ECCS Recommendations Simple Joints

(1)

ECCS TC10

Connections

European Recommendations for the Design of

Simple Joints in Steel Structures

(2)

European Recommendations for the Design of Simple Joints in Steel Structures

2

European Recommendations for the Design of Simple Joints

in Steel Structures

Nº126, 1st edition, 2009 Published by:

ECCS – European Convention for Constructional Steelwork [email protected]

www.eccspublications.eu

All rights reserved. No parts of this publication may be reproduced, stored in a retrieval sys-tem, or transmitted in any form or by any means, electronic, mechanical, photocopying, re-cording or otherwise, without the prior permission of the copyright owner

ECCS assumes no liability regarding the use for any application of the material and informa-tion contained in this publicainforma-tion.

(3)

3

TC10

Connections

European Recommendations for the Design of

Simple Joints in Steel Structures

J.P. Jaspart

J.F. Demonceau

S. Renkin

(4)

(5)

5

PREFACE

This document intends to provide European recommendations for the design of simple joints in steel structures.

Eurocode 3 Part 1-8 “Design of Connections” gives precise guidelines for the design of structural joints aimed at transferring bending moments. But for simple joints, information is only provided in Eurocode 3 for some specific failure modes. The way on how internal forces distribute amongst the various components within the joints is also not explicitly described.

The present publication fills this gap by proposing practical guidelines for the design of simple joints commonly used in Europe. The design rules presented in this document are in full agreement with the principles of Eurocode 3, and in particular of Eurocode 3 Part 1-8.

This document has been prepared at Liège University, editorially checked by Prof. D. Anderson from Warwick University and approved by the Technical Committee TC10.

The members of TC10 who contributed to the document are: Bijlaard F.S.K. (chairman) The Netherlands

Brettle, M. (secretary) United Kingdom

Aasen B. Norway

Anderson D. United Kingdom

Arda T.S. Turkey Bayo E Spain Beg D. Slovenia Braham M. Luxembourg Bucak Ö Germany Calado L. Portugal Dubina D. Romania Grecea D. Romania

Gresnigt A.M. The Netherlands

Girao A.M. Portugal / The Netherlands

Iglesias G Spain

Jaspart J.P. Belgium

Karamanos S.A. Greece

Kouhi J. Finland

Malik A United Kingdom

Moore D.B. United Kingdom

Nethercot D.A. United Kingdom

Puthli R.S. Germany

Ryan I. France

Sedlacek G. Germany

(6)

6

Steurer A Switzerland

Silva L.A.P.S. Portugal

Taylor J.C. United Kingdom

Ungermann D Germany

Veljkovic M. Sweden

Verhoeven J The Netherlands

Wald F. Czech Republic

Weynand K. Germany

(7)

7

With the aim of establishing a full design approach according to the general design principles stated in Eurocode 3, some design sheets for header plate and fin plate connections were prepared at the University of Liège and discussed at several meetings of Technical Committee 10 « Connections » of the European Convention for Constructional Steelwork (ECCS). The present report contains all these design rules. Explanations about these rules as well as indications on their range of validity are available in [10].

In a few years, it is expected that the practical design recommendations presented in this publication or in its eventual revised version will replace, in every country, the national normative documents or recommendations. In this way, it will simplify the free trade between the different European countries.

(10)

10

2. SCOPE AND FIELD OF APPLICATION

2.1 Types of structure

Simple structural joints are commonly met in steel framed buildings but they can be used also in other types of structures to connect steel elements (for example in bridges).

2.2 Types of connected elements

The shape of the structural connected elements which are considered in this report are: - I or H beams;

- I or H columns (with a possible extension to RHS and CHS).

2.3 Types of loading

The design methods are intended for joints subject to predominantly static or quasi-static loading. Fatigue aspects are not considered.

The resistance of the joints is checked under shear and tying forces. The shear forces correspond to usual loading conditions of the structure during its life; tying forces may de-velop when the frame is subjected to an explosion or when a supporting column is lost under exceptional events (Fig. 2.1).

Figure 2.1: Tying forces

2.4 Steel grades

(11)

11 2.5 Possible joint configurations

The configurations of simple joints addressed in the present publication are the following: • Beam-to-column (Fig. 2.2):

a) Single-sided joint configurations

Major axis Minor axis b) Double-sided joint configurations

Major axis Minor axis Figure 2.2: Beam-to-column joint configurations • Beam-to-beam (Fig. 2.3):

a) Single-sided joint configurations

Un-notched supported beam Single notched supported beam Double notched supported beam

(12)

12

b) Double-sided joint configurations

Un-notched supported beam Single notched supported beam Double notched supported beam Figure 2.3: Beam-to-beam joint configurations

• Beam splice (Fig. 2.4 a and b):

Figure 2.4 a: Beam splice joint Possible locations for such joints are shown in Fig. 2.4 b.

Figure 2.4 b: Possible locations of simple joints • Column splice (Fig. 2.5):

Figure 2.5: Column splice joint

joint position + + _ _ _ _ _ + + +

(13)

13 • Braced connection (Fig. 2.6):

Figure 2.6: Braced configuration • Column base (Fig. 2.7):

Figure 2.7: Column base joint configuration

Amongst these joint configurations, only the two first ones will be explicitly covered: beam-to-column and beam-to-beam configurations. The others are expected to be covered in a revised edition of the present publication.

2.6 Types of fasteners 2.6.1 Bolts

There are two classes of bolts: normal bolts and high strength bolts. The second class can be used for preloaded bolts which are characterized by a slip-type resistance mode in shear.

In this document, only non-preloaded bolts are explicitly covered. Their design geo-metrical and mechanical characteristics are given in the tables 2.1 and 2.2 respectively.

Column-concrete "connection"

Concrete-ground "connection"

(14)

14

d (mm) 8 10 12 14 16 18 20 22 24 27 30

A (mm²) 50 78 113 154 201 254 314 380 452 573 707

As (mm²) 36 58 84 115 157 192 245 303 353 459 561

with d = nominal diameter of a bolt shank A = nominal area of a bolt

As = tensile stress area of a bolt

Table 2.1: Bolt areas

Bolt grade 4.6 5.6 6.8 8.8 10.9

fyb (N/mm²) 240 300 480 640 900

fub (N/mm²) 400 500 600 800 1000

Table 2.2: Nominal values of yield strength fyb and ultimate tensile strength fub for bolts

2.6.2 Welds

In Eurocode 3, various types of weld are considered: fillet welds, fillet welds all round, butt welds, plug welds and flare groove welds. Only fillet welds are explicitly considered here.

2.7 Types of connections

Three connections types, used in the present design recommendations to connect a beam to a column or a beam to a beam, are specified below.

• Header plate connections

The main components of a header plate connection are shown in Fig. 2.8: a steel plate, a fillet weld on both sides of the supported beam web, and two single or two double vertical bolt lines. The plate is welded to the supported member and bolted to a sup-porting element such as a steel beam or column. Its height does not exceed the clear depth of the supported beam .The end of the supported steel beam may be un-notched, single notched or double notched.

(15)

15 Figure 2.8: Header plate connection

• Fin plate connections

The main components of a fin plate connection are shown in Fig. 2.9.: a fin plate, a fil-let weld on both sides of the plate, and a single or double vertical bolt line. The plate is welded to a supporting member such as a steel beam or column and bolted to web of the supported beam. The end of the supported steel beam may be un-notched, single notched or double notched.

Figure 2.9: Fin plate connection • Web cleat connections

A web cleat connection is characterised (see Fig. 2.10) by two web cleats and three single or double vertical bolt lines (two on the supporting element and one on the sup-ported member). The cleats are bolted to the supporting and supsup-ported members. Un-notched, single notched or double notched supported beams may be considered.

Supporting element Supported beam Plate Fillet weld Single-vertical row bolt group

Double-vertical row bolt group Single vertical bolt line Double vertical bolt line Supported beam Fillet weld Fin plate Supporting element Single-vertical row bolt group

Double-vertical row bolt group

(16)

16

Figure 2.10: Web cleat connection Note:

Traditionally, other types of beam-to-column connections are considered as hinges. But nowadays Eurocode 3 Part 1-8 classifies them as semi-rigid. Two examples are given in Fig. 2.11.

Figure 2.11: Other simple connections

2.8 Reference code

The design rules presented in this publication are based on the resistance formulae pro-vided by Eurocode 3 Part 1-8, at least as far as information is available. When this is not the case, the basic design principles prescribed by Eurocode 3 are followed.

Web cleat _Web

cleat Supported beam Supporting element OR WITH OR Single-vertical row bolt group

Double-vertical row bolt group

Single-vertical row bolt group

Double-vertical row bolt group Single vertical

bolt line

Double vertical bolt line

Single vertical bolt line

Double vertical bolt line

(17)

17 3. JOINT MODELLING FOR FRAME ANALYSIS AND DESIGN

REQUIRE-MENTS 3.1 General

The effects of the actual response of the joints on the distribution of internal forces and moments within a structure, and on the overall deformations, should generally be taken into account; but when these effects are sufficiently small, they may be neglected.

To identify whether the effects of joint behaviour on the analysis need be taken into ac-count, a distinction should be made between the three following types of joint modelling:

- simple, in which the joint may be assumed not to transfer bending moments;

- continuous, in which the behaviour of the joint may be assumed to have no effect on the analysis;

- semi-continuous, in which the behaviour of the joint needs to be explicitly taken into account in the analysis.

The appropriate type of joint modelling depends on the classification of the joint and on the selected procedure for structural analysis and design.

3.2 EC 3 classification system

The joints can be classified according to the values of their main structural properties, i.e. rotational stiffness, strength in bending and rotational capacity (or ductility). The struc-tural properties of all the joints need to correspond to the assumptions made in the strucstruc-tural frame analysis and in the design of the members. In particular, as far as simple joints are con-cerned, the available rotation capacity of the joints should be sufficient to accept the rotations evaluated in the analysis process.

In Eurocode 3 Part 1-8, joints are classified by stiffness and by strength. Ductility as-pects are also to be considered; they will be more especially addressed in Section 4 below.

3.2.1 Classification by stiffness

This classification is only applicable to beam-to-column joint configurations. Through the comparison of its actual rotational stiffness Sj,ini with classification boundaries (Fig. 3.1), a joint may be considered as:

(18)

18

Figure 3.1: Boundaries for stiffness classification of joints - Nominally pinned

The joint shall be capable of transmitting the internal forces, without developing significant moments which might adversely affect the structural members. It shall be also capable of accepting the resulting rotations under the design loads.

⇒ Boundary: Sj,ini ≤ 0,5 EIb / Lb

- Rigid

The joint behaviour is assumed not to have significant influence on the distribution of internal forces and moments in the structure, nor on its overall deformation. ⇒ Boundaries: Sj,ini ≥ kb EIb / Lb

where kb = 8 for frames where the bracing system reduces the horizontal displacement by at

least 80%;

kb = 25 for other frames. - Semi-rigid

The joint provides a predictable degree of interaction between members, based on the design moment-rotation characteristics of the joint. It should be able to trans-mit internal forces and moments.

Pinned Rigid Semi-rigid Mj φ Sj,ini

Initial rotational stiffness Stiffness boundaries

(19)

19 ⇒ Boundaries: A joint which doesn't meet the criteria for a rigid or a nominally pinned joint shall be classified as a semi-rigid joint.

Key values: E is the elastic modulus of the beam material; Ib is the second moment area of the beam;

Lb is the beam span (distance between the axes of the supporting columns).

3.2.2 Classification by strength

Through the comparison of its actual design moment resistance Mj,Rd with the design moment resistances of the members that it connects ( Fig. 3.2), a joint may be classified as:

Figure 3.2: Boundaries for strength classification of joints - Nominally pinned

The joint shall be capable of transmitting the internal forces, without developing significant moments which might adversely affect the members of the structure. It shall also be capable of accepting the resulting rotations under the design loads. ⇒ Boundary: Mj,Rd ≤ 0,25 M full-strength (see Fig. 3.3) - Full-strength

The design resistance of a full strength joint shall be not less than that of the con-nected members.

⇒ Boundary: Mj,Rd ≥ M full-strength (see Fig. 3.3) Partial-strength Full-strength Mj Pinned φ Mj,Rd

Joint moment resistance Strength boundaries

(20)

20

Key values: Mb,pl,Rd is the plastic moment resistance of the beam (possibly reduced by axial or shear forces in the beam);

Mc,pl,Rd is the plastic moment resistance of the column (possibly reduced by axial or shear forces in the column).

Figure 3.3: Full-strength resistance - Partial-strength

A joint which doesn't meet the criteria for full-strength or nominally pinned joints should be considered to have a partial-strength resistance.

3.3 EC 3 joint modelling

The joint modelling depends on the joint classification (see above) and on the selected process for structural analysis and design. As said before, Eurocode 3 considers three types of joint modelling (simple, continuous and semi-continuous) dependent on whether or not the effects of joint behaviour on the analysis can be neglected. The appropriate type of joint mod-elling should be determined from the Table 3.1.

METHOD OF GLOBAL

ANALYSIS CLASSIFICATION OF JOINT

Elastic Nominally pinned Rigid Semi-rigid

Rigid-Plastic Nominally pinned Full-strength Partial-strength

Elastic-Plastic Nominally pinned Rigid and full-strength

Rigid- and partial-strength Semi-rigid and partial-strength Semi-rigid and full-strength

TYPE OF JOINT MODEL Simple Continuous Semi-continuous

Table 3.1: Type of joint model

Top column:

M full-strength = min ( Mb,pl,Rd , Mc,pl,Rd )

Within column height:

M full-strength = min ( Mb,pl,Rd , 2 Mc,pl,Rd )

(21)

21 So, in the global analysis, the joint behaviour can be replaced by (Fig. 3.4):

- a hinge, for the simple modelling;

- a rotational spring, for the semi-continuous modelling [10];

- an infinitely rigid and resistant rotational spring, for the continuous modelling.

Figure 3.4: Local joint modelling

In the global structural analysis, the hinge or spring which models the joint is assumed to be located at the intersection of the axes of the connected elements.

3.4 Simple joint modelling

The design rules in this guide are given for joints which are assumed not to transmit bending moments. Thus, the joints should be modelled by hinges. Unfortunately, many joints which are traditionally considered as a hinge do not fulfil the stiffness and/or strength limita-tions required by Eurocode 3 for nominally pinned joints.

Two different attitudes may be adopted in such a case:

- According to the Eurocode 3 requirements, the joint is modelled by a rotational spring and is therefore considered as semi-rigid (what it is in reality). Its rotational stiffness, design bending resistance and shear resistance have to be evaluated and the actual properties of the joint have to be explicitly taken into consideration in the structural analysis and in the design phase. This approach is the more

scientifi-TYPE OF JOINT

MODEL CONFIGURATION SINGLE-SIDED CONFIGURATION DOUBLE-SIDED BEAM SPLICE

Simple

Continuous

Semi-continuous

(22)

22

cally correct one but it needs more complex calculations as far as the global analy-sis and joint design are concerned.

- Despite its actual properties, the joint is considered as a hinge and the design rules presented in this present publication for simple joints can be applied, but under some strict conditions which ensure the safe character of the approach. The global analysis and the joint design are more simple in this case as they are based on a more traditional hinged (simple) approach.

If the second option is chosen, the joint is assumed not to transfer bending moments even if it is not the truth. So bending moments develop in the joints although they are de-signed to resist only shear forces. This is potentially unsafe and at first sight is not basically acceptable.

But a careful examination of this problem leads to the conclusion that the "hinge as-sumption" is safe if the two following requirements are fulfilled:

- the joint possesses a sufficient rotation capacity; - the joint possesses a sufficient ductility.

The first requirement relates to the rotational capacity that the joint should have, in or-der to "rotate" as a hinge, without developing too high internal bending moments.

The second requirement is there to ensure that the development of combined shear and bending forces into the joint is not leading to brittle failure modes (for instance, because of a rupture of a bolt or a weld). In other words, the design of the joint should allow internal plas-tic deformations instead of brittle phenomena.

If these two requirements (sufficient rotation capacity and ductility) are fulfilled, it can be demonstrated that to consider an actually semi-rigid joint as a nominally pinned one is safe for design purposes and, in particular, for the evaluation of:

- the frame displacements:

the stiffness of the actual structure is always greater than that of the hinged one, and all the actual displacements are therefore lower than the calculated ones; - the plastic failure loading:

as the actual bending strength of the joint is higher than the considered one (equal to zero), the first order plastic resistance of the frame is higher than the one evalu-ated on the basis of a hinge behaviour;

(23)

23 - the critical loading of linear elastic instability:

the transversal stiffness of the actual structure is larger than the one of the structure with nominally pinned joints, and the rotational restraints at the end of the columns in the actual structure are higher than these calculated with a hinge assumption; this ensures the safe character of the hinge assumption as far as global and local instability are concerned;

- the elastic-plastic phenomena of instability:

the actual stiffness of the structure is greater than the considered one but the actual loading conditions are more important than those acting on the structure with nominally pinned joints; nevertheless, various studies ([14], [15] and [16]) show that the “hinged” approach is safe.

For further explanations, see [10].

In this guide, the design recommendations relate to the "hinge model". Specific design requirements ensuring safety are presented for each of the connection types considered.

3.5 Summary of design requirements

As said before, the internal forces in the joint are here determined by a structural analy-sis based on simple joint modelling. The hinge is assumed to be located at the intersection of the axes of the connected elements. As a result of this structural analysis, the maximum ap-plied shear force and rotation in the joint, respectively VEd and φrequired, are obtained.

From the geometrical properties of the joint and the mechanical properties of its consti-tutive materials, the available rotation capacity of the joint, φavailable, can be estimated, as well as its design shear resistance, VRd. To ensure the validity of this approach, some ductility re-quirements have to be satisfied and the available rotation of the joint has to be higher than the required one. Finally, the joint will be considered as acceptable if the applied shear force does not exceed the design shear resistance.

Sometimes, the evaluation of the resistance to tying forces is requested for robustness purposes.

(24)

24

4. PRACTICAL WAYS TO SATISFY THE DUCTILITY AND ROTATION RE-QUIREMENTS

4.1 General principles

A simple joint is nothing else than an idealisation of the reality. Joints like those studied in the present document undergo a significant internal rotation but transfer some bending moments. As explained above, to ensure the safety of the simple joint model, some require-ments for sufficient ductility and rotation capacity are necessary.

These requirements can be written for each considered connection type, in the form of simple criteria based on the mechanical and geometrical characteristics of the different com-ponents forming the connection.

The rotation capacity requirements provide to the hinge a sufficient rotation without de-veloping too significant bending moments which might adversely affect the members of the structure. These criteria are often expressed as geometrical limitations.

The ductility requirements avoid the occurrence of brittle failures, especially in bolts and welds, and buckling. Their derivation is more complex. In the "hinged" structural analy-sis, the joint is assumed to be only subjected to a shear force. In reality, a bending moment and a shear force are acting simultaneously in the joint. In an "applied shear force – applied bending moment" graph (Fig. 4.1), the evolution of the actual and idealised loading types can be represented by two paths. The first is a horizontal one (MEd = 0) and the second an oblique one. The inclination of the actual loading path depends on the relative stiffness between the joint and the connected elements.

Figure 4.1: Loading paths

M

Sd

Design loading path

Actual loading path

V

Sd

M

Ed

(25)

25 Note: For fin plate connections, two different cross-sections inside the joint have to be

con-sidered separately. The first is located at the external face of the supporting member; while the second is through the centre of the bolt group (Fig. 5.2).

The actual loading situation is different in these two sections, so leading to two dis-tinct MEd – VEd paths in the diagram shown on Figure 4.2.

If a "hinge" model is considered, the first section is assumed to transfer only shear forces (MEd = 0) while the second one, in accordance with equilibrium, transfers the same shear force VEd and a bending moment MEd equal to VEd . z.

z is defined as the distance between the external face of the supporting element and the centre of the bolt group.

M

Sd

V

Sd

Design loading path for the external face of the supporting member Design loading path for the section of the bolt group centre Actual loading path for the external face of the supporting member Actual loading path for the section of the bolt group centre

1 z

Figure 4.2: Loading paths for a fin plate connection

The design resistance of each component of the joint can be represented in a "shear force – bending moment" graph. Dependent on whether this resistance is influenced by the applied bending moment, its representation will be a curve or a vertical line. Figure 4.3 illus-trates it for three possible failure modes in a fin plate connection. The relative positions of the different resistance curves or lines depend on the geometrical and mechanical characteristics of the joint components.

Figure 4.3: Design resistances for some components of a fin plate connection and principle for the derivation of the shear resistance of the joint

V

Ed

M

Ed

M

Sd

V

Sd Fin plate in shear (gross section) Fin plate in bearing

Bolts in shear VRa VRd z .

VEd

M

Ed

(26)

26

In reality, the actual shear resistance, VRa, of the joint could be defined at the intersec-tion between the actual loading path, in the appropriate cross-secintersec-tion, and the design resis-tance curves or lines of the weakest component (Fig. 4.3). If a similar principle is applied to the design loading path, a design shear resistance, VRd, is then obtained.

If the failure mode corresponding to the VRa value is a brittle one, the design shear re-sistance VRd is seen as to be an unsafe estimation of the joint resistance (Fig. 4.4 a). The only way to reach the design shear resistance VRd is to rely on a plastic redistribution of internal forces inside the joint, as shown on Figure 4.4 b.

a) Premature brittle failure

b) Possible plastic redistribution of internal forces Figure 4.4: Determination of the shear resistance of the joint

As a conclusion, the ductility requirements will aim to ensure that the move from the actual to the design shear resistances may occur, as a result of a plastic redistribution of inter-nal forces inside the joint.

Bolts in shear

Fin plate in bearing

M

Sd

Fin plate in shear (gross section)

V

Sd VRd VRa Brittle failure No possible redistribution of internal forces

V

Ed

M

Ed Ductile failure Fin plate in bearing

VRa

M

Sd

Bolts in shear

V

Sd

Fin plate in shear (gross section) Possible redistribution of internal forces VRd

VEd

M

Ed

(27)

27 In the next paragraphs, the design requirements to be fulfilled to allow sufficient rota-tion capacity and ductility are specified for all the connecrota-tion types covered in the present publication.

4.1.1 Header plate connection

4.1.1.1 Design requirements for sufficient rotation capacity

To enable rotation without increasing too much the bending moment which develops into the joint, contact between the lower beam flange and the supporting member has to be strictly avoided. So, it is imperative that the height hp of the plate is less than that of the sup-ported beam web (Fig. 4.5):

hp ≤ db where db is the clear depth of the supported beam web.

If such a contact takes place, a compression force develops at the place of contact; it is equilibrated by tension forces in the bolts and a significant bending moment develops (Fig. 4.5). Compression force Bending moment Bending moment Rotation

φ

available Contact between the supported beam and the supporting element

Tension forces in the bolts

Figure 4.5: Contact and evolution of the bending moment

The level of rotation at which the contact occurs is obviously dependent on the geo-metrical characteristics of the beam and of the header plate, but also on the actual deforma-tions of the joint components.

In order to derive a simple criterion that the user could apply, before any calculation, to check whether the risk of contact may be disregarded, the following rough assumptions are made (see Fig. 4.6):

(28)

28

- the supporting element remains un-deformed;

- the centre of rotation of the beam is located at the lower extremity of the header plate.

On the basis of such assumptions, a safe estimation (i.e. a lower bound) of the so-called "available rotation of the joint" φavailable may be easily derived:

Figure 4.6: Geometrical characteristics of the joint and illustration of contact between the beam and the supporting element

This available rotation has to be greater than the "required rotation capacity" which varies according to the structural system and loading. A simple criterion ensuring the suffi-cient joint rotation capacity may be written as:

φavailable > φrequired

For instance, the required rotation capacity, for a beam (length L and inertia I) simply supported at its extremities and subjected to an uniformly distributed load (factored load γ p at ULS), is given by:

By expressing that φavailable > φrequired , a simple criterion ensuring a sufficient joint rota-tion capacity may be derived:

hp he tp hb db φavailable e p available h t = φ φrequired EI 24 L p 3 γ = EI 24 L p h t 3 e γ >

(29)

29 Similar criteria may be derived for other load cases (Annex 1).

4.1.1.2 Design requirements for sufficient joint ductility

As bending moments develop in the joint, the bolts and the welds are subjected to ten-sion forces in addition to shear forces. Premature failure of those elements which exhibit a brittle failure and which are more heavily loaded in reality than in the calculation model has therefore to be strictly avoided. Simple related criteria should therefore be proposed.

Criterion to avoid premature bolt failure because of tension forces

In Eurocode 3, a criterion based on the T-stub approach ensures that a yield lines mecha-nism develops in the plate before the strength of the bolts is exhausted (see [3]); its back-ground is given in [12]. This criterion, initially developed for end plates and column flanges, is here safely extended to column (weak axis beam-to-column joints) or beam (beam-to-beam joint configurations) webs.

According to this criterion, at least one of the two following inequalities (1) and (2) has to satisfied:

(1)

(2) for a supporting column flange

yw

w ub

2,8 f

d

t ≥ f for a supporting column or beam web (or faces of hollow sections)

Note:

This criterion is expected to be satisfied by most of the supporting webs because of their slenderness.

where:

d is the nominal diameter of the bolt shank; tp is the thickness of the header plate;

tcf is the thickness of the supporting column flange; tw is the thickness of the supporting column or beam web; fyp is the yield strength of the steel constituting the header plate;

fycf is the yield strength of the steel constituting the supporting column flange; fyw is the yield strength of the steel constituting the supporting column or beam

web;

fub is the ultimate strength of the bolt. p t d _≥ 2,8 ub yp f f cf t d ≥ 2,8 ub ycf f f

(30)

30

Such a criterion does not ensure that the whole shear capacity of the bolt may be consid-ered when evaluating the shear resistance of the joint. In fact, when this requirement is satisfied, it may be demonstrated:

- that the tension force in the bolts may amount 0,5 Bt.Rd, i.e. 50% of the design

ten-sion resistance Bt,Rd of the bolts;

- that, for such a tension force, the actual shear resistance only amounts 64% of the full shear resistance of the bolts (according to the EC 3 resistance formula for bolts in shear and tension).

This looks at first sight to be disappointing as the user tries to maximise the shear resis-tance of the joint. It may be argued though that only the bolts located in the upper half of the header plane are affected by such a reduction, as the others are located in a compres-sion zone, and are therefore not subjected to tencompres-sion forces.

So finally a reduction is taken into consideration by multiplying the total resistance of the bolts in shear by a factor 0,8 (i.e. a reduction factor of 0,64 for half of the bolts located in the upper half of the header plate – 0,5.[1 + 0,64] ≈ 0,8).

Criterion to avoid premature weld failure because of tension or shear forces

The welds must be designed according to EC3 Part 1-8. In the case of relatively small loads in relation to the capacity of the web, application of the rules in 4.5.3.2 of Part 1-8 may lead to rather thin welds. If the rupture strength of those thin welds is lower than the yield strength of the weakest of the connected parts, the connection has so little deforma-tion capacity that it usually is not sufficient to accommodate effects due to imposed de-formations etc. In such a case the connection will behave in a brittle way.

To avoid this, the welds can be designed "full strength". The rupture strength of full strength welds is greater than the rupture strength of the adjacent plate; so, in the case of overloading, the plate will fail before the welds. This is a safe design but not always nec-essary, taking into account the requirement that the welds should at least be able to ensure yielding of the plate before rupture in the welds. In the IIW recommendations of 1976, it is stated that, if the welds are designed at 70 % of the full strength, yielding of the plate is ensured before rupture of the welds. After the re-evaluation of weld design formulae in-cluded in the ENV version of EC3, which gave some smaller weld sizes than in IIW rules, it was decided in the Dutch standard NEN 6770 [7] to modify the 70 % to 80 %.

Unfortunately this rule does not exist in Part 1-8 of EC3, what means that designers have to decide for themselves how to ensure adequate deformation capacity. Obviously, to adopt full strength welds is safe, but not really necessary.

For the case of the header plate it should be noted that, especially at the extremities of the welds, local stresses and strains may be very high and some strain hardening may occur. Therefore it is recommended to design these welds "full strength".

(31)

European Recommendations for the Design of Simple Joints in Steel Structures 31 2 weld

σ

τ

σ

_⊥= _⊥=

According to clause 4.5.3.2 of Eurocode 3 Part 1-8, using the directional method it fol-lows: Mw w u c f γ β τ τ σ σ = _⊥+ _⊥+ 2 ≤ // 2 2 ₃ ₃ _and Mw u f γ σ_⊥ ≤ where:

fu = the nominal ultimate tensile strength of the weaker part joined

γMw = partial safety factor for welded connections (γMw = 1,25)

βw = correlation factor (βw = 1,0 for steel grades S420 and S460, see Table 4.1)

a a l t σ⊥ τ⊥_σ las Fkop Fzij l σ_z_σ_x σweld

F

end

F

side

A

A b t

Figure 4.7: End fillet and side fillet welds For end fillet welds is and τ_// =0. From the first formula reported above, it follows:

Mw w u weld weld c f γ β σ σ σ ⎟ ≤ ⎠ ⎞ ⎜ ⎝ ⎛ + ⎟ ⎠ ⎞ ⎜ ⎝ ⎛ = 2 2 2 3 2 end u w Mw w u weld f f . . 2 = ≤ γ β σ

For double end fillet welds:

end u w x end u w end f t f F a . . . . 2 2 ⋅ = ≥ σ A

The greatest weld size is found for σx = fy if in the connected plate. In Table 4.1 the

re-quired weld sizes are given for this case.

For side fillet welds is σ_⊥ =τ_⊥ =0 and τ_// =τ_weld. From the first here-above reported formula, it follows:

side u w Mw w u weld f f . . 3 = ≤ γ β τ

(32)

European Recommendations for the Design of Simple Joints in Steel Structures 32 Steel grade S235 S275 S355 S420 M S420 N S460 M S460 N fy (N/mm2) 235 275 355 420 420 460 460 ft (N/mm2) 360 430 510 520 550 550 580 βw 0,80 0,85 0,90 1,00 1,00 1,00 1,00 fw.u.end (N/mm2) 255 286 321 294 311 311 328 fw.u.side (N/mm2) 208 234 262 240 254 254 268

Full strength double end fillet welds (design stress:σx = fy) a ≥ 0,46 t a ≥ 0,48 t a ≥ 0,55 t a ≥ 0,71 t a ≥ 0,68 t a ≥ 0,74 t a ≥ 0,70 t Full strength double side

fillet welds

(design stress: τplate =

fy/√3) a ≥ 0,33 t a ≥ 0,34 t a ≥ 0,39 t a ≥ 0,50 t a ≥ 0,48 t a ≥ 0,52 t a ≥ 0,50 t Double end fillet welds

to ensure yield in the plate before rupture in the welds (design stress: σx = 0,8fy) a ≥ 0,37 t a ≥ 0,38 t a ≥ 0,44 t a ≥ 0,57 t a ≥ 0,55 t a ≥ 0,59 t a ≥ 0,56 t

Table 4.1 - Values of βw and fw.u.end and fw.u.side for steels according to EN 10025 and EN 10113 and weld

thickness in case of double fillet welds. Plate thickness smaller than 40 mm.

4.1.1.3 Conclusions

If the rotation capacity and ductility requirements specified in 4.1.1.1 and 4.1.1.2 are satisfied, the shear resistances of all the constitutive components are evaluated and the design shear resistance of the connection corresponds to the weakest one, as illustrated in Figure 4.8. This is allowed as all the possible detrimental effects linked to “bending-shear” interaction phenomena are integrated into the ductility requirements.

In reality, the first component to yield is not necessarily the weakest one, in terms of shear resistance, and two different situations may occur (Fig. 4.8). In the first case (Fig. 4.8 a), the same failure mode is obtained by following the actual and design loading paths. For the second case (Fig. 4.8 b), the failure mode obtained with the actual loading path is not the weakest one, but is ductile enough to allow a plastic redistribution of internal forces to take place until the design shear resistance is reached.

Finally – and this is of importance for practice - it has to be noted that the rotation ca-pacity and ductility requirements may be checked before any resistance calculation.

(33)

33 a) one single failure mode

b) different failure modes

Figure 4.8: Possible failure modes for a header plate connection

Plastic mechanism in the header plate

Design shear resistance B ea m w eb in s he ar MSd S up po rt in g el em en t i n be arin g H ea de r pl at e in b ea ri ng H ea de r pl at e in s he ar (ne t s ec tio n) H ea de r pl at e in s he ar ( sh ea r bl oc k) H ea de r pl at e in s he ar ( gr os s se ct io n) B ol ts in s he ar VSd

Plastic mechanism in the header plate

Design shear resistance B ea m w eb in s he ar MSd S up po rt in g el em en t i n be ar in g H ea de r pl at e in b ea ri ng H ea der pl at e in s he ar ( ne t s ec ti on ) H ea de r pl at e in s he ar ( sh ea r bl oc k) H ea de r pl at e in s he ar ( gr os s se ct io n) B ol ts in s he ar VSd

Design loading path Actual loading path

(34)

34

4.1.2 Fin plate connection

4.1.2.1 Design requirements for sufficient rotation capacity

So as to permit a rotation without increasing too much the bending moment which de-velops into the joint, contact between the lower beam flange and the supporting member has to be strictly avoided. To achieve it, the height hp of the fin plate should be lower than that of the supported beam web (Fig. 4.9):

hp ≤ db where db is the clear depth of the supported beam web

If such a contact takes place, a compression force develops at the place of contact; it is equilibrated by tension forces in the welds and in the plate, and additional shear forces in the bolts.

Compression force

Shear forces in the bolts

Bending moment Bending moment Rotation φavailable Contact between the supported beam and the supporting element

Figure 4.9: Contact and evolution of the bending moment

The level of rotation at which the contact occurs is obviously dependent on the geo-metrical characteristics of the beam and of the fin plate, but also on the actual deformations of the joint components.

In order to derive a simple criterion that the user could apply, before any calculation, to check whether the risk of contact may be disregarded, the following rough assumptions are made (see Fig. 4.10):

- the supporting element and the fin plate remain un-deformed;

- the centre of rotation of the beam is located at the centre of gravity of the bolt group.

(35)

35 On the basis of such assumptions, a safe estimation (i.e. a lower bound) of the so-called "available rotation of the joint" φavailable may be easily derived:

- if z >

(

)

2 e p 2 h h 2 h g z _⎟⎟ ⎠ ⎞ ⎜⎜ ⎝ ⎛ + + − : " " available = ∞ φ - else:

Figure 4.10: Geometrical characteristics of the joint and illustration of the contact between the beam and the supporting element

This available rotation has to be greater than the "required rotation capacity" which varies according to the structural system and loading. A simple criterion ensuring the suffi-cient joint rotation capacity may be written as:

φavailable > φrequired

Expressions forφrequired are given 4.1.1.1 and Annex 1.

(

)

⎟⎟ ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎜ ⎜ ⎝ ⎛ + − − ⎟ ⎟ ⎟ ⎟ ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎜ ⎜ ⎜ ⎜ ⎝ ⎛ ⎟⎟ ⎠ ⎞ ⎜⎜ ⎝ ⎛ + + − = φ e p h 2 e p 2 h available h 2 h g z arctg h 2 h g z z arcsin Centre of rotation Centre of rotation hp he gh hb db z z φavailable φavailable

(36)

36

4.1.2.2 Design requirements for sufficient joint ductility

As previously explained, the design shear resistance of the joint may be reached, as a result of a plastic redistribution of internal forces amongst the different constitutive compo-nents. This requires that no local brittle failure modes or instabilities develop during this re-distribution. The failure modes which could prevent redistribution of internal forces to take place are, for fin plate connections: the bolts and the welds in shear on account of their brittle nature, and the buckling of the fin plate which is assumed to be non-ductile in terms of plastic redistribution.

Criterion to avoid premature weld failure because of tension forces

A similar criterion as the one established for the header plate connection, may be written. For fin plates also high local stresses are to be expected, but of less severity than in the case of the header plate. It is considered acceptable that in the check for ductility, weld sizes referring to the “80 % rule” are applied, see Table 4.1. The procedure is the follow-ing one: first, the weld size should be determined on the basis of the design loads; and secondly the deformation capacity should be checked. So, if the design loads require a 90 % full strength weld, that weld size should be applied.

Criterion to permit a plastic redistribution of internal forces between the "actual" and "design" resistance points

(1) First of all, the design shear resistance of the connection should be associated with a ductile mode. Failure by bolts in shear or by buckling of the fin plate is therefore excluded. A first criterion can be written:

min( VRd 1; VRd 7 ) > VRd where: VRd 1 is the shear resistance of the bolts; VRd 7 is the buckling resistance of the fin plate; VRd is the design shear resistance of the connection.

(2) Secondly, the component which yields under the "actual" loading in the connec-tion has also to ductile (so, no bolts in shear or buckling of the fin plate). To en-sure this, different criteria have to be fulfilled dependent on the failure mode ob-tained through treating the connections as “hinged”:

• Failures by bolts in shear or buckling of the fin plate: Excluded by the first criterion (1).

(37)

37 For one vertical bolt row, at least one of the following two inequalities has to be satisfied:

Fb,hor,Rd ≤ min ( Fv,Rd; VRd 7 β) for the beam web Fb,hor,Rd ≤ min ( Fv,Rd; VRd 7 β) for the fin plate

For two vertical bolt rows, at least one of the following three inequalities has to be satisfied:

(3) Lastly, during the redistribution process, the "bolts in shear" failure mode should not be met. To avoid that, simple criteria can be written that again depends on the failure mode resulting from treating the connection as a “hinge”:

• Failure by bolts in shear or buckling of the fin plate: Excluded by the first criterion (1).

• Failure by fin plate or beam web in bearing:

If the two first criteria (1) and (2) are fulfilled, no additional criterion is nec-essary.

• All the other failure modes:

VRd 1 > min ( VRd 2; VRd 8 ) where VRd 1 is the shear resistance of the bolts; VRd 2 is the bearing resistance of the fin plate; VRd 8 is the bearing resistance of the beam web.

max (

(

2 2

)

2 Rd , v F 1 _α ₊_β _; 2 7 Rd V 1 ₎_≤ 2 Rd , hor , b 2 Rd , ver , b F F ⎟⎟_⎠ ⎞ ⎜ ⎜ ⎝ ⎛ _β + ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ _α

for the beam web

max (

(

2 2

)

2 Rd , v F 1 β + α ; ₂ 7 Rd V 1 ₎_≤ 2 Rd , hor , b 2 Rd , ver , b F F ⎟⎟_⎠ ⎞ ⎜ ⎜ ⎝ ⎛ _β + ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝

⎛ _α _{for the fin plate}

VRd 6 ≤ min( 2 2 3 2 β + α Fv,Rd; 3 2 VRd 7 )

(38)

38

Notation used in the above requirements is given in the part "Design sheets for fin plate connections" of the present publication.

The criteria (1), (2) and (3) can be only checked after the evaluation of the design shear resistance of the joint.

For further explanations about the derivation of these requirements, see [10].

4.1.3 Web cleat connection 4.1.3.1 General

The behaviour of a web cleat connection may be considered as the combination of the behaviours of header and fin plates connections. The design rules and requirements for a safe approach may be simply deduced from those established for the two previous connection types.

4.1.3.2 Design requirements

They are also easily deduced from the previous requirements expressed for header and fin plate connections.

(39)

39 5. GEOMETRY OF THE THREE CONNECTION TYPES

5.1 Symbols

5.1.1 General notation

• For the bolts:

n Total number of bolts A Nominal area of a bolt As Tensile stress area of a bolt d Nominal diameter of a bolt shank d0 Diameter of a bolt hole

fu,b Ultimate strength of a bolt fy,b Yield strength of a bolt • For the welds:

a Throat thickness of the welds

βw Correlation factor for the evaluation of the weld resistance • For the supporting and supported elements:

t Thickness of the supporting plate (tcf and tcw for respectively a column flange and web, tbw for a beam web)

tw Thickness of the supported beam web Ab,v Gross shear area of the supported beam Ab,v,net Net shear area of the supported beam

fu Ultimate strength of a steel element (index bw for beam web, cf and cw for respec-tively column flange and web)

fy Yield strength of a steel element (index bw for beam web, cf and cw for respectively column flange and web)

• Safety coefficients:

γM0 Partial safety factor for steel sections; it is equal to 1,0

γM2 Partial safety factor for net section at bolt holes, bolts, welds and plates in bearing; it is equal to 1,25

Note: The value of the partial safety factors reported here are those recommended in Eurocode 3 but other values may be assigned in National Annexes

• Loading:

(40)

40

• Resistance:

VRd Shear resistance of the joint Fv.Rd Design resistance in shear

5.1.2 Particular notation for header plate connections

e1 p1 e2S p1 e1 e2 mp p2' e1 p1 e1 p1 p2' p2 e2S e2 mp

Figure 5.1: Header plate notations

hp Height of the header plate tp Thickness of the header plate Av Gross shear area of the header plate Avnet Net shear area of the header plate fyp Yield strength of the header plate n1 Number of horizontal rows n2 Number of vertical rows e1 Longitudinal end distance e2 Transverse end distance p1 Longitudinal bolt pitch p2 Transverse bolt pitch

(41)

41

mp Distance between the inner vertical bolt row and the toe of the weld connecting the header plate to the beam web (definition according to EN 1993 Part 1-8)

5.1.3 Particular notation for fin plate connections

Figure 5.2: Fin plate notations

hp Height of the fin plate tp Thickness of the fin plate Av Gross shear area of the fin plate Avnet Net shear area of the fin plate fyp Yield strength of the fin plate n1 Number of horizontal rows n2 Number of vertical rows

e1 Longitudinal end distance (fin plate) e2 Transverse end distance (fin plate) e1b Longitudinal end distance (beam web) e2b Transverse end distance (beam web) p1 Longitudinal bolt pitch

p2 Transverse bolt pitch

zp Horizontal distance from the supporting web or flange to the first vertical bolt-row zp = z for connections with one bolt-row

zp = z –p2/2 for connections with two bolt-rows I Moment of inertia of the bolt group

(42)

42

5.1.4 Particular notation for cleat web connections

Figure 5.3: Web cleat notations

hc Height of the cleat tc Thickness of the cleat Av Gross shear area of the cleat Avnet Net shear area of the cleat

Supported beam side:

dsb Nominal diameter of a bolt shank d0sb Diameter of a bolt hole

nb Total number of bolts n1b Number of horizontal rows n2b Number of vertical rows

e1b Longitudinal end distance (cleat) e2b Transverse end distance (cleat) p1b Longitudinal bolt pitch

p2b Transverse bolt pitch

e2bb Transverse end distance (beam web) e1bb Longitudinal end distance (beam flange)

(43)

43

z Lever arm

I Moment of inertia of the bolt group

Supporting element side:

ds Nominal diameter of a bolt shank d0s Diameter of a bolt hole

n_s Total number of bolts n1s Number of horizontal rows n2s Number of vertical rows

e1s Longitudinal end distance (cleat) e2s Transverse end distance (cleat) p1s Longitudinal bolt pitch

p2s Transverse bolt pitch

e2ss Transverse end distance (supporting element)

e22s Longitudinal distance between the inner vertical bolt row and the beam web

5.2 Geometrical requirements

The design rules may only be applied if the positioning of holes for bolts respects the minimum spacing, end and edge distances given in the following table (Eurocode 3 require-ments).

Maximum 1) 2) 3) Structures made of steels according to EN 10025 except steels acc. to EN

10025-5 Structures made of steels according to EN 10025-5 Distances and spacings, see figure 5.4 Minimum Steel exposed to the weather or other corrosive in-fluences

Steel not exposed to the weather or other corrosive in-fluences

Steel used unpro-tected End distance e1 1,2 d0 4t + 40 mm The larger of 8t or 125 mm End distance e2 1,2 d0 4t + 40 mm Spacing p1 2,2 d0 The smaller of 14t or 200 mm The smaller of 14t or 200 mm The smaller of 14tmin or 175 mm Spacing p2 2,4 d0 The smaller of 14t or 200 mm The smaller of 14t or 200 mm The smaller of 14tmin or 175 mm

(44)

44

1) Maximum values for spacing, edge and end distances are unlimited, except in the following cases: - for compression members in order to avoid local buckling and to prevent corrosion in exposed

members and;

- for exposed tension members to prevent corrosion.

2) The local buckling resistance of the plate in compression between the fasteners should be calculated according to EN 1993-1-1 as column-like buckling by using 0,6 pi as buckling length. Local buckling

between the fasteners need to be checked if p1/t is smaller then 9 ε. The edge distance should not

ex-ceed the maximum to satisfy local buckling requirements for an outstand element in the compression members, see EN 1993-1-1. The end distance is not affected by this requirement.

3) t is the thickness of the thinner outer connected part.

Table 5.1: Minimum spacing, end and edge distances

(45)

45 6. DESIGN SHEETS

6.1 General

The forces applied to joints at the ultimate limit state result from a structural analysis and shall be determined according to the principles given in EN 1993-1-1. The resistance of the joint is determined on the basis of the resistances of the individual fasteners, welds and other components, as shown below.

6.2 Design sheet for connections with a header plate 6.2.1 Requirements to ensure the safety of the approach

To apply the design rules presented in section 6.2.2, all the following inequalities have to be satisfied. (1) hp ≤ db (2) _required e p h t φ >

(3) If the supporting element is a beam or column web:

p t d _{≥ 2,8} ub yp f f OR yw w ub 2,8 f d t ≥ f

If the supporting element is a column flange:

p t d _{≥ 2,8} ub yp f f OR cf td ≥ 2,8 _ub ycf f f (4) a > 0,4 tbw βw 3 0 M 2 M ubw ybw f f γ γ (βw is given in Table 4.1)

(46)

46

6.2.2 Resistance to shear forces

FAILURE MODE _V_ERIFICATION

Bolts in shear VRd 1 = 0,8 n Fv,Rd 2 M ub v Rd , v A f F γ α =

• where the shear plane passes through the threaded portion of the bolt:

A = As (tensile stress area of the bolt) - for 4.6, 5.6 and 8.8 bolt grades:

v

α = 0,6

- for 4.8, 5.8, 6.8 and 10.9 bolt grades: α = v 0,5

• where the shear plane passes through the un-threaded portion of the bolt:

A (gross cross area of the bolt) α = 0,6 v

(according Table 3.4 in EN 1993 Part 1-8)

Header plate in bearing VRd 2 = n Fb,Rd

2 M p up b 1 Rd , b t d f k F γ α = where αb = min ( ou 1,0 f f ; 4 1 d 3 p ; d 3 e up ub 0 1 0 1 ₋ ₎ k1 = min ( 1,7;2,5 d p 4 , 1 ; 7 , 1 d e 8 , 2 0 2 0 2 ₋ ₋ ₎

(47)

European Recommendations for the Design of Simple Joints in Steel Structures 47 Supporting member in bearing VRd 3 = n Fb,Rd 2 M u b 1 Rd , b t d f k F γ α =

• where the supporting element is a column flange: t = tcf fu = fucf αb = min ( ou1,0 f f ; 4 1 d 3 p u ub 0 1 ₋ ₎ k1 = min ( 1,7;2,5 d e 8 , 2 ; 7 , 1 d p 4 , 1 0 s 2 0 2 ₋ ₋ ₎

• where the supporting element is a column web: t = tcw fu = fucw αb = min ( ou1,0 f f ; 4 1 d 3 p u ub 0 1 ₋ ₎ k1 = min ( 1,7;2,5 d p 4 , 1 0 2 ₋ ₎

• where the supporting element is a beam web: t = tbw fu = fubw αb = min ( ou1,0 f f ; 4 1 d 3 p u ub 0 1 ₋ ₎ k1 = min ( 1,7;2,5 d p 4 , 1 0 2 ₋ ₎

Formula as written here apply to major axis beam-to-column joints (connection to a beam-to-column flange), to sin-gle-sided minor axis joints and to sinsin-gle-sided beam-to-beam joint configurations. In the other cases, the bearing forces result from both the left and right con-nected members, with the added problem that the num-ber of connecting bolts may differ for the left and right connections. The calculation procedure may cover such cases without any particular difficulty. It could just bring some more complexity in the final presentation of the design sheet.

(48)

48

Header plate in shear:

Gross section VRd 4 = 0 M yp p p 3 f 27 , 1 t h 2 γ (2 sections)

Header plate in shear:

Net section VRd 5 = 2 M up net . v 3 f A 2 γ (2 sections) with Av,net = tp ( hp – n1 d0)

Header plate in shear: Shear block VRd 6 = 2 Feff,Rd (2 sections) • if hp < 1,36 p22 and n1 > 1: Feff,Rd = 0 M nv yp 2 M nt up Rd , 2 , eff A f 3 1 A f 5 , 0 F γ + γ = • else: Feff,Rd

=

0 M nv yp 2 M nt up Rd , 1 , eff A f 3 1 A f F γ + γ = with p22 = p2' for n2 = 2 = p2' + p2 for n2 = 4

Ant = net area subjected to tension

- for one vertical bolt row (n2 = 2): Ant = tp ( e2 –

2 d₀

)

- for two vertical bolt rows (n2 = 4): Ant = tp ( p2 + e2 – 3

2 d₀

)

Anv = net area subjected to shear = tp ( hp – e1 – (n1 – 0,5) d0 ) (see clause 3.10.2 in EN 1993 Part 1-8)

(49)

49 Header plate in

bend-ing • if hp ≥ 1,36 p22: VRd 7 = ∞ • else: VRd 7 = 0 M yp w 22 el f 2 ) t p ( W 2 γ − with p22 = p2' for n2 = 2 = p2' + p2 for n2 = 4 6 h t W 2 p p el =

Beam web in shear

VRd 8 = 3 f h t 0 M ybw p bw γ (clause 5.4.6 in Eurocode 3) Shear resistance of the joint Rdi 8 1 i Rd V V

_min

= = NOTE:

The design shear resistance of the joint can only be considered if all the requirements (section 6.2.1) are satisfied.

(50)

50

6.2.3 Resistance to tying forces

FAILURE MODE VERIFICATION

Bolts in tension Nu 1 = n Bt,u

with: Bt,u = fub As/γMu

Header plate in bend-ing

Nu 2 = min ( Fhp,u,1; Fhp,u,2 )

Fhp,u,1 = ) n m ( e n m 2 m l ) e 2 n 8 ( p p w p p p . u 1 , t . p . eff w p + − − Fhp,u,2 = p p p u . t p . u 2 , t . p . eff n m n B n m l 2 + + where np = min ( e2; 1,25 mp ) mu.p = Mu up p f t γ 4 2 leff.p1 = leff.p2 = hp

(usually safe value; see EC3 – table with effective lengths for end plates, case “Bolt-row outside tension flange of beam” – for more precise values; the effective lengths given in the table have however to be multiplied by a factor 2 before being introduced in the two expressions given above)

Supporting member in bending

Nu 3 = See EN 1993 Part 1-8 for column flanges (with substitu-tion of Bt.Rd by Bt,u, fy by fu and γM0 by γMu).

See published reference documents for other supporting members (for instance [12])

Beam web in tension Nu 4 = tw hp fubw/γMu

Welds The full-strength character of the welds is ensured through rec-ommendations for weld design given in the design sheet for shear resistance.

Tying resistance of the

joint ui 4 1 i u

N

_min

=

(51)

51 6.3 Design sheet for connections with a fin plate

6.3.1 Requirements to ensure sufficient rotation capacity The two following inequalities has to be fulfilled.

(1) hp ≤ db (2) φ_available >φ_required where: • if z >

(

)

2 e p 2 h h 2 h g z _⎟⎟ ⎠ ⎞ ⎜⎜ ⎝ ⎛ + + − : " " available = ∞ φ • else:

(

)

⎟⎟ ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎜ ⎜ ⎝ ⎛ + − − ⎟ ⎟ ⎟ ⎟ ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎜ ⎜ ⎜ ⎜ ⎝ ⎛ ⎟⎟ ⎠ ⎞ ⎜⎜ ⎝ ⎛ + + − = φ e p h 2 e p 2 h available h 2 h g z arctg h 2 h g z z arcsin

6.3.2 Requirements to avoid premature weld failure The following inequality has to be fulfilled.

a > 0,4 tp βw 3 0 M 2 M up yp f f γ γ (βw is given in Table 4.1)

ECCS Recommendations Simple Joints

ECCS TC10

Connections

European Recommendations for the Design of

Simple Joints in Steel Structures

European Recommendations for the Design of Simple Joints

in Steel Structures

TC10

Connections

European Recommendations for the Design of

Simple Joints in Steel Structures

J.P. Jaspart

J.F. Demonceau

S. Renkin

PREFACE

CONTENTS

M

V

M

M

V

V

M

M

V

VEd

M

M

V

V

M

M

V

VEd

M

φ

σ

τ

σ

F

F

A

(

)

(

)

(

)

(

)

=

min

N

N

min

=

(

)

(

)

_min

_min