Eurocodes pour

les Structures Mixtes

Élaboration d'une
Approche Transnationale

Formation: Eurocode 4

Lecture 9 : Composite joints

Annex A

Summary:

 Traditionally structural joints are considered as rigid or pinned

 The lecture is intended to introduce the concept of advanced, semi-rigid joints
 The requirements relating to stiffness, strength and rotation capacity are explained
 The steps of joint representation are shown
 The principles of the component method (including component identification, component
characterisation and component assembly) for joint characterisation are presented
 Design provisions for composite joints

Pre-requisites:

 Basic knowledge about frame analysis and design

 Basic definitions and concepts on semi-rigid joints
 Plastic response of frames
 Knowledge of SSEDTA-1 Module 5

Notes for Tutors:

 This is the extended, detailed version with furthermore additional information and background
knowledge for designing composite joints.
 In Annex B there are calculation procedures and worked examples for composite joints

© SSEDTA 2001 Last modified jj/06/18

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

References:
 [1] EN 1994-1-1: Design of composite steel and concrete structures Part 1.1
General rules and rules for buildings; Draft No. 2; April 2000
 [2] prEN 1993-1-8: Eurocode 3: Design of steel structures
Part 1.8: Design of joints; Draft No.1; 26. February 2000
 [3] SSEDTA (1): Structural Steelwork Eurocodes Development of A Trans-national Approach
Module 5: Structural Joints
 [4] Gerald Huber: Non linear calculations of composite sections and semi-continuous joints
Doctoral Thesis, Verlag Ernst & Sohn, December1999
 [5] COST C1: Composite steel-concrete joints in frames for buildings: Design provisions
Brussels, Luxembourg 1999
 [6] COST C1: Control of the semi-rigid behaviour of civil engineering structural connections
Proceedings of the international conference Liège, 17 to 19 September 1998
 [7] COST C1: Composite steel-concrete joints in braced frames for buildings
Brussels, Luxembourg 1996
 [8] Teaching Modules of A-MBT,1999/2000
Application Centre Mixed Building Technology , Innsbruck -Austria,
 [9] ESDEP: European Steel Design Education Programme; CD-ROM 1996
 [10] EN 1992-1: Design of concrete structures
 [11] EN 1993-1: Design of steel structures; April 1992

Contents:

1 Introduction
1.1 Definition and terminology
1.1.1 Types of joints
1.1.1.1 Conventional joints
1.1.1.2 Advanced joints
1.2 Composite joints for simple framing
1.3 Joints in semi-continuous construction
1.4 Scope

2 Joint Representation (General)

2.1 Joint characterisation
2.1.1 Component method
2.1.2 Component models (spring models)
2.1.3 Component assembly
2.1.3.1 Linear grouping
2.1.3.2 Rotational grouping
2.1.3.3 Rotational stiffness
2.1.3.4 Design moment resistance
2.1.3.5 Transformation of joint characteristics
2.1.3.6 Moment resistance
2.1.3.7 Stiffness
2.1.4 Basic components of a joint
2.2 Classification of beam-to-column joints
2.2.1 Classification by strength
2.2.2 Classification by rotational stiffness
2.3 Idealisation
2.4 Joint modelling of beam-to-column joints
2.4.1 General
2.4.2 Joint modelling reflecting the actual behaviour
2.4.3 Simplified joint modelling

3 Joint Representation of Composite Beam-to-Column Joints, Design Provisions

© SSEDTA 2001 Last modified jj/06/18 2

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

3.1 Framing arrangements in composite construction

3.1.1 Introduction
3.1.2 General factors influencing framing arrangements
3.1.3 Composite construction
3.1.4 Sizing of composite beams
3.1.5 Framing arrangements and composite joints
3.2 Design provisions
3.2.1 Basis of design
3.2.2 Design moment resistance
3.3 Resistance of basic components
3.3.1 General
3.3.2 Column web panel in shear
3.3.3 Column web in compression
3.3.4 Column web in tension
3.3.5 Column flange in bending
3.3.6 End-plate in bending
3.3.7 Beam flange and web in compression
3.3.8 Beam web in tension
3.3.9 Longitudinal slab reinforcement in tension
3.3.10 Contact plate in compression
3.4 Design moment resistance of joints with full shear connection
3.4.1 General
3.4.2 Beam-to-column joints with contact plate
3.4.3 Beam-to-column joints with partial-depth end-plate
3.4.4 Beam-to-column joints with bolted end-plate connections
3.5 Rotational stiffness
3.5.1 Basic model
3.5.2 Initial stiffness Sj,ini
3.5.2.1 Connections with one layer of components in tension
3.5.2.2 Connections with more than one layer of components in tension
3.5.3 Stiffness coefficients for basic joint components
3.5.4 Deformation of the shear connection
3.6 Rotation capacity
3.7 Detailing
3.7.1 Longitudinal slab reinforcement
3.7.2 Transverse slab reinforcement
3.7.3 Anchorage of reinforcement

4 Summary
4.1 General
4.2 Field of application of the calculation procedures
4.3 Summary of key steps
4.4 Conclusion

© SSEDTA 2001 Last modified jj/06/18 3

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

1 Introduction
This lecture concerns building frames with composite steel-concrete beams. The primary aim
is to explain how to design joints under hogging bending moment as composite elements.

1.1 Definition and terminology

 Composite joint [1] 1.4.2

A joint between composite members, in which reinforcement is intended to contribute to
the resistance and the stiffness of the joint
 Basic component (of a joint)
specific part of a joint that makes an identified contribution to one or more of its structural
properties
 Connected member
member that is supported by the member to which it is connected
 Connection
location at which two members are interconnected, and the means of interconnection
 Joint
assembly of basic components that enables members to be connected together in such a
way that the relevant internal forces and moments can be transferred between them
 Joint configuration
type or layout of the joint or joints in a zone within which the axes of two or more inter-
connected members intersect
 Structural properties (of a joint)
its resistance to internal forces and moments in the connected members, its rotational
stiffness and its rotation capacity

Beam-to-column joint

beam-to-beam joint splice

Figure 1 Types of joints

© SSEDTA 2001 Last modified jj/06/18 4

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

The key feature is the provision of continuous slab reinforcement to act in tension across the
joint. This increases substantially both resistance and stiffness for little increase in work on
site.

Generally there is a lack of full continuity between the floor system and the column. Frames
with composite joints are therefore semi-continuous in nature. This type of framing enables
advantage to be taken of the stiffness and moment resistance inherent in many forms of
connection whilst avoiding the expense of rigid and full-strength steelwork connections. The
recognition of this approach is widely regarded as one of the advances in prEN 1993-1-8 [2],
compared to earlier standards.
Conventionally, joints have been treated as nominally pinned without any strength or stiffness
(simple joints) or as rigid with full strength (continuous joints). The differences between
conventional design and semi-continuous construction can be seen from Figure 2 to Figure 4.
The types of framing depend on the joint characteristics, particularly on the initial stiffness and
the moment resistance relative to the connected members. The semi-continuous approach
provides greater freedom, enabling the designer to choose connections to meet the particular
requirements of the structure.

© SSEDTA 2001 Last modified jj/06/18 5

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

1.1.1 Types of joints

The construction of composite joints used to be based on the already well-known conventional
steel joints (see Figure 1 and Figure 5). The composite floor system was connected to the
columns by hinged or semi-rigid steel connections, as e.g. welded, angle cleats and flush
endplate joints. However a gap separated the concrete res. composite slab from the column to
prevent an interaction with the joint. This constructional separation was necessary due to a
lack of knowledge in view of modelling the interaction. As the gap had to be provided by
expensive constructional means and furthermore the joint’s stiffness and strength were
relatively low in comparison to the adjacent composite members this solution was not
economic.

Nowadays in advanced composite joints the floor system is integrated into the joint (Figure 5),
leading to much higher values of stiffness and strength. This very efficient solution became
possible, because the interaction problem between the slab and the column has been solved in
comprehensive studies.

Generally one has to distinguish between beams located below the slab (conventional floors)
and beams which are already integrated into the slab (slim floors). A further distinction
concerns the type of connection before concreting of the slab. In that cases where already a
semi-rigid steel connection is provided (e.g. welded, angle cleats, flush or partial depth
endplates) the slab supplies additional stiffness and strength after hardening of the concrete. In
the very economic case of hinged steel joints during erection, by adding contact pieces and due
to the integration of the slab into the joint a high bearing capacity and stiffness can be gained
without any further bolting or welding on site (joints with brackets and additional contact
plates for the compression transfer or hinged steel joints with fins or angle cleats, see Figure
5). For slim floor decks the authors in any case recommend to design hinged steel joints, which
then by concreting are automatically converted to semi-rigid composite joints with
considerable resistance and rotation ability.

A further increase of joint stiffness and strength can be obtained by concreting the column
sections (encased composite columns). Figure 5 gives an overview of all mentioned joint types
using H-shaped column sections. Regarding the fire resistance and also the appearance,
hollow column sections, which can be filled with concrete, turned out to be forward-looking.

1.1.1.1 Conventional joints:

 Simple joints

© SSEDTA 2001 Last modified jj/06/18 6

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

pinned

Figure 2 Simple joint (pinned)

 Continuous joints

rigid, full strength

Figure 3 Rigid joint (continuous)

1.1.1.2 Advanced joints

 Semi continuous joints

© SSEDTA 2001 Last modified jj/06/18 7

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

M
Resistance

stiffness

rotation capacity 

rigid or semi-rigid,
full or partial strength,
specific rotation capacity

Figure 4 Advanced composite joint (semi continuous)

© SSEDTA 2001 Last modified jj/06/18 8

 Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

C O N V E N T IO N A L - s te e l jo in ts

gap gap gap

w e ld e d a n g le c le a ts f lu s h e n d p la te

A D V A N C E D - c o m p o s ite jo in ts
b e a m lo c a te d b e lo w s la b
c o n v e n tio n a l flo o r

w e ld e d a n g le c le a ts p a r t ia l d e p th e n d p la t e

fin s + a n g le c le a ts + b ra c k e t +
b e a m in te g r a te d in to s la b

c o n ta c t p la t e s c o n ta c t p la t e s c o n ta c t p la t e s
s lim f lo o r

Figure 5 Types of joints- H- shaped column

© SSEDTA 2001 Last modified jj/06/18 9

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

As semi continuous joints influence the response of the frame to load, they should be modelled
in the global frame analysis. Chapter 3 describes suitable methods for doing this.
Traditionally the joints simply have not been considered in the global calculation as a separate
element. However as a joint strictly consists of parts of the column, parts of the beams and
parts of the slabs, connecting elements and sometimes also includes stiffening elements, so the
real behaviour can only be taken into consideration by defining the joint as a separate element
(Figure 6) within the structure, additional to the beams and column elements.

storey building

beam to column joint

Figure 6 Joint as a separate element within the structure

This enables more efficient constructions but the influence of the joints on the global behaviour
is so important that the old-fashioned philosophy of perfect hinges or fully continuous
restraints does not describe the real behaviour of a semi continuous joint. (See Figure 7)

It is conventional to simplify the framing as simple or continuous framing, which has the
advantage of straight-forward calculation. However with modern design approaches it is
appropriate to replace the conventional calculation methods by more advanced ones which
treat the joints in a realistic manner. As already noted, semi-continuous construction enables
advantage to be taken of the stiffness and moment resistance inherent in many forms of
connection, without the expense of forming the rigid and/or full-strength joints necessary for
continuous construction.

Although structural behaviour is three-dimensional, the usual presence of stiff floor slabs
normally allows the designer to neglect out-of-plane and torsional deformations of the joint.
The joint characterisation is therefore usually in the form of a moment rotation relationship
(Figure 7).

According to this new approach, joints may be assessed with regard to the following three
main characteristics.

 The initial rotational stiffness Sj,ini:

A joint with a very small rotational stiffness and which therefore carries no bending moment is
called a hinge. A rigid joint is one whose rigidity under flexure is more or less infinite and
which thus ensures a perfect continuity of rotations. In between these two extreme boundaries
we speak about semi-rigid joints.

 The design moment resistance M j,Rd:

© SSEDTA 2001 Last modified jj/06/18 10

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

In contrast to a hinge, a joint whose ultimate strength is greater than the ultimate resistance
(ultimate strength) of the parts whose linkage it ensures is called a full strength joint. Again a
partial strength joint represents a middle course between these extremes. (For simplicity from
now on “resistance” will mostly be used for the ultimate resistance value; the terms
“resistance” and “strength” are used in the Eurocodes with an identical meaning.)

 The design rotation capacity φ Cd:

Brittle behaviour is characterised by fracture under slight rotation, usually without plastic
deformations. Ductile behaviour is characterised by a clear non-linearity of the moment-
rotation curve with a large plateau before fracture. It usually indicates the appearance of plastic
deformations. The ductility coefficient is the ratio between the ultimate rotation and the elastic
rotation limit. Semi-ductility falls in between brittle and ductile behaviour.

Figure 7 Joint response [4]

Chapter 3 describes how these can be calculated. As is usual in the Eurocodes, partial safety
factors are applied to loads and strengths during ultimate limit state verifications, but not to
the modulus of elasticity nor to limiting strains.

By analysing full-scale joints it could be seen very quickly that the number of influencing
parameters is too large. So world-wide the so-called component method is accepted universally
as the best method to describe the joint behaviour analytically. In contrast to the common finite
element method (FEM), which often fails to consider local load introduction problems, the
joint here is divided into logical parts exposed to internal forces. So while the FEM works on
the level of strains and stresses, the component method concentrates on internal forces and
deformations of the component springs.

In recent years all over the world extensive testing programs have been performed worldwide
for studying the non-linear behaviour of individual components and their assembly to gain the
non-linear moment-rotation characteristic of the whole joint formed by these components.

Traditionally, engineering judgement has been used to ensure that joint behaviour
approximates to that required for simple and continuous construction. However, the concept of
semi-continuous construction requires a more precise statement of what constitutes each joint.
In prEN 1993-1-8 [2] this is provided by a classification system based on joint resistance and
stiffness. This is extended to composite constructions. Advice is also given on types of
connections suitable for each type of construction and on detailing rules.

A consistent approach for structural joints

The rotational behaviour of actual joints is well recognised as being often intermediate between
the two extreme situations, i.e. rigid or pinned.
Later in this lecture, the difference between joints and connections will be introduced. For the
time being, examples of joints between one beam and one column only will be used.
Consider now the bending moments and the related rotations at a joint (See Figure 8):

(a) Rigid joint (b) Pinned joint (c) Semi-rigid joint

Figure 8 Relationship of bending moment and rotations at a joint

When all the different parts in the joint are sufficiently stiff (i.e. ideally infinitely stiff), the
joint is rigid, and there is no difference between the respective rotations at the ends of the

© SSEDTA 2001 Last modified jj/06/18 11

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

members connected at this joint (Figure 8a). The joint experiences a single global rigid-body
rotation which is the nodal rotation in the commonly used analysis methods for framed
structures.
Should the joint be without any stiffness, then the beam will behave just as simply supported
whatever the behaviour of the other connected member (Figure 8b). This is a pinned joint.
For intermediate cases (non zero and non infinite stiffness), the transmitted moment will result
in a difference φ between the absolute rotations of the two connected members (Figure 8c). The
joint is semi-rigid in this case.
The simplest means for representing the concept is a rotational (spiral) spring between the
ends of the two connected members. The rotational stiffness S of this spring is the parameter
that links the transmitted moment Mj to the relative rotation φ  which is the difference
between the absolute rotations of the two connected members.
When this rotational stiffness S is zero, or when it is relatively small, the joint falls back into
the pinned joint class. In contrast, when the rotational stiffness S is infinite, or when it is
relatively high, the joint falls into the rigid joint class. In all intermediate cases, the joint
belongs to the semi-rigid joint class.
For semi-rigid joints the loads will result in both a bending moment Mj and a relative rotation
φ between the connected members. The moment and the relative rotation are related through a
constitutive law which depends on the joint properties. This is illustrated in Figure 9, where,
for the sake of simplicity, the global analysis is assumed to be performed with linear elastic
assumptions.
At the global analysis stage, the effect of having semi-rigid joints instead of rigid or pinned
joints is to modify not only the displacements, but also the distribution and magnitude of the
internal forces throughout the structure.
As an example, the bending moment diagrams in a fixed-base simple portal frame subjected to
a uniformly distributed load are given in Figure 10 for two situations, where the beam-to-
column joints are respectively either pinned or semi-rigid. The same kind of consideration
holds for deflections.

(a) Rigid joint (b) Pinned joint (c) Semi-rigid joint

(φ = 0) (Mj = 0) (Mj and φ  0)

Figure 9 Modelling of joints (case of elastic global analysis)

© SSEDTA 2001 Last modified jj/06/18 12

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

(a) Pinned joints (b) Semi-rigid joints

Figure 10 Elastic distribution of bending moments in a simple

portal frame

1.2 Composite joints for simple framing

Most of the joints in composite frames are currently treated as nominally pinned and the
frames are therefore of “simple” construction. This provides the advantage that the global
analysis is straightforward because the structure is easier to calculate.
Nominally pinned joints are cheap to fabricate and easy to erect. They may be the most
appropriate solution when significant ground settlement is expected. However, deeper beam
sections will be required compared to other forms of construction, leading to an increased
height of the structure and greater cladding costs. Due to the beam end rotations, substantial
cracking can occur in the joint area if the slab is continuous over internal supports.

The joint has to be designed to transfer safely the vertical shear and any axial forces. The
resistance of the slab to vertical shear is small and is neglected. As any continuity in the slab is
also neglected, a nominally pinned joint for a composite beam is therefore designed in the
same way as a simple steelwork connection.

If the slab is constructed as continuous, uncontrolled cracking is permitted by Eurocode 2 if the

exposure conditions are Class 1. According to EN 1992-1 [10], most interiors of buildings for
offices or normal habitation are in this class, which implies that there is no risk of corrosion of
reinforcement. Appearance requires a floor finish with ductile behaviour or provision of a
covering. Even so, minimum areas of reinforcement are specified to prevent fracture of the
bars or the formation of very wide cracks under service loading. Where cracks are avoided by
measures such as the provision of joints in the slab, these should not impair the proper
functioning of the structure or cause its appearance to be unacceptable.

In industrial buildings, no additional finish or covering is included and the slab itself provides
the floor surface. In such cases continuous or semi-continuous construction is preferable, so
that crack widths can be adequately controlled.
Although joint ductility is essential in simple construction, it is not usual for designers to
calculate either the required or available rotation capacity. For steel connections it is sufficient
to rely on the observed behaviour of joint components.

Rotation capacity can be provided by bolt slip and by designing the components in such a way

© SSEDTA 2001 Last modified jj/06/18 13

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

that they behave in a ductile manner. Local yielding of thin steel end-plates and the use of a
wide transverse bolt spacing are examples of this approach. Fracture of bolts and welds should
not be the failure mode. However, as the slab reinforcement is assumed to have no influence on
the behaviour at the ultimate limit state, fracture of this does not limit the rotation capacity of
the joint.

1.3 Joints in semi-continuous construction

Composite moment-resisting joints provide the opportunity to improve further the economy of
composite construction. Several possible arrangements for composite joints are shown in
Figure 5, demonstrating the wide variety of steelwork connection that may be used. In the
interest of economy though, it is desirable that the steelwork connection is not significantly
more complicated than that used for simple construction of steel frames. With conventional
unpropped construction, the joints are nominally pinned at the construction phase. Later the
steelwork connection combines with the slab reinforcement to form a composite joint of
substantial resistance and stiffness.

A particularly straightforward arrangement arises with "boltless" steelwork connections. At the

composite stage all of the tensile resistance is provided by the slab reinforcement, and no bolts
act in tension. The balancing compression acts in bearing through plates or shims, inserted
between the end of the lower flange of the beam and the face of the column to make up for
construction tolerances. Additional means to resist vertical shear must also be provided, for
example by a seating for the beam.

The benefits which result from composite joints include reduced beam depths and weight,
improved service performance, including control of cracking, and greater robustness. Besides
the need for a more advanced calculation method and the placement of reinforcement, the
principal disadvantage is the possible need for transverse stiffeners to the column web, placed
opposite the lower beam flange. These are required if the compression arising from the action
of the joint exceeds the resistance of the unstiffened column web. Alternatively, stiffeners can
be replaced by concrete encasement.

1.4 Scope
In the following, detailed provisions are given for:
 Joints with flush or partial-depth end-plates;
 Joints with boltless connections and contact plates.

The joints are intended for conventional composite beams in which the structural steel section
is beneath the slab. It is assumed that the connections are subject to hogging bending moment
due to static or quasi-static loading. Profiled steel decking may be used both to form the in-situ
concrete and to act as tensile reinforcement, creating a composite slab; alternatively, the slab
may be of in-situ reinforced-concrete construction or may use precast units.
Steel sections for columns may be H- or I-shaped and may act compositely with concrete
encasement.

2 Joint Representation (General)

Conventionally beam-to-column joints have been treated either as pinned without any strength
or stiffness or as fully rigid with full strength due to lack of more realistic guidance in view of
joint representation. In reality both assumptions may be inaccurate and uneconomic and do
only represent the limiting cases of the real moment-rotation behaviour. They may lead to a
wrong interpretation of the structural behaviour in terms of load resistance and deflections. So
whereas up to now the joint construction expensively has been adopted to the possibilities of
calculation, the new approach is to develop efficient joint types first and to take their realistic
behaviour into consideration within the frame analysis afterwards. Due to the interaction

© SSEDTA 2001 Last modified jj/06/18 14

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

between joints and members an overall cost-optimisation is only possible if both design tasks
are taken over by the same party. So the designer at least should be able to bring in a first good
guess of the joint characteristics depending on the chosen joint configuration.

First discoveries on the importance of joint representation released a real innovation boom,
which can be seen by a tremendous number of research activities all over the world.
So e.g. comprehensive research projects on representation and design of structural steel and
composite connections and their effect on the frame response have been co-ordinated by the
“European Co-operation in the Field of Scientific and Technical Research (COST) [5],[6],[7].

Moment resisting joints have to transfer moments and forces between members with an
adequate margin of safety. Their behaviour obviously influences the distribution of moments
and forces within the structure. Therefore the list of construction elements as beams, slabs and
columns has to be extended by the joints. An overall account of the behaviour would need to
recognise its three-dimensional nature. However the presence of rather stiff continuous floor
slabs usually allows to neglect out-of-plane and torsional deformations of the joint.
That is why the attempt to describe the connections response can be reduced to a description of
the in-plane behaviour. In contrast to the idealising assumptions of beam-to-column
connections as hinges or full restraints, with the main attention aimed on their resistance, the
interest of actual research activities lies in the assessment of the non-linear joint response of
the whole M-φ curve with the following three main characteristics:
 Initial rotational stiffness
 Moment resistance
 Rotation capacity

It is true that the conventional simplifications of continuous or simple framing bring the
advantage of a very simple calculation, but reality again lies in between the extreme limiting
cases, what especially affects the serviceability limit state and the stability of the whole
structure.

The joint representation covers all necessary actions to come from a specific joint
configuration to its reproduction within the frame analysis. These actions are the
 Joint characterisation: determination of the joint response in terms of M-φ curves
reflecting the joint behaviour in view of bending moments and shear (in case of moment
imbalance)
 Joint classification: if aiming at conventional modelling the actual joint behaviour may
be compared with classification limits for stiffness, strength and ductility; if the respective
requirements are fulfilled a joint then may be modelled as a hinge or as a rigid restraint.
 Joint idealisation: for semi-continuous joint modelling the M-φ behaviour has to be taken
into consideration; depending on the desired accuracy and the type of global frame
analysis the non-linear curves may be simplified as bi- or tri linear approximations.
 Joint modelling: reproduction (computational model) of the joint´s M-φ behaviour within
the frame modelling for global analysis.

Figure 11 shows how the moment-rotation behaviour of joints can be represented in global
structural analysis in a practical way.

Figure 11 Practical procedure of joint representation

© SSEDTA 2001 Last modified jj/06/18 15

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

2.1 Joint characterisation

This chapter describes how to derive the M-φ curves, representing the necessary input data for
the joint models.
To determine these, several possibilities exist:
 Joint tests
 Finite element calculations
 An analytical approach

The general background for any joint model comprises three separate curves:
 One for the left hand connection
 One for the right hand connection
 One for the column web panel in shear

Figure 12 Left connection, column web panel in shear, right

connection

2.1.1 Component method

A description of the rotational behaviour of a joint has to take into account all sources of
deformations within the joint area. Furthermore all possibilities of local plastic deformations
and instabilities have to be covered by such an analytical model. The multitude of influencing
parameters was the reason, why the first attempts to develop a component method directly
based on full-scale joint tests have been doomed to failure. Also a second research tendency
with the aim of finite element modelling did not really succeed because of local detailing
problems. So for the analytical determination of the non-linear joint response a macroscopic
inspection by subdividing the complex finite joint into so-called simple “components” proved
to be successful. In contrast to the finite element method (FEM) the components - as logical
subsystems of the joint – are exposed to internal forces and moments and not to stresses. As it
will be described later in detail these components can be understood as translational springs
with a non-linear force-deformation behaviour. All considered components can be tested in
clear, relatively cheap so-called “component tests”, on the basis of which theoretical models
can be developed. Finally the total joint response (reproduced in the “joint model for global
frame analysis”) can be derived by assembling all influencing components in the correct
manner based on the so-called “component model”. This method offers the advantages of a
minimum of testing costs, clear physical calculation models for the components and a
maximum of flexibility for the designer, who is able to combine a multitude of available
components to a most economic joint configuration. Numerous full-scale joint tests confirm
this approach.

In Figure 13 the development of joint modelling is shown. Conventionally rigid or hinged

joints of infinite small size have been assumed in the global analysis. But as any real joint has
a finite size, deformations occur under relevant member forces. So in the traditional view the
beam and column stubs within the realistic joint area (b j, hj according to Figure 13) already can
be understood as an attempt to model the deformation of the real joint. The first step of
improvement was to regard the joint as a separate element of finite size. The second step was
to describe the force-deformation behaviour of all individual components by non-linear
translational springs. Setting together a joint out of these components, taking into
consideration their location and compatibility conditions, results in the component model.

© SSEDTA 2001 Last modified jj/06/18 16

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

Figure 13 Modelling of joints – conventional and advanced

For a further treatment it is important to define three different locations within the finite joint
area:
 C represents the Centre point of the joint located at the intersection between beam and
column axes.
 L is the Loading point located at the centre of rotation in view of tension and compression
at the front of the column flange (simplifying L can be assumed at the level of the beam
axis).
 S is the Shear point located at the top and bottom of the joint’s lever arm z along the
column axis.

In the third step the components are assembled to rotational springs in L and S forming the
finite joint model, which then can be used in the advanced global analysis. There the joint
region (bj, hj) is set infinite stiff and all deformations are assigned to the flexural springs at the
joint edges (L and S). A further possibility in advanced global analysis is to concentrate the full
joint’s flexibility within flexural springs at the axes intersection points C (separately for the left
and right hand side). Doing so the joint again is reduced to an infinite small point within the
global analysis, however this requires a transformation from the finite joint model to this
concentrated joint model.

An analytical description of the behaviour of a joint has to cover all sources of deformabilites,
local plastifications, plastic redistribution of forces within the joint itself and local instabilities.
Due to the multitude of influencing parameters, a macroscopic inspection of the complex joint,
by subdividing it into “components”, has proved to be most appropriate. In comparison with
the finite element method, these components, which can be modelled by translational spring
with non-linear force-deformation response, are exposed to internal forces and not to stresses.
The procedure can be expressed in three steps:

 Component identification
determination of contributing components in compression, tension and shear in view of
connecting elements and load introduction into the column web panel.

 Component characterisation
determination of the component´s individual force-deformation response with the help of
analytical mechanical models, component tests or FEM-simulations.

 Component assembly
assembly of all contributing translational component springs to overall rotational joint
springs according to the chosen component model.

2.1.2 Component models (spring models)

The component model for steel joints and later that for composite joints, has been developed
on the basis of considerations of how to divide the complex finite joint into logical parts
exposed to internal forces and moments and therefore being the sources of deformations. So in
horizontal direction one has to distinguish between the connecting and the panel zones, in
vertical direction between the tension, the compression and the shear region, all together
forming six groups as illustrated in Figure 14.

© SSEDTA 2001 Last modified jj/06/18 17

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

Figure 14 Groups of component model [4]

Table 1 Groups of component model

According to prEN 1993-1-8 [2] a basic component of a joint is a specific part of a joint that
makes-dependent on the type of loading-an identified contribution to one or more of its
structural properties. When identifying the contributing components within a joint one can
distinguish between components loaded in tension (or bending), compression and shear. Beside
the type of loading one can distinguish between components linked to the connecting elements
and those linked to load-introduction into the column web panel (both are included in the
connection) and the component column web panel in shear. The nodal subdivision for the most
general case of a composite beam-to-column joint leads to a sophisticated component model
like that developed in Innsbruck (Figure 15). There, the interplay of the several components is
modelled in a very realistic way. Any composite joints (beam–to-column, beam-to-beam or
beam splice), even steel joints can be derived as being only a special case of this general
model. The sophisticated component model leads to a complex interplay of components and
therefore to iterations within the joint characterisation itself. For simplifications concerning
the component interplay a simplified component model has been developed (Figure 15).

Figure 15 Sophisticated and simplified spring models

Figure 16 Refined component model

Through familiarity with the spring model (component model), which represents the interplay
of all deformation influences, the principles of economic joint construction with respect to
stiffness, resistance failure modes and ductility can be logically derived. Assuming a specific
component with a given resistance and deformation ability, it is self-evident that by increasing
its lever arm the joint´s moment resistance increases whilst the rotation capacity decreases.
Providing resistance in the compression region far beyond the resistance of the tension region
is uneconomic. Similarly strengthening the connecting elements will not make the joint
capable of sustaining more load, if failure is already dominated by the column web in tension,
compression or shear. It also can be recognised immediately that it makes no sense to combine
a relatively weak and ductile slab reinforcement with a stiff and brittle steelwork connection.
These examples give ideas on how to use the component model directly for a qualitative plastic
redistribution within the joint itself, all components are used up to their plastic resistance and
provide a further yield plateau for redistribution of moments within the whole frame.

© SSEDTA 2001 Last modified jj/06/18 18

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

Figure 16 shows the real joint situation on the left side and the corresponding component
model on the right. The method of component modelling started with beam-to-column joints in
steel. The bolt-rows in tension are modelled by corresponding spring rows, separating the
influences of load-introduction within the column web panel and the connection to the column
flange. The same philosophy has been followed for the compression. For an unbalanced loaded
joint, the deformations due to the shear have to be modelled by an additional shear spring.

It has been proved by tests on full scale joints that the measured moment-rotation behaviour is
in good agreement with the calculated curves using the component method. Out of this
excellent experience the component method has been extended to composite joints, where only
the additional concrete components had to be analysed. Thus the composite component model
can be seen as the most general tool covering all considerable joint types. In this sense steel
joints or special types of joints like splices, beam-to-beam joints or weak axis joints can be
regarded as special cases of this general model (see Figure 17).

Figure 17 Modelling of joints – examples

2.1.3 Component assembly

The transfer from the force-deformation curves of the individual basic joint components to the
moment-rotation curves representing the connection or the shear panel has to be done based on
the component modes fulfilling the requirements of compatibility and equilibrium. Doing so it
is assured that the joint model behaves exactly in the same way than the complex component
model with respect to applied moments. Depending on the intended level of accuracy the
assembly can be done for the main rotational key values only (initial rotational stiffness,
moment resistance, rotation capacity) or for the full shape of the resulting M-φ curves.

As already mentioned the assembly of the sophisticated component model leads to iteration
loops due to the complex interplay of components.
For simplification, iterations can be avoided by using the simplified component model used in
the Eurocodes, where the sum of all basic component springs can be derived by adding them
step by step acting parallel or in series. (See Figure 18and Figure 19)

Figure 18 Assembly of parallel spring groups

Figure 19 Assembly of serial spring groups

A component model for a composite beam-to-column joint configuration is shown in Figure

16. Sixteen different components have been identified and these may be grouped as
compression, tension and shear. The assembly of components to determine the structural
properties is now explained.

© SSEDTA 2001 Last modified jj/06/18 19

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

2.1.3.1 Linear grouping

Individual components are represented by translational springs. The first step is therefore to
group components which act together directly in parallel or in series. Translational stiffness,
resistance and deformation capacity are considered separately. Each of the regions in
compression, tension or shear is also considered separately. The assembly is illustrated in
Figure 20, in which the translational stiffness is denoted by ci, defined by:

ci=Fi / wi (0)

The stiffness coefficients ki (see 3.5.3) and used elsewhere in this document, are related to c i
by:

ki=ci / Ei (0)

For composite joints it would be more convenient to use a stiffness ci rather than a coefficient
ki. This is because a composite joint includes more than one material. Despite this, stiffness
coefficients are used elsewhere in this lecture, to avoid differing from prEN 1993-1-8 [2].
For components acting in parallel - for example, the column web of a steel column section and
concrete encasement- the initial stiffness and resistances can be added. However, the smallest
of the deformation capacities governs that of the group. Such behaviour is represented by the
diagrams in the lower part of Figure 20.

For components effectively acting in series - for example, a single bolt-row comprising the
end-plate in bending, the bolts themselves in tension and the column flange in bending- the
initial stiffness is obtained by a reciprocal relationship, but the resistance is that of the weakest
component. The deformation capacity is the sum of the weakest component´s capacity and the
corresponding deformations of the other components at that load level. Such behaviour is
represented by the diagrams in the upper part of Figure 20.

For the regions in compression and shear, the linear grouping results in one effective
translational spring per group, each with its own translational stiffness, resistance and
deformation capacity. However, the tension region may consist of several rows of springs, each
representing a layer of longitudinal slab reinforcement or a bolt-row. This may also be reduced
to one equivalent translational spring, as shown in Figure 21 by considering the rotational
behaviour of the joint.

© SSEDTA 2001 Last modified jj/06/18 20

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

Figure 20 Linear grouping

© SSEDTA 2001 Last modified jj/06/18 21

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

2.1.3.2 Rotational grouping

It is assumed in the simplified Eurocode component model that the centre of rotation for all
tension rows is located at the centre of the flange of the beam. This is valid for connections
with relatively stiff load-introduction in compression. It can be readily shown that the
rotational stiffness Seff,i is related to the translational stiffness ceff,i by:

Seff,i = ceff,i zi2 (0)

where:
zi is the distance from the centre of rotation to the effective spring i.

It is also required that the moment-rotation behaviour of each of the systems shown in Figure
21 is equal. An additional condition is that equilibrium of forces is maintained. As a result the
effective stiffness of the tension rows can be replaced by a single equivalent translational
stiffness ceq at an equivalent lever arm z. The formulae are given in Figure 21, where the
derivation is also summarised.
This figure also gives the resulting expression for resistance and deformation capacity of the
tension region, assuming as a particular case that the second tension row limits the
deformation capacity.
The second (and final) stage in the rotational grouping converts the translational properties
into moment-rotation relationships at S and L (the edges of the finite joint model defined in
Figure 13). The conversion is illustrated in Figure 21 in which the resulting formulae are also
given at the bottom left. The effective translational stiffness for the shear region is converted
into a flexural spring by multiplication with z 2. The effective translational stiffness for the
compression region and the equivalent translational stiffness for the compression region are
added in series; the resulting translational stiffness is then converted to a rotational spring by
multiplication with z 2.

Figure 21 Grouping of tension springs

2.1.3.3 Rotational stiffness

It will be noted that these rotational stiffnesses represent the deformability of the column web
panel in shear separately from the other sources of deformation. For a single-sided joint
configuration, the total rotational stiffness can be expressed directly in terms of the effective
translational stiffness in shear and compression and the equivalent translational stiffness in
tension:
(0)

where ci represents the effective or equivalent stiffness of the region i.

For a double-sided joint configuration, the extent of the shear in the column web panel is
influenced by the ratio of unbalance between the moments at the two connections. In prEN
1993-1-8 [2] this influence is included through a parameter β (see Table 4).

2.1.3.4 Design moment resistance

For resistance calculation, Figure 21, has shown the tension region as limited by the

© SSEDTA 2001 Last modified jj/06/18 22

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

deformation capacity of the second row. Recommendations to ensure ductile behaviour in steel
joints are given in prEN 1993-1-8 [2]. These are applicable to steelwork parts of composite
joints.
Detailing rules given in this document avoid brittle failure modes associated with composite
action, provided that high-ductility reinforcement is allocated for the composite joint.
Provided a plastic distribution of bolt forces is allowed by prEN 1993-1-8 [2], the design
moment of resistance MRd may be expressed as:

(0)

where the summation is over all rows of longitudinal slab reinforcement and bolt-rows in the
tension region.
In practice though the resistance of the connection in compression or of the web panel in shear
may be lower than that of the group of components in tension. For equilibrium the total tensile
force must not exceed the design resistance of the compression group F Lc,Rd and the shear
resistance VS,Rd / β. If this condition is reached at a tension row i, the contribution to the
moment resistance of all other tension rows closer to the centre of compression is neglected.

2.1.3.5 Transformation of joint characteristics

The procedures described above provide moment-rotation relationships at S and L (the edges of
the finite joint model defined in Figure 13). In such models the connection springs are
arranged at the edges of the finite joint area, as shown in Figure 29. In the simplified joint
model, the combined joint springs for shear and connections are located at the intersection C
of the beam and column axes. It is obvious that the acting bending moment usually increases
from the joint edge to the axes intersection point. The consideration of the higher value for the
applied moment therefore leads to a more conservative design. On the other hand, the
weakness of the “extended” beam and column stubs within the finite joint area adds an extra
rotation to the joints, leading to an overestimation of the global frame deformation.
These effects can be compensated by transforming the properties which apply at L and S to
properties at C.

2.1.3.6 Moment resistance

It has to be demonstrated that the internal moments due to the design loads, obtained by global
analysis, do not exceed the corresponding resistance values. However, the moment resistance
of a joint relates to the point L at the periphery of the finite joint area (see Figure 29). So
strictly only the internal moment at the face of the column, not that at the column centre-line,
has to be compared with the joint's resistance. The difference in internal moment might be
important for short-span beams with deep column profiles.

2.1.3.7 Stiffness
Two approaches have been proposed to allow for the additional flexibility which results from
extending the beam and column into the joint area:

 Transformation of rotational stiffness

By assembling appropriate components, rotational stiffnesses are obtained to represent load-
introduction and the connection at L and shear deformation at S. These are then transformed
separately from L to C and S to C, taking account of the beam and columns stubs introduced by
the simplified model. The transformed stiffness are then combined to obtain the total rotational
stiffness at C.

The transformation from S to C:

(0)
where:
f=1 for a joint at the top of the column;

© SSEDTA 2001 Last modified jj/06/18 23

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

f=2 for a joint within a continuous length of the column;

Ic is the second moment of area of the adjacent column section;
z is the lever arm of the joint.

For transformation from L to C:

(0)
where:
Ib is the second moment of area of the adjacent beam section in hogging bending;
Lj is half of the depth of the column section (see Figure 30).

 Transformation of component stiffness

As a further simplification, stiffness of individual components can be calibrated to include the
flexibility of the beam and column stubs. Thus quasi-transformed stiffness values can be given
in design codes, leaving the designer to assemble the components according to the joint
configuration. The result is a total rotational stiffness for the concentrated joint model at C.

2.1.4 Basic components of a joint

The design moment-rotation characteristic of a joint depends on the properties of its basic [2] 5.1.4
components.

The following basic joint components are identified in this lecture:

 Column web panel in shear
 Column web in compression
 Column web in tension
 Column flange in bending
 End-plate in bending
 Beam flange and web in compression
 Beam web in tension
 Bolts in tension
 Longitudinal slab reinforcement in tension
 Contact plate in compression

A major-axis beam-to-column composite connection consists of a combination of some of the

basic components listed above, excluding the column web panel in shear. In particular, it
always includes the following:
 Column web in compression
 Longitudinal slab reinforcement in tension

Methods for determining the properties of the basic components of a joint are given: [1] 8.1.2
 For resistance in 3.3 [1] 8.3.3
 For elastic stiffness in 3.5.3 [1] 8.4.2

Relationships between the properties of the basic components of a joint and the structural
properties of the joint are given:
 For moment resistance in 3.2.2 [1] 8.3.4
 For rotational stiffness in 3.5 [1] 8.4.1
[1] 8.5
 For rotation capacity in 3.6

2.2 Classification of beam-to-column joints

In conventional design of building structures, the framing is treated as simple or continuous,
even though practical joints always possess some moment resistance and show some flexibility.

© SSEDTA 2001 Last modified jj/06/18 24

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

Traditionally, engineering judgement has been used to ensure that joint behaviour
approximates to that required for these forms of construction.

However, the concept of semi-continuous construction requires a more precise statement of the
joint behaviour. In prEN 1993-1-8 [2] this is provided by a classification system based on joint [2] 7.2.2
resistance and stiffness. Table 2 shows how the types of joint model (representing behaviour), [2] 7.2.3
the form of construction and the method of global analysis are all related.

Table 2 Type of joint model

2.2.1 Classification by strength

A beam-to-column joint may be classified as full-strength, nominally pinned or partial strength
by comparing its moment resistance with the moment resistances of the members that it joins.
For this purpose, the moment resistance of the joint is compared with the moment resistance of
the connected members. The purpose is to indicate whether the joint or the adjacent member
cross-sections will limit resistance. In the first case, the joint is either “partial strength” or
“nominally pinned”. In the second case the joint is “full-strength”. Thus a beam-to-column
joint will be classified as full-strength if the following criteria are satisfied.

For a joint at the top of a column:

Mj,Rd ≥ Mb,pl,Rd (0)

or:
Mj,Rd ≥ Mc,pl,Rd (0)

where:
Mb,pl,Rd design plastic moment resistance of the composite beam in hogging bending
immediately adjacent to the joint;
Mc,pl,Rd design plastic moment resistance of the column.

For a joint within a continuous length of column:

Mj,Rd ≥ Mb,pl,Rd (0)

or:
Mj,Rd ≥ 2 Mc,pl,Rd (0)
A joint may be classified as nominally pinned if its moment resistance M j,Rd is not greater than
0,25 times the moment resistance required for a full-strength joint, provided that it also has
sufficient rotation capacity.
With composite beams the moment resistance depends on whether the member is in sagging or
hogging bending. With braced construction, beam end connections are subject to hogging
bending and the moment resistances of joints and beams should therefore be those applicable
to this situation. Under such moment, EN 1994-1-1 [1] requires full shear connection and [1] 5.3
therefore the beam´s resistance should be calculated on this basis. Plastic analysis should be
used for sections Class 1 and Class 2, including beam sections with Class 3 webs upgraded to
Class 2 using the “hole-in-the web” approach.

The classification of the joint has implications for its design criteria. With a partial-strength or
nominally pinned joint, it is the joint rather than the member section which requires rotation
capacity. Tests have shown that rotation capacity may be limited.

© SSEDTA 2001 Last modified jj/06/18 25

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

2.2.2 Classification by rotational stiffness

A beam-to-column joint may be classified as rigid, nominally pinned or semi-rigid according
to its stiffness, by determining its initial rotational stiffness S j,ini and comparing this with
classification boundaries.
As shown in Figure 22, the classification given by prEN 1993-1-8 [2] compares stiffness of the
joint with that of the connected member. The purpose is to indicate whether account has to be
taken of the influence of joint flexibility on the frame response. The classification boundaries
were determined from consideration of ultimate limit states.

Figure 22 Boundaries for stiffness classification of beam-to-

column joints

Zone 1: rigid, if Sj,ini ≥ 8 (0)

Zone 2: semi-rigid, all joints in zone 2 should be classified as semi-rigid. More

accurately, joints in zones 1 or 3 may also be treated as semi-rigid.

Zone 3: nominally pinned, if Sj,ini ≤ 0,5 (0)

where: E Ib is the uncracked flexural stiffness for a cross-section of a composite beam;
Lb is the span of a beam (centre-to-centre of columns).

For composite joints, it is necessary to decide whether the classification should be related to the
cracked or uncracked flexural rigidity of the beam´s cross-section. The boundary for rigid
joints has been determined by providing restraint against column collapse. In a braced frame
the hogging beam end moments then decrease. As the unloading stiffness is taken as equal to
the initial stiffness, the uncracked properties of the equivalent steel section should be used. To
be strictly consistent with elastic global analysis, the modular ratio should be determined in
accordance with EN 1994-1-1 [1]. For simplification, it is recommended that the short-term [1] 5.1.4
modular ratio should be taken as 7,0 irrespective of the grade of concrete.

2.3 Idealisation

In prEN 1993-1-8 [2], it is considered that the full non-linear M-φ curve consists of three
parts, as shown in Figure 23. Up to a level of 2/3 of the design moment resistance (M j,Rd), the
curve is assumed to be linear elastic. The corresponding stiffness is the so called initial
stiffness Sj,ini . Between 2/3 (Mj,Rd) and Mj,Rd the curve is non-linear. After the moment in the
joint reaches Mj,Rd, a yield plateau could appear. The end of this M-φ curve indicates the
rotation capacity (φCd) of the joint.

Figure 23 Non-linear M- φ curve

The characterisation adopted by Eurocode 3 assumes a fixed ratio between the initial stiffness
Sj,ini and the secant stiffness at the intersection between the non-linear part and the yield
plateau (Sj at the level Mj,Rd), see Figure 23. For composite joints with bolted end-plates, this
ratio is taken equal to 3,0. Contact plate joints have a less gradual decrease in stiffness and the
ratio is taken as 2,0.
The shape of the non-linear part for a bending moment M j,Sd between 2/3 (Mj,Rd ) and Mj,Rd can [1] 8.4.1(2)
be found with the following interpolation formula:

(0)

© SSEDTA 2001 Last modified jj/06/18 26

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

where ψ= 2,7 for joints with bolted flush end-plates and 1,7 for joints with contact plates. In
this interpolation formula, the value of Sj is therefore dependent on Mj,Sd.

Depending on the available software either the full non-linear shape of all joint-curves
discussed until now or multi -linear simplifications of them can be assigned to the respective
flexural springs. Figure 24 shows curve idealisations proposed in prEN 1993-1-8 [2]. It is
evident that the required input as well as the capability required from the software both
increase if high accuracy is intended.

Figure 24 Possibilities for curve idealisation

The actual moment-rotation response of joints usually is described by means of a non-linear

curve, see (Figure 24a). However, the use of such non–linear curves requires sophisticated
frame analysis programs. In order to enable a (more simple) linear calculation, e.g. an elastic
global frame analysis (this is still the current practice in most European countries), the non-
linear curve may be simplified by straight lines. As a conservative assumption each curve lying
below the non-linear curve in general may be used in the frame analysis. For example a tri-
linear curve is shown in Figure 24c. The most simple curve is a bi-linear curve. In order to use
a bi-linear curve which provides the most efficient solutions (i.e. the highest stiffness),
comparative studies were performed to calibrate an idealised joint stiffness S j*. The idealised
joint stiffness Sj* is constant for all values of applied moments smaller than the design moment
resistance of the joint. prEN 1993-1-8 [2] gives guidelines on how to derive such a simplified
bi-linear curve as shown in Figure 24b. The idealised joint stiffness S j* can easily be calculated
by dividing the initial joint stiffness Sj,ini with a stiffness modification factor η:

Sj*= Sj,ini /η (0)

The stiffness modification factor η depends on the type of connection (contact plate, bolted [1] 8.2.1.2 Table 8.1
end-plate) (See Table 3).

When elastic global analysis is used, the joints should be classified according to their stiffness.
In the case of a semi-rigid joint, its rotational stiffness S j for use in the global analysis should
generally be taken as equal to the value of Sj , corresponding to the bending moment Mj,Sd.

As a simplification, the procedure illustrated in Figure 22 may be adopted, as follows:

Provided that the moment Mj,Sd does not exceed 2/3 Mj,Rd the initial rotational stiffness of the
joint Sj,ini may be used in the global analysis.

Where the moment Mj,Sd exceeds 2/3 Mj,Rd the rotational stiffness should be taken as S j,ini/η,
where η is the stiffness modification coefficient from Table 8.1 of EN 1994-1-1 [1] or Table 3 [1] Table 8.1
below

Table 3 Stiffness modification coefficient η

As a further simplification, the rotational stiffness may be taken as S j,ini/η in the global analysis,
for all values of the moment Mj,Sd, as shown in Figure 25.

© SSEDTA 2001 Last modified jj/06/18 27

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

Figure 25 Rotational stiffness to be used in elastic global analysis

When rigid-plastic global analysis is used, joints should be classified according to their [1] 8.2.2.2
strength.

When elastic-plastic global analysis is used, the joints should be classified according to both
stiffness and strength.

2.4 Joint modelling of beam-to-column joints

As already mentioned there are three types of joint modelling which are closely linked to the
method of global analysis and to the classification of the joint:
A continuous joint ensures full rotation continuity, a semi continuous joint ensures only partial
rotational continuity and a simple joint prevents any rotational continuity between the
connected members.

2.4.1 General
To model the deformation behaviour of a beam–to-column joint, account should be taken of the
shear deformation of the web panel and the rotational deformation of the connection.
Beam-to-column joint configurations should be designed to resist the internal bending
moments Mb1,Sd and Mb2,Sd , normal forces Nb1,Sd and Nb2,Sd and shear forces Vb1,Sd and Vb2,Sd
applied to the connections by the connected beams (see Figure 26).

Figure 26 Forces and Moments acting on a joint

The resulting shear force Vwp,Sd in the web panel should be obtained using:

Vwp,Sd =(Mb1,Sd – Mb2,Sd) / z - (Vc1,Sd – Vc2,Sd) / 2 (0)

where:
z is the lever arm.
For global structural analysis, single-sided and double-sided beam to column joint
configurations may be modelled either as indicated in 2.4.2 or using the simplified method
given in 2.4.3.

2.4.2 Joint modelling reflecting the actual behaviour

To model a joint in a way that closely reproduces the expected behaviour, the web panel and
each of the connections should be modelled separately, taking account of the internal moments
and forces in the members, acting at the periphery of the web panel (See Figure 26).
This may be achieved as shown in Figure 27a), in which the shear deformation of the web
panel is reproduced by a separate translation spring and the rotational deformation of each
connection is reproduced by flexural springs.

© SSEDTA 2001 Last modified jj/06/18 28

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

Figure 27 Modelling the joint deformability

Within a joint two separate sources of deformability have to be distinguished – those due to the
connection (load-introduction into the column through the connecting elements) and those due
to the shear panel (column web panel in shear due to the moment imbalance between the left
and right hand sides). To reflect actual joint behaviour a cruciform joint configuration has to
be modelled by two separate M L-φL curves representing the left and the right connection and
one additional Ms-φs curve for the shear panel, see Figure 28. It has to be noted that the panel
is actually loaded in shear which results in a shear deformation, and therefore, the M s-φs curve
has to be derived from a V-γ relationship for the shear panel. For accuracy a single-sided joint
has to be modelled by two separate M-φ curves, one for the connection and one for the shear.
Figure 29 illustrates equally good possibilities of “joint modelling reflecting the actual
behaviour by so-called finite–joint models”

Figure 28 Joint modelling reflecting the actual behaviour

Figure 29 Application of joint modelling by finite joint models

If the software for frame analysis does not support flexural springs they may be substituted by
equivalent beam stubs with appropriate bending stiffness. The derivation of the equivalent stub
stiffness is shown in Figure 30 assuming a constant bending moment along the stub. Thus less
deviation to the actual behaviour is achieved if the stub is short.

Figure 30 Equivalent stub instead of rotational spring

For sake of simplicity a simplified joint modelling has been introduced in prEN 1993-1-8 [2]
for ever-day design, where the nodal zone is no longer taken into account with finite
dimension. The aim is to represent the overall joint behaviour in one single rotational spring
for each joint. Therefore simplifications are introduced as described below.

2.4.3 Simplified joint modelling

As a simplified alternative to the method of modelling, a single sided joint configuration may
be modelled as a single joint, and a double-sided joint configuration may be modelled as two
separate but inter-acting joints, one for each connection.
Each of these joints should be modelled as a separate rotational spring, as shown in Figure 31,
in which each spring has a moment-rotation characteristic that takes into account the
behaviour of the web panel as well as the influence of the relevant connection.
The finite size of the joint, neglected by this model, may be compensated by transforming the
moment-rotation characteristic.
When determining the moment resistance and rotational stiffness for each of the joints, the
possible influence of the web panel in shear should be taken into account by means of the
transformation parameters β1 and β2 where:
β1 is the value of the transformation parameter β for the right-hand side joint

© SSEDTA 2001 Last modified jj/06/18 29

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

β2 is the value of the transformation parameter β for the left-hand side joint
Conservative values for β1 and β2 based on the values of the beam moments M b1,Sd and Mb2,Sd at
the periphery of the web panel may be obtained from Table 4:
[2] Table 7.3

Table 4 Approximate values for the transformation parameter β

Alternatively more accurate values of β1 and β2 based on the values of the beam moments
Mj,b1,Sd and Mj,b2,Sd at the intersection of the member centrelines, may be determined from the
simplified model as follows:

β1=│1- (Mj,b2,Sd / Mj,b1,Sd)│ (0)

β2=│1- (Mj,b2,Sd / Mj,b2,Sd)│ (0)

where:
Mj,b1,Sd is the moment at the intersection from the right hand beam
Mj,b2,Sd is the moment at the intersection from the left hand beam

For sake of simplicity a simplified joint modelling has been introduced in prEN 1993-1-8 [2]
for “every – day design”, where the nodal zone is no longer taken into account with finite
dimensions. The aim is to represent the overall joint behaviour in one single rotational spring
for each joint. Therefore simplifications are introduced as described below

Figure 31 Modelling the web panel and the connections as

combined rotational springs
The load–deformation (V-γ curve of the column web panel in shear has to be converted into a
moment-rotation (Ms-φs) curve, using the lever arm z. However, the actual shear force V wp in
the column web panel is not equal to the “local” tension and compression forces within the
connection (Fc=Ft=M / z), see Figure 31, due to the global shear forces acting in the column.
For the simplified modelling according to prEN 1993-1-8 [2] this effect is neglected. In case of
a single-sided beam-to-column joint configuration, both curves representing the web panel and
the connection are added in series to one combined M-φ curve for the joint, see Figure 32.
In case of a double-sided joint configuration, two separate but interacting joints may be
modelled. Each of these joints takes into account the behaviour of the column web panel as
well as the behaviour of the relevant connection. Thus, the web deformation would be taken
into account twice but a so-called transformation factor β (Table 4)can be introduced to
consider the actual loading of the shear panel in relation to the loading of the individual joints.
The validity and accuracy of this procedure in demonstrated in extensive background studies.

Figure 32 Derivation of a combined rotational spring for the

simplified modelling

The second difference concerns the location of the rotational spring. In finite joint modes the
connection springs are arranged at the edges of the finite joint area, as shown in Figure 29.
In the simplified joint modelling the combined joint springs for shear and connections are
located at the beam-to-column axes intersection point, see Figure 31. It is obvious that the
acting consideration of the higher value for the applied moment therefore leads to a more
conservative design. On the other hand the weakness of the “extended” beam and column
overestimates the global frame deformation. This simplification is the same as that, made in
the conventional design where simple or continuous modelling is used. The influence of both

© SSEDTA 2001 Last modified jj/06/18 30

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

effects becomes more important with increasing joint dimensions in comparison to the beam
span and column height. These effects can be compensated by a so-called “spring
transformation”, which leads to additional refining terms within the formulae characterising
the joint properties.

Figure 33 Simplified joint modelling

© SSEDTA 2001 Last modified jj/06/18 31

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

3 Joint Representation of Composite Beam-to-Column

Joints, Design Provisions

3.1 Framing arrangements in composite construction

3.1.1 Introduction
In buildings, the structural frame provides the skeleton around which other elements are
constructed. These include the external cladding, internal finishes and the services.
Consequently, the frame should not be designed in isolation and the most efficient solution for
the structure may not be the most appropriate for the total building. It is not possible therefore
to prescribe standard framing arrangements for all buildings. However, guidance can be given
on those factors which should be considered in developing a suitable layout.

3.1.2 General factors influencing framing arrangements

In determining the most appropriate arrangement, the structural engineer will need to consider
the following factors.
 Site-dependent factors, including ground conditions and planning constraints
 Building use
 Design loading
 Building services and finishes
 External wall construction
 Lateral stiffness
 Structure costs, fabrication and construction.

Building use has a strong influence, which in recent years has led to wide clear spans. Studies
have shown that composite construction (steel beams acting with the slab above or Slimfloor
options) generally provides the most cost-effective solution for the floors. Greater stiffness and
resistance means that beams can be shallower for the same span, leading to lower storey
heights and reduced cladding costs. Alternatively, wider spans are possible within a given floor
depth. The speed of construction is an important influence on the popularity of these
structures.

3.1.3 Composite construction

Composite action within the slab itself is achieved by the well-known use of profiled steel
sheeting. Although deep profiles are available, most floors use relatively shallow decking
(typically in the range of 45-70 mm) spanning about 3m between beams. For these spans the
decking does not normally require propping during concreting.

With slabs spanning this distance, secondary beams are needed. Alternative layouts for wide
spans are shown in Figure 34. The need to consider together the steel framing and the
particular services can be seen by considering these alternatives.

As the heavily-loaded primary beams in Figure 34a) have the shorter span, both these and the
secondary beams will be of similar depth. The governing depth of floor construction is
therefore minimised. Space to accommodate services exists between the beams. However if
uninterrupted service space is required then that can only be provided by a void between the
underside of the beams and the ceiling, or by making use of cellular beams. The overall depth
for structure and services then increases.

© SSEDTA 2001 Last modified jj/06/18 32

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

Figure 34 Alternative framing systems for floors

In Figure 34b) the secondary beams will be of less depth, but at the expanse of deep primary
members. However, substantial zones now exist for services beneath the secondary members
but within the overall floor depth governed by the size of the primary beams.

3.1.4 Sizing of composite beams

A simply supported composite beam is generally governed by one of the following conditions:
 Resistance to the ultimate design loads;
 Total deflection under serviceability loading;
 Vibration.

For a beam with composite end connections, the load capacity depends on the sagging moment
resistance of the beam section in the mid-span region and also on the resistance of the
connections to hogging moment. For the same load capacity, this leads to a smaller steel
section than in simple construction.

It must also be remembered though that with unpropped beams the construction condition
could govern. As the steelwork part of a composite joint resembles a nominally-pinned steel
connection, the beam is simply-supported during construction. There is more likelihood
therefore that this will govern. This does not mean that there is no benefit from the composite
joints – by making the composite stage less critical a smaller steel section is more likely.
Furthermore, over-design at the composite stage can be minimised by use of partial shear
connection for sagging bending. For simply-supported composite beams, typically around 50%
of the total deflection is due to sagging of the steel section at the construction stage. With
composite connections, this proportion increases, because smaller loading may also affect the
design. Detailed analysis of the dynamic response is beyond the scope of this document and
reference should be made to specialised information.

When sections are governed by deflection, pre-cambering can be used to reduce the final
sagging displacement. It should be noted though that the deflection at the end of construction
may not be important. If deflection does not impair appearance, the reference level for
displacement should be taken as the upper side of the composite beam. The deflection at the
end of construction can therefore be neglected provided that:
 A false ceiling is provided;
 The top surface of the floor slab is level at the end of construction (as would normally be
the case because of the flow of wet concrete);
 Consideration is given to the loading effect of ponding (increased depth of concrete due to
deflection).

At the composite stage, composite joints are clearly beneficial because of the substantial
increase in stiffness. It has been shown that even connections of modest stiffness cause
significant reduction in beam deflection. A substantial increase in natural frequency is also
obtained. In addition, composite joints permit cracking to be controlled, where it is required
for appearance and/or durability.

3.1.5 Framing arrangements and composite joints

Considering the factors given above, the designer should determine:
 The layout of the beams and columns;

© SSEDTA 2001 Last modified jj/06/18 33

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

 The types of beams, for example rolled steel section acting compositely with the slab
above or a Slimfloor option; non-composite steel beams may be included;
 The types of column, for example rolled H-sections or hollow sections; composite sections
may be included;
 The orientation of the column sections (if appropriate);
 Types of joint, for example composite joints, moment-resisting steel joints, nominally-
pinned joints;
 Types of connection, for example composite joints with contact plate, composite joints
with full-depth steel end-plates, “simple” steel fin plates.

The choices given as examples show the freedom available to the designer to meet the
particular requirements of the structure. This is characteristic of semi-continuous construction
and is illustrated by the example shown in Figure 35. In the discussion which follows, it is
assumed that the designer has sought to achieve beams of similar depth. It is assumed that for
the building concerned composite connections are not required purely for crack control. The
numbers refer to the joint types in the lower part of the figure and are also used to identify
particular beams.

All internal beams are to be designed as composite members, but as the edge beams 1-2 and 3-
3 carry less load, they may be designed as non-composite. In that case it follows that
connections 1, 2 and 3 are also non-composite.

Connections to perimeter columns can be composite, provided that a region of slab exists
beyond the column in which to anchor the tension reinforcement or reinforcement is looped
around the column. Here it is assumed that connections to the perimeter columns will be bare
steel, to avoid any difficulty in achieving this.

Each of connection types 1 and 2 may be either nominally-pinned or moment-resisting, as

required to achieve the aim of beams of similar depth. In the interests of overall economy
though, a stiffened connection should be avoided. For ease of fabrication and erection,
connection 3 should be nominally-pinned. In any case, a column web has only very limited
resistance under unbalanced moment. Joint 4 is also nominally pinned, because its one-sided
nature.

The remaining connections, 5-7, may be designed compositely. By adjusting the reinforcement,
the designer is able to vary at will the structural properties of the joint. The moment resistance
and stiffness of the steel connection can also be varied by changing its details. Indeed, joint
types 5-7 can be replaced by contact-plate joints, as shown in Figure 35. This freedom helps
the designer to achieve economically a uniform large floor grid, the beams and their
connections would be repeated many times. The resulting reductions in beam sections would
therefore be widespread and there would be substantial repetition in fabrication and erection.
Finally, it will be noted that composite connections are used about both axes of the internal
column. With a decking say 70 mm deep and a thin overall slab depth (say 120 - 130mm),
problems could arise in accommodating the two layers of reinforcement in a limited thickness
of slab. In such circumstances one connection could be bare steel. In the example (Figure 35)
this should be connection 6; as it connects to the column flange, it could still be designed as
moment-resistant. Alternatively, a greater thickness of concrete could be provided by using
shallower decking.

© SSEDTA 2001 Last modified jj/06/18 34

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

Figure 35 Choice of joint type

3.2 Design provisions

3.2.1 Basis of design

Design provisions in this lecture are based on the component method for steel joints described
in prEN 1993-1-8 [2]. The simplified model for a composite joint is shown in Figure 36.

Figure 36 Simplified model according to EN 1994-1-1 [1]

The prEN 1993-1-8 [2] already provides expressions for the design resistance and initial
stiffness of the following components:

Compression region:
 Column web in compression
 Beam flange and web in compression

Tension region:
 Column flange in bending
 Column web in tension
 End-plate in bending
 Beam web in tension
 Bolts in tension

Shear region:
 Column web panel in shear

For composite joints, the following additional basic components are relevant:
 Longitudinal slab reinforcement in tension;
 Contact plate in compression.

Although not regarded as a separate basic component, account may also need to be taken of
concrete encasement to the column. This is treated as a form of stiffening.
Expressions for design resistance and initial stiffness are given in 3.3. The stiffness coefficients
ki for components affected by concrete encasement are transformed into equivalent all-steel
values, using a modular ratio. This enables a single value for modulus of elasticity E to be used
to determine the rotational stiffness of the joint, in the same way as for steel joints in prEN
1993-1-8 [2]. A similar transformation enables the modulus of elasticity of reinforcement to
have a different value to that for structural steel.

Unlike the sophisticated model, the simplified Eurocode model shown in Figure 36 does not
make explicit allowance for the following:

 Slab concrete in bearing against the column;

 Transverse slab reinforcement;
 Slip of the beam´s shear connection.

Account is taken of these actions either through detailing rules to exclude their influence or

© SSEDTA 2001 Last modified jj/06/18 35

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

(for slip) by reduction factor on the stiffness.

3.2.2 Design moment resistance

To simplify calculation, plastic theory is used to determine the design moment resistance. This
moment is therefore taken as the maximum evaluated on the basis of the following criteria:
 The internal forces are in equilibrium with the forces applied to the joint
 The design resistance of each component is not exceeded
 The deformation capacity of each component is not exceeded
 Compatibility is neglected

Contact plate joint:

In such joints, the steelwork connection provides no resistance to tension arising from bending.
The distribution of internal forces is therefore easy to obtain. As can be seen from Figure 37,
the compression force is assumed to be transferred at the centroid of the lower beam flange
and the tension force at the centroid of the reinforcement.

Figure 37 Contact plate joint with one row of reinforcement

The design moment resistance of the joint M j,Rd is dependent on the design resistance F Rd of the
weakest joint component; for this joint, the relevant components are assumed to be the
reinforcement in tension, the column web in compression, the beam flange and web in
compression or the column web panel in shear. So:

Mj,Rd= FRd .z (0)

where z is the lever arm of the internal forces. [1] 8.2.4.1

For connections with contact plates, the centre of compression should be assumed to be in line
with the mid-thickness of compression flange.

For connections with contact plates and only one row of reinforcement active in tension, the
lever arm z should be taken as the distance from the centre of compression to the row of
reinforcement in tension.

For connections with contact plates and two rows of reinforcement active in tension the lever
arm z should be taken as the distance from the centre of compression to a point midway
between these two rows, provided that the two rows have the same cross-sectional area.(See
Figure 38)

Figure 38 Determination of the lever arm z

For connections with other types of steelwork connections the lever arm z should be taken as
equal to zeq obtained using the method given in 5.3.3.1 of prEN 1993-1-8 [2]. [2]5.3.3.1

For the assumption to be correct, detailing rules need to ensure that, under unbalanced loading,
failure does not occur by crushing of concrete against the column section. These are given in
3.7.1.

© SSEDTA 2001 Last modified jj/06/18 36

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

Joints with steelwork connection effective in tension:

A composite flush end-plate connection, such as that shown in Figure 39, is an example of
such a joint. Tension arising from bending is resisted by the combined action of the
reinforcement and the upper part of the steelwork connection. As there is more than one row of
components in the tension region, the distribution is now more complex. It is assumed that as
the moment increases the reinforcement bars reach their design resistance before the top row
of bolts. High-ductility reinforcement, as defined in EN 1992-1 [10], is to be used and [10]
therefore redistribution of internal forces can take place. Thus each bolt-row in turn may reach
its resistance, commencing with the top row.

Figure 39 Composite flush end-plate joint

Within the bolt-rows, three types of internal force distribution may be identified: [3] Module 5: Lecture
 An elastic distribution; 15
 A plastic distribution;
 An elasto-plastic distribution.

The appropriate one will depend on the deformation capacity of the joint components. The
distributions are illustrated in Figure 40.

Figure 40 Internal force distribution of reinforcement and bolt rows

In Figure 40, a plastic redistribution of the internal forces takes place progressively from the
upper bolt-row towards the lower ones because of their sufficient deformation capacity. prEN
1993-1-8 [2] considers that a bolt-row possesses sufficient capacity to allow this when:
 FRd,i is associated to the failure of the beam web in tension or
 FRd,i is associated to the failure of the bolt-plate assembly and
 FRd,i ≤ 1,9 Bt,Rd where Bt,Rd is the tension resistance of the bolt-plate assembly.

The design moment resistance is given by:

(0)

where the summation includes the reinforcement in tension.

In some cases (Figure 40b) , the plastic redistribution of forces is interrupted, because of the
lack of deformation capacity in the last bolt-row which has reached its design resistance. (F Rd,i
> 1,9 Bt,Rd and linked to the failure of the bolts or of the bolt-plate assembly). In the bolt-rows
located lower than bolt-row k, the forces are then linearly distributed according to their
distance to the centre of compression i.e. the centre of the lower beam flange. The design
moment resistance then equals:

(0)

where the first summation includes the reinforcement and all bolt-rows down to row k, and the
second summation is over the bolt-rows below row k.
In the elastic distribution (Figure 40c), the forces in the bolt-rows are proportional to the

© SSEDTA 2001 Last modified jj/06/18 37

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

distance to the centre of compression. This distribution applies to joint configurations with
rather stiff end-plates and column flanges. The above formula applies with k corresponding to
the top bolt-row.

The three distributions may be interrupted because the compression force F c (Figure 39) is
limited by the design resistance of the beam flange and web in compression, the column web in
compression or the column web panel in shear. The moment resistance M Rd is then evaluated
with similar formulae to those given above but now only a limited number of bolt-rows or
amount of reinforcement is taken into consideration.

These are determined such that:

(0)
where:
n is the number of the last row permitted to transfer a tensile force;
Fℓ is the tensile force in row number l;
Fc,Rd is the least design resistance of the beam flange and web in compression, the
column web in compression and (if appropriate) the column web panel in shear.

[1] 8.3.3
3.3 Resistance of basic components

3.3.1 General
In Table 5 a list of components covered by prEN 1993-1-8 [2] is shown.

For composite joints, the following additional basic components are relevant:
 Longitudinal slab reinforcement in tension
 Contact plate in compression
[3] Module 15

[2] 5.2.6

Table 5 List of components covered by prEN 1993-1-8 [2]

[1]8.3.3.1
3.3.2 Column web panel in shear
For a single sided joint, or for a double sided joint in which the beam depths are similar, the
shear resistance Vwp,Rd of the steel column web panel, subject to a shear force V wp,Sd should be
obtained using:
(0)
where:
Vwp,a,Rd is the design shear resistance of the steel column web panel, see prEN 1993-1-8 [2]
5.2.6.1;
Avc is the shear area of the column, see EN 1993-1 [11] 5.4.6. [11] 5.4.6

The resulting shear force Vwp,Sd in the web panel should be obtained using:

Vwp,Sd =(Mb1,Sd – Mb2,Sd) / z - (Vc1,Sd – Vc2,Sd) / 2 (0)

see Figure 41

© SSEDTA 2001 Last modified jj/06/18 38

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

Figure 41 Forces and moments acting on a joint

Where the steel column web is encased in concrete the design shear resistance of the panel,
Vwp,Rd , may be increased to:
(0)

For a single-sided joint, or a double-sided joint in which the beam depths are similar, the
design shear resistance of concrete encasement to the column web panel V wp,c,Rd is given by:

Vwp,c,Rd=  (0,85 fck / γc) Ac sin θ (0)

where:
Ac=[0,8 ( hc-2tfc)cos θ] [bc-twc] (0)

θ= arctan [(h c-2 tfc)/z]

bc is the width of the flange of the column section;

hc is the depth of the column section;
tfc and twc are defined in Figure 42;
z is the lever arm.

The reduction factor  to allow for the effect of longitudinal compression in the column on the
design resistance of the column web panel in shear is given by:

=0,55[1+2(NSd / Npl,Rd))≤ 1,1 (0)

where: [1] 6.8.3.2

NSd is the design axial compressive force in the column;
Npl,Rd is the design plastic resistance of the column´s cross-section including the
encasement, see also EN 1994-1-1 [1] 6.8.3.2.
[1] Figure 8.2

Figure 42 Transverse compression on a column

[1] 8.3.3.2
3.3.3 Column web in compression
The resistance of an unstiffened steel column web subject to transverse compression F c,wc,a,Rd [2] 5.2.6.2
should be determined in accordance with prEN 1993-1-8 [2] 5.2.6.2, using an effective width
beff,c,wc.

For a contact plate connection the effective width beff,c,wc of column web in compression, see
Figure 42, should be obtained using:

beff,c,wc =tfb + 5 (tfc + s) + sp (0)

in which:
- for a rolled I or H section: s = r c
- for a welded I or H section: (0)

sp is the length obtained by dispersion at 45˚ through the contact plate (at least t cp and,
provided that the depth of contact plate below the flange is sufficient, up to 2 t cp).

© SSEDTA 2001 Last modified jj/06/18 39

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

For a bolted end-plate connection the effective width b eff,c,wc of column web in compression
should be obtained using:

beff,c,wc= tfb +2 ap + 5(tfc + s) + sp (0)

in which:
sp is the length obtained by dispersion at 45˚ through the end-plate (at least t p and, provided
that the length of end-plate below the flange is sufficient, up to 2 t p).

Where an unstiffened column web is encased in concrete, the design resistance in transverse
compression, Fc,wc,Rd may be increased to:

Fc,wc,Rd= Fc,wc,a,Rd + Fc,wc,c,Rd (0)

where:
Fc,wc,a,Rd is the design resistance of the unstiffened steel column web subject to transverse
compression;
Fc,wc,c,Rd is the design resistance to transverse compression of the concrete encasement to the
column web.
The design resistance to transverse compression of concrete encasement to a column web
Fc,wc,c,Rd is given by:

Fc,wc,c,Rd = kwc,c teff,c (bc - twc) (0,85 fck / γc) (0)

The effective thickness of concrete t eff,c is given by:

teff,c = ℓo + 5tfc (0)

For a contact plate connection, the load introduction length ℓ o (see Figure 42) is given by:

ℓo = tfb + sp (0)

where sp is defined above.

For a bolted end-plate connection, the load introduction length ℓ o is given by:

ℓo = tfb + 2 ap + sp (0)

where sp is defined above.

Where in a steel web the maximum longitudinal compressive stress σ com,Ed due to axial force [2] 5.2.6.2(8)
and bending moment in the columns exceeds 0,5 fy,wc in the web (adjacent to the root radius for
a rolled section or the toe of the weld for a welded section), its effect on the resistance of the
column web in compression should be allowed for by multiplying the value of F c,wc,Rd by a
reduction factor kwc,c given in prEN 1993-1-8 [2] 5.2.6.2(8). The factor k wc,c to allow for the
effect of longitudinal compression in the column on the design resistance of the concrete
encasement to transverse compression Fc,wc,c,Rd is given by:

kwc,c = 1,3+3,3 σcom,c,Ed /(fck / γc) 2,0 (0)

where σcom,c,Ed is the longitudinal compressive stress in the encasement due to the design axial
force NSd.

3.3.4 Column web in tension [2] 5.2.6.3

The resistance of an unstiffened column web subject to transverse tension should be

© SSEDTA 2001 Last modified jj/06/18 40

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

determined from:

(0)

where ω is a reduction factor to allow for the possible effects of shear in the column web panel.

For a bolted connection, the effective width beff,t,wc of column web in tension should be taken as [2] 5.2.2
equal to the effective length of equivalent T-stub representing the column flange. (See prEN [3] Module 5
1993-1-8 [2] 5.2.2 as well as SSEDTA 1 [3] Module 5.
The reduction factor ω allow for the possible effects of shear in the column web panel should
be determined from prEN 1993-1-8 [2] Table 5.2.

3.3.5 Column flange in bending

See prEN 1993-1-8 [2] 5.2.6.4 [2] 5.2.6.4

3.3.6 End-plate in bending

See prEN 1993-1-8 [2] 5.2.6.5 [2] 5.2.6.5

3.3.7 Beam flange and web in compression [1] 8.3.3.3

For a composite joint, the compression resistance of a beam´s steel flange and the adjacent
compression zone of the beam´s steel web, may be assumed to act at the level of the centre of
compression. It may be assumed to be given with sufficient accuracy by:

Fc,fb,Rd = Mc,Rd / (h - tfb) (0)

where:
h is the depth of the connected beam´s steel section;
Mc,Rd is the moment resistance of the steel section, reduced if necessary to allow for shear, [11] 5.4.5
see also EN 1993-1 [11] 5.4.5 and 5.4.7; [11] 5.4.7
tfb is the thickness of the connected beam´s steel flange (See Figure 42).

3.3.8 Beam web in tension

In a bolted end-plate connection, the tension resistance of the beam web should be obtained
using:

Ft,wb,Rd =beff,t,wb twb fy,wb / γM0 (0)

The effective width beff,t,wb of the beam web in tension should be taken as equal to the effective [2] 5.2.6.5
length of the equivalent T-stub representing the end-plate in bending, obtained from prEN
1993-1-8 [2] 5.2.6.5 for an individual bolt-row or a bolt-group.

[1] 8.3.3.4
3.3.9 Longitudinal slab reinforcement in tension
The design tension resistance of a row r of reinforcing bars is given by:

Ftr,s,Rd = Ar,s fsk /γs (0)

where:
Ar,s is the cross-sectional area of the longitudinal reinforcement in row r within the total

© SSEDTA 2001 Last modified jj/06/18 41

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

effective width of concrete flange beff, as given by EN 1994-1-1 [1]5.2.2; [1] 5.2.2
fsk is the characteristic strength of the reinforcement;
γs is the partial safety factor for reinforcement.
EN 1994-1-1 [1] of 5.2.1(4) is applicable. [1] 5.2.1(4)
For ductile failure the total cross-sectional area of longitudinal reinforcement should not
exceed the limit given in EN 1994-1-1 [1] 8.6.1
[1] 8.6.1

3.3.10 Contact plate in compression

The design compression resistance of a contact plate is given by the lesser of:

Fc,cp,Rd =bcp h cp fy / γM0 (0)

where:
bcp is the breadth of the contact plate;
hcp is the height of the contact plate;
γM0 is the partial safety factor for structural steel;

and:
Fc,cp,Rd =bcp (tfb + sp) fy / γM0 (0)
where:
sp is the length obtained by dispersion at 45˚ through the contact plate.

[1] 8.3.4
3.4 Design moment resistance of joints with full shear connection

3.4.1 General [2] 5.2.7

The methods for determining the design moment resistance of a joint M j,Rd do not take account
of any co-existing axial force NSd in the connected member. They should not be used if the
axial force in the connected member exceeds 10% of the plastic resistance N pl,Rd of its cross-
section.
The shear connection should be designed in accordance with EN 1994-1-1 [1] 6.7.2.

3.4.2 Beam-to-column joints with contact plate

The design moment resistance Mj,Rd of a beam-to-column joint may be determined from:

(0)
where:
Ftr,Rd is the effective design tension resistance of row r of reinforcing bars;
hr is the distance from row r to the centre of compression;
r is the number of a particular row.

The centre of compression should be assumed to be in line with the mid-thickness of the
compression flange of the connected member.

The effective design tension resistance F tr,Rd for each row of reinforcing bars should be
determined in sequence, starting with row 1, the row furthest from the centre of compression.

When determining the value of Ftr,Rd of row r of reinforcing bars, all other rows of
reinforcement closer to the centre of compression should be omitted.

The effective design tension resistance Ftr,Rd of row r of reinforcing bars should be taken as its
design tension resistance Ftr,s,Rd as an individual row, (See also 3.3.9), reduced if necessary to

© SSEDTA 2001 Last modified jj/06/18 42

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

satisfy the conditions specified below.

The effective tension resistance Fr,Rd of row r of reinforcing bars should, if necessary, be
reduced below the given value in order to ensure that, when account is taken of all rows up to [2] 5.2.7.2(7)
and including r:
The total resistance / β (β see Table 4) (0)
The total resistance does not exceed the smallest of:
The resistance of the column web in compression F c,wc,Rd (see 3.3.3)
The resistance of the beam flange and web in compression F c,fb,Rd (see 3.3.7)
The resistance of the contact plate in compression F c,cp,Rd (see 3.3.10)

3.4.3 Beam-to-column joints with partial-depth end-plate

The design moment resistance may be determined in accordance with 3.4.2 therefore it is
necessary that the contribution of the bolt-rows to the design moment resistance may be
neglected.

3.4.4 Beam-to-column joints with bolted end-plate steelwork

The design moment resistance Mj,Rd of a beam-to-column joint may be determined from:

(0)

where:
Ftr,Rd is the effective design tension resistance of row r of reinforcing bars or bolts;
hr is the distance from row r of reinforcing bars or bolts to the centre of compression;
r is the number of a particular row.

For bolted end-plate connections, the centre of compression should be assumed to be in line
with the mid-thickness of the compression flange of the connected member.

The effective design tension resistance Ftr,Rd for each row of reinforcing bars or bolts should be
determined in sequence, starting with row 1, the row furthest from the centre of compression.
When determining the value of Ftr,Rd for row r of reinforcing bars, all other rows of
reinforcement closer to the centre of compression and all bolt-rows should be omitted.

The effective design tension resistance Ftr,Rd of row r of reinforcing bars should be taken as its
design tension resistance Ftr,s,Rd as an individual row, (see 3.3.9) reduced if necessary to satisfy
the conditions specified below.
When determining the value of F tr,Rd for bolt-row r all other bolt-rows closer to the centre of
compression should be omitted.

The tension resistance Ftr,Rd of bolt-row r as an individual bolt-row should be taken as the
smallest value of the tension resistance for an individual bolt-row of the following basic
components:
 The column web in tension Ft,wc,Rd (see prEN 1993-1-8 [2] 5.2.6.3) [2] 5.2.6.3
 The column flange in bending Ft,fc,Rd (see prEN 1993-1-8 [2] 5.2.6.4) [2] 5.2.6.4
 The end-plate in bending Ft,ep,Rd (see prEN 1993-1-8 [2] 5.2.6.5) [2] 5.2.6.5
 The beam web in tension Ft,wb,Rd (see prEN 1993-1-8 [2] 5.2.6.8) [2] 5.2.6.8

The effective tension resistance Ftr,Rd of row r of reinforcing bars or bolts should, if necessary,
be reduced below the given value Ft,Rd in order to ensure that, when account is taken of all rows
up to and including r:
the total resistance [2] 7.3.3
(0)
with β from Table 4;
the total resistance does not exceed the smaller of:
the resistance of the column web in compression Fc,wc,Rd

© SSEDTA 2001 Last modified jj/06/18 43

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

(see prEN 1993-1-8 [2] 5.2.6.2)

the resistance of the beam flange and web in compression F c,fb,Rd [2] 5.2.6.2
(see prEN 1993-1-8 [2] 5.2.6.7)

[2] 5.2.6.7
The effective tension resistance Ftr,Rd of bolt-row r should, if necessary, be reduced below the
value of Ft,Rd in order to ensure that the sum of the resistances taken for the bolt-rows up to and
including bolt-row r that form part of the same group of bolt-rows, does not exceed the
resistance of that group as a whole. This should be checked for the following basic
components.

 The column web in tension Ft,wc,Rd (see prEN 1993-1-8 [2] 5.2.6.3) [2] 5.2.6.3
 The column flange in bending Ft,fc,Rd (see prEN 1993-1-8 [2] 5.2.6.4) [2] 5.2.6.4
[2] 5.2.6.5
 The end-plate in bending Ft,ep,Rd (see prEN 1993-1-8 [2] 5.2.6.5) [2] 5.2.6.8
 The beam web in tension Ft,wb,Rd (see prEN 1993-1-8 [2] 5.2.6.8)

Where the effective tension resistance F tx,Rd of one of the previous bolt-rows x is greater than
1,9Bt,Rd then the effective tension resistance F tr,Rd for bolt-row r should be reduced, if necessary,
in order to ensure that:

Ftr,Rd ≤ Ftx,Rd h r / h x (0)

where:
hx is the distance from bolt-row x to the centre of compression;
x is the bolt-row farthest from the centre of compression that has a tension resistance
greater than 1,9 Bt,Rd.

The same method may be applied to bolted beam-to-beam joints with welded end-plates by
omitting the items relating to the column.
The method given in prEN 1993-1-8 [2] 5.2.7.2 for a single-sided beam-to-column joint is [2] 5.2.7.2
described in the flow chart below (Figure 43).
[2] Figure 5.30

Figure 43 Unstiffened end-plate joint with two bolt-rows in tension

3.5 Rotational stiffness

3.5.1 Basic model [2] 5.3

The rotational stiffness of a joint should be determined from the flexibilities of its basic
components, each represented by its elastic stiffness coefficient k i . These elastic stiffness
coefficients are of general application. The numbering of stiffness coefficients is consistent
with that in prEN 1993-1-8 [2] . The elastic translational stiffness of a component i is obtained
by multiplying ki with Ea.
For connections with more than one layer of components in tension, the stiffness coefficients k i
for the related basic components should be combined (see 3.5.2.2).
In a bolted connection with more than one bolt-row in tension, as a simplification, the
contribution of any bolt-row may be neglected, provided that the contributions of all other bolt-
rows closer to the centre of compression are also neglected. The number of bolt–rows retained
need not necessarily be the same for the determination of the moment resistance.

© SSEDTA 2001 Last modified jj/06/18 44

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

Provided that the axial force N Sd in the connected member does not exceed 10% of the
resistance Npl,Rd of its cross-section, the rotational stiffness S j of a joint, for a moment Mj,Sd less
than the moment resistance Mj,Rd of the joint, may be obtained with sufficient accuracy from:

(0)

where:
Ea modulus of elasticity of steel;
ki stiffness coefficient for basic joint component i;
z lever arm, see Figure 38;
μ stiffness ratio Sj,ini / S j , see below;
Sj,ini initial rotational stiffness of the joint, given by the expression above with
μ=1,0.

The stiffness ratio  should be determined from the following:

- if Mj,Sd ≤ 2/3 Mj,Rd μ=1 (0)
- if 2/3 Mj,Rd < Mj,Sd ≤ Mj,Rd (0)
in which the coefficient ψ is obtained from Table 8.2 of EN 1994-1-1 [1] or Table 6 below.

Table 6 Value of the coefficient ψ

3.5.2 Initial stiffness Sj,ini

The initial rotational stiffness Sj,ini is derived from the elastic translational stiffness of the joint
components. The elastic behaviour of each component is represented by a spring. The force-
deformation relationship of this spring is given by:

Fi = E ki wi (0)

Fi the force in the spring i;

E the modulus of elasticity of structural steel;
ki the translational stiffness coefficient of the spring i;
wi the deformation of spring i.
Separate rotational stiffness can be derived for the connection and the web panel in shear. In
simplified joint modelling a value for the overall joints is all that is required. The derivation of
this is now explained.

3.5.2.1 Connections with one layer of components in tension

Contact plate joint:
Figure 44 shows the spring model for a contact plate joint in which tensile forces arising from
bending are carried only by one layer of reinforcement.

Figure 44 Spring model for a contact plate composite joint

Here, k1 represents the column web panel in shear, k 2 the unstiffened column web subject to
compression from the contact plate and k 13 the longitudinal reinforcement bars in tension. The

© SSEDTA 2001 Last modified jj/06/18 45

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

contact plate itself is assumed to have infinite stiffness.

The force in each spring is equal to F. The moment M j acting in the spring model is equal to
F.z, where z is the distance between the centroid of the reinforcement in tension and the centre
of compression (assumed located in the centre of the lower beam flange). The rotation φ j in the
joint is equal to (w1+w2+w13)/z. In other words:

(0)

3.5.2.2 Connections with more than one layer of components in

tension
Joints with steelwork connection effective in tension:
Figure 45a) shows the spring model adapted for a more complicated case where tensile forces
arising from bending are carried not only by a layer of reinforcement but also by a bolt-row in
tension within the steelwork connection. The reinforcement is assumed to behave like a bolt-
row in tension, but with different deformation characteristics. It is assumed that the
deformations are proportional to the distance to the point of compression, but that the elastic
forces in each row are dependent on the stiffness of the components.

Figure 45 Spring model for a composite flush end-plate connection

Figure 45b) shows how the deformations per bolt-row of the end-plate in bending, the bolts in
tension, the column flange in bending and the column web in tension are added to form an
effective spring per bolt-row, with an effective stiffness coefficient k eff,r (r is the index of the
row number). Figure 45c) indicates how these effective springs at each level are replaced by an
equivalent spring acting at a corresponding lever arm z. The stiffness coefficient of this
equivalent spring is keq and this can be directly applied in the preceding equation for the
stiffness Sj,ini .

The formulae given below to determine k eff,r , z and keq can be derived from the sketches of
Figure 45. The basis for these formulae is that the moment-rotation behaviour of each of the
systems in Figure 45 is equal. An additional condition is that the compressive force in the
lower rigid bar is equal in each of these systems.

For composite connections with more than one layer of components in tension, the basic
components related to all of these layers should be represented by a single equivalent stiffness
coefficient keq determined from:

(0)

where:
hr is the distance between layer r and the centre of compression;
keff,r is the effective stiffness coefficient for layer r taking into account the stiffness k i for
the basic components listed below as appropriate;
zeq is the equivalent lever arm.

For a composite joint with more than a single layer of reinforcement considered effective in
tension, the above provisions are applicable provided that the layers are represented by a
single layer of equivalent cross-sectional area and equivalent distance to the centre of
compression.

© SSEDTA 2001 Last modified jj/06/18 46

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

The effective stiffness coefficient keff,r for bolt-row r should be determined from:
(0)
where:
ki,r is the stiffness coefficient representing component i relative to bolt-row r.

The equivalent lever arm z eq should be determined from:

(0)
In the case of a bolt-row r of a beam-to-column joint with an end-plate connection, k eff,r should
be based upon (and replace) the stiffness coefficients k i for:

 The column web in tension (k3)

 The column flange in bending (k4)
 The end-plate in bending (k5)
 The bolts in tension (k10).

In the case of a bolt-row r of a beam-to-beam joint with bolted end-plates, k eff,r should be based
upon (and replace) the stiffness coefficients ki for:
 The end-plates in bending (k5)
 The bolts in tension (k10).

In calculating the stiffness of joint configurations with bolted end-plate connections, the
stiffness coefficients ki for the basic components listed in Table 8 should be taken into account.

3.5.3 Stiffness coefficients for basic joint components

In stiffness calculations, the stiffness coefficients k i for basic components taken into account
should be as listed in Table 7 for joint configurations with contact plate connections, and in
Table 8 for bolted end-plate connections.

Table 7 Stiffness coefficients for joints with contact plate

connections

Table 8 Stiffness coefficients for joints with bolted end-plate

connections
The basic joint components are identified and listed in this lecture. Other basic components are [2] 5.3.2
identified in prEN 1993-1-8 [2] 5.3.2. [1] 8.4.2
For an unstiffened column web panel in shear, the stiffness coefficient k 1 for a single-sided
joint, or a double-sided joint in which the beam depths are similar, should be obtained from the
following formulae.

For a joint with a steelwork contact plate or a partial depth end-plate:

[1]8.4.2 (2)
(0)

where:
Avc is the shear area of the column, see EN 1993-1 [11] 5.4.6; [11] 5.4.6
z is the lever arm from EN 1994-1-1 [1] 8.2.4; [1] 8.2.4
β is the transformation parameter from Table 4 and prEN 1993-1-8 [2] 7.3.3. [2] 7.3.3

© SSEDTA 2001 Last modified jj/06/18 47

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

For a joint with a flush end-plate:

[1] 8.4.2 (3)
(0)

For an unstiffened column web in compression, in a contact plate connection, the stiffness
coefficient k2 should be obtained from:
[1] 8.4.2 (4)
(0)

where:
beff,c,wc is the effective width of the column web in compression from prEN 1993-1-8 [2] [2] 5.2.6.2
5.2.6.2;
dc is the clear depth of the column web;
twc is the thickness of the column web.

For an unstiffened column web in compression, in a bolted end-plate connection, the stiffness
coefficient k2 should be obtained from:
[1] 8.4.2 (5)
(0)

Where concrete web encasement is used, the stiffness coefficients for the relevant basic joint [1] 8.4.2 (6)
components k1 and k2 should be increased as described below.

For a column web in shear in a joint with a steelwork contact plate connection, the stiffness [1] 8.4.2 (7)
coefficient k1 should be obtained from:

(0)

where:
Ea is the modulus of elasticity for structural steel;
Ecm is the modulus of elasticity for concrete;
bc is the breadth of the column´s steel section;
hc is the height of the column´s steel section.
For a column web in shear in a joint with a flush end-plate connection, the stiffness coefficient
k1 should be obtained from:
[1] 8.4.2 (8)
(0)

For a column web in compression in a contact plate connection, the stiffness coefficient k 2
should be obtained from:
[1] 8.4.2 (9)
(0)

where:
bel = tfb + 2(tfc + s) + sp (0)
[2] 5.2.6.2
and s and sp are defined in prEN 1993-1-8 [2] 5.2.6.2.

For a column web in compression in a bolted end-plate connection, the stiffness coefficient k 2
should be obtained from:
[1] 8.4.2 (10)
(0)

For a column web in tension, in a stiffened or unstiffened bolted connection with a single bolt-
row in tension, the stiffness k 3 should be obtained from:

© SSEDTA 2001 Last modified jj/06/18 48

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

(0)

where:
beff,t,wc is the effective width of the column web in tension (See 3.3.4). For a joint with a single
bolt-row in tension, beff,t,wc should be taken as equal to the smallest of the effective lengths l eff
(individually or as a part of a group of bolt-rows).

For a column flange in bending, for a single bolt-row in tension the stiffness coefficient k 4
should be obtained from:

(0)

where:
leff is the smallest of the effective lengths (individually or as part of a bolt group) for this
bolt-row for an unstiffened column flange or a stiffened column flange;
m is defined in Figure 46;
tfc is the thickness of the column flange.

For an end-plate in bending, for a single bolt-row in tension the stiffness coefficient k 5 should
be obtained from:

(0)

where:
leff is the smallest of the effective lengths (individually or as part of bolt-rows);
m is defined in Figure 46.
For a beam flange and web in compression, the stiffness coefficient k 7 should be taken equal to
infinity. This component need not be taken into account when calculating the rotational
stiffness Sj.

Figure 46 Dimensions of an equivalent T-stub flange

For a beam web in tension, the stiffness coefficient k 8 should be taken as equal to infinity and
needs also not be taken into account when calculating the rotational stiffness S j.

For bolts in tension, the stiffness coefficient k10 for a single bolt-row should be obtained from:

k10=1,6 As /Lb (0)

where:
As is the tensile stress area of the bolt;
Lb is the bolt elongation length, taken as equal to the grip length (total thickness of
material and washers), plus half the sum of the height of the bolt head and the height
of the nut.

For longitudinal reinforcing bars in tension, the stiffness coefficient k 13 for a row r should be
obtained from formulae below:

For a double-sided joint configuration under balanced loading, M b1,Sd = Mb2,Sd (see Figure 41) [1] 8.4.2 (12)

(0)

where:
Ar,s is the cross-sectional area of the longitudinal reinforcement within the total effective

© SSEDTA 2001 Last modified jj/06/18 49

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

width of concrete flange beff, as given by EN 1994-1-1 [1] 5.2.2; [1] 5.2.2
Es is the modulus of elasticity for steel reinforcement;
hc is the height of the column´s steel section.

For a beam-to-beam joint, the breadth of the flange of the supporting primary beam replaces
the height of the column section.

For a single-sided joint configuration:

(0) [1] 8.4.2 (13)
where:
Kβ=2,6.

For a double-sided joint configuration under unbalanced loading, with M b1,Sd Mb2,Sd (see Figure
41)
for the more heavily loaded joint: [1] 8.4.2 (14)

(0)

where:

Kβ = β (4,3 β2 - 8,9 β + 7,2) (0)

for the less heavily loaded joint:

(0)

For a contact plate in compression, the stiffness coefficient k 14 should be taken as equal to [1] 8.4.2 (15)
infinity. This component need not be taken into account when calculating the rotational
stiffness Sj.

3.5.4 Deformation of the shear connection

Unless account is taken of deformation of the shear connection by a more exact method, the [1] 8.4.3
influence of slip on the stiffness of the joint should be determined by following:
For:
 a double-sided joint configuration under balanced loading;
 a single-sided joint configuration;
 the more heavily loaded joint in a double-sided configuration under unbalanced loading:
the stiffness coefficient k13 (see EN 1994-1-1 [1] 8.4.2) should be multiplied by the reduction [1] 8.4.2
factor kr

(0)

where:

(0)

hs is the distance between the longitudinal reinforcing bars in tension and the centre of
compression;
ds is the distance between longitudinal reinforcing bars in tension and the centroid of the
beam´s steel section;

(0)

© SSEDTA 2001 Last modified jj/06/18 50

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

(0)

Ia is the second moment of area of the beam´s steel section;

ℓ is the length of the beam in hogging bending adjacent to the joint, which in a braced
frame may be taken as 15% of the length of the span;
N is the number of shear connectors distributed over the length ℓ;
ksc is the stiffness of one shear connector.

The stiffness of the shear connector ksc may be taken as (0,7 PRk /s), where: [1] 8.4.3 (3)

PRk is the characteristic resistance of the shear connector

s is the slip, determined from push tests in accordance with EN 1994-1-1 [1] 11.2 at a [1] 11.2
load of 0,7 PRk.

Alternatively, for a solid slab or for a composite slab in which the reduction factor k t is unity, [1] 8.4.3 (4)
see EN 1994-1-1 [1] 6.7.5.2, the following approximate values may be assumed for k sc:

for 19 mm diameter headed studs: 100 kN/mm

for cold-formed angles of 80 mm to 100 mm height: 70 kN/mm.

For a composite joint with more than a single layer of reinforcement considered effective in [1] 8.4.3 (5)
tension, the above formulae are applicable provided that the layers are represented by a single
layer of equivalent cross-sectional area and equivalent distances from the centre of
compression and the centroid of the beam´s steel section.

3.6 Rotation capacity

It is not usual for designers to calculate either the required or available rotation capacity of
structural elements. Rotation capacity is essential though if redistribution of bending moments
is assumed in the global analysis. For members, well-known classification systems are used to
ensure that adequate rotation capacity is available.
With semi-continuous construction, rotation capacity may be required of the joints rather than
the members. As a result, prEN 1993-1-8 [2] gives guidance on joint ductility.

Component models can be used to calculate the rotation capacity of a joint, provided the
limiting deformation capacity of each active component is known. For steel joints though it is
often sufficient to rely on the observed behaviour of critical joint components. Thus for bolted
joints the prEN 1993-1-8 [2] permits the designer to assume sufficient rotation capacity for
plastic global analysis provided that the moment resistance of the joint is governed by the
resistance of one of the following:
 The column web panel in shear;
 The column flange in bending;
 The beam end-plate in bending.

In the latter two cases, the thickness of the flange or the end-plate must also be limited to avoid
fracture of the bolts.
For composite joints, yielding of the slab reinforcement in tension is the main source of
predictable deformation capacity. The rotation capacity corresponding to this failure mode can
be calculated from a simplified component model.

When plastic global analysis is used, the partial-strength joints should have sufficient rotation [1] 8.5 (1)
capacity. Where necessary, see EN 1994-1-1 [1] 8.2.3.3, full-strength joints should also have [1] 8.2.3.3
sufficient rotation capacity.

© SSEDTA 2001 Last modified jj/06/18 51

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

When elastic global analysis is used, joints should have sufficient rotation capacity if the [1] 8.5 (2)
conditions given in EN 1994-1-1 [1] 8.2.3.2 are not satisfied. [1] 8.2.3.2

A joint with a bolted connection, in which the moment resistance M j,Rd is governed by the
resistance of bolts in shear, should not be assumed to have sufficient rotation capacity for
plastic global analysis.

In the case of members of steel grades S235, S275 and S355, the provisions given below may
be used for joints in which the axial force N Sd in the connected member does not exceed 10% of
the resistance Npl.Rd of its cross-section. However, these provisions should not be applied in the
case of members of steel grades S420 and S460.

A beam-to-column joint in which the moment resistance of the joint M j,Rd is governed by the [1] 8.5 (3)
resistance of the column web panel in shear, may be assumed to have sufficient rotation
capacity for plastic global analysis.
The steelwork parts of a composite joint with a bolted connection with end-plates may be
assumed to have sufficient rotation capacity for plastic analysis, provided that both of the [1] 8.5 (4)
following conditions are satisfied:
 The moment resistance of the steelwork connection is governed by the resistance of either:
 The column flange in bending;
 The beam end-plate in bending.
 The thickness t of either the column flange or beam end-plate satisfies:
(0)

where:
d is the nominal diameter of the bolts;
fub is the ultimate tensile strength of the bolts;
fy is the yield strength of the relevant basic component.

The rotation capacity of a composite joint may be determined by testing. Alternatively, [1] 8.5 (5)
appropriate calculation models may be used.

The design rotation capacity determined from a tested structure or element should be adjusted [1] 8.5 (6)
to take account of possible variations of the properties of materials from specified characteristic
values.

3.7 Detailing

3.7.1 Longitudinal slab reinforcement

For ductile failure, the total cross-sectional area of longitudinal reinforcement A s should not [1] 8.6 1 (1)
exceed the following limit:

(0)

where:
for a solid slab deff is the overall depth of the slab;
for a composite slab deff = hc (see Figure 47).

Figure 47 Sheet and slab dimensions

Longitudinal reinforcement for a composite joint shall be positioned so that the distance e L [1] 8.6 1 (2)
from the axis of the column web to the centre of gravity of the longitudinal reinforcement
placed each side of the column is within the following limits:

© SSEDTA 2001 Last modified jj/06/18 52

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

0,7 bc ≤ eL ≤ 2,5 bc (0)

where:
bc is the width of the column´s steel section.

3.7.2 Transverse slab reinforcement

Transverse slab reinforcement shall be provided adjacent to the column.
Transverse reinforcement shall be positioned so that the distance e T from the face of the [1] 8.6 1 (2)
column´s steel section to the centre of gravity of the transverse reinforcement placed each side
of the column is within the following limits:

eL ≤ eT ≤ 1,5 eL (0)

Sufficient transverse reinforcement should be provided so that the design tensile resistance of [1] 8.6 1 (3)
the transverse reinforcement placed each side of the column fulfils the condition:

(0)

where:
AT is the area of transverse reinforcement placed each side of the column;
fsk,T is the characteristic tensile strength of that reinforcement, and:

(0)

[1] 8.6 1 (3)

3.7.3 Anchorage of reinforcement
Bars shall be anchored in accordance with EN 1992-1 [10] so as to develop their design
tension resistance.
For a single-sided joint configuration, the longitudinal slab reinforcement in tension shall be
anchored sufficiently well beyond the span of the beam to enable the design tension resistance
to be developed.

4 Summary

4.1 General

Following configurations and connection types are described

 Double-sided beam-to-column joint configurations
 Single-sided beam-to-column joint configurations

Types of steelwork connections:

 Connections with contact plates
 Connections with partial-depth end-plates
 Connections with flush end-plates

4.2 Field of application of the calculation procedures

 Static loading
 Small normal force N in the beam
(N/Npl < 0,1 where Npl is the squash load of the beam)
 Strong axis beam-to-column joints
 In the case of beam-to-beam joints and two-sided beam-to-column joints, only a slight

© SSEDTA 2001 Last modified jj/06/18 53

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

difference in beam depth on each side is possible

 Steel grade: S235 to S355
 H and I hot-rolled cross-sections
 Flush end-plates: one bolt-row in tension zone
 All bolt property classes as in prEN 1993-1-8 [2], it is recommended to use 8.8 or 10.9
grades of bolts
 One layer of longitudinal reinforcement in the slab
 Encased or bare column sections

4.3 Summary of key steps

The calculation procedures are based on the component method which requires three steps:
 Definition of the active components for the studied joint;
 Evaluation of the stiffness coefficients (k i) and/or strength (F Rd,i) characteristics of each
individual basic component;
 Assembly of the components to evaluate the stiffness (S j) and/or resistance (Me, MRd)
characteristics of the whole joint.

The stiffness (k i) and the design resistance (F Rd,i) of each component are evaluated from
analytical models. The assembly is achieved as follows:

Initial stiffness:

(0)

where:
z relevant lever arm;
n number of relevant components;
Ea steel elastic modulus.

Nominal stiffness:
for beam-to-column joints with contact plates
for beam-to-column joints with flush end –plates

Plastic design moment resistance:

FRd= min [FRd,i] (0)

MRd = FRd . z (0)

Elastic design moment resistance:

Me=2/3 MRd (0)

4.4 Conclusion
Conventionally joints have been treated either as pinned or as fully rigid due to a lack of more
realistic guidance in view of modelling. In reality both assumptions may be inaccurate and
uneconomic and do only represent the boundaries of the real moment-rotation behaviour. They
may lead to a wrong interpretation of the structural behaviour in terms of load resistance and
deflections. So whereas up to now the joint construction expensively has been adopted to the
possibilities of calculation, the new approach is to develop efficient joint types first and to take
their realistic behaviour into consideration within the global frame analysis afterwards.

© SSEDTA 2001 Last modified jj/06/18 54

Structural Steelwork Eurocodes –Development of a Trans-National Approach
Composite Joints: Annex A

In contrast to the idealising assumptions of beam-to-column connections as hinges or full

restraints with the main attention to their resistance, the interest of recent research activities
lies in the assessment of the whole non-linear joint response with the three main
characteristics.
 Initial rotational stiffness
 Moment resistance
 Rotation capacity

The moment-rotation behaviour of joints can be represented in structural analysis in a practical

way:
The joint representation covers all necessary actions to come from a specific joint
configuration to its reproduction within the frame analysis. These actions are the:
 Joint characterisation: determination of the joint response in terms of M-φ curves
reflecting the joint behaviour in view of bending moments and shear (in case of moment
imbalance).
 Joint classification: if aiming at conventional modelling the actual joint behaviour may
be compared with classification limits for stiffness, strength and ductility; if the respective
requirements are fulfilled a joint then may be modelled as a hinge or as a rigid restraint.
 Joint idealisation: for semi-continuous joint modelling the M- φ behaviours have to be
taken into consideration; depending on the desired accuracy and the type of global frame
analysis the non-linear curves may be simplified as bi- or tri linear approximations.
 Joint modelling: reproduction (computational model) of the joint´s M- φ behaviour within
the frame modelling for global analysis.

An analytical description of the behaviour of a joint has to cover all sources of deformabilities,
local plastifications, plastic redistribution of forces within the joint itself and local instabilities.
Due to the multitude of influencing parameters, a macroscopic inspection of the complex joint
by subdividing it into “components” has proved to be most appropriate. In comparison with the
finite element method, these components, which can be modelled by translational spring with
non-linear force-deformation response, are exposed to internal forces and not to stresses.
The procedure of the COMPONENT METHOD can be expressed in three steps:
 Component identification
determination of contributing components in compression, tension and shear in view of
connecting elements and load introduction into the column web panel.
 Component characterisation
determination of the component´s individual force-deformation response with the help of
analytical mechanical models, component tests of FE-simulations.
 Component assembly
assembly of all contributing translational component springs to overall rotational joint
springs according to the chosen component mode.

Basic components of a joint are described and detailed design provisions for resistance and
rotational stiffness are given.

© SSEDTA 2001 Last modified jj/06/18 55