will be transferred to the outer layers of the concrete core, causing their temperature to rise. However, concrete is not a good conductor of heat, and the rate of heat flow through the core will be slow.

As the temperature of the outer layers increases, the concrete strength itself will begin to fall as the heat degrades the concrete. The degradation includes the driving off of water, which is present both as free moisture and from the hydrated constituents of the mix. This produces a marked plateau in the concrete’s temperature-time profile as a considerable amount of heat is absorbed in converting this moisture to steam.

37

It is imperative that venting is provided in the steel shell to allow any steam to escape. BS 5950-8 and EC4-1-2 recommend that the sections should contain one vent hole, with a minimum diameter of 20 mm, at the top and bottom of each storey (see Section 4.1.1). The longitudinal spacing of these holes should never exceed 5 m. Care must also be taken to ensure that these vent holes are positioned such that they are not within the depth of the floor construction.

The steel shell contains the concrete and prevents direct flame impingement, both preventing progressive spalling and reducing the rate of degradation of the core.

Failure of the column will occur when the combined strength of the steel and concrete has reduced to the level of the applied load.

Parameters affecting fire performance The most significant design parameters affecting the performance of concrete filled columns in fire are detailed below.

Material strength From solely a fire resistance perspective, the most efficient column is obtained by using a high strength concrete with a thin-walled grade S275 steel section. This gives the most advantageous ratio of concrete load capacity to overall column strength.

Column size As the external size of a column increases, the cross-sectional area of the concrete core will increase at a faster rate than that of the steel. Accordingly, the core of a larger sized column will support a greater proportion of the total load than a smaller one.

External protection (if applied) Externally applied protection will reduce the rate of heating and will therefore increase the fire resistance. The use of external protection will normally eliminate the need to reinforce the concrete core.

Applied load The lower the level of the applied load, the lower the stresses produced, and the longer the period of fire stability of the column.

Effective length For short columns (i.e., columns with effective lengths up to approximately 12 times the column width), failure will occur when the combined strength of the materials reduces to a level that is less than the applied load (i.e., a crushing failure). As the column length increases, failure will become progressively more related to instability considerations. EC4-1-2 allows a reduction in the effective length of columns in fire (see Section 3.6.1 for further details).

Bending moments and eccentricity When the flexural stiffness of the steel shell is lost, the bending resistance of the column is significantly reduced, because of the relatively low value of the elastic modulus for concrete and its poor flexural strength compared to steel.

The effect of accidental eccentricity and out-of-straightness in a column subjected to nominal axial loads is not significant for short columns. However, this can be a significant factor for slender columns in fire.

38

The presence of moments in the column produced by either end moments or an eccentrically applied end load (cleat loading) also has a significant effect on the fire stability.

Reinforcement The presence of reinforcement in the concrete will improve the flexural and axial properties of the core and so improve the fire resistance, particularly where buckling stability and/or bending moments are major factors.

The Eurocodes use the parameter ‘axis distance’ and not concrete cover to describe the position of the reinforcement within the concrete core. Axis distance is measured from the centre of the bar to the inside of the steel tube, as shown in Figure 3.1.

3.2 Protected columns Concrete-filled structural hollow section columns with a design capacity based on the room temperature properties of the full cross-section may be protected against fire with externally applied insulating materials. The presence of the concrete core will substantially add to the heat sink of the column and so markedly reduced the effective Section Factor of the composite column. The Section Factor VAm / is the ratio of the exposed perimeter to the area of steel and is commonly used in assessing the required thickness of fire protection.

The thickness of fire protection material required for a concrete-filled structural hollow section column may be determined as follows:

1. Determine the resistance of the fully composite column according to EC4-1-1.

2. Calculate the utilization factor in fire, defined as:

C20ºat Resistancestatelimit fire at Load

=η

3. Using the utilization factor, assess the critical temperature for the section using the design procedures for a simple steel section according to ENV 1993-1-2: 1995[27] (EC3-1-2).

4. The effect of the heat sink, from the presence of the concrete core, is to increase the Section Factor of the steel section. As a result, the effective Section Factor may be calculated based on an increased wall thickness of the steel section as follows[31, 32]:

cesse ttt +=

Axisdistance

Axisdistance

Figure 3.1 Illustration of axis distance for CHS and SHS sections

39

with ice bt 15.0= for Tbi 12<

Ttce 8.1= for Tbi 12≥

where: ts is the wall thickness (mm).

tce is the effective increase in wall thickness due to the concrete core (mm). bi is the minimum dimension of the concrete core (mm). T is the fire resistance time (mins).

5. Calculate the effective Section Factor.

se

m

tVA

onbasedSHSofAreaSHSofarea Surface

=

6. The thickness of fire protection material required can then be obtained in the conventional manner.

The above method for determining the thickness of externally applied fire protection may be used for any passive insulating materials. The same approach may also be adopted for intumescent coatings, provided that suitable test evidence is available.

3.3 Unprotected columns The method proposed in this section directly follows the principles of Eurocode 4. EC4-1-2 gives general design methods for composite columns and, in Annex G, specific rules for concrete filled hollow sections. Following a calibration exercise, SCI and Corus concluded that a combination of the two methods was the best approach, which is fully in line with the principles of this Eurocode.

For the combined buckling resistance and bending check, SCI and Corus have adopted a more conservative approach than that recommended by EC4-1-2. This approach takes into account the slenderness of the column; a factor presently ignored in the Eurocode.

Tables to allow an estimate of the column size to achieve the required fire resistance are given in Section 6.

The Eurocode states, in general principles, that the buckling resistance of a column may be computed by using buckling curve ‘c’. The squash resistance and Euler buckling resistances are calculated by integrating across the cross section taking into account the temperature distribution and corresponding material properties.

Thus, the buckling resistance is given by:

Rdpl,fi,Rdfi, NN χ=

40

where: χ is the reduction coefficient for buckling obtained from curve ‘c’ of Section 5.5.1, of

EC3-1-1 and is dependant on the non-dimensional slenderness ratio θλ .

Rdpl,fi,N is the design value of the plastic resistance to axial compression in the fire situation.

The calculation is carried out in five stages, as follows:

1. Carry out a thermal analysis to establish the temperature distribution throughout the cross-section.

2. Calculate the properties of the cross-section. 3. Calculate the buckling resistance of the column. 4. Calculate the bending resistance of the column. 5. Calculate the effect of interaction between axial load and bending.

DD ENV 1991-2-2: 1996[26] (EC1-2-2) gives an equation for determining the heat transfer, which includes two partial factors. These are applied to the radiation and convection terms. They are included to allow heating rates in different furnaces to be modelled.

Rate of heat input (W/m²) = ( ) ( ) ( )[ ]4s

4grrsgc 273273 +θ−+θσφεγ+θ−θαγ

where:

sθ is the steel temperature (ºC)

gθ is the gas temperature (ºC)

rε is the effective emmissivity σ is the Stefan Boltzmann constant (5.67 × 10-8 W/m² ºK4)

Values for the other parameters are given in EC1-2-2, as follows:

α is the convection coefficient = 25 W/m2 ºK cγ is the partial factor for convection (a boxed value) = 1.0

rγ is the partial factor for radiation (a boxed value) = 1.0 φ is the orientation (view) factor = 1.0

The UK NAD to EC1-2-2 specifies 0.45 for rγ . From a consideration of test data, SCI and Corus consider this value to be too low[4] for calculating the heat input to concrete filled hollow sections. It is therefore recommended that a value of 1.0 for rγ is more appropriate for concrete filled hollow sections.

3.4 Partial safety factors During 2002 the UK National Application Document (NAD) for EC4-1-2 is expected to be published. This will specify the partial safety factors and combination factors that should be used in the UK. The relevant factors that are due to be given in the NAD are:

41

Partial material factors:

Steel, M,fi,aγ 1.0 Concrete, M,fi,cγ 1.1 Reinforcement, sM,fi,γ 1.0 Combination factors, 1,1ψ , for imposed floor loading:

Storage 0.9 Escape stairs and lobbies 0.9 Plant 1.0 All other areas 0.7 The NAD will also give more conservative values than the Eurocode for the effective length of columns in fire (see Section 3.6.1).

3.5 Properties of the cross-section It is generally not feasible to carry out design calculations by hand and use of the ConcFill software is advised (see Section 6). The first step is to divide the cross-section into a number of elements, as part of a finite difference thermal analysis. The structural section properties are then calculated by summing the properties of all the elements, with due account being paid to the effect of the temperature on each element.

3.5.1 Squash (plastic) resistance, Nfi,pl,Rd The squash resistance of the composite column is calculated from the following summation:

∑∑∑ γ+

γ+

γ=

m M,fi,c

c,c,

k sM,fi,

max,s,s,

j M,fi,a

max,a,a,Rdpl,fi,

θθθθθθ fAfAfAN

where:

θa,A is the area of an element of the steel section.

θs,A is the area of an element of the reinforcement.

θc,A is the area of an element of the concrete.

θmax,a,f is the strength of the steel section element at temperature θ (see Section 3.8.1).

θmax,s,f is the strength of a steel reinforcement bar at temperature θ (see Section 3.8.2).

θc,f is the strength of the concrete element at temperature θ (see Section 3.8.3).

ifi,M,γ is the partial material safety factor for the element.

3.5.2 Effective flexural stiffness The effective flexural stiffness of the composite column ( ) efffi,EI can be determined from:

( ) θθ,θθ,θ c,m

c,s,k

s,a,j

,a,efffi, IEIEIEEI ∑∑∑ σσσθ ++=

where:

42

θa,I , θs,I and θc,I are the second moment of area of each element.

σθ,a,E , σθ,s,E and σθ,c,E are the tangent moduli of the stress-strain relationship for the material of each element at its temperature θ and stress σ.

3.6 Column buckling resistance 3.6.1 Effective length of composite columns in fire The effective (buckling) length of composite columns in fire can be determined from Clause 4.3.6 of EC4-1-2, which is shown schematically in Figure 3.2. For columns within an intermediate storey of a building, the buckling length of the column is 0.5 times the system length. For columns on the top storey, the buckling length should be taken as 0.7 times the system length.

In the draft UK NAD for EC4-1-2 the factors of 0.5 and 0.7 have been conservatively increased to 0.7 and 0.85 respectively.

3.6.2 Buckling resistance The buckling resistance is calculated from the squash resistance and the Euler (elastic) critical load using the EC3-1-1 buckling curve ‘c’. The Euler buckling load is given by:

where:

θl is the buckling length of the column in fire The non-dimensional slenderness ratio is given by:

crfi,

Rdpl,fi,

NN

=λθ

L

L

L

Shear wall or other bracing system

L

0.7L

0.85L

Figure 3.2 Effective length of columns in fire conditions (N.B., the values of 0.7 and

0.85 are taken from the draft UK NAD)

2 ( ) 2

eff fi, cr fi,

θ

π =

l

EI N

43

The resistance of a composite column in axial compression (buckling load) is obtained from:

Rdpl,fi,Rdfi, NN χ=

where: χ is the reduction coefficient for buckling obtained from curve ‘c’ of Section 5.5.1, of

EC3-1-1 and is dependant on the non-dimensional slenderness ratio θλ .

The reduction factor may be determined from:

[ ] 5.022

1

θλφφχ

−+= but 0.1≤χ

where:

φ ( )[ ]22.015.0 θθ λ+−λα+=

α is an imperfection parameter depending on the buckling curve considered. For buckling curve ‘c’ the imperfection factor, α, is 0.49.

3.7 Combined compression and bending The design method can be applied to columns in multi-storey steel frames. These columns should carry predominately axial loads. It is recommended that the ratio of applied moment to moment capacity should not exceed 0.67.

The calculation of the plastic bending resistance is broadly similar to the calculation of the squash load and flexural stiffness. Two steps have to be carried out.

Find the plastic neutral axis such that:

∑∑∑ =γ

+γ

+γ

0cfi,M,

c,c,

sfi,M,

max,s,s,

afi,M,

max,a,a, θθθθθθ fAfAfA

where the stresses are considered as positive in tension or negative in compression.

Calculate the plastic moment resistance:

∑∑∑ γ+

γ+

γ= y

fAy

fAy

fAM

cfi,M,

c,c,

sfi,M,

max,s,s,

afi,M,

max,a,a,Rdfi,

θθθθθθ

where y represents the distance from the centroid of each element to the plastic neutral axis.

The interaction of axial compression and bending is checked using the expression below. This is a more conservative approach than that given in the Eurocode.

0.1Rdfi,

fyy

Rdfi,

fxx

Rdfi,

f ≤++=MM

kMM

kN

NR

44

where:

fN is the applied load.

fxM is the maximum applied moment about the major axis.

fyM is the maximum applied moment about the minor axis.

xk , yk are modification factors based on slenderness (see Section 3.7.1)

For a circular section, the moments about orthogonal axes can be added vectorially.

In cases where there are no applied moments, or the moments are small, it is recommended that a nominal moment about the minor axis is assumed.

The nominal moments can be assessed according to existing practice in BS 5400-5: 1979[30] as follows:

For a CHS ffx 03.0 DNM = For a RHS ffy 03.0 BNM = where: D is the diameter of the CHS B is the width of the RHS

3.7.1 Determination of k factor The factor k is a moment multiplier (k ≥ 1.0) used to take account of any possible second order effects in slender columns in accordance with ambient temperature composite column design to EC4-1-1.

The factor should be assessed separately for each axis, since it is dependent on the variation of end moments on an axis, as well as the slenderness on that axis.

Second order effects are only to be considered significant if both of the following conditions are satisfied:

1) 1.0Rdfi,

f >N

N

2) ( )r−>λθ 22.0

where:

r is the ratio of the smaller to the larger end moment (see Figure 2.5). If there is any transverse loading, r should be taken as 1.0.

If second order effects are not significant then k = 1.0.

Otherwise:

0.11

cr,

≥−

=

f

f

NN

k β

45

where: β is the equivalent moment factor.

For columns with transverse loading within the column length, the value for β should be taken as 1.0. For pure end moments, β can be determined as follows:

r44.066.0 +=β but 44.0≥β

3.8 Material properties For the design of the resistance of concrete filled structural hollow sections in fire, the design method of EC4-1-2 Annex G uses a special set of material properties. These are slightly different from those used for the design of other composite elements (see Section 2.2).

During the development of EC4-1-2, the calibration of the Annex G method was carried out using a set of material properties that were different from those generally adopted for other elements in the Eurocode. Due to the fact that there was insufficient time to recalibrate the method, the material properties for the design of concrete filled sections have remained unchanged.

The material properties used in Annex G are reproduced below.

3.8.1 Hot rolled structural steel

The ratio Cay,20

ay,θ

°ff

is given by:

If 0 < θ ≤ 600ºC

θ

θ+=

°

1750log900

0.1

eCay,20

ay,

ff θ

If 600 < θ ≤ 1000ºC

( )( )240

34.0340

Cay,20

ay,

−θθ−

=°f

f θ

The ratio Ca,20

a,

°EE θ is given by:

If 0 < θ ≤ 600ºC

θ

θ+=

°

100log2000

0.1

eCa,20

a,

EE θ

If 600 < θ ≤ 1000ºC

46

( )

( )3.5369.0690

Ca,20

a,

−θθ−

=°E

E θ

where: Cay,20°f is the yield stress of the structural steel at 20ºC.

θay,f is the yield stress of the structural steel at temperature θ.

Ca,20°E is the modulus of elasticity of the structural steel at 20ºC.

θa,E is the modulus of elasticity of the structural steel at temperature θ.

3.8.2 Steel reinforcement bars

The ratios Csy,20

sy,

°ff θ and

Cs,20

θs,

°EE

are obtained from the following table:

Table 3.1 Strength reduction factors for steel reinforcing bars

Temperature sθ (ºC) 0 400 580 750

Csy,20

sy,

°ff θ and

Cs,20

θs,

°EE

1.00 1.00 0.15 0

3.8.3 Structural concrete

The ratios Cc,20

θc,

°ff

and Cc,20

c,

°EE θ are obtained from the following table:

Table 3.2 Strength reduction factors for concrete

Temperature cθ (ºC) 0 50 200 250 400 600 1000

Cc,20

θc,

°ff

1.00 1.00 1.00 1.00 0.76 0.45 0

Cc,20

c,

°EE θ 1.00 1.00 0.50 0.41 0.15 0.05 0.05

Cc,20°E is obtained from Table 2.2

47

4 FABRICATION AND CONNECTIONS

4.1 General This section illustrates the more salient points to be considered when detailing the connections to concrete filled columns so as to ensure economy and efficiency in design. The details will follow standard structural engineering practice and will depend on the nature of the loads and the type of beam or flooring present.

4.1.1 Column Venting It is necessary to provide ventilation holes in the column walls to prevent the dangerous build up of steam pressure inside the column in the event of a fire. Two 20 mm diameter holes placed diametrically opposite each other both at the top and bottom of each storey height have been used in testing and have proved to be adequate. Care must be taken that the holes are positioned outside the level of any floor slab or screed.

A drain hole should be also provided at the base of a column to prevent water collecting if it is left standing empty on site prior to filling (Figure 4.1). This hole may be one of the steam vent holes.

4.1.2 Welding to Concrete Filled Columns Primary load bearing structural components should be welded to the column before filling with concrete. The welding of relatively small lugs and fittings after filling is permissible.

4.2 Column Splices and Flange plate connections Column splices are generally made either by butt-welding lengths together, or by providing bolted flange connections at the ends of column lengths. Some common examples of these connections are shown in Figure 4.2.

Drain andvent hole

Figure 4.1 Column base plate and drain hole

48

Column Splice Connections Flange plate connection Section through floor with flange plates on RHS or CHS columns cast into concrete slab. The bottom plate has projecting studs welded to it; for flange plates connected above the slab, in a raised floor space, a normal bolting arrangement may be used. In general, the flange plate connection provides transfer of axial forces but only limited moment transfer.

Welded splice connection Elevation on RHS column with weld backing strip and angle brackets for erection; brackets are removed after some welding completed; backing strip also used as alignment spigot. For CHS Columns, vertical plates are used instead of brackets.

Welded splice connection with long bolts The use of long bolts with double nuts permits accurate alignment of the SHS and backing material before site welding.

Figure 4.2 Column splice connections

Particular details that should considered when using these two connection types are as follows:

(i) Welded Splices

To allow welding to take place, the concrete filling should be stopped at a level approximately 250 mm below the top of the column. The next column length can then be spliced into place using a full penetration butt weld with a backing ring. Following this, concrete filling can be resumed.

49

(ii) Bolted Flange Joints

When a flanged connection is used, the flange should be welded to the column before filling. When flanges are used at both ends of a column length, at least one flange will have to incorporate an aperture for filling (see Figure 4.3). In order to ensure complete filling under this flange, the maximum size of any lip should not exceed 15 mm.

4.3 Beam-to-column connections Simple Connections Simple connections are normally assumed to give vertical support but to provide only limited restraint against rotation: these connections are assumed to be able to rotate without damage. A selection of some common examples is shown in Figure 4.4 below.

Drain and vent hole

Vent holeA A

Section A - A

15 mm max.

Detail 1

Detail 1

Figure 4.3 Column cap plate and splice detail for flange plate connections

50

Reverse channel connection Section through floor with steel beams connected to CHS columns by a channel fixing; a standard flexible end plate is normal at the beam end, which then develops only nominal moments. A similar detail is used for RHS columns.

Web cleat connection Section through floor with secondary and primary beams which are connected to RHS columns by a T-section web cleat; this connection can have greater stiffness and robustness than the equivalent fin plate connection.

Fin plate/Shear tab connection Section through composite deck floor with steel beam connected to RHS or CHS column by a finplate; a very economic joint; a seating cleat may be used to help erection, with removal afterwards if required.

Hollobolt or Flowdrill connection Section though floor with steel beam connected to face of RHS column by Hollobolts or fully threaded bolts in Flowdrill holes; only nominal moments develop with flush or partial depth flexible end plates; thicker plates can develop between 10 to 15 per cent of beam capacity.

Figure 4.4 Beam-to-column joints – simple connections

51

The most common arrangements for beam-to-column joints make use of bolted connections via attachments welded to the faces of the hollow section column. By far the most common connection of this type is the fin plate connection, using a flat plate welded to each column face. For RHS columns, an alternative is the web cleat connection, using single angle sections, or T-sections, welded to the column face. The use of double angles is a further option and provides greater capacity than a single angle would. An increasingly popular option for CHS or RHS columns is the use of the reverse channel connection. Of these, the most economic is the fin plate connection or the web cleat connection using a single angle. Note also that it is not necessary to have a continuous weld of the bracket to the column when under internal conditions of exposure.

Simple steel connections using flexible end plates or double angle cleats, which are bolted direct to the column, are also possible with RHS columns. These joints use either expanding bolt types, such as Hollobolt, or fully threaded bolts in tapped holes produced by the Flowdrill system.

Moment connections Moment Connections are those that are assumed to give vertical support, provide a degree of restraint against rotation and develop some moment capacity. A selection of some common examples is shown in Figure 4.5 below.

Rigid or semi-rigid moment connections are feasible with all types of hollow section column. These may use flange plates or beam stubs, which are usually of the same section as the beam being connected. Through-plated connections are another popular type of moment connection; this is similar in appearance to the fin plate connection but has slots in the column to allow a single plate to be taken through it. In almost all cases, moment connections are more expensive than simple connections, but the extra cost of the connection can be more than offset by savings in beam sizes, or by provision of more usable floor space.

Beam-to-column connections in fire Connections generally need to be either fire protected or shielded by other elements such as floor slabs.

Questions sometimes arise as to how the loads are transferred into a concrete filled column from an incoming beam, when the outer steel tube is exposed to fire below the connection. If bolts or studs protrude into the concrete core, then some degree of direct load transfer can be envisaged. However, it will often be the case that the beam-to-column connection takes the form of a fin plate, resulting in no obvious direct load path into the concrete core. In this case, the load transfer is via the cold part of the steel tube above the connection. The beam reaction is therefore resisted by tension in the tube, which is transferred into the concrete core by a combination of shear, bond and direct axial load via the column cap.

52

Composite beam moment connection Section through floor with composite beam connected to RHS or CHS columns by fin plates and bottom flange plate to provide continuous or semi-continuous joints; the top reinforcement in the concrete slab is designed to provide the other arm of the moment couple.

Steel beam moment connection Section through floor with secondary steel beams on top of primary beams which are connected to RHS or CHS columns by fin plates and flange plates to provide continuous or semi-continuous joints.

Stub connection Section through floor with stub beam sections welded to face of RHS columns; moment capacity usually limited by yielding of column face; large moments require internal diaphragms or external flange plates.

Through-plate connection Section through floor with through-plate passed through slots and welded to each face of RHS or CHS column; the through plate connection allows significant axial forces and bending moments to be transferred from the beam, if this is required.

Figure 4.5 Beam-to-column joints – moment connections

4.4 Shearhead connections to slabs The use of steel decking or precast concrete floor slabs, supported on steel beams, can lead to economical designs. In each of these cases, the fixing of the steel beams to the concrete filled hollow sections can be achieved by using the connections shown in Figure 4.4 and Figure 4.5.

53

In situ concrete beams and floors may also be used with concrete filled hollow sections. Figure 4.6 illustrates some shearhead connection details to provide shear and moment transfer between reinforced concrete flat slabs and composite columns.

Centre stub shearhead Section through concrete flat slab floor with column stubs welded to RHS or CHS columns; a stiffening collar may be required around the column top when large moments are produced in it from unbalanced loads.

Grid shearhead Section through concrete flat slab floor with grid welded to faces of RHS columns; the connection is designed for transfer of tensile and compressive forces from the slab to the shearhead elements.

Figure 4.6 Shearhead connections to in situ concrete slabs

4.5 Base plate connections With the exception of the need to provide a drain/ vent holes in the wall of the hollow section (see Figure 4.1), no specialised details are required for base plate connections. A typical example of a base plate for a concrete filled column using SHS is shown in Figure 4.7.

Baseplate with loose bolts Section through base with cast-in bolts in boxes; bolts are loosened soon after concrete has set; underside of base and bolt boxes are grouted after baseplate is packed up and column is plumb.

Figure 4.7 Base plate connections

54

5 PRACTICAL CONSIDERATIONS

5.1 Concrete Filling

5.1.1 General This section has been written on the assumption that the hollow sections are to be filled with concrete on site. For sections filled under factory conditions, the same basic requirements need to be met; however, the more closely controlled conditions in the factory could have advantages, particularly when high strength concretes are being used.

It must be emphasised that the concrete plays an important structural role and it will only do this if a good standard of concrete design, concreting procedure, control and site supervision is used.

In many cases, grout is an acceptable alternative to concrete for filling hollow sections and in most cases the following recommendations also apply to grout filling.

5.1.2 Concrete materials, testing and control Good control over concrete materials and testing is required, especially for higher strength mixes. The testing of materials for concrete, and of the concrete itself, should be in accordance with the standards and regulations generally accepted or laid down in the country in which the work is to be carried out. General information on concrete and concreting may be found in textbooks on the subject.

5.1.3 The concrete mix The concrete must fill the steel section completely. Because it is not possible to inspect the concrete visually after placing, there must be no doubts about the ability of the concrete and the concreting procedure to meet these requirements. Due consideration of the following recommendations will ensure, as far as possible, that these requirements are satisfied.

The concrete should have the following properties:

(i) sufficient workability to ensure proper compaction;

(ii) sufficient cohesiveness to reduce the likelihood of segregation and bleeding;

(iii) the specified 28-day strength;

(iv) when filling a column with a plain concrete mix, the maximum size of aggregate should be 10 mm for 100 mm minimum section columns, 20 mm for 140 to 180 mm minimum section columns and 40 mm for 200 mm minimum section columns; and

(v) when a reinforcement cage is inserted into the concrete for additional capacity, or to improve the fire performance, the clearance between the reinforcement bar outer fibre and the SHS inner wall should be at least 5 mm larger than the maximum aggregate size.

The workability of the concrete should be high enough to ensure its proper compaction, but not too high otherwise most of the other desirable properties of the mix will suffer. With the exception of bottom pressure filling (see below), an average slump of about 50 mm will probably suffice in most instances. The attainment of high strength may make stiffer mixes necessary, but this will increase the difficulties of placing and compaction. A lower slump will be possible where large sizes or short column lengths make placing and compaction easier.

55

Given the desired workability and knowledge of the available aggregates, a mix to give the required strength may be designed in accordance with the recommendations of any of the standard publications.

In order to obtain the required cohesiveness and to reduce bleeding it is advisable to increase the proportion of fine aggregate in the total aggregate content by up to about 5% more than would normally be used, especially for smaller size columns. Fine sands produce more cohesive mixes than coarse sands.

A water-reducing admixture may help throughout the strength range and may be very useful when high strength concretes are specified.

5.1.4 Preparation for placing The internal surface of the hollow section needs no special preparation. However if cutting oils have been used, any deposits of oil in the region of the cut should be removed. It is important that all loose material and debris be removed from the bottom of the column. Any free water, which may have collected at the bottom, must also be removed.

5.1.5 Placing and compacting the concrete Conventional top filling Special plant will not normally be required. If a good cohesive mix has been obtained it may quite satisfactorily be dropped into the column from the top. However, care must be taken to ensure that the first 100 mm or so of concrete in the bottom of the column does not contain an excessive amount of coarse aggregate, which would lead to ‘honeycombing’, or voids between aggregate particles.

The concrete is best compacted by an internal (or poker) vibrator: this type should be used whenever possible. The vibrator should be lowered to the bottom of the column and switched on immediately before any concrete is placed. The concrete should be placed slowly around the vibrator, which must be left running continuously, and raised as the concrete level rises, such that its position is just below the surface of the concrete; its position may be judged by raising and lowering it, and noting the change in sound as it emerges from the concrete.

Placing the concrete requires careful consideration because of the cost of the plant required to handle relatively small quantities. The majority of concrete used on site is ready-mixed, and this is most economical if it can be used in 6 m³ loads, which need to be placed within half an hour if waiting charges are to be avoided. A crane and skip could handle concrete at this rate into the large columns. A mobile pump with a boom can place the concrete continuously as it is being compacted and, provided that the poker diameter is at least 50 mm, this rate of placing could be achieved in medium sized columns. However, concrete in smaller columns should be placed at a proportionally slower rate, in order to achieve satisfactory compaction. Mortar pumps with 50 mm diameter rubber hoses could be used, provided that the maximum aggregate size does not exceed 10 mm. Generally, each time the maximum size of aggregate is halved, the cement content needs to be increased by 20%.

If it is decided to mix the concrete in small batches on site, it is possible to manhandle the concrete in buckets and to compact it with a 40 to 50 mm square timber tamper. With care, compaction by this method is at least as good as vibration.

External vibration is also possible; however, a lot of energy will be absorbed by the steel column. For this reason, a very powerful clamp-on vibrator is necessary; this should be moved up as the concrete surface rises.

56

The concrete may settle during the first hour in hot weather, or perhaps three hours in cold weather. This may be overcome by adding extra concrete during placing, and trowelling it off flush with the top of the cap plate after the concrete has settled, but before it is too stiff to trowel.

Reworking concrete is not harmful provided that the concrete is not ‘loosened’ in the process, and no extra water is added.

Drain holes that are present in the column to prevent the build up of water on site, and to allow the escape of moisture in the event of fire (see Section 4.1.1), should be temporarily plugged whilst the concrete is being placed. If this is not done, grout and fine material will escape from the mix, leaving an area of weak honeycombed concrete inside the column. The plugs should be removed after the concrete has achieved its initial set.

Bottom pressure filling Bottom pressure filling of composite columns using hollow sections has, in the past, been proved to be very successful in Australia, which has led to it being widely employed in Japan and North America.

The significant features of this method of filling are:

• Pumping heights of 24 m can be readily achieved with normal pumps, and up to a maximum height of 200 m for special pumps.

• Pumping from the base of the column avoids aggregate segregation; vibration of the concrete, whether it being internal or external, is not normally required.

• Mixes need to be of a high workability, with slumps exceeding 150 mm. (However, it is recommended that the workability of a mix be established from the flow table test, in accordance with BS 1881-105[33]).

• Peak pumping rates of 50 m³/hour are achievable.

• Although a typical concrete pump pressure is 60% greater than the hydrostatic head (using a 125 mm standard pump), this pressure may rise depending on the number of obstructions within the column (e.g., reinforcement cages).

• No surface preparation is necessary at breaks in the concrete, prior to pouring the next (upper) section.

Figure 5.1 shows a typical cut-off valve used as an entry point for bottom pressure filled sections. The valve is usually located at between 300 to 450 mm above the column base, and is welded to the wall of the hollow section (thereby resulting in a permanent fitting). The size of valve appropriate for a standard 125 mm diameter supply pipe will usually dictate that, for practical purposes, a column with a minimum cross-sectional dimension of 300 mm is needed. As a consequence, for columns with smaller cross-sectional dimensions, conventional top filling is preferred.

57

5.1.6 Curing the concrete The main objects of curing are to prevent early drying out of the concrete, and to protect it from frost. The concrete can only dry out at the top surface and, in most cases, a polythene sheet or bag placed over the top and tied or taped around the column will provide the necessary protection.

5.1.7 Concreting in cold weather Concrete must not be allowed to freeze during the early stages of setting and hardening. The steel section has negligible insulating properties so, if cold weather is expected, concreting should be suspended, or insulation and heating (if necessary) provided.

In frosty weather the steel section should be warmed above 0°C to melt any ice on, or in the column. Care should be taken to ensure that all constituents of the concrete are free of ice. After placing, the columns should be wrapped in insulating quilts to prevent the temperature of the concrete dropping below 10°C for the first day, or even two days for 20 N/mm2 concrete with a low initial temperature. These temperatures ensure that the concrete matures sufficiently before it is allowed to freeze.

Figure 5.1 Cut-off valve used in bottom pressure filled composite columns

58

6 SOFTWARE

ConcFill2 is a program that will check a composite column at the normal and the fire limit state. All calculations are in accordance with EC4-1-1 and EC4-1-2, as described in Sections 2 and 3. In fire, applied protection is not considered. ConcFill2 replaces ConcFill, which was an earlier version distributed on CD by Corus in 1999.

The program uses a sophisticated load specification in which dead and imposed loads may be applied centrally and at user defined offsets at the top and bottom of a storey height. Moments may also be applied at the top and bottom of the column. Loads applied at offsets are used to generate additional moments. Moments are considered in two directions and effective lengths may be specified in two directions separately for both normal and fire conditions.

Facilities are included which make the positioning of reinforcement very straightforward. It is also possible to rotate the section without having to re-enter applied loads.

The user may select the section size from the large range supplied by Corus.

The extensive output may be viewed on the screen before being printed. This includes: details of all factored and unfactored applied loads and moments generated by load offsets; cross-section resistances and column buckling resistances for both normal and fire conditions; and unity factors, which give an indication of critical sensitivities and efficiency.

ConcFill2 is an analysis program and the user has to iteratively arrive at an appropriate section size. Table 6.1 has been developed to assist in the initial selection process.

The table is based on the following assumptions.

Concrete grade 30/37 Steel grade Celsius 355 Column length 3.5 Effective length (normal) 0.85 Effective length (fire) 0.7 Reinforcement Approximately 4% Axis distance 60 minutes 40 mm

90 minutes 50 mm 120 minutes 60 mm

Enquiries about the availability of ConcFill2 should be addressed directly to Corus†

† Corus Tubes, Structural & Conveyance Business, PO Box 6024, Weldon Road, Corby, Northants, NN17 5ZN , the United Kingdom UK Freephone Technical Helpline : +44 (0) 1724 405060 Fax: +44 (0) 1536 404127 Email: corustubes.s-c@corusgroup.com

59

Table 6.1 Column buckling resistances and bending resistances for initial sizing in ConcFill2

Fire resistance (mins) Section type

Section size Rebar Normal design 60 90 120

168.3x 10 4 x 16 dRN

dRM

1858 96

481 18

178 7

- -

193.7x10 4 x 16 dRN

dRM

2361 132

691 26

293 10

- -

219.1x12.5 4 x 20 dRN

dRM

3394 209

1162 47

589 23

221 11

244.5x16 4 x 20 dRN

dRM

4666 320

1636 75

899 35

355 15

273x16 4 x 25 dRN

dRM

5593 418

2240 111

1426 61

711 29

323.9x16 4 x 25 dRN

dRM

7022 604

2940 158

2062 92

1386 58

355.6x16 6 x 25 dRN

dRM

8276 758

3876 233

2906 149

2290 114

406.4x16 6 x 32 dRN

dRM

10299 1054

5590 389

4501 276

3869 228

457x16 6 x 32 dRN

dRM

12040 1358

6627 479

5423 337

4817 283

Circular Hollow Section

508x16 6 x 32 dRN

dRM

14300 1768

8504 674

7137 493

6521 425

160x160x12.5 4 x 16 dRN

dRM

2663 161

552 22

142 5

- -

180x180x16 4 x 16 dRN

dRM

3747 244

810 33

280 10

- -

200x200x16 4 x 20 dRN

dRM

4473 318

1183 53

540 22

107 5

250x250x16 4 x 25 dRN

dRM

6320 546

2134 112

1439 67

777 35

300x300x16 4 x 32 dRN

dRM

8334 846

3504 220

2627 146

1835 89

350x350x16 4 x 32 dRN

dRM

10201 1169

4494 294

3389 186

2621 121

Square Hollow Section

400x400x16 8 x 32 dRN

dRM

12906 1691

7008 560

5628 404

4744 309

Note: dRN is the design buckling resistance (kN) and dRM is the design bending moment resistance (kNm).

60

7 REFERENCES

1. BRITISH STANDARDS INSTITUTION BS EN 10210-1: 1994 Hot finished structural hollow sections of non-alloy and fine grain structural steels. Technical delivery conditions.

2. BRITISH STANDARDS INSTITUTION BS EN 10210-2: 1997 Hot finished structural hollow sections of non-alloy and fine grain structural steels. Tolerances, dimensions and sectional properties.

3. BRITISH STANDARDS INSTITUTION DD ENV 1994-1-2: Eurocode 4 Design of composite steel and concrete structures: Part 1.2: General rules Structural fire design (in preparation, available late 2000).

4. NEWMAN, G.M. and SIMMS, W.I. The fire resistance of concrete filled tubes to Eurocode 4 The Steel Construction Institute, P259, 2000.

5. BRITISH STEEL TUBES & PIPES Fire resistant design. A guide to evaluation of structural hollow sections using BS 5950 Part 8 TD408, 1998.

6. BRITISH STEEL TUBES & PIPES Fire resistant design. A guide to evaluation of structural hollow sections using BS 5950 Part 8 TD409, 1998.

7. BRITISH STEEL TUBES & PIPES Intumescent coatings & SHS concrete filled columns TD410, 1998.

8. CORUS TUBES Hot finished SHS for multi-storey columns TD416, 1999.

9. WEBB, J. and PEYTON, J.J. Composite concrete filled steel-tube columns Proceedings of the 2nd National Structures Conference Institution of Engineers Australia, 1990, pp. 181-185.

10. SEWELL, J.S. Columns for buildings Engineering News, Vol. 48, No. 17, 1902.

11. KLÖPPEL, K. and GODER, D. Collapse load tests on concrete filled steel tubes and the derivation of design formulae Der Stahlbau, 1957.

61

12. BOUÉ, P. Concrete filled steel stanchions Acier Steel, No. 9, 1957, pp.351-356.

13. NEOGI, P.K., SEN, H.K. AND CHAPMAN, J.C. Concrete filled tubular steel columns under eccentric loading The Structural Engineer, Vol. 47, No. 5, 1969, pp. 187-195.

14. KNOWLES, R.B. AND PARK, R. Strength of concrete filled steel tubular columns Journal of the Structural Division, ASCE, Vol. 95, No. 2, pp. 2565-2587.

15. KNOWLES, R.B. AND PARK, R. Axial load for concrete filled steel tubes Journal of the Structural Division, ASCE, Vol. 96, No. 10, pp. 2125-2155.

16. BRITISH STANDARDS INSTITUTION DD ENV 1992-1-1: 1992 Eurocode 2 Design of concrete structures: Part 1.1: General rules and rules for buildings.

17. EUROPEAN COMMITTEE FOR STANDARDIZATION Draft prEN 1994-1-1: 2001 Eurocode 4 Design of composite steel and concrete structures: Part 1.1: General rules and rules for buildings.

18. BRITISH STANDARDS INSTITUTION prEN 1994-1-1:2002 Eurocode 4 Design of composite steel and concrete structures: Part 1.1: General rules and rules for buildings.

19. BRITISH STANDARDS INSTITUTION BS 4449:1997 Specification for carbon steel bars for the reinforcement of concrete.

20. BRITISH STANDARDS INSTITUTION DD ENV 10080: 1996 Steel for the reinforcement of concrete. Weldable ribbed reinforcing steel B500. Technical delivery conditions for bars, coils and welded fabric.

21. COUCHMAN, G.H. and WAY. A. Joints in steel construction: Composite connections The Steel Construction Institute/The British Constructional Steelwork Association, P213/98, 1998.

22. BRITISH STANDARDS INSTITUTION BS 5950-8: 1990 Structural use of steelwork in building: Part 8: Code of practice for fire resistant design.

23. BRITISH STANDARDS INSTITUTION BS EN 10210-1: 1994 Hot finished structural hollow sections of non-alloy and fine grain structural steels: Part 1: Technical delivery requirements.

62

24. BASU, A.K. and SOMERVILLE, W. Derivation of formulae for the design of rectangular composite columns Proceedings of the Institution of Civil Engineers Supplementary Volume 1969, pp. 233-280

25. VIRDI, K.S. and DOWLING, P.J. Composite columns in axial bending Axially compressed structures (Vol. 1 of Stability and Strength Series – Editor: R. Narayanan) Applied Science Publishers, 1982

26. BRITISH STANDARDS INSTITUTION DD ENV 1991-2-2: 1996 Eurocode 1 Basis of design and actions on structures: Part 2.2: Actions on structures exposed to fire (together with United Kingdom National Application Document).

27. COMITÉ EUROPÉEN DE NORMALISATION (CEN) ENV 1993-1-2: 1995 Design of steel structures: Part 1.2: General rules Structural fire design.

28. BRITISH STANDARDS INSTITUTION DD ENV 1993-1-1: 1992 Eurocode 3: Design of steel structures: Part 1.1: General rules and rules for buildings.

29. COMITÉ EUROPÉEN DE NORMALISATION (CEN) prEN 1993-1-8: 2002 Eurocode 3: Design of steel structures: Part 1.8: Design of joints.

30. BRITISH STANDARDS INSTITUTION BS 5400-5: 1979 Steel, Concrete and composite bridges: Part 5: Code of practice for design.

31. EDWARDS, M. The performance in fire of concrete filled SHS columns protected by intumescent paint Proceedings of the 8th International Symposium on Tubular Structures, Balkema, Rotterdam, Netherlands, 1998

32. EDWARDS, M. The performance in fire of fully utilised concrete filled SHS columns with external fire protection Proceedings of the 9th International Symposium on Tubular Structures, Balkema, Rotterdam, Netherlands, 2000

33. BRITISH STANDARDS INSTITUTION BS 1881-105: 1984 Methods for the determination of flow

34. BECK, H. Nailed Shear Connection in Composite Tube Columns Paper No,35, Proceedings of the 1999 Eurosteel Conference, CVUT, Prague, ISBN 80-01-01963-2

This page intentionally

blank

www.corusgroup.com

Corus TubesStructural & Conveyance BusinessSales Enquires contact:UK Sales officePO Box 6024, Weldon RoadCorby, Northants NN17 5ZNUnited KingdomT +44 (0)1536 402121F +44 (0)1536 404127www.corustubes.comcorustubes.s-c@corusgroup.comTechnical Helpline (UK Freephone)0500 123 133 or +44 (0) 1724 405060

Corus TubesStructural & Conveyance BusinessSales Enquires contact:Netherlands Sales officePostbus 39 4900 BB OosterhoutThe Netherlands

T +31 (0)162 482300F +31 (0)162 466161corustubes.s-c@corusgroup.com

CT26/4:PDF:UK:04/2005

Care has been taken to ensure that the contentsof this publication are accurate, but Corus UKLimited and its subsidiary companies do notaccept responsibility for errors or for informationwhich is found to be misleading. Suggestionsfor or descriptions of the end use or applicationof products or methods of working are forinformation only and Corus UK Limited and itssubsidiaries accept no liability in respectthereof. Before using products supplied ormanufactured by Corus UK Limited customersshould satisfy themselves of their suitability.

English Language