0 Chapter 0

Introduction to the course

Design of concrete structures
0.1 Aim of the course notes
These course notes are intended to help the student reach the course goals. After the end
of the course, the students are expected to be able to calculate independently structures
in reinforced concrete in both the design and control phases, in accordance with the
regulations in Eurocode 2 (EN1992-1-1).
This is an academic-level course in full accordance with the concept of university level
education; this means that the course notes focus on the development and understanding
of calculation methods. Yet, as later contacts between the students and the professional
environment are important, the course notes stay close to the Eurocode 2 regulations.

0.2 Eurocode 2: a document in evolution

Eurocode 2 has been reworked several times. The present course notes are based on
version EN 1992-1-1:2004, which has been published in English at the end of 2004. The
official introduction of the code in Belgium depends on the publication in Dutch and in
French of the so called National Annex (Annexe National Nationale Bijlage - ANB),
the approval of the ANB by the General Commission of the Eurocodes and by the
Board of the Belgian Institute for Normalisation (Belgisch Instituut voor Normalisatie
Institut Belge de Normalisation BIN-INB). The introduction of the Belgian ANB took
place in 2009 with the indication: NBN EN 1992-1-1 ANB:2009.

0.3 Introduction of Eurocode 2

The following paragraphs reproduce important passages taken from the Foreword in
EN 1992-1-1:2004 and from the first chapter of the code (Section 1 - General).
0.3.1

Background to the Eurocode programme

In 1975, the Commission of the European Community decided on an action programme

in the field of construction. The objective of the programme was the elimination of
technical obstacles to trade and the harmonization of technical specifications. Within
this action programme, the Commission took the initiative to establish a set of
harmonized technical rules for the design of construction works which, in a first stage,
would serve as an alternative to the national rules in force in the Member States and,
ultimately, would replace them.
For fifteen years, the Commission, with the help of a Steering Committee with
Representatives of Member Sates, conducted the development of the Eurocodes
programme, which led to the first generation of European codes in the 1980s.
In 1989, the Commission and the Member States of the EU decided, on the basis of an
agreement between the Commission and CEN, to transfer the preparation and the
publication of the Eurocodes to CEN through a series of Mandates, in order to provide
them with a future status of European Standard (EN).
The Structural Eurocode programme comprises the following standards generally

0-1

consisting of a number of Parts :

EN 1990 Eurocode 0 : Basis of Structural Design
EN 1991 Eurocode 1 : Actions on structures
EN 1992 Eurocode 2 : Design of concrete structures
EN 1993 Eurocode 3 : Design of steel structures
EN 1994 Eurocode 4 : Design of composite steel and concrete structures
EN 1995 Eurocode 5 : Design of timber structures
EN 1996 Eurocode 6 : Design of masonry structures
EN 1997 Eurocode 7 : Geotechnical design
EN 1998 Eurocode 8 : Design of structures for earthquake resistance
EN 1999 Eurocode 9 : Design of aluminum structures
Eurocode standards recognize the responsibility of regulatory authorities in each
Member State and have safeguarded their right to determine values related to regulatory
safety matters at national level where these continue to vary from State to State.
The Eurocodes standards provide common structural design rules for everyday use for
the design of whole structures and component products of both a traditional and an
innovative nature. Unusual forms of construction or design conditions are not
specifically covered and additional expert consideration will be required by the designer
in such cases.
0.3.2

National Standards implementing Eurocodes

The national Standards implementing Eurocodes will compromise the full text of the
Eurocode (including any annexes), as published by CEN, which may be preceded by a
National title page and National foreword, and may be followed by a National annex.
The National annex may only contain information on those parameters which are left
open in the Eurocode for national choice, known as Nationally Determined Parameters,
to be used for the design of buildings and civil engineering works to be constructed in
the country concerned, i.e. :
values and/or classes where alternatives are given in the Eurocode,
values to be used where a symbol only is given in the Eurocode,
country specific data (geographical, climatic, etc.), e.g. snow map,
procedure to be used where alternative procedures are given in the Eurocode.
It may contain
decisions on the application of informative annexes,
references to non-contradictory complementary information to assist the user to
apply the Eurocode.
Furthermore, all the information accompanying the CE Marking of the construction
products which refer to Eurocodes should clearly mention which Nationally Determined
Parameters have been taken into account.
0.3.3

Additional information specific to EN 1992-1-1

EN 1992-1-1 describes the principles and requirements for safety, serviceability and
durability of concrete structures, together with specific provisions for buildings.
It is based on the limit state concept used in conjunction with a partial factor method.
This standard gives values with notes indicating where national choices may have to be
0-2

made. Therefore the national Standard implementing EN 1992-1-1 should have a

National annex containing all Nationally Determined Parameters to be used for the
design of buildings and civil engineering works to be constructed in the relevant
country.
0.3.4

Scope of Eurocode 2

Note : the code makes distinction between Principles (indicated by the letter P) and
Application rules.
(1) P (= Principle) Eurocode 2 applies to the design of buildings and civil engineering
works in plain, reinforced and prestressed concrete. It complies with the principles
and requirements for the safety and serviceability of structures, the basis of their
design and verification that are given in EN 1990 : Basis of structural design.
(2) P Eurocode 2 is only concerned with the requirements for resistance,
serviceability, durability and fire resistance of concrete structures. Other
requirements, e.g. concerning thermal or sound insulation, are not considered.
(3) P Eurocode 2 is intended to be used in conjunction with :
EN 1990 :
Basis of structural design
EN 1991 :
Actions on structures, subdivded in the following parts:
EN 1997
Geotechnical design
EN 1998
Design of structures for earthquake resistance
hENs :
Construction products relevant for concrete structures
EN 197-1 : Cement : Composition, specification and conformity criteria
for common cements
EN 206-1 : Concrete : Specification, performance, production and
conformity
EN 10080 : Steel for the reinforcement of concrete
EN 10138 : Prestressing steels
EN 12390 : Testing hardened concrete
ENV 13670 : Execution of concrete structures
EN 13791 : Testing concrete
(4)

0.3.5
(1)

(2)

P Eurocode 2 is subdivided into the following parts :

EN 1992-1-1 :
General rules and rules for buildings
EN 1992-1-2 :
Fire resistance of concrete structures
EN 1992-2 :
Reinforced and prestressed concrete bridges
EN 1992-3 :
Liquid retaining and containment structures
Scope of Part 1-1 of Eurocode 2
P Part 1-1 of Eurocode 2 gives a general basis for the design of structures in plain,
reinforced and prestressed concrete made with normal and light weight aggregates
together with specific rules for buildings.
P The following subjects are dealt with in Part 1-1.
Section 1 : General
Section 2 : Basis of design

0-3

(3)
(4)

Section 3 : Materials
Section 4 : Durability and cover to reinforcement
Section 5 : Structural analysis
Section 6 : Ultimate limit state
Section 7 : Serviceability limit states
Section 8 : Detailing of reinforcement and prestressing tendons General
Section 9 : Detailing of members and particular rules
Section 10 : Additional rules for precast concrete elements and structures
Section 11 : Lightweight aggregate concrete structures
Section 12 : Plain and lightly reinforced concrete structures
P Sections 1 and 2 provide additional clauses to those given in EN 1990 Basis of
structural design
P Part 1-1 does not cover :
resistance to fire;
particular aspects of special types of building (such as tall buildings);
particular aspects of special types of civil engineering works (such as
viaducts, bridges, dams, pressure vessels, offshore platforms or liquidretaining structures);
no-fines concrete and aerated concrete components, and those made with
heavy aggregate or containing structural steel sections (see Eurocode 4 for
composite steel concrete structures).

0.4 Table of content of the course notes

Part 1

Fundamentals of the design of structures in reinforced concrete

Chapter 1
Design methods: elastic design, method using partial safety
factors, probabilistic design and failure design.
Chapter 2
Method with partial safety factors design loads. This
chapter deals with the determination of the load combinations
as design parameters. The chapter also includes fundamental
aspects on the identification of structural schemes and the
performance of structural analysis of structures in reinforced
concrete.
Chapter 3
Method with partial safety factors design properties of
materials. This chapter discusses material behavior and
identifies design values to be used in analysis.
Part 2
Basic theory of design of beams in reinforced concrete
Chapter 4
Determination of the main reinforcement in reinforced
concrete sections in the Ultimate Limit State (ULS), for
bending and normal forces.
Chapter 5
Verification of sections in the Serviceability Limit State
(SLS): verification of stress levels and simplified criteria for
avoiding unacceptable crack opening and deformation.
Chapter 6
Practical detailing of the main reinforcement.
Chapter 7
Design for shear, with analysis of shear strength of members
without and with shear reinforcement and detailing of shear
0-4

reinforcement.
Chapter 8
Torsion: the chapter is limited to some principal aspects of
this problem.
Chapter 9
Determination of crack opening and deflections of beams by
means of simplified models.
Part 3
Design of slabs
Chapter 10
Design of one-way, two-way and continuous slabs by means
of simplified design methods (Marcus, Czerny). Practical
detailing of reinforcement and handling of the torsion
problem in corners. The equivalent framework method for
flat slab design is discussed and is complemented by an
introduction to the punching shear design problem.
Chapter 11
The strip method is presented as a versatile tool for solving
various design problems: openings in slabs, concentrated
loads, complex shapes and boundary conditions.
Part 4
Design of columns
Chapter 12
Determination of the sensitivity of columns for buckling.
Detailing of reinforcement in columns. Buckling analysis is
not part of this course.
Part 5
Strut and tie method
Chapter 13
This chapter presents the strut and tie method for the design
of walls, foundation blocs, deep beams, corbels and all
regions with discontinuity in geometry or actions.

0.5 Other references (next to the Eurocodes) used as basis for the
elaboration of the course notes
BOND, A.J., HARRISON, T., NARAYAN, R.S., BROOKER, O., MOSS, R.M.,
WEBSTER, R., HARRIS, A.J., How to design concrete structures using Eurocode 2,
The Concrete Centre (www.concretecentre.com), 2006
BRAAM, C.R., DEES, W.C., Gewapend beton Constructieleer en Ontwerpen en
Compendium Eurocode 2, uitgegeven door Aeneas (www.aeneas.nl) in samenwerking
met Cement&BetonCentrum (www.cementenbeton.nl) , 2010
OBRIEN, E.J., DIXON, A.S., Reinforced and prestressed concrete design The
complete process, Longman Scientific & Technical, 1995

0-5

Part I
FUNDAMENTALS OF THE DESIGN OF
STRUCTURES IN REINFORCED CONCRETE

1 Chapter 1
Design methods
1.1 Introduction
The aim of this course is the calculation of beams, slabs, columns, walls, ... in
reinforced concrete for different load cases. In practice, this means that sections have to
be designed for different combinations of internal forces M, N, V and T, and thus for
tensile loading, compressive loading, bending, torsion, combined load cases, etc.
Starting from the internal forces M, N, V and T, one can calculate stresses and strains in
construction elements with quasi-isotropic materials, by application of simple strength
of materials formulas. Stress and strain distribution in reinforced concrete is much more
complex than in quasi-isotropic materials. Yet, in general and in particular for design
purposes (where it is asked that construction elements are able to transmit loads with a
certain safety with respect to failure), it is not necessary to know the exact distribution
of stresses and strains. The design purpose will be reached by application of formulas
based on the fundamentals of equilibrium and on simplifying assumptions. One should
thus be aware of the fact that stress distributions obtained by means of these formulas,
are approximate.

1.2 Elastic design

1.2.1

Principle

The elastic design method is based on the assumption of elastic material behaviour for
steel and concrete. On the basis of the externally applied loads, the shape of the
construction element and its boundary conditions, the internal forces M, N, V, are
determined and subsequently the stresses in the critical zones, taking into account linear
elastic material behaviour. It is assumed that failure occurs as soon as plastic
deformations appear. The design equation may thus be written as:
( , ) Re
(1.2.1-1)
s
with:
Re = yield value (limit of the domain of elastic material behaviour);
s = safety factor with respect to the crossing of the yield value.

1-1

The equation is similar to the traditional failure criteria (such as the VON MISES
criterion) used in strength of materials. The principle idea of the elastic design method
is presented in a schematic way in figure 1.2.1-1.

x
(BERNOULLI)

x=E.x

M z = x . y.dA

dA

(EQUIVALENCE
INTERNAL/EXTERNAL
FORCES)

Example :
Pure bending

(HOOKE)

x
y

x =

Mz
y
Iz

DESIGN

STRUCTURAL

CALCULATION

ANALYSIS

M, V,
N, T

(, )

F
Re (, )
s

Figure 1.2.1-1
Principle of the elastic design method
1.2.2

Criticisms on the elastic design method

Some fundamental criticisms on the elastic design method may be formulated.

1.
A sufficient safety s on the level of stresses (which are in fact reasoning tools of
the engineer) does not necessarily lead to the same safety on the level of applied

1-2

loads. This is illustrated in figure 1.2.1-1, for the case of a column loaded in pure
compression, where a non negligible 2nd order bending moment appears.
The elastic design method, which is sometimes called the method of acceptable
stresses, does not give an indication of the real safety with respect to the failure
of the construction element. For a construction element that has been designed
with this method, one does not have any idea of the failure load, because the
ultimate deformation capacities of steel and concrete up to failure have not been
exploited. This is in essence an uneconomical design philosophy.
Today, there is a tendency to use more performing materials, such as high strength
concrete and steel with high yield limit. Application of elastic design with high
performing materials leads to reduced thicknesses of slabs, for example, but gives
rise to problems when the serviceability criteria are verified: thin constructions
may present larger deformations or unacceptable cracks.

2.

3.

F
Re

Re
s

F{ =Re/s} F{ =Re}
F{ =Re/s}

F{ =Re}

Figure 1.2.2-1
2nd order effect; the safety factor s with respect to the stresses does not correspond to
the same level of safety with respect to the loads

1.3 Limit state design

1.3.1

Definitions and philosophy

Reference: EN 1990:2002; 3.5

1-3

1.3.1.1 Limit state

Generally speaking, a limit state is a state beyond which the structure no longer fulfils
the relevant design criteria; these criteria may be seen in a broad sense, dealing with
security, stability, comfort, durability, cost, etc.
1.3.1.2 Classification of limit states
Two classes of limit states are considered:
1.
Ultimate Limit State (ULS)
The limit states that concern the safety of people or the safety of the structure, are
considered as ULS. In general, ULS are related to structural collapse or
catastrophic failure of the structure. EN 1990 considers also the following ULS:
- loss of static equilibrium of the structure or part of it, considered as a rigid
body (tilting or turning bottom up, sliding, lifting, etc.);
- rupture of the structure or any part of it, including supports and foundations;
- failure by excessive deformation, transformation of the structure or part of it
into a mechanism, loss of stability of the structure or any part of it (buckling,
web-buckling, folding, aero elastic instability, etc);
- failure caused by fatigue or other time-dependent effects.
2.
Serviceability Limit State (SLS)
In case SLS are reached, problems arise concerning the functioning of the
structure or structural members under normal use, the comfort of people, the
appearance or its durability. The verification of SLS deals with the following
aspects:
- unacceptable deformations that affect the appearance, the comfort of the
users, or the normal functioning of the structure (including the functioning of
machines or services) or that cause damage to finishes or non-structural
members: excessive deflections and rotation angles at supports, inclined
position of furniture or machines, cracks in non bearing members);
- unacceptable vibrations that cause discomfort to people or that limit the
functional effectiveness of the structure.
1.3.1.3 Design situation
Limit states are related to design situations, which are situations during which the
basic variables see further are considered to be constant. Four design situations
are considered:
persistent design situations, which refer to conditions of normal use;
transient design situations, which refer to temporary load conditions, for
example during execution or repair;
accidental design situations, which refer to exceptional conditions, such as
fire, explosions, impact, localized failure;
seismic design situations.
1.3.1.4 Design working life
The limit states of structural systems are verified for one or several reference periods
called design working life (expressed in a number of years) which corresponds to the
duration of a design situation. The design working life is in practice nothing else than
the expected lifetime of the structural system. Yet, in certain cases, it can also be the

1-4

duration of the construction process or the duration of a particular phase in the

construction process or the duration of a particular utilization phase of the structure.
1.3.1.5 Basic philosophy of reliability management: the probabilistic method accepts
the idea of the probability that a limit state is exceeded
The essential point of the method is that the theoretical probability of exceeding a limit
state during the design working life is accepted. As constructions have various lifetime
durations, there is a large variability of the acceptable probability of exceeding a limit
state.
EN 1990:2002 presents a classification of structures in design working life categories,
each one corresponding to an indicative design working life: see table 1.3.1-1.
Table 1.3.1-1
Indicative design working life according to EN 1990:2002
(Table 2.1 in EN 1990:2002)
Examples
Design working Indicative design
life category
working life
(years)
1
10
Temporary structures (structures or parts of
structures that can be dismantled with a view to
being re-used should not be considered as
temporary)
2
10 25
Replaceable structures
3
15 30
Agricultural structures
4
50
Building structures
5
100
Monumental building structures, bridges, civil
engineering structures
Each structure also fits in a so-called consequence class (CC): see table 1.3.1-2. The
CC is identified by considering the consequences of failure or malfunctioning of the
structure.

Consequence
Class
CC3

CC2

CC1

Table 1.3.1-2
Consequence classes according to EN 1990:2002
(Table B1 in EN 1990:2002)
Description
Examples
High consequence for loss of human
life, or economic, social or
environmental consequences are very
important

Public buildings, such as

for example a concert
hall,

Medium consequence for loss of human

life; economic, social or environmental
consequences are considerable
Low consequence for loss of human life,
and economic, social or environmental

Residential and office

buildings,..

1-5

Agricultural buildings
where people do not

consequences are small or negligible

normally enter (storage

buildings, greenhouses,...)

Each CC is associated with a so-called Reliability Class (RC). The reliability classes
are defined by the reliability index concept: each RC is characterized by a minimum
value of the reliability index ; see table 1.3.1-3. Assuming a standard normal
distribution (which leads to the formula P() hereafter), each corresponds to a value of
P, which indicates the probability of failure.
P ( ) =

exp 2
2
2
1

Table 1.3.1-3
Reliability classes according to EN 1990:2002
(Table B2 in EN 1990:2002)
1 year reference period
50 years reference period
Reliability class
Minimum
Probability of
Minimum
Probability of
values for
failure P
values for
failure P
-7
RC3
5,2
10
4,3
3,9 . 10-5
RC2
4,7
10-6
3,8
2,9 . 10-4
RC1
4,2
10-5
3,3
1,7 . 10-3
If it is appropriate to attribute another level of safety than RC2 to a structure, then the
multiplication factor KFI should be used; see table 1.3.1-4. This factor should be applied
to the actions in their fundamental combination (see further). In practice, application of
this factor leads to a reduction or an increase of the design loads on the structure.
Table 1.3.1-4
The multiplication factor KFI to be applied on the actions, according to EN 1990:2002
(Table B3 in EN 1990:2002)
Reliability class (RC)
RC1
RC2
RC3
KFI factor for actions
0,9
1,0
1,1
A concrete example: according to EN 1990:2002, a building must be situated in the
design working life category 4 and has an indicative design working life of 50 years; a
building must also be situated in the consequence class 2 which corresponds to the
reliability class 2. The RC 2 is characterized by minimum equal to 3,8 and the
probability of failure P is 2,9.10-4.
1.3.1.6 Basic variables
For each limit state, structural design is performed by means of calculation models in
which a series of basic variables have to be implemented. The basic variables are
measurable physical properties. The following basic variables are identified by EN
1992-1-1:2004:

1-6

 the actions: the loads and deformations (thermal effects, differential settlements,
movements) that are applied onto the structure;
the material properties: yield strength of steel, compression strength of concrete,
tensile strength of concrete, stiffness (modulus of elasticity), shrinkage, creep,
etc.;
the geometrical parameters: shape and dimensions of the load bearing system and
of its structural elements and sections, as well as imperfections.
The probabilistic design method consists of the determination of the probability of the
exceeding of a limit state on the basis of the statistical distributions of the basic
variables. These distributions are often not known exactly and are withdrawn from
hypothetical extrapolations. This is the reason why, in most cases, it is not possible to
determine exactly the probability of exceeding a limit state, and thus why the values in
table 1.3.1-3 have to be considered as conventional values. Moreover, in most common
cases, simplifications are accepted which makes that accurate probabilistic design is not
justified anymore.
In practice, the partial safety factor method, which is a semi-probabilistic method, is
used.
1.3.2

Fundamentals of the partial safety factor method (semi-probabilistic method)

Figure 1.3.2-1 presents the basic idea of the partial safety factor method in a schematic
way. The method considers two streams of effects (= internal forces) of the design
values: the effects of the actions (the loads) and the effects related to the material
properties.
1st step
The characteristic value of the action Fk is determined on the basis of a statistical
analysis of the applied action F. In general, a normal distribution can be adopted, which
allows to define the characteristic value Fk in a simplified way as the value of F which
will not be exceeded with a probability of 95%. The concept of characteristic value thus
takes into account the uncertainty about the prediction of the value of the action that
will be applied on the moment of the structural design.

In order to take into account additional uncertainties, Fk is multiplied by the partial

safety factor for the action F in order to obtain the design value of the action Fd
(d < design):
Fd = F Fk

(1.3.2-1)

The additional uncertainty for the action takes into account the possibility of
unfavourable deviations of the actions from the characteristic values (thus exceptional
high loads which should normally not determine the design).
Structural analysis then leads to the determination of the design value of the effect of
the action Ed, which is essentially a bending moment Md, a normal force Nd, or a shear
force Vd, etc. which is due to the application of the applied force Fd onto the structure.

1-7

s
s

choice -diagram

non linear constitutive law

c c s s

100/00

-diagram
z

Equivalence
internal/external forces

dA

M
= f

x
y

Rd

Ed

f
95%

fk

95%

fd= fk/m

Fk

Fd=F.Fk

Figure 1.3.2-1
Limit state design; partial factor method
2nd step
The statistical analysis of the material properties of concrete and steel permits to
determine the characteristic values fk :
the characteristic value of the compressive strength of concrete fck (c < concrete);
the characteristic value of the tensile yield strength of the steel fyk (y < to yield).
In general, a normal distribution can be adopted, which allows to define the
characteristic value fk in a simplified way as the value of f which will be available in
95% of the cases. The concept of characteristic value thus takes into account the
uncertainties due to the spread of the strength measurements performed on real
specimens.
In order to take into account additional uncertainties, fk is divided by a partial safety
factor for materials m (m < materials), in order to determine the design resistance fd : fcd
and fyd.
fk
(1.3.2-2)
fd=
m

The additional uncertainty for the material takes into account the possibility of
unfavorable deviations of the material resistance from the characteristic value:
exceptional low values of the resistance should normally not determine the design.

1-8

A strength of materials approach is then used (via the expression of static equilibrium
and by means of the constitutive laws) to determine the design value of the resistant
internal force Rd , which is essentially the resistant bending moment Md, the resisting
normal force Nd, or the resisting shear force Vd, etc.; these are the internal forces that a
section can generate from the point-of-view of the materials strength.

Note 1 : next to the group of uncertainties about the actions and the group of
uncertainties about the material resistance, there is a third group of
uncertainties to be considered, related to the values of the internal forces. These
uncertainties are related to the inevitable approximations due to the
assumptions in the strength of materials theory, dimensional uncertainties and
uncertainties related to the execution: for example, strength of materials on site
may be slightly different from that measured in a carefully prepared test
specimen, tested in laboratory conditions (this is particularly true for concrete
where placing, compaction and curing are very important to the strength).
These uncertainties are covered by the factors F en m.

Note 2 : the partial safety factor method is a so-called semi probabilistic

method; indeed: the method starts from statistically determined characteristic
values, but uses chosen partial safety factors to cover several other types of
uncertainties.

Note 3 : the partial safety factor method is designated as partial factor

method in EN 1992-1-1:2004. This designation is used in the following text.
Partial safety factors are thus called partial factors.

3rd step
For each limit state, the comparison is made between the internal forces as effects from
the actions and the internal forces related to the material resistance. For each limit state,
the following condition should be verified:

Ed Rd
or

Ed Fid Rd ,lim ( f yd , f cd )
u

(1.3.2-3)

This condition expresses that, in each section of the construction element to be

designed, the internal force as a result from the most unfavourable (u < unfavourable)
combination of the design actions, should be smaller (or equal) than the ultimate design
value of the resisting internal force, determined on the basis of the materials design
strength values.

1.4 Design assisted by testing

Design may be based on a combination of tests and calculations. By the realization of a
large number of tests up to failure, on a large number of identical construction elements,
and by statistical interpretation of the results, it is possible to establish formulas that
express the failure load in function of the geometrical characteristics (dimensions) and
1-9

mechanical characteristics of the used materials. This is a global and probabilistic

approach. This approach could be applied for example in the case of large series
production of standardized precast elements.

1-10

2 Chapter 2
Partial factor method the actions
2.1 Aim of chapter 2
This chapter focuses on the right side of figure 1.3.2-1, with two major topics:
the identification of the actions F, including the determination of the
characteristic values Fk and the design values Fd;
the structural analysis that leads to the determination of the internal forces in a
structural member in function of the applied actions.

2.2 Classification of actions

Reference: EN 1991-1-1:2002; EN 1992:2004
A distinction is made between:
2.2.1

Permanent actions (or loads) G

These are actions that can occur throughout a given design situation, and for which the
variation of the value with time can be neglected in comparison with their mean value.
The main permanent actions are:
self-weight of the structures;
self-weight of non-structural parts of the building, such as roofing, surfacing and
coverings, partitions, hand rails, wall cladding, suspended ceilings, thermal
insulation, fixed services, fresh concrete, etc;
actions resulting from soil pressures;
prestressing actions P;
imposed deformations due to creep, shrinkage or expansion of the used materials;
actions resulting from settlements of soils.
2.2.2

Variable actions Q

These are actions that do not occur continuously throughout a given design situation,
and for which the variations of the value with time can not be neglected in comparison
with their mean value. The main variable actions are:
service loads (imposed loads);
loads applied during the construction process, resulting from the actions of
forklifts, due to the storage of materials, due to variations in the self-weight of
elements during the construction process, etc;
soil pressure resulting from loads moving on the ground surface;
wind loads;
loads resulting from rain or snow, hail storms and frost, as well as effects of
moisture (actions due to deformations such as swelling or shrinkage);
the variable part of the imposed or restrained deformations due to temperature or
moisture.

2-1

2.2.3

Accidental actions A

These are actions with a very low probability of occurrence during the considered
reference period. The main accidental actions are:
collisions and explosions;
fire;
accidental soil settlements;
ground or snow sliding;
earthquakes;
tornados;
floods and accidental erosion.

2.3 Characteristic values of the actions

2.3.1

Permanent actions

According to their definition, permanent actions do not change much in time; yet, some
uncertainty may exist about their real value, for example due to variations in the
workers methods for application of materials during the construction process. EN 1990
assumes that a permanent action is characterized by a normal distribution and thus can
be represented by its characteristic value Gk, with a probability P = 5% that it will be
exceeded (Gkmax) or that it will not be reached (Gkmin) :
Gkmax=Gm+1,64.sg

Gkmin=Gm-1,64.sg

where
Gm is the mean value of the permanent action;
sg is its standard deviation.
In case of insufficient statistical data, the nominal value Gn having the same level of
safety, can be chosen as characteristic value. Nominal values for self-weight are
presented in EN 1991-1-1 (Densities, self-weight and imposed loads).

2.3.2

Variable actions

If enough statistical data are available (and with the assumption of a normal
distribution), the characteristic value of a variable action Qk is defined as the value of Q
characterized by the probability P=5% to be exceeded during a reference period tref (the
design working life).
Without sufficient statistical data, the nominal value Qn, determined on the basis of all
available information and having at least the same safety level, may be chosen as
characteristic value. Nominal values (and thus characteristic values) are presented in the
following standards:
- EN 1991-1-1 : Imposed loads or service loads
- EN 1991-1-2 : Actions on structures exposed to fire
- EN 1991-1-3 : Snow load
- EN 1991-1-4 : Wind actions

2-2

EN 1991-1-5 :
EN 1991-1-6 :
EN 1991-1-7 :
EN 1991-2 :
EN 1991-3 :
EN 1991-4 :

Thermal actions
Actions during execution
Accidental actions
Traffic loads on bridges
Actions induced by cranes and machinery
Silos and tanks

2.4 Design values of the actions

2.4.1

Introduction

The partial factors are determined in function of the nature of the actions and the type of
"combinations"; the term "combination" is used in the standards to indicate a group of
actions occurring simultaneously.
2.4.2

Partial factors for the actions

Table 2.4.2-1 presents the partial factors for the actions according to EN 1990:2002,
specifically to be used for the design of structural members. The table makes a
difference between permanent and transient design situations (P/T) and accidental
design situations (A).
Table 2.4.2-1
Partial factors for the actions (Reference = Table A1.2(B) in EN 1990:2002)
Action
Permanent action
- favourable
- unfavourable
Variable action
- unfavourable
Accidental action

Symbol

Situation
P/T

G,inf
G,sup

1,00
1,35

1,00
1,00

Q
A

1,50
-

1,00
1,00

Note : EN 1990:2002; A1.3 proposes slightly different partial factors for the
other limit states such as static equilibrium, design of structural members
involving geotechnical actions and the resistance of the ground. EN 1992-11:2004; 2.4.2 also presents partial factors for shrinkage actions (SH), for
prestressing (P) and for fatigue loads (F,fat).

2.4.3

Combinations of actions

For each design situation, all combinations of actions that can occur simultaneously
should be considered in order to identify the combination that leads to the most
unfavourable effect on the design. Each combination of actions includes:
all the permanent actions G (except prestressing P);
the eventual effect of prestressing P;
2-3

 one variable action (Q) or accidental action (A) taken as main action;
the other additional variable actions Q.
In order to take into account the limited probability that all actions take place
simultaneously with their maximum (unfavourable) intensity, the characteristic values
of the variable actions are multiplied by combination factors , leading to the following
reduced values :
combination value of a variable action 0.Qk
frequent value of a variable action 1.Qk
quasi-permanent value of a variable action 2.Qk
For ULS design, the following combinations are considered:
the fundamental combination in which the main action is one of the variable
actions, and
the accidental combination in which the main action is the accidental action.
For SLS verification, the following combinations are considered:
the characteristic combination in which the main variable action is taken with its
characteristic value;
the frequent combination in which the main variable action is taken with its
frequent value;
the quasi-permanent combination in which the main variable action is taken with
its quasi-permanent value.
Table 2.4.3-1 presents a summary of the combinations of actions according to EN
1990:2002). It should be observed that for all SLS verifications, all partial factors
F = 1.
Table 2.4.3-2 presents the combination factors 0, 1 et 2 according to EN 1990:2002.

Limit
state

Table 2.4.3-1
Summary of the combinations of actions (EN 1990:2002; 6)
Combination Permanent actions Main action
Additional actions
Fundamental
combination

g .Gk + p .Pk

Accidental
combination

g .Gk + p .Pk

Characteristic
combination

Gk + Pk

Frequent
combination

Gk + Pk

Quasipermanent

Gk + Pk

ULS

SLS

2-4

Variable
action
q .Q1k

q . 0i .Qik

Accidental
action
A
Variable
action
Q1k

11.Q1k + 2i .Qik

i >1

i >1

0i

.Qik

Variable
action
11.Q1k

2i

.Qik

Variable
action

2i

.Qik

i >1

i >1

i >1

combination

21.Q1k

Table 2.4.3-2
Partial combination factors for buildings (Table A1.1 in EN 1990:2002)
Action
0
1
2
Imposed loads in buildings
Category A: domestic, residential areas
0,7
0,5
0,3
Category B: office areas
0,7
0,5
0,3
Category C: congregation areas (theatres, etc.)
0,7
0,7
0,6
Category D: shopping areas
0,7
0,7
0,6
Category E: storage areas
1,0
0,9
0,8
Category F: traffic area
0,7
0,7
0,6
vehicle weight < 30 kN
Category G: traffic area
0,7
0,5
0,3
30 kN < vehicle weight <160 kN
Category H: roofs
0
0
0
Snow loads on buildings
Finland, Iceland, Norway, Sweden
0,7
0,5
0,2
Remainder of CEN Member States, for sites located
0,7
0,5
0,2
at altitude H > 1000 m a.s.l.
Remainder of CEN Member States, for sites located
0,5
0,2
0
at altitude H < 1000m a.s.l.
ANB (Belgium), annex to EN 1990 (2002)
0,5
0
0
Wind loads on buildings
0,6
0,2
0

Note: There are two possibilities to describe the effects of the combination of
actions :
by considering the effect E of the combinations of de design values of
actions; this is an expression of the type: E(sum of F..Fk), or
by considering the sum of the effects E of the characteristic values of
actions, multiplied by the appropriate partial factors; this is an expression of
the type: sum of F..E(Fk).
These two expressions are equivalent on the condition that the effect E is
proportional to the action F; this corresponds to a linear relationship such as
shown in figure 2.4.3-1 (straight line identified by (a)).
The expression sum of F..E(Fk) should be used when the effect E increases
more slowly than the actions, such as illustrated by curve (b) in figure 2.4.3-1. In
this case, the first formulation would not be on the safe side.
The expression E(sum of F..Fk) should be used when the effect E increases
faster than the actions, such as illustrated by curve (c) in figure 2.4.3-1. In this
case, the second formulation would not be on the safe side.

2-5

E
(b)

(c)

E(F1+F2)

E(F1+F2)
E(F2)
E(F1)

E(F2)

F1 F2

F1+F2

examples with :
F = 1
=1

E(F1+F2) < E(F1) + E(F2)

E(F1)
F1 F2

F1+F2

E(F1+F2) > E(F1) + E(F2)

Figure 2.4.3-1
Evolution of the effect E in function of the action F

2.5 Structural analysis

2.5.1

Introduction : the different methods of analysis

Reference : EN 1992-1-1:2004; 5.1.1

The purpose of structural analysis is to establish the distribution of internal forces Ed in
all constitutive parts of a structure. This topic is typically discussed in the course
"Stability of structures". The analysis is performed taking into account simplifying
assumptions concerning:
the geometry of the structure; the structure is usually simplified into a scheme
composed of linear elements (beams, columns), 2-dimensional elements (walls,
slabs) and occasionally shells ;
the behaviour of the structure, for which one of the four following idealizations is
commonly used:
linear elastic behaviour ;
linear elastic behaviour with limited redistribution ;
plastic behaviour ;
non-linear behaviour.
Linear elastic structural analysis is based on the assumption that the relationship
between stresses and strains is linear: = E.. For a beam with homogeneous crosssection, loaded in bending, this leads to a linear relationship between the bending
moment and the curvature: M = E.I.K. In linear elastic analysis, one may adopt the
bending stiffness E.I of the uncracked cross-section; however, if the cracking in
concrete has a non-negligible unfavourable effect on the structural behaviour, it must be
taken into account in the design. Linear analysis of elements based on the theory of
elasticity may be used for both the serviceability and limit states.
Taking into account the specific nature of the load bearing system, the limit state, and
particular design or execution requirements, the analysis of structural members for the
verification of ULS may be performed by means of linear analysis without or with
limited redistribution or by means of non-linear or plastic analysis. The term non-linear
analysis refers to the analysis where adequate non-linear behaviour for materials is
considered. The term 2nd order analysis is used where non-linear behaviour is the
result of excessive deformations of the element. Linear elastic analysis with
redistribution, non-linear and plastic analysis are topics that are discussed briefly in

2-6

separate chapters in the present course notes. These methods may lead to a reduction of
reinforcement quantities, but care has to be taken of the verification of the rotation
capacity or ductility conditions.
Finally, complementary analyses methods are necessary in local areas where the
assumption of BERNOULLI, the linear strain distribution, is not verified; this is the
case for support areas, areas with concentrated loads, areas with a discontinuous
variation of the cross section, anchorage zones, corbels, etc. In these cases, plastic
methods are available, such as the strut and tie method.
2.5.2

Idealisation of the structure : structural models

2.5.2.1 Nomenclature
Reference: EN 1992-1-1:2004; 5.3.1
Distinction is made between the following structural elements:
beam: a horizontal member is considered as a beam if the span is not less than 3
times the overall section depth (l 3depth). Otherwise, the member should be
considered as a deep beam, for which plastic design methods are necessary;
slab: is a member for which the minimum panel dimension is not less than 5 times
the overall slab thickness;
column: is a member for which the section depth does not exceed 4 times its
width and the height is at least 3 times the section depth. Otherwise, the member
should be considered as a wall, for which plastic design methods are necessary.
2.5.2.2 Geometric data: effective span of beams and slabs in buildings
References : EN 1992-1-1:2004; 5.3.2.2
The effective span leff of a member is calculated as follows (formula (5.8) in EN 19921-1:2004):
leff = ln + a1 + a2

(2.5.2-1)

with:
ln the clear distance between the faces of the supports;
a1 and a2 (at both ends of the span) are determined from the appropriate ai values
in figure 2.5.2-1 (figure 5.4 in EN 1992-1-1:2004).

2-7

Figure 2.5.2-1
Determination of the effective span (leff) for different support conditions, with
leff = ln + a1 + a2 (figure 5.4 in EN 1992-1-1:2004).

Note: The transmission of forces from the beam to the column (see figure 2.5.22) is not only function of the dimensions of the column section, but is also
function of the depth of the beam. The disturbed zone (according to the
principle of Barr de St Venant) and the transmission of forces is inevitably
determined by the depth of the supported member.
hA
hB

Transmission of forces

Figure 2.5.2-2
Transmission of forces is inevitably determined by the depth of the supported member
2.5.2.3 Geometric imperfections
Reference: EN 1992-1-1:2004; 5.2
The unfavourable effects of possible deviations in the geometry of the structure must be
taken into account in the structural analysis: the parasitic internal forces should be
added to the normal internal forces. The effects may be important for members with
axial compression and structures with vertical loads.

2-8

Note 1: Deviations in cross section dimensions are not at stake here: they are
covered by the partial factors and should not be included in structural analysis.

Note 2: Imperfections must be taken into account in ULS; they need not be
considered for SLS.
Imperfections in the load bearing system may be represented by an inclination: it is
assumed that the load bearing system is not perfectly vertical but presents an inclination
angle i with respect to the vertical axes. The inclination angle is determined by
(formula (5.1) in EN 1992-1-1:2004):

i = 0 . h . m
with

0 = the basic value of the inclination angle = 1/200

h = a reduction factor to take into account the length or height ( ) via the
formula:

h =

; 2 / 3 h 1

m = a reduction factor to take into account the number m of vertical members

that are collaborating:

m = 0,51 +

Note: Taking into account the imposed limit values for h, one finds that for an
isolated column (m = 1), the inclination angle varies between 1/200 and 1/300,
whatever the length of the column. The following values for the length give rise
to the following values for the inclination angle:
Length of the vertical element (m)
1
2
4
8
12
16

i
1/200
1/200
1/200
1/282,8
1/300
1/300

For an isolated column, EN 1992-1-1:2004 allows two alternative methods to take into
account the imperfections; see figure 2.5.2-3:
the imperfection may be taken into account as an eccentricity ei caused by the
inclination: ei = i . l0 / 2, where l0 is the buckling length of the model column
(that is equivalent to the real column see further in the chapter on buckling). For
walls and isolated columns in braced systems, ei = l0 / 400 may be used as
simplification;
the imperfection may be taken into account by calculating the equivalent
horizontal force (or transverse load) Hi, applied in the most unfavourable position
2-9

(that gives maximum moment). The expressions for the transverse load are the
following (with N = axial force):
- for unbraced members: Hi = i . N (figure a1 in figure 2.5.2-3);
- for braced members: Hi = 2 . i . N (figure a2 in figure 2.5.2-3).
EN 1992-1-1:2004 recommends using the inclination method only for statically
determinate members; the equivalent transverse load method can be used for both
determinate and indeterminate members.
For structures, the inclination may be represented by transverse forces Hi to be included
in the analysis together with other actions, by means of the formulas:
braced structure (figure b in figure 2.5.2-3): Hi = i . (Nb Na) ;
floor slab (figure c1 in figure 2.5.2-3): Hi = i . (Nb + Na) / 2;
roof slab (figure c2 in figure 2.5.2-3): Hi = i . Na .
-

Figure 2.5.2-3
Examples of the effects of geometric imperfections (figure 5.1 in EN 1992-1-1: 2004)
2.5.3

Simplifications

EN 1992-1-1:2004 proposes several possibilities for simplification of the structure and

thus of the structural analysis. The most important simplifications are summarized in the
following.

2-10

Reference: EN 1992-1-1:2004; 2.3.3

In building structures, temperature and shrinkage effects may be omitted in
global analysis provided joints are incorporated at every 30m distance to
accomodate resulting deformations.
Reference: EN 1992-1-1:2004; 5.1.1 (8)
In buildings, the effects of shear and axial forces on the deformations of linear
elements and slabs may be ignored where these are likely to be less than 10% of
those due to bending.
Reference: EN 1992-1-1:2004; 5.3.2.2 (2)
Continuous slabs and beams may be analysed on the assumption that the
supports provide no rotational restraint.
Reference: EN 1992-1-1:2004; 5.3.2.2 (4)
Regardless of the method of analysis used, where a beam or slab is continuous
over a support which may be considered to provide no restraint to rotation, the
design support moment calculated on the basis of a span equal to the centre-tocentre distance between supports, may be reduced by an amount MEd as
follows (formula (5.9) in EN 1992-1-1:2004):
M Ed = FEd ,sup .t / 8

(2.5.3-1)

with
FEd,sup = design support reaction;
t
= breadth of the support.
Reference: EN 1992-1-1:2004; 5.3.2.2 (3)
Where a beam or slab is monolithic with its supports, the critical design moment
at the support should be taken equal to the moment at the face of the support.
The moment at the face of the support should not be less than 65% of the full
fixed end moment.

Note: Specific information is presented by EN 1992-1-1:2004 for the structural

analysis of beams, frames, slabs, walls, deep beams, corbels, etc. These
recommendations are discussed in further chapters in these course notes.

2.5.4

Identification of load paths in structures

2.5.4.1 A good view on load paths throughout the structure is necessary

The identification of the routes by which loads are carried through the members of a
structure to its foundations, is essential for the understanding of structural behaviour.
Load path identification is indispensable in the pre-design and design phase of a project:
it helps to define the role of each member in the overall context of the structure as well
as the important internal forces. In the case of concrete structures, load path analysis
helps to identify the position of the major reinforcement and the most loaded sections.
Some examples are taken from O'BRIEN, 1995.

2-11

2.5.4.2 Example 1: interpretation of bending moment diagram

Figure 2.5.4-1 shows a reinforced concrete frame, with articulated supports and loaded
by a horizontal force. The bending moment diagram allows deducing the scheme for the
longitudinal reinforcement.

Figure 2.5.4-1
Frame with articulated supports ; study of the effect of a horizontal load
(a) deflected shape, (b) bending moment diagram, (c) scheme with longitudinal
reinforcement
2.5.4.3 Example 2: exploitation of relative stiffness
The structure in figure 2.5.4-2 is composed of the beam ABC and the column BD; the
concentrated load Q is applied on top of the column. The question is what part of the
action Q is transmitted by the beam ABC towards the supports A and C, and what part of
the load Q is transmitted by the column BD towards the fixed support D?

2-12

Figure 2.5.4-2
Example of a structure composed of a column and a continuous beam. The question:
what construction element is responsible for the load transfer?
The problem proposed in figure 2.5.4-2 can easily be solved by the comparison of the
stiffness of both constitutive elements. The analysis of relative stiffness is (next to the
consideration of force equilibrium) a powerful method for load path identification. The
concept of relative stiffness is that applied loads in statically indeterminate structures
tend to be distributed between adjacent structural members in proportion to their relative
stiffness; the stiffer members carry the larger proportion of the load. The stiffness k of a
member can be defined as either the force Q which is required to cause a unit
displacement ( = 1) or the moment M required to cause a unit rotation ( = 1):
k=

or

k=

(2.5.4-1)

For a particular member, stiffness depends on where the force or moment is applied;
this is shown in the examples presented in figure 2.5.4-3: each figure mentions the
stiffness associated with the member and the applied load (OBRIEN, 1995).
1

2-13

Figure 2.5.4-3
Stiffness for some common arrangements of load on members (OBRIEN, 1995)
As mentioned above, the question asked in figure 2.5.4-2 can be answered by
comparison of the stiffness of the beam and the column; the stiffness can be determined
from figure 2.5.4-3:
stiffness of the beam (figure 2.5.4-3, case 6):
48.E.I 6.E.I
kb =
= 3
l
(2.l )3

stiffness of the column (figure 2.5.4-3, case 7):

A.E
kc =
h
with A = the cross section of the column.
Taking into account the assumption that all constitutive elements are characterized by
the same value of the elastic modulus E, and considering typical values for A, E, I, h and
l, one finds that kc >> kb.
For example, with the following assumptions:
a beam with rectangular section (height = 0,4 m x width = 0,3 m),
a column with square section (0,3 m x 0,3 m),
the height of the column h = 3 m,
the length of the beam 2.l = 6 m,
one finds the ratio kc / kb = 250. For practical cases, the column carries over 90% of the
load Q. The distribution of the load can be calculated by the following formulas:
2-14

 load Qb carried by bending and shear in the beam:

kb
1
Q=
Q
Qb =
1 + kc / kb
kb + kc
load Qc carried by compression in the column:
k /k
kc
Q= c b Q
Qc =
1 + kc / kb
kb + kc

2.5.4.4 Three-dimensional structures

An example of structural analysis of a three-dimensional structure, in the pre-design
phase, is presented in annex A2.5.4.4.
2.5.5

Example: structural analysis of a continuous beam

2.5.5.1 Introduction
In practice, in order to simplify calculations, three-dimensional concrete structures may
be broken up into smaller manageable two-dimensional substructures. The following
example, which is taken from OBRIEN (1995), shows that modeling and structural
analysis of concrete structures goes together with engineering judgment, because of the
different possible choices that can be made.
Figure 2.5.5-1 presents a floor system composed of a slab supported by three parallel
beams which are themselves supported by three columns. The structure is realized with
cast-in-place concrete and is thus a monolithic structure. The total load on the slab is
evaluated as a uniformly distributed load of 10 kN/m2.

Figure 2.5.5-1
Analysis of a slab-beam-column three-dimensional concrete structure (dimensions in
mm)

2-15

2.5.5.2 Analysis of the bending moments in the slab

EN 1992-1-1:2004; 5.3.1(5) stipulates that a slab subjected to dominantly uniformly
distributed loads may be considered to be one-way spanning if it possesses two free
(unsupported) and sensibly parallel edges (see also further in the chapter on slabs). The
slab in figure 2.5.5-1 may thus be considered as one-way spanning: the slab spans
between beams! Each of the three beams acts as a support to the slab and transfers the
reactions from the slab to the columns. The slab may thus be calculated by considering
strips (beams) with unit width (1m). A clear distinction can be made between strip XX
and YY, because of the presence or not of the columns. For strip XX, one may neglect
the torsional stiffness of the beams in spite of the observation that the rotation of the
cross sections in J, K and L is undoubtedly somewhat influenced by the torsional
stiffness of the beams. With that assumption, the strip may be considered as a
continuous beam simply supported by the three beams: figure 2.5.5-2 (a). The
corresponding bending moment diagram is presented in figure 2.5.5-2 (b). It can be seen
that the maximum moment in the strip occurs at point K and is equal to 43,8 kNm. Yet,
since the ends of the slab are not really simply supported, there will in fact be a small
hogging moment along the edge and, as a result, a nominal amount of top reinforcement
should be provided along the edges of the slab.
For the strip XX, mid-way between the supporting columns, the model in figure 2.5.5-2
is sufficiently accurate. However, for strips nearer the supporting columns, such as strip
YY in Figure 2.5.5-1, the resistance to rotation at points B, E and H is more significant.
If the columns were very stiff, a suitable model for strip YY would be that illustrated in
figure 2.5.5-3 (a), which leads to the bending moment diagram shown in figure 2.5.5-3
(b).
The actual bending moment diagram for strip YY will be something between those
corresponding to the two models, but for preliminary purposes, the envelope of
moments could be used for design.
It can be seen that for both hogging and sagging moments, the model with clamped
edges leads to smaller values than the model with simple supports. The results for the
simply supported model can thus be used for the design of all 1m strips XX and YY; yet,
for the YY strip, one should use the hogging moment equal to 20,8 kNm (edge ABC) and
equal to 40,8 kNm (edge GHI).

2.5.5.3 Analysis of the vertical load in column B (figure 2.5.5-1)

In order to find the maximum vertical load in column B, it is first necessary to find the
reactions from each 1m wide strip of slab (XX and YY strips).
Starting from the bending moment diagram for strip XX, shown in figure 2.5.5-2 (b),
one can easily find the reaction force on the left RXXL (L < left) (reasoning illustrated in
figure 2.5.5-4) :
5 m . R XXL + 43,8 kNm = 10 kN/m . 5 m . 2,5 m
R XXL = 16,4 kN

For the strip YY, the following reaction force on the left RXXL (L < left) is identified:

2-16

RYYL =

10 kN/m . 5 m
= 25 kN
2

The reaction transferred to beam ABC from any strip is something between 16,4 and
25 kN (for strips with 1m width). For the calculation of the maximum force in column
B, it is conservative to assume the greater of these values for all strips: the resulting load
on beam ABC is a uniformly distributed load of 25 kN/m, which then allows to
determine the reaction in column B.
If the columns were assumed to have infinite rigidity, each span (AB and BC) would be
fixed at either end and the reaction at B would equal half the load from each span:
RB = 2.

25 kN/m . 5 m
= 125 kN
2

If on the other hand, the column stiffness were ignored, the beam would be simply
supported at A, B and C and the analysis of the continuous beam ABC would lead to the
reaction force in B equal to:
RB = 156 kN

Figure 2.5.5-1
Model for strip XX in Figure 2.5.5-1;
(a) continuous beam, simply supported; (b) bending moment diagram (M in kNm)

2-17

Figure 2.5.5-2
Model for the strip YY in Figure 2.5.5-1; (a) continuous beam with all fixed supports;
(b) bending moment diagram (M in kNm)

Figure 2.5.5-3
Determination of the reaction force RXXL, starting from the bending moment diagram for
the strip XX in Figure 2.5.5-1
2.5.6

Example : structural analysis of frames

A lot of concrete buildings are composed of beams and columns which, when rigidly
connected (usually the case with concrete, unless a precast building system is used)
make up a continuous frame. In annex A2.5.6. to these course notes, one can find an
example of structural analysis of a three-dimensional framed building ; the example is
taken from OBRIEN (1995). Today, user-friendly commercial software is available for
the structural analysis of such types of structures. However, the annex A2.5.6. shows an

2-18

example of modeling and simplification with the aim of obtaining approximate results
permitting the verification of numerical results.
For approximate calculations, assumptions and simplifications are needed, for example
in order to reduce the number of combinations of actions that has to be considered for a
complex structure. The greater the number of members in the frame, the greater the
number of possible combinations of applied load that have to be considered.
A common simplification is to represent the three-dimensional frame with smaller, twodimensional sub-frames. This is highly determined by the eventual presence of bracing
against horizontal loads. A frame which is braced against horizontal loads using
substantial bracing members such as cores and shear walls is termed a non-sway
frame. Such a frame is essentially designed to resist the applied vertical loads, while the
bracing members are designed to carry the horizontal loads. A sway frame is
characterized by the absence of bracing members; in that case, the horizontal loads are
transmitted to the foundations by frame action.
The fully worked out example is presented in annex A2.5.6.

2-19

3 Chapter 3
Partial factor method strength of materials
3.1 Aim of chapter 3
This chapter focuses on the left side of figure 1.3.2-1, and discusses in particular:
the strength properties of materials and the determination of the characteristic
values fk and the design values fd;
the design characteristics of materials with, in particular the constitutive equations
to be used in design calculations. HOOKE's law is applied in the elastic design
method. But in the partial factor design method, non linear relationships are used
for steel and concrete. The effects of the use of the constitutive equations will
become clear in chapter 4, but the principle idea may already be formulated: an
ULS is characterized by a specific strain distribution in the cross-section, where
the ultimate deformation capacity of the materials is reached; the strains
correspond to stresses according to the constitutive equations; on the basis of this
stress distribution, and by application of the equilibrium equations (translation,
rotation), the resisting (stabilizing) internal forces are determined, which are then
compared to the applied (destabilizing) internal forces. By means of the same
reasoning steps, but in the inverse order, a cross-section can be designed in
function of the applied internal forces and thus in function of the applied loads.

3.2 Characteristic strength

The properties of materials are usually described by their characteristic values. These
values are determined on the basis of a sufficient number of tests performed on a
homogeneous sample of the material. They correspond to a prescribed probability P that
this property will not be not reached. The prescriptions regarding the taking of samples
and the realization of the tests are defined in specific standards. It is usually assumed
that the statistical distribution of the strength of the material corresponds to a normal
curve.
The characteristic value fk is defined by the equation:
fk = fm kf . sf = fm . (1 kf . f)

(3.2-1)

where
fm
sf

= mean value of the strength, as a result of the tests;

= standard deviation of the strength resulting from the tests, equal to:

( f

fm )2

n 1

f
kf

= variation coefficient: f = sf / fm
= number that is function of the chosen probability P; in general, P = 5%
which leads to kf = 1,64.

3-1

If there are not enough statistical results available, the nominal value defined by
standards and guaranteed by the manufacturer, may be considered as characteristic
value.
In the context of reinforced concrete, this leads to the following:
for steel, the characteristic value is chosen to be equal to the yield strength, the
real one or the conventional one (0,2% offset); the used symbol is fyk (y < to
yield). EN 1992-1-1:2004 accepts three levels of steel strengths:
S400 : fyk = 400 MPa
S500 : fyk = 500 MPa
S600 : fyk = 600 MPa
for concrete, the characteristic compressive strength fck (c < concrete) is the value
of the strength for which it is expected that only 5% of the measurements
obtained from a range of compression tests will be inferior to this value. The
compression tests are performed on cylindrical specimens, with diameter =
150mm and height = 300mm, at the age of 28 days; before the tests, the
specimens are kept in strict preservation conditions: under water at a temperature
of 20C. fck is thus defined by:
fck = fcm . (1 1,64 . )
with
fcm

(3.2-2)

= mean value of the strength measured on cylinders;

= variation coefficient. For ordinary concrete that is produced on a
construction site (cast-in-place), the value = 15% is usually chosen,
resulting in:
fck = 0,754 fcm

(3.2-3)

For concrete with a better quality (precast structures), = 0,10 is more

appropriate, leading to:
fck = 0,836 fcm

Note 1: The compressive strength of concrete depends on the type of cement

used, on the temperature, on the conditions of hardening and on the time. EN
1992-1-1:2004; 3.1.2 (5) and (6) indicate that it can be useful to know the
characteristic strength at different times t corresponding to important phases in
the life of a structure: the stripping of the formwork, the transfer of prestress
force, etc. The following formulas are available (for a mean temperature of 20C
and curing conditions in accordance with EN 12390):
fck (t) = fcm (t) 8 (MPa)
fck (t) = fck

for 3 < t < 28 days

for t 28 days

3-2

f cm (t ) = e

s 1 28 / t

. f cm

with
fcm =

s=

mean compressive strength after 28 days (corresponding to the

strength classes presented in table 3.1 in EN 1992-1-1:2004 see
further in these course notes).
coefficient depending on the type of cement:
s = 0,20 for cement of strength classes CEM 42,5R, 52,5N and
52,5R (Class R)
s = 0,25 for cement of strength classes CEM 32,5R and 42,5N
(Class N)
s = 0,38 for cement of strength class CEM 32,5N (Class S)

Note 2: Reminder of some conversion formulas:

fc cub 150
=
fc cub 200 . 1,05
=
fc cyl 300 150 . 1,265
=
fc cyl 100 113 . 0,93
fc cyl 300 150

=
=
=

fc cub 200 . 0,830

fc cub 150 . 0,791
fc cyl 100 113 . 0,736

3.3 Design strength

The partial factors m to be applied to the characteristic value of the strength of
materials, are presented in EN 1992-1-1:2004; 2.4.2.4, in function of the type of
materials and of the design situation; see table 3.3-1.
Table 3.3-1
Partial factors m for materials
(Table 2.1N from EN 1992-1-1:2004)

m for ultimate limit states ULS

Materials

c for concrete

s for reinforcing

s for prestressing

Persistent and transient

(P/T)

1,5

1,15

1,15

Accidental (A)(*)

1,2

1,0

1,0

Design situations

steel

steel

(*) NBN EN 1992-1-2-ANB : fir fire, partial factors should be = 1,0

m for serviceability limit states SLS

P/T and A

1,0

1,0

3-3

1,0

Note 1: In previous versions of Eurocode 2, different values of m were applied

in function of the control method used during the manufacturing: a distinction
was for example made between concrete mixed on site or in factory. The
increased introduction of quality verification standards (such as the BENOR
certification) has made an end to the use of the so called self-controlled
materials (at least for concrete).

Note 2: EN 1992-1-1:2004; 2.4.2.4 (3) specifies that smaller values for c and s
may be used if measures are taken to reduce uncertainties concerning the
material properties.

This leads to the following values of design strength:

for reinforcing steel
for ULS: s = 1,15; the design value of the yield strength, as well in tension
as in compression, is:
f yd =

f yk
1,15

(3.3-1)

for SLS: s = 1,0

for concrete
for compressive strength in ULS: c = 1,5.
EN 1992-1-1:2004; 3.1.6 (1) defines the design value of the compressive
strength of concrete as follows:
fcd = cc . fck / c

(3.3-2)

with
c = partial factor for concrete = 1,5;
cc = coefficient which takes into account:
- long term effects on the compressive strength: the creep-effect,
as mentioned in the diagrams obtained by RSH (see course on
"technology of concrete");
- unfavourable effects resulting from the way the load is
applied, as mentioned in EN 1992-1-1:2004; 3.1.6 (1). This
wants to take into account that the largest compressive stresses
due to bending frequently occur at the upper level of a crosssection, where the quality of concrete is in general lower
because of possible segregation during compaction.
EN 1992-1-1:2004; 3.1.6 (1) indicates that cc must have a value between
0,8 and 1, but recommends to use the value 1.
In Belgium, the National Annex (NBN EN 1992-1-1 ANB) recommends
the use of the value 0,85 for verification in ULS for axial loads, bending
and combined axial force with bending; for other loading types (shear and
3-4

torsion), cc = 1 should be used. This means in practice that for

calculations in accordance with NBN EN 1992-1-1 ANB, the following
design values have to be used for the compressive strength of concrete:
for ULS design of the main reinforcement (thus for normal stresses
due to axial loads and bending moments):
fcd = 0,85 . fck / 1,5
(3.3-3)
for ULS design of shear reinforcement (necessary to take up shear
loads and torsion):
fcd = fck / 1,5
(3.3-4)
for compressive strength in SLS: c = 1,0 and cc = 1; the characteristic
value is used;
for tensile strength in ULS, the design tensile strength fctd is defined as:
fcd = ct . fctk,0,05 / c
(3.3-5)
with c = 1,5 and ct = 1.

3.4 Material properties for concrete to be used in design calculations:

design assumptions
3.4.1

General

The following models regarding the properties of materials are idealizations adapted to
the specific needs of the design of structures in normal and high strength concrete. The
models are not applicable to lightweight aggregate concrete, for which the designer is
referred to EN 1992-1-1:2004; section 11.
3.4.2

Physical properties

Density
- = 2400 kg/m3 for plain concrete (that is unreinforced concrete);
- = 2500 kg/m3 for reinforced concrete with normal percentages of
reinforcement.
Thermal expansion coefficient
- If the thermal expansion does not have major consequences on the structural
behaviour, the thermal expansion coefficient can be taken equal to 10.10-6 /C.
- If the thermal expansion has a big influence and if the nature of the used
aggregates is known, the thermal expansion coefficient between 0C and 150C
can be taken equal to :
- 12.10-6 /C for concrete made with gravel;
- 10.10-6 /C for concrete made with crushed porphyry (igneous rock);
- 8.10-6 /C for concrete made with limestone aggregates.
3.4.3

Mechanical properties

Reference: EN 1992-1-1:2004; 3.1

3.4.3.1 Strength classes
Table 3.4.3-1 presents an overview of the mechanical properties of the different strength

3-5

classes that should be taken into account for design purposes. Concrete is specified in
terms of the 28-day characteristic strength. C30/37 concrete has a characteristic cylinder
strength of 30 MPa and a characteristic cube strength of 37 MPa; it is to be observed
that there is some rounding off in the values.
In earlier versions of Eurocode 2, strength classes were limited to the class C50/60. EN
1992-1-1:2004 also takes into account the high strength concrete classes, in accordance
with the standard EN 206-1:2001 "Concrete: specification, performance, production
and conformity".

Notes:
- EN 206-1:2001 allows the use of class C100/115, while EN 1992-11:2004 is limited to class C90/105;
- some codes of practice allow the concrete strength to be used in design
to be varied according to the age of the concrete when the design loads
are applied. EN 1992-1-1:2004 does not accept this: design calculations
are based on the 28-day value.

3-6

(0/00) 3,50 3,50 3,50 3,50 3,50 3,50 3,50 3,50 3,50 3,13 2,88 2,66 2,60 2,60

cu3

6,6

(0/00) 1,75 1,75 1,75 1,75 1,75 1,75 1,75 1,75 1,75 1,82 1,89 2,03 2,16 2,30

6,3

3,5

c3

6,0

3,4

5,0

98

105

90

2,00 2,00 2,00 2,00 2,00 2,00 2,00 2,00 2,00 1,75 1,59 1,44 1,40 1,40

5,7

3,2

4,8

88

95

80

5,5

3,0

4,6

78

85

70

(0/00) 3,50 3,50 3,50 3,50 3,50 3,50 3,50 3,50 3,50 3,13 2,88 2,66 2,60 2,60

5,3

3,0

4,4

68

75

60

cu2

4,9

2,9

4,2

63

67

55

(0/00) 2,00 2,00 2,00 2,00 2,00 2,00 2,00 2,00 2,00 2,20 2,29 2,42 2,52 2,60

4,6

2,7

4,1

58

60

50

c2

4,2

2,5

3,8

53

55

45

(0/00) 3,50 3,50 3,50 3,50 3,50 3,50 3,50 3,50 3,50 3,21 3,02 2,84 2,80 2,80

3,8

2,2

3,5

48

50

40

cu1

3,3

2,0

3,2

43

45

35

(0/00) 1,77 1,87 1,97 2,07 2,16 2,25 2,32 2,40 2,46 2,53 2,59 2,70 2,80 2,80

2,9

1,8

2,9

38

37

30

c1

2,5

fctk,0.95 (MPa) 2,0

1,5

2,6

33

30

25

(GPa) 27,1 28,6 30,0 31,5 32,8 34,1 35,2 36,3 37,3 38,2 39,1 40,7 42,2 43,6

1,3

fctk,0.05 (MPa) 1,1

2,2

28

25

20

Ecm

1,9

(MPa) 1,6

24

fctm

20

20

(MPa)

15

fck,cube (MPa)

16

fcm

12

(MPa)

fck

Strength classes for concrete

Table 3.4.3-1
Strength classes and mechanical properties of concrete
(Table 3.1 from EN 1992-1-1:2004)

The following relationships go together with table 3.4.3-1:

fcm = fck + 8 (MPa)

fctm = 0,30 . fck2 /3

fctm = 2,12 . ln (1 + fcm/10)

fctk,0.05 = 0,7 . fctm

fctk,0.95 = 1,3 . fctm

Ecm = 22 . (fcm/10)0,3 with fcm in MPa

for strength classes C50/60

for strength classes > C50/60

5% fractal
95% fractal

3-7

c1 and cu1 are used in the idealized stress-strain diagram proposed by EN 19921-1:2004 for non-linear structural analysis (see further in paragraph 3.4.3.2).
c2 , cu2 and n are used in the parabola-rectangle stress-strain diagram proposed
by EN 1992-1-1:2004 for the design of cross-sections (see further in paragraph
3.4.3.3).
c3 and cu3 are used in the bi-linear stress-strain diagram proposed by EN 19921-1:2004 as an alternative for the parabola-rectangle diagram for the design of
cross-sections (see further in paragraph 3.4.3.4).

3.4.3.2 Stress-strain diagram to be used in non-linear structural analysis

Figure 3.4.3-1 shows the idealized stress-strain diagram c c that should be used for
non-linear structural analysis. The curve corresponds to the equation:

c
f cm

k 2
1 + (k 2)

( 0 < | c | < | cu1 | )

with

= c / c1
k = 1,05 . Ecm .| c1 | / fcm
and
c1 () = 0,7 . (fcm)0,31 < 2,8
cu1 () = 2,8 + 27 . (98-fcm)/1004 for fck 50 MPa
Figure 3.4.3-2 shows the idealized stress-strain diagrams for the different strength
classes of concrete.

Figure 3.4.3-1
Idealized stress-strain diagram to be used for non-linear structural analysis; the limit
value 0,4 . fcm is used in the definition of the secant modulus Ecm (Figure 3.2 in EN
1992-1-1:2004)

3-8

 c [Mpa]
100

C90/105
C80/95

80

C70/85
C60/75
C55/67

60

C50/60
C45/55
C40/50

40

C35/45
C30/37
C25/30

20

C20/25
C16/20
C12/15

-3

0,0

0,5

1,0

1,5

2,0

2,5

3,0

3,5

c [10 ]

Figure 3.4.3-2
Idealized stress-strain diagrams for the different concrete strength classes
3.4.3.3 Stress-strain diagram to be used for the design of cross-sections: the parabolarectangle diagram
Figure 3.4.3-3 shows the parabola-rectangle stress-strain diagram that should be used
for the design of cross-sections. The parabolic part of the diagram corresponds to the
equation:
n

c
for 0 c c 2
c = f cd 1 1
c 2

with
for fck < 50 MPa

for fck 50 MPa

c2 () = 2,0

c2 () = 2,0 + 0,085 . ( fck - 50 )0,53

cu2 () = 3,5

cu2 () = 2,6 + 35 . ( 90 - fck )/1004

n = 2,0

n = 1,4 + 23,4 . ( 90 - fck )/1004

fcd = cc . fck / c ( = 0,85 . fck / 1,5 according to NBN EN 1992-1-1:2009 ANB

for the ULS design of the main reinforcement).
For the rectangular part of the diagram, the equation is:

c = f cd

for c 2 c cu 2

Figure 3.4.3-4 shows the parabola-rectangle diagrams for the different concrete strength
classes.

3-9

Figure 3.4.3-3
Parabola-rectangle diagram of concrete in compression, to be used for the design of
cross-sections (Figure 3.3 in EN 1992-1-1:2004)

c [Mpa]
100
C90/105
C80/95

80

C70/85
C60/75

60

C55/67
C50/60
C45/55
C40/50

40

C35/45
C30/37
C25/30
C20/25

20

C16/20
C12/15

-3

0,0

0,5

1,0

1,5

2,0

2,5

3,0

3,5

c [10 ]

Figure 3.4.3-4
Parabola-rectangle diagrams for concrete in compression, corresponding to the different
concrete strength classes
3.4.3.4 Alternative stress-strain diagram that can also be used for the design of crosssections: the bi-linear diagram
Figure 3.4.3-5 shows the bi-linear stress-strain diagram c c, proposed as a valuable
alternative for the parabola-rectangle diagram, which can also be used for the design of
cross-sections. The bi-linear diagram is characterized by the following equations:
for fck < 50 MPa

for fck 50 MPa

3-10

c3 () = 1,75

c3 () = 1,75 + 0,55 . ( fck 50 )/40

cu3 () = 3,5

cu3 () = 2,6 + 35 . ( 90 - fck )/1004

Figure 3.4.3-5
Bi-linear stress-strain diagram for concrete in compression, which can also be used for
the design of cross-sections as an alternative for to the parabola-rectangle diagram
(Figure 3.4 in EN 1992-1-1:2004)
3.4.3.5 Second alternative stress-strain diagram that can also be used for the design of
cross-sections: the rectangular stress distribution
EN 1992-1-1:2004 allows the use of a rectangular stress distribution as an alternative
for the parabola-rectangle and bi-linear diagrams: see figure 3.4.3-6. The parameter ,
defining the effective height of the compression zone, and the parameter , defining the
effective strength, follow from:
for fck < 50 MPa

for 50 < fck 90 MPa

= 0,8

= 0,8 ( fck 50 )/400

= 1,0

= 1,0 ( fck 50 )/200

Note: If the width of the compression zone decreases in the direction of the
extreme compression fibre, the value . fcd should be reduced by 10%. In the
previous version of the standard, this reduction was only 5%.

3-11

Figure 3.4.3-6
Rectangular stress distribution for concrete in compression, which can also be used for
design of cross-sections, as an alternative an alternative for the parabola-rectangle and
bi-linear diagrams (Figure 3.5 in EN 1992-1-1:2004)
3.4.3.6 Remark about the idealized stress-strain diagrams
The earlier presented stress-strain diagrams, as well as the properties mentioned in table
3.4.3-1, are valid for concrete with an age of minimum 28 days. The mean compressive
strength at a different age (t < 28 days) can be determined by means of the formulas
mentioned in paragraph 3.2; one can also use experimental values.
3.4.3.7 Tensile strength
As was already mentioned before, EN 1992-1-1:2004; 3.1.6 (2) defines the design
tensile strength (under uni-axial tension) by fctd = ct . fctk,0.05 / c where ct takes into
account the long term effects; both EC2 and ANB (Belgium) consider the value ct = 1.
Formulas are available to determine the value of the tensile strength at different ages;
see EN 1992-1-1:2004; 3.1.2 (9).
EN 1992-1-1:2004; 3.1.8 defines the flexural tensile strength by the following
expression (where h = height of the cross-section):

h
f ctm, fl = max 1,6
f ctm
1000

f ctm

3.4.3.8 Modulus of elasticity

The modulus of elasticity is not only function of the strength class of concrete, but
depends also on the properties of the used aggregates. If the real value of the modulus of
elasticity is not known, or if high accuracy is not required, table 3.4.3-1 can be used to
estimate the value of the mean secant modulus Ecm (the secant modulus is defined in
figure 3.4.3-1). The values of Ecm in table 3.4.3-1 are specific for concrete made with
quartz aggregates; if other aggregates are used, the following modifications are
recommended:
- limestone 10% reduction of the value of Ecm
- sandstone 30% reduction of the value of Ecm
- basalt 20% increase of the value of Ecm
- porphyry 10% increase of the value of Ecm (prescription of Belgian ANB).
Formulas are available to determine the modulus of elasticity at other ages than 28 days;
3-12

see EN 1992-1-1:2004; 3.1.3 (3).

3.4.3.9 Poisson's ratio
Reference: EN 1992-1-1:2004; 3.1.3(4)
Poisson's ratio may be taken equal to 0,2 in the case of elastic deformations.
If cracking is allowed in concrete under tension, Poisson's ratio may be taken
equal to 0.
3.4.3.10 Creep and shrinkage
These properties will be discussed in the chapter on deformations of concrete beams
(see further in these course notes).

3.5 Design values for reinforcing steel

Reference: EN 1992-1-1:2004; 3.2 and annex C.
EN 1992-1-1:2004; 3.2 refers to the standard EN 10080 for the definition of the
properties of the steel types that can be used in reinforced concrete; If the Belgian ANB
is applied, the standards NBN A24-301:1986 to NBN A24-304:1986 should be used.
All the properties regarding reinforcing steel that are important for the design are
summarized in the Annex C to EN 1992-1-1:2004; although this is just an annex to
EC2, it still is valid as standardisation document.
3.5.1

General

The properties detailed in EN 1992-1-1:2004 ; annex C are valid for temperatures

between -40C and 100C (instead of 200C, as was mentioned in the older versions of
the standard; the reason for the reduced limit temperature is the increased use of thermal
treatments during the production process of steel).
The standard only refers to ribbed bars (also called bars with improved adherence
because of the additional mechanical bond with the concrete) and to weldable bars. The
standard pays much attention to the bendability (in order to avoid tension cracking of
the steel, the radius of the bend should not be less than a critical value) and to the
weldability of the reinforcement; EN 1992-1-1:2004; 3.2.5 discusses some welding
methods.
3.5.2

Different types of reinforcing steel and main mechanical properties

3.5.2.1 Production and types of reinforcement

Steel quality is determined by the chemical composition and by treatment after rolling.
In general, steel is composed of
iron;
various impurities, such as phosphorus (P), sulphur (S) and nitrogen(N);
alloying components, with
o carbon (C), as the main non-metallic component;
o a whole series of ferro-alloying components such as manganese (Mn),
silicon (Si), chromium (Cr), etc.
Depending on the content of alloying components, steel types are divided into nonalloyed, low-alloyed and alloyed steels. All types of steel for ordinary reinforcement
contain in principle no substantial amount of alloying components (less than 2,5% of

3-13

alloying components), hence they are non-alloyed (FIB, 2009). For prestressing
reinforcement, non-alloyed and low alloyed steel is used.
The mechanical properties of steel, such as strength, deformability, weldability and
suitability to heat treatment depend nearly entirely on the carbon content, which is
usually:
0,15-0,20% for ordinary reinforcing steel (which leads to tensile strengths up
to 600 MPa);
0,50-0,80% for prestressing steel (which leads to tensile strengths up to 1800
MPa).
Due to the high content of carbon, prestressing steel is not weldable in contrast to
ordinary reinforcement.
The first steps in the production process are identical for all steel reinforcement types:
steel making, casting of semi-finished products and hot rolling. During the hot rolling
process, heated steel is squeezed when it passes through a set of steel rolls. The rods are
often coiled up before they are used for drawing into wire or for fabricating into bars.
The production routine for non-weldable steel is hot rolling without additional
treatment; the production of weldable steel can be executed by three process routes:
o hot rolling after micro-alloying;
o hot rolling followed by heat treating of the steel in solid state; a well
known example is the Termpcore process which allows to produce very
ductile steel with excellent weldability without adding of expensive
alloying components;
o hot rolling followed by cold-forming. This process leads to increase of
the yield strength with reduction of ductility. Stretching, drawing,
twisting and cold-rolling are the most common methods of cold-forming.
The main types of products used to reinforce concrete are:
bars; the European standard EN 10080 provides the following scale: 6,8,10,
12, 14, 16, 20, 25, 28, 32 and 40 mm. Bars can be produced as plain (not
recommended in the standard), indented or ribbed;
decoiled rods and wires; coils are produced with rebar sizes between 6 and
16 mm and with wire sizes between 4 and 12 mm;
welded wire fabrics; is a common type of industrially fabricated twodimensional reinforcement which is suitable for reinforcing slabs;
lattice girders; this is a two or three-dimensional metallic structure
comprising an upper chord, a lower chord and continuous or discontinuous
diagonals, which are welded or mechanically assembled to the chords.
Lattice girders are often used to pre-manufacture precast lattice girder plates
for constructing floor slabs.
3.5.2.2 Main mechanical properties for design purposes
In older standards on reinforced concrete, grade 220 (fyk = 220 MPa) was commonly
used for hot-rolled mild steel bars; these bars usually had a smooth surface (bond with
the concrete only by adhesion). This grade is not mentioned anymore in the actual
Eurocode 2, which only considers high yield bars manufactured with a ribbed surface.
Distinction is made between (see figure 3.5.2-1):

3-14

hot rolled high yield steel, for which the limit of the elastic domain (the yield
stress) is characterized by the presence of a real threshold. Designation: BE500
indicates hot rolled steel with characteristic yield value fyk = 500 MPa;
cold worked high yield steel, for which the yield stress is conventionally defined
for a residual elongation of 0,2% (the 0,2% offset method). Designation: DE500 is
a cold worked steel with a characteristic value of the yield stress equal to
fyk = f0,2k = 500 MPa.

Figure 3.5.2-1
Sketch of the stress-strain diagram of reinforcing steel: hot rolled steel (designation:
BE) and cold worked steel (designation: DE)
(Figure 3.7 from EN 1992-1-1:2004)
EN 1992-1-1:2004 only considers steel with a yield strength fyk between 400 and
600 MPa. ANB (Belgium) prescribes a maximum value equal to fykmax = 500 MPa.

Note: It was already mentioned that steel grade 220 is not used anymore for
design purposes. It can still be found on a construction site as auxiliary bars for
lifting purposes. However, as grade 220 was used in the 20th century, engineers
will still be confronted with it in the context of maintenance, renovation and
reassignment of buildings. This is the reason why hot-rolled mild steel 220 MPa
is still mentioned in the rest of these course notes and appears in different tables.

Conclusion: the following grades of steel are considered in these course notes: BE220,
BE400, BE500, BE600 and DE400, DE500 and DE600.
3.5.3

Ductility

Steel for rebars is not only characterized by its yield strength, but also by its tensile
strength ftk and by the associated value of the strain at maximum force uk (u < ultimate):
see figure 3.5.2-1.
EN 1992-1-1:2004 considers three ductility classes which are characterized by the ratio
k = (ft / fy)k and by the characteristic strain at maximum force uk :

3-15

class A : uk 2,5 % and k 1,05; this class is associated with small diameter (
12mm) cold-worked bars used in mesh (or fabric). This is the lowest ductility
category; application of moment redistribution is limited (see further in these
course notes, in the chapter on plastic methods);
class B : uk 5 % and k 1,08; this class is commonly used for reinforcing bars;
class C uk 7,5 % and 1,15 k < 1,35; high ductility steel, which is used in
earthquake design or similar situations.
Figure 3.5.3-1 shows in a schematic way the stress-strain diagram of hot rolled high
yield steel BE500, for the three ductility classes; figure 3.5.3-2 presents the same kind
of diagram for cold worked high yield steel.

700

(Mpa)

Hot rolled steel

600
500

class A

class B

class C

400
300
200
100
0
0

2,5

7,5

Figure 3.5.3-1
Stress-strain diagram of hot rolled steel BE500
for the three ductility classes

3-16

10

(%)

 (Mpa)

700

Cold Worked steel

600
500

class A

class B

class C

400
300
200
100
0
0

2,5

7,5

10

(%)

Figure 3.5.3-2
Stress-strain diagram of cold worked steel DE500
for the three ductility classes

3.5.4

Physical properties

density (mass per unit volume): = 7850 kg/m3

thermal expansion coefficient: = 10 . 10-6 /C

3.5.5

Design assumptions

3.5.5.1 Idealized stress-strain diagram to be used for the design of cross sections
Figure 3.5.5-1 shows the idealized stress-strain diagram to be used for design
calculations.

Figure 3.5.5-1
Idealized stress-strain diagram of reinforcing steel, for tension and compression

3-17

(figure 3.8 from EN 1992-1-1:2004).

EN 1992-1-1:2004; 3.2.7 (2) adds to the former the following specifications for normal
design:
the design value of the ultimate strain ud can be put equal to 0,9 . uk , if the
inclined top branch in figure 3.5.5-1 is used;
if the horizontal top branch is used, the strain in the steel should not be limited.

Note: taking into account the horizontal branch instead of the inclined top branch
corresponds to a conservative approach, in particular when cold worked steel is
used. Furthermore, it should be noted that modern steel grades are more and
more characterized by a very slowly climbing plastic branch in the tensile curve.

The Belgian ANB is more restrictive about this and introduces additional limitations;
one may choose between:
the limitation of the design value of the ultimate strain ud to 0,8 . uk , for both the
horizontal and the inclined branches. For example, in the case of steel in class A
with uk = 2,5 %, this corresponds to an ultimate strain of 2% ;
the limitation of the ultimate strain ud to 1%.

Note: In many European countries, it is common practice to limit the design

value ud of the ultimate strain in steel to 1%; it is assumed that when this strain
is reached, the concrete is already heavily cracked. Some countries, such as
Germany, traditionally permit larger strains in design calculations of reinforced
concrete, which leads to smaller amounts of reinforcement (this can be verified
by the students after the study of chapter 4).

3.5.5.2 Modulus of elasticity

The modulus of elasticity is equal to Es = 200.000 MPa = 200 GPa.
3.5.5.3 Conclusion: simplified stress-strain diagram for reinforcing steel, to be used in
the context of this course
Figure 3.5.5-2 shows the idealized stress-strain diagram for reinforcing steel that is used
for the design of cross-sections in the following of these course notes.

fyd = fyk / s

Es = 200 GPa

s
10 0/00

3-18

Figure 3.5.5-2
Idealized stress-strain diagram for reinforcing steel, that is used in the following text of
these course notes

3-19

Part II
BEAMS IN REINFORCED CONCRETE
4 Chapter 4
Design of cross-sections at ULS,
for bending and axial force
4.1 Introduction: practical elaboration of the principal idea of the
partial factor method overview of basic assumptions
4.1.1

Practical working-out of the principal idea

The aim of this chapter is to design cross-sections that withstand the imposed internal
forces in the context of an ULS. It is thus necessary to develop relationships,
specifically for the chosen ULS, between geometrical parameters and internal forces,
taking into account the material properties. The following method is used to develop the
relationships (this is the scheme that will be applied in all following paragraphs in this
chapter):
step 1: choice of an ULS for the selected cross-section; the ULS is defined by a
specific strain distribution in the cross-section for which the ultimate deformation
capacity of steel and/or concrete is reached;
step 2: determination of the corresponding stress distribution in the cross-section
by application of the simplified design constitutive relationships;
step 3: expression of static equilibrium (translation, rotation) permits to determine
the stabilizing (resisting) internal forces, which have to be compared with the
destabilizing (imposed) internal forces.
Chapter 4 is limited to the resistance to bending and axial forces, and is thus limited to
the problem of normal stresses in the cross-section (shear stress is considered in the
chapter on design for shear loads and torsion). The aim of the chapter is to design a
cross-section (that is to determine the necessary surface of concrete and percentage of
reinforcement) to resist to imposed bending moments and axial forces, and thus to a
distribution of normal stress.
4.1.2

Basic assumptions

Reference: EN 1992-1-1:2004; 6.1

When determining the ultimate moment resistance of reinforced concrete cross-sections,
the following assumptions are made:
1. Plane sections remain plane; the assumption of BERNOULLI is applied. This
assumption is not valid for short and compact construction elements and for
discontinuity regions (zones nearby concentrated loads or discontinuous changes
in dimensions); adapted design methods are needed here (plastic methods, strut
and tie method).
4-1

2. The strain in bonded reinforcement, whether in tension or in compression, is the

same as that in the surrounding concrete: c = s.
3. The tensile strength of concrete is ignored.
4. The stresses in the concrete in compression are derived from the (simplified)
design stress-strain relationships discussed in chapter 3: the parabola-rectangle
diagram, the bi-linear diagram or the rectangular stress distribution (see figures
3.4.3-3, 3.4.3-5 and 3.4.3-6 in chapter 3 in these course notes).
5. The stresses in the reinforcing steel are derived from the (simplified) design curve
discussed in chapter 3: the bi-linear stress-strain relationship (see figures 3.5.5-1
and 3.5.5-2 in chapter 3 in these course notes).
6. The compressive strain in the concrete shall be limited to cu2 or cu3 depending on
the stress-strain diagram used. With the parabola-rectangle diagram, the strain is
limited to cu2. This means that for sections that are not loaded in pure
compression, but in simple or combined bending, the limit value cu2 has to be
respected (- 0,0035 for the concrete classes up to C50/60).
7. For sections which are subjected to pure compression or to approximately
concentric loading (e/h < 0,1), the compressive strain should be limited to c2 or
c3 depending on the stress-strain diagram used. In practice, with the parabolarectangle diagram, the strain in compression is limited to c2 (- 0,002 for the
concrete classes up to C50/60).
8. For the intermediate cases between pure compression and combined bending
(compression + bending), the ULS strain diagram is defined starting from the
strain c2 at the distance (cu2 - c2) . h / cu2 from the extreme compression fibre of
the section (see figure 4.1.3-1). For concrete classes up to C50/60, the strain is 0,002 at the distance equal to 3/7th of the depth, from the extreme compression
fibre of the section.
9. The ultimate strain of the steel reinforcement is ud (according to EN 1992-11:2004, 3.3.6(7) only when the inclined branch of the diagram is used). The
Belgian ANB imposes ud also when the horizontal branch is used. Moreover, the
Belgian ANB specifies also that the strain in the steel reinforcement may be
limited to 1%.
4.1.3

Possible strain distributions in ULS

Starting from the basic assumptions, a series of ULS strain distributions can be
identified: see figure 4.1.3-1.
Meaning of symbols used in figure 4.1.3-1
h
depth of the cross-section;
d
is called the "effective" depth of the cross-section; the effective depth is
defined as the depth from the extreme compression fibre of the section to the
centre of gravity of the tension steel;
As1 cross sectional area of steel reinforcement in the tensile zone (which normally
occurs when bending is considered tension reinforcement);
As2 cross sectional area of steel reinforcement in the compression zone (which
normally occurs when bending is considered compression reinforcement);
s1 strain in the tension reinforcement;
s2 strain in the compression reinforcement;

4-2

d1
d2
c

distance of the centre of gravity of the tension steel towards the nearest-by
concrete surface;
distance of the centre of gravity of the compression steel towards the nearestby concrete surface;
compression strain in concrete.
section

c2 cu2

d2

h1 c 2
cu 2

As2
1
d

h
2

As1
d1

ud
Elongation
Tensile strain

Compression strain

Figure 4.1.3-1
Strain distributions in ULS (Reference: figure 6.1 in EN 1992-1-1:2004)
Discussion of figure 4.1.3-1
The ultimate values of the strains to be respected for the different loading situations,
lead to the identification of three domains in figure 4.1.3-1; the limits between the
domains are defined by particular positions of the lines that materialize the strain
distributions:
Domain 1: is characterized by the tensile strain of the steel reinforcement equal to
ud ; these strain distributions may occur for the following load cases: pure
tension, tension with small eccentricity, simple bending and combined bending,
without full exploitation of the deformation capacity of the concrete.
Domain 2: is characterized by the ultimate compression strain of concrete equal
to cu2; these strain distributions may occur for the following load cases: simple
bending or combined bending with full exploitation of the deformation capacity of
the concrete.
Domain 3: is characterized by the ultimate compression strain of concrete varying
between cu2 and c2; these strain distributions may occur for the following load
cases: compression with small eccentricity and axial compression.
In each of the three domains, one can consider several sub-domains: see figure 4.1.3-2.
This figure is translated into figure 4.1.3-3 for the concrete classes up to C50/60, with
cu2 = - 0,0035; c2 = - 0,002 and ud = 0,010.

4-3

c2 cu2
D

O'
d

cu 2
cu 2 + ud

h1 c 2
cu 2

1b

1a

C(c2)

2a

2b

ud

c2

Figure 4.1.3-2
ULS strain diagrams; detailed elaboration of figure 4.1.3-1

2 3,5
O' D B

3
h
7

0,259d
1b
1a

C(2)

2a

2b

ud

E
2

Figure 4.1.3-3
ULS strain diagrams for the concrete classes up to C50/60, with cu2 = - 0,0035; c2 = 0,002 and ud = 0,010.
Discussion of figure 4.1.3-2
Domain 1: the ULS strain diagrams turn around point A which corresponds to the
ultimate tension strain in the steel reinforcement, equal to ud (1% in figure 4.1.3-3).
Two sub-domains can be identified:
Domain 1a: tensile load with small eccentricity. The whole cross-section is
loaded in tension.
Domain 1b: simple bending or combined bending without full exploitation
of the deformation capacity of the concrete.

4-4

Domain 2: the ULS strain diagrams turn around point B which corresponds to the
ultimate compression strain in the concrete, equal to cu2 (0,35% in figure 4.1.3-3). Two
sub-domains can be identified:
Domain 2a: simple or combined bending; the tension strain in the steel
reinforcement is in between 0 and ud ; this domain is characterized by the
full exploitation of the deformation capacity of the concrete.
Domain 2b: combined bending; all reinforcements are in compression. Only
a small part of the cross-section is still in tension. This domain is
characterized by the full exploitation of the deformation capacity of the
concrete.
Domain 3: the ULS strain diagrams turn around point C. In this domain, the whole
cross-section is in compression. Point C is identified by the intersection of the line BO
(which corresponds to the passage from a partially tended section towards a fully
compressed section) with the line DE (which corresponds to the strain diagram in axial
compression). The distance of C towards the extreme compression fibre is equal to (1
c2 / cu2 ) . h (that is 3/7th of the total depth of the section in figure 4.1.3-3).
4.1.4

Conclusions

All load cases will be discussed in the following paragraphs of chapter 4. For each load
case, all possible ULS strain distributions will be considered. It should be noted that for
each load case, studies will be necessary in different domains. Table 4.1.4-1 presents an
overview of the different load cases to be considered and the corresponding domains in
which these load case are to be examined.
As was already mentioned in paragraph 4.1.1, the aim is to develop in each domain the
equations that will permit to solve the design problem: the determination of the
reinforcement for a selected concrete cross-section and for imposed internal forces.

4-5

Load case

Axial tension

Possible domains
N

1a (left limit)
s

ud

cu2

1a
1b
2a

Eccentrically
applied tension
(bending combinde
with tension)

ud

cu2
c

1b
2a

Simple bending

Infinite
distance

ud

N=0

cu2

N'=0

Simple bending

Infinite
distance

1b
2a
s

ud

c2 cu2

N'

Eccentrically
applied
compression
(bending combined
with compression)

1b
2a
2b
3

s
ud

c2

c2

Axial compression

N'

3 (right limit)
s
c2

Table 4.1.4-1
Overview of load cases and corresponding domains to be analyzed

4-6

4.2 Bending rectangular cross-section

4.2.1

Introduction

For the case of simple bending, the depth of the neutral axis is at most equal to the
effective depth; the ULS strain diagrams are situated in the domains 1b and 2a.
It is common to start with the study of domain 2a, because this allows a better
understanding of the small difficulties encountered in the neighbouring domains.
Equations are first developed for the case of a singly reinforced cross-section, which
is characterized by the presence of reinforcement only near the tensile face.
4.2.2

Description of the design problem

Given
the quality of the selected materials; the design stress-strain curves are thus
known, as well as fyd and fcd ;
the dimensions of the concrete cross-section: width b and depth h:
- in general, the width b is not unknown; several criteria are available for the
choice of a good starting value:
- architectural and esthetical considerations (the width of a beam may
be determined by the diameter of the cylindrical columns supporting
the beam, for example!);
- the resistance against shear load (see further in the chapter on shear);
- economical aspects of the design solution, etc.
- in general, first approximate values are chosen for the total depth h or the
effective depth d, again on the basis of several possible criteria:
- serviceability conditions: a certain minimum depth is necessary to
avoid excessive deformations of the beam (see further in chapter on
SLS);
- architectural and esthetical considerations;
- functional conditions;
- economical aspects of the design solution, etc.
the design bending moment Md (imposed or destabilizing internal force).
Question
Determine the area of the tensile reinforcement As1 (or designated by As in this case of
singly reinforced cross-section).
4.2.3

Domain 2a

4.2.3.1 Basic figure

The equations that are needed to solve the design problem are developed on the basis of
figure 4.2.3-1, where one finds the schematic representation of:
the singly reinforced section;
one ULS strain diagram, situated in domain 2a. The position of the neutral axis
(NA) with respect to the extreme compression fibre, is indicated by the symbol x ;
the stress diagram that is deduced from the strain diagram by means of the design
stress-strain relationships for steel and concrete;

4-7

 the cross-section with the resultant forces. Fc is the resultant force of the
compression stresses in the concrete (in fact, it is the resultant of the elementary
internal forces). The integral of the elementary compression forces can easily be
calculated by considering the surface under the parabola-rectangle diagram and by
multiplying this surface by the width b. The surface under the parabola-rectangle
diagram can be calculated because the equation of the parabolic part is known (see
chapter 3); it is expressed as . (fcd . x), where the filling coefficient expresses
the degree of filling-up of the rectangle (fcd . x) by the parabola-rectangle diagram.
Fc is thus expressed as:
Fc = . (fcd . x) . b
The position of the resultant force Fc is also known; therefore, the position of the
centre of gravity of the parabola-rectangle diagram. The distance of the centre of
gravity towards the extreme compression fibre, is expressed as: G . x, where G is
called the coefficient of the centre of gravity.

Note: For the concrete classes up to C50/60, with cu2 = - 0,0035 and c2 = 0,002, one finds:
Fc = . (fcd . x) . b = 0,81 . (fcd . x) . b with G . x = 0,416 . x.
Fs is the resultant of the tensile stresses in the steel; in a simplified way (taking s
constant over the whole area of reinforcement), Fs is expressed as:
Fs = As . s
the cross-section with the imposed internal force, which is the design bending
moment Md.
Two types of equations may be developed on the basis of figure 4.2.3-1:
the compatibility equations , which describe relationships between strains at
different levels in the cross-section, and which translate in fact the linearity of the
strain diagram (BERNOULLI);
the equilibrium equations (translation and rotation), which describe the
equivalence between the imposed (destabilizing) internal forces on one hand, and
the resisting (stabilizing) internal forces on the other hand.

cu2

1 c 2 x
cu 2

fcd

G.x

c2

Fc=
.x.fcd.b

Md

z=d-G.x
As

Fs=As.s

0sud
(a)

(b)

(c)

Figure 4.2.3-1
Principle figure for the study of bending in domain 2a; (a) strain diagram;

4-8

(d)

(b) stress diagram; (c) resultant forces;

(d) imposed load (internal bending moment)
4.2.3.2 Compatibility equations
These equations express that the cross-section remains plane after deformation:

s
dx
=
cu 2
x

(4.2.3-1)

Introduction of the symbol :

=x/d
permits to rewrite (4.2.3-1) as:

s = cu 2

(4.2.3-2)

Inversely, can be expressed in function of s:

cu 2
cu 2 + s

(4.2.3-3)

Notes:
- For the concrete classes up to C50/60, cu2 = - 0,0035, and the formulas
(4.2.3-2) and (4.2.3-3) can be written as:
3,5 1
1000

(4.2.3-4)

3,5
3,5 + 1000 s

(4.2.3-5)

s =
and

- Limits of domain 2a
The limits of domain 2a are defined by:
- on the left side: s = ud; for the concrete classes up to C50/60 and adopting
s = ud = 1%, one finds = 0,259;
- on the right side: s = 0% and consequently = 1.
- Several particular values of may be highlighted within domain 2a,
especially those corresponding with the steel tensile strain s = fyd; see figure
4.2.3-2. For the concrete classes up to C50/60 (with cu2 = - 0,0035), one
finds via (4.2.3-5) the limit values lim which correspond with s = fyd / Es for

4-9

different steel grades: see table 4.2.3.-1. The significance of lim will be
examined later on, but it can be said already that strain distributions with >
lim, and thus s < fyd / Es, are to be avoided.
Table 4.2.3-1
Limit values lim for different steel grades, calculated for the concrete classes up to
C50/60 (with cu2 = - 0,0035)
S220
S400
S500
S600
fyk (MPa)
220
400
500
600
fyd (MPa)
191,3
347,8
434,8
521,7
0,785
0,668
0,617
0,573
lim

cu2
O'

voor
= ud

=1

2a

lim
s
A

ud

s
s
sy = fyd/Es
Figure 4.2.3-2
Domain 2a; strain diagram corresponding with the entering of the steel reinforcement
into the domain of plastic materials behaviour: = lim

4.2.3.3 Equilibrium equations

1. Translations equilibrium; see figure 4.2.3-1:

Fc = Fs
. (fcd . x) . b = As . s
With = x / d, one finds:

4-10

 . . fcd .b .d = As . s

(4.2.3-6)

2. Rotation equilibrium; see figure 4.2.3.-1:

Md = Fc . z ( = Fs . z )
Md = ( . fcd . x . b) . (d G . x)
With = x / d, one finds:

Md = . . fcd .b .d2 . (1 G . )

(4.2.3-7)

4.2.3.4 The concept of reduced moment d

The concept of reduced moment d is introduced by means of the following
expression:

d =

Md
b.d 2 . f cd

(4.2.3-8)

Introduction of d in (4.2.3-7) leads to:

d = . . (1 G . )

(4.2.3-9)

Note:
One can now also calculate the values of d that correspond with the limits of
domain 2a, by introduction of the values of lim in equation (4.2.3-9).
Table 4.2.3-2 presents the limit values of d for the concrete classes up to
C50/60 and taking into account the choice to put ud = 1% .
Table 4.2.3-2
Limit values of d for domain 2a for the concrete classes up to C50/60 (with
cu2 = - 0,0035), putting ud = 1%

d
Limit on the left side (s = ud)
0,259
0,187
Limit on the right side (s = 0%)
1
0,473
S220 (s = fyd / Es = 0,096 %)
lim = 0,785
d,lim = 0,428
S400 (s = fyd / Es = 0,174 %)
lim = 0,668
d,lim = 0,391
S500 (s = fyd / Es = 0,217 %)
lim = 0,617
d,lim = 0,371
S600 (s = fyd / Es = 0,261 %)
lim = 0,573
d,lim = 0,353

4-11

4.2.3.5 Calculation of the area of reinforcement

Given: the design bending moment Md permits to determine the reduced moment:

d =
Control:

Md
b.d 2 . f cd

f cd = 0,85.

with

f ck
1,5

left-limit d right-limit domain 2a!

The rotation equilibrium equation (4.2.3-9), written in inverse way as = f(d) leads to
the quadratic equation:
. G . 2 - . + d = 0

which gives as useful solution:

1 1

4 G

2 G

(4.2.3-10)

This permits then to deduce the value of x and thus the position of the NA.
With and cu2 and the compatibility equations (4.2.3-2), one can calculate s.
The steel stress-strain diagram then permits to determine s.
The translation equilibrium (4.2.3-6) finally leads to As:
As =

. . f cd .b.d
s

(4.2.3-11)

4.2.3.6 Geometrical and mechanical reinforcement ratios

The quantity of reinforcement in a cross-section is often expressed by means of the

geometrical reinforcement ratio and the mechanical reinforcement ratio .
The geometrical reinforcement ration is defined by:
= As / b.d

Equation (4.2.3-11) may be rewritten:

. . f cd
s

4-12

(4.2.3-12)

The mechanical reinforcement ratio is defined by:

As f yd
.
b.d f cd

Equation (4.2.3-11) may be rewritten as:

. . f yd
s

(4.2.3-13)

Note concerning expression (4.2.3-13) :

For d dlim , one has s = fyd , which leads to = .

Thanks to formulas (4.2.3-13) and (4.2.3-10), one observes the direct relationship
between and d:

. f yd
s .2 G

4
1 1 G d

(4.2.3-14)

4.2.3.7 Practical design table

In order to assist fast calculations by hand, annex A4.2.3.7 presents the tables A4.2.3.7a to x, which are elaborated on the basis of the equations mentioned above; the tables
present the values of in function of d, as well as other properties such as (and thus
the position of the NA), defined by z/d with z = the lever arm between Fc and Fs, and
s (which depends on the steel grade). The filling coefficient and the coefficient of the
centre of gravity G are also presented.
Tables A4.2.3.7-a -b, -c and -d are elaborated for the concrete classes up to C50/60
(thus with cu2 = - 0,0035 and c2 = - 0,002); table a is realized with ud = 0,010 and
table b, -c and d are realized with ud = 0,8 . uk (uk = 2,5 % for steel of class A,
uk = 5,0 % for steel of class B and uk = 7,5 % for steel of class C).
Tables -e to -x are elaborated for the high strength concrete classes.
Each table is composed as follows:
s (MPa)
d

=x/d

=z/d

c ()

s ()

with

d defined by d =

4-13

Md
b.d 2 . f cd

S220

S400

S500

S600

 defined by =

As f yd
.
b.d f cd

and

. f yd
s .2 G

4
1 1 G d

for d > lim (s < fyd)

4 G

1
1

2 G

for d lim (s = fyd)

4 G

1 1

2 G

and G = f(concrete class)

Note concerning formula (4.2.3-13).

In domain 2a, the complete parabola-rectangle diagram is used; the following
values are determined by numerical integration:
Concrete class

Up to C50/60
C55/67
C60/75
C70/85
C80/95
C90/105

0,810
0,744
0,694
0,627
0,598
0,583

0,416
0,393
0,377
0,360
0,355
0,353

z d G .x
=
= 1 G .
d
d

c = c,ultimate = cu2 (in domain 2a !)

s = cu 2 .

s = f(s) via the design stress-strain diagram for the selected steel grade

Auxiliary figure for table A4.2.3.7

4-14

4.2.4

Domain 1b

4.2.4.1 Introduction: the particular problem in domain 1b

The maximum allowable strain (cu2; 0,0035 for the concrete classes up to C50/60) is
not reached and the equilibrium is obtained with a maximum value of the concrete stress
which can be smaller than fcd.
On the other hand, the steel strain is always equal to ud (1% or 0,8.uk); see figure 4.2.41.

c
O'

cu2

1b
d

s
A

ud

s,ult

Figure 4.2.4-1
Strain diagram in domain 1b; 0 < c cu2 and s = ud
4.2.4.2 Compatibility equation

c
x
=
s ,ult d x

(4.2.4-1)

With the introduction of:

s,ult = ud and = x / d
one finds:

c = ud .

x
= ud .
dx
1

And thus:

c
c + ud
4-15

(4.2.4-2)

Note: the limits of domain 1b

The limits are defined by:
- on the left side: c = 0% = 0 ;
- on the right side: c = cu2; for the commonly used concrete classes up to
C50/60 and assuming s = 0,01, the limit on the right is defined by c =
0,0035 = 0,259.
Table 4.2.3-2 presents the limit values of d for the concrete classes up to
C50/60 and taking into account the choice to put ud = 1% .

A particular case is found for c = c2 ; for c < c2 the parabola-rectangle diagram is

reduced to a portion of the parabolic part. One thus has to make the distinction between
two cases in function of the extreme concrete strain c with respect to c2 (smaller or
larger than c2): see figure 4.2.4-2. For the concrete classes up to C50/60 (with
c2 = 0,002) and assuming s = 0,01, one finds = 0,167 which corresponds to the limit
between two sub-domains in domain 1b.
c2

c2

(a)

(b)

Figure 4.2.4-2
Domain 1b; schematic representation of the stress diagram in the concrete for extreme
strain c smaller than c2 (a) or larger than c2 (b)
4.2.4.3 First case: 0 c c2 (left sub-domain in domain 1b)
1. Auxiliary figure
The auxiliary figure, on the basis of which the equations can be elaborated, is presented
in figure 4.2.4-3.
c2

fcd
xG

As

.x.fcd.b

d-G.x

ud

s (s=ud)

4-16

As.s

Figure 4.2.4-3
Auxiliary figure for domain 1b;
case in which extreme concrete strain is smaller than c2
2. Equilibrium equations
On the basis of figure 4.2.4-3, one can write the following relationships:
translation equilibrium

. fcd . x . b = As . s
or (introducing = x/d):

. fcd . d . b . = As . s

(4.2.4-3)

or (introducing the definition of ):

f
As f yd
.
= . . yd
b.d f cd
s

(4.2.4-4)

Moreover, s = s (for s = ud) = fyd (steel is certainly yielding for ud!)

rotation equilibrium
Md = . fcd . x . b . (d G . x)
or (introducing = x/d):
Md = . fcd . b . d2 . . (1 G . )
Taking into account that:

d =

Md
b.d 2 . f cd

one writes:
d = . . (1 G . )

Note 1:

4-17

(4.2.4-5)

 and G are functions of c on the one hand, and c is related to via the

compatibility equation (4.2.4-2) on the other hand, which thus leads to a direct
link between d and .

Note 2: limits of the sub-domain in 1b, to the left of c2

- on the left side: c = 0% = 0 (4.2.4-5) d = 0.
- on the right side: c = c2 ; the limit values of d are mentioned in the design
tables A4.2.3.7; for the commonly used concrete classes up to C50/60 and
assuming s=0,01, one finds = 0,167; = 2/3; G = 3/8 (4.2.4-5) d =
0,105

3. Solution scheme
The solution scheme is similar to the one in domain 2a.
Given: Md

d; verification: is d situated in domain 1b? If yes, then:

rotation equilibrium equation (4.2.4-5)

compatibility equation (4.2.4-2) c which allows verification and eventual

adjustment of and G

translation equilibrium equation (4.2.4-3)

4. Practical design table
On the basis of the equations and methodology presented above, the design tables
A4.2.3.7 can now be completed for the left sub-domain in 1b.
4.2.4.4 Second case: c2 c cu2 (right sub-domain in domain 1b)
1. Auxiliary figure
The auxiliary figure, on the basis of which the equations can be elaborated, is presented
in figure 4.2.4-4.
c2

fcd
xG

x
d

As

.x.fcd.b

d-G.x

ud

s (s=ud)

As.s

Figure 4.2.4-4
Auxiliary figure for domain 1b;
case in which extreme concrete strain c is situated between the limits: c2 c cu2
2. Equilibrium equations

4-18

On the basis of figure 4.2.4-4, the equilibrium equations may be developed, which are
fully similar to the ones developed for the left sub-domain in 1b.

Note: limits of the sub-domain in domain 1b, to the right of c2

- on the left side: c = c2 ; the limit values of d are mentioned in the design
tables A4.2.3.7; for the concrete classes up to C50/60 and assuming s = 0,01,
one finds = 0,167; = 2/3; G = 3/8 (4.2.4-5) d = 0,105.
- on the right side: c = cu2 ; the limit values of d are mentioned in the design
tables A4.2.3.7; for the concrete classes up to C50/60 and assuming s = 0,01,
one finds = 0,259 ; = 0,81; G = 0,416 (4.2.4-5) d = 0,187.

3. Solution scheme and practical design tables

The solution scheme is identical to the one for the left sub-domain in 1b. On the basis of
the equations and methodology presented above, the design tables A4.2.3.7 can now be
completed for the right sub-domain in 1b.
4.2.5

Practical design calculation in ULS of a singly reinforced rectangular crosssection by means of design tables A4.2.3.7

4.2.5.1 Calculation of the resisting moment Md

= verification of ULS performance of a given section.
Given:
the dimensions b, d, As;
the materials characteristics:
- resistance class of the concrete and thus fcd (with eventually the inclusion of the
long term effects), c2 and cu2 (if one selects the parabola-rectangle diagram);
- steel grade and thus fyd and ud ; here, the decision must be taken to work with
ud = 0,8. uk or with ud = 0,01.
Question: Md = ?
Solution :
is calculated by means of the formula:

As f yd
.
b.d f cd

with

f cd = 0,85.

Table A4.2.3.7 d Md = d . b . d2 . fcd

4.2.5.2 Calculation of the area of reinforcement As
= design problem
Given:
certain dimensions: b, h (of d);
4-19

f ck
1,5

the material characteristics:

- resistance class of the concrete and thus fcd (with eventually the inclusion of the
long term effects), c2 and cu2 (if one selects the parabola-rectangle diagram);
- steel grade and thus fyd and ud ; here, the decision must be taken to work with
ud = 0,8. uk or with ud = 0,01.
the imposed load: Md

Question: As = ?
Solution:
d is calculated by means of formula

d =

Md
b.d 2 . f cd

with

f cd = 0,85.

f ck
1,5

and one looks for the calculated value of d in the adequate table A4.2.3.7.
1st case: d lim (see values of lim in table A4.2.3.7)
- the tensile steel works at the yield value and is thus used efficiently (s = fyd);
- only one reinforcement (on the tensile side) is necessary to resist to the imposed
moment, for the given dimensions of the cross-section; this is called a singly
reinforced section;
- calculation scheme: d table A4.2.3.7 As = . b . d . fcd / fyd

Note: alternative formula for the determination of the area of reinforcement As

The rotation equilibrium (see auxiliary figure in support of table A4.2.3.7) may
also be expressed as follows (with rotation centre in the centre of gravity of the
compression stress bloc):
Md = Fs . z = As . s . z
with s = fyd for d lim
Introducing = z/d, this equation is rewritten as follows:
Md = . d . As . fyd
which gives
Md
(4.2.4-6)
As =
.d . f yd
As can be observed in table A4.2.3.7, for the commonly used concrete classes up
to C50/60 and for d lim , the value of is about 0,9. One may thus determine
the area of reinforcement with expression (4.2.4-6) with the assumption that =
0,9. The parameter , which is representative for the lever arm between the
resultant forces, is thus useful for quick preliminary design calculations.

2nd case: d > lim (see values of lim in table A4.2.3.7)

This case should be avoided for two reasons:

4-20

- the steel is not used in an efficient way: s < fyd. The bending moment Md must be
equivalent to the couple of resultant forces Fc and Fs: see figure 4.2.5-1; a large
value of d (d > lim) leads to a large value of Fc, and thus to a large value of Fs
(one cannot change much to the lever arm); with a small s, this leads to an
excessive value for As (which, on top of that, is not used efficiently);
- the ULS of the cross-section is obtained by brittle failure of the compressive
concrete, without plastic deformation of the steel reinforcement. The plastic
deformation of the reinforcement is an essential condition regarding the safety of
buildings, because large plastic strains in the tensile steel go together with large
crack openings in the adjacent concrete, which gives observable warnings before
failure. This is also called the ductility condition.
cu2
O'

Md
d

Fc

lim
lim

Fs

s
O

fyd

s
ud

fyd/Es

Figure 4.2.5-1
Auxiliary figure for the reasoning about d > lim
In order to avoid d > lim, steel reinforcement must be put in the compression zone
of the cross-section; this permits to realize Fc with a smaller area of compressed
concrete (figure 4.2.5-1). Adding steel in the compressive zone permits to get a
higher position of the NA; this leads to a smaller value of x, to a larger value of s
and to a strain diagram characterized by d lim.

Important note concerning the ductility condition

In the previous version of Eurocode 2 (NBN B15-002:1998; 2.5.3.4.2.(5)), an
additional condition was formulated in order to assure sufficient ductility in
design: for linear elastic analysis, the ration = x/d was not accepted to be larger
than the following values:
0,45 for concrete classes C12/15 up to C35/45;
0,35 for concrete classes C40/45 up to C50/60.
The additional condition was introduced in order to avoid a too low position of
the NA in the cross-section, and thus to be sure to have sufficient ductility.

4-21

It should be noted that the actual version of Eurocode 2 (EN 1992-1-1:2004)

does not include this condition anymore for linear elastic analysis. On the other
hand, the Belgian ANB (NBN EN1992-1-1-ANB:2008; 5.4) has now included
even more severe conditions:
x
for fck 50 MPa
0,45
d
x
0,37

for fck > 50 MPa

d 0,6 + 0,0014 / cu 2
One observes that the condition is more severe for high strength concrete
because of its more ductile character.
4.2.5.3 Detailing of the reinforcement
Chapter 6 in these course notes will discuss the detailing of the reinforcement, focusing
on the translation of the calculated area of reinforcement As (a number of cm2) into a
real reinforcement with respect for rules concerning the choice of the diameter of the
bars, the spacing between bars, the cover of the reinforcement, etc.
Yet, at this early stage of the course, it is sufficient to take notice of the auxiliary table
4.2.5-1, that presents the area of reinforcement As in function of a chosen diameter and
the number n of bars.

Note: nominal diameter

The nominal diameter of a bar (which is ribbed!) is the diameter of an equivalent
circular area.
Table 4.2.5-1
Auxiliary table presenting the area of reinforcement As
in function of the nominal diameter and the number n of bars

n
1
2
3
4
5
6
7
8
9
10
11
12

10

28,3
56,6
85,0
113
141
170
198
226
254
283
311
339

50,3
101
151
201
251
302
352
402
452
503
553
603

78,5
157
236
314
393
471
550
628
707
785
864
942

Nominal diameter (mm)

12
14
16
20
25
2
Section As (mm ) composed of n bars
113 154 201 314 491
226 308 402 628 982
339 462 603 942 1437
452 616 804 1257 1963
565 770 1005 1571 2454
679 924 1206 1885 2945
792 1078 1407 2199 3436
905 1232 1608 2513 3927
1018 1385 1810 2827 4418
1131 1539 2011 3142 4909
1244 1693 2212 3456 5400
1357 1847 2413 3770 5890

4-22

28

32

40

616
1232
1847
2463
3079
3695
4310
4926
5542
6158
6773
7389

804
1608
2412
3216
4020
4824
5628
6432
7236
8040
8844
9648

1257
2513
3770
5027
6283
7540
8796
10053
11310
12566
13823
15080

4.2.6

Dimensions of the cross-section of beams

4.2.6.1 Width (or breadth) b

As was already pointed out in paragraph 4.2.2, b is chosen in most of the cases in
function of:
architectural considerations; for example, the width of a beam may be chosen
equal (or not) to the diameter of the supporting cylindrical column;
the resistance to shear (see further in chapter 7 in these course notes);
economical aspects of the design solution.
4.2.6.2 Depth h (and thus also the effective depth d)
As was already pointed out in paragraph 4.2.2, the depth h (or total depth of the crosssection) is mostly chosen in function of:
serviceability conditions: a certain minimum depth is necessary to avoid
excessive deformations of the beam (see further in chapter on SLS);
architectural and esthetical considerations;
functional conditions;
economical aspects of the design solution.
The choice can also be made to adopt the minimum depth or the optimum depth ;
these two notions are explained hereafter.

Note : if h is known, then d is also known

If one knows the depth h of the cross-section, then the effective depth d is also
known, and vice-versa. Indeed, figure 4.2.6-1 shows that the difference between
h and d is determined by
- the thickness of the cover,
- the diameter of the link or stirrup (which may be necessary in the beam
to reinforce its shear load capacity, and which is useful anyway from a
constructive point of view to realize the reinforcement cage),
- the diameter of the main bars,
- the vertical distance between the different layers of reinforcement.
Without entering in the discussion of the detailing of reinforcement at this point,
it can be observed that in general, the difference between h and d is about 10%
of h.

4-23

armature
constructive

d
h = 1000

8
beugel
d
h = 600

28

28

30

20
= 552
d = 600 30 8
2
d 0,92.h

50

pingle
8

20

40

10
d = 1000 40 10 28

58
= 893
2

d 0,89.h

Figure 4.2.6-1
Schematic representation of two possibilities of reinforcements of a crosssection, with indication of the difference between h and d
4.2.6.3 Optimum depth h
"Optimum" = full exploitation of the ultimate resistance of steel and concrete at the
same time; this means that the ULS of the section to be designed, corresponds to the
strain diagram just in between the domains 1b and 2a.
Starting with:

d =

Md
b.d 2 . f cd

one finds:
d=

Md
1
=
d .b. f cd
d

Md
b. f cd

Note:
For concrete classes up to C50/60, and assuming ud = 0,010, the limit between
the domains 1b and 2a corresponds to = 0,259 and d = 0,187. The optimum
depth is then:
d = 2,31

Md
b. f cd

One also finds the reinforcement area associated with that depth:
4-24

(4.2.6-1)

As = .b.d .

f
f cd
= 0,209.b.d . cd
f yd
f yd

or, by replacement of = z/d = 0,892 in (4.2.4-6):

Md
As =
0,892.d . f yd

(4.2.6-2)

4.2.6.4 Minimum depth h

One finds the minimum depth of a simply reinforced section starting from:

d =

Md
b.d 2 . f cd

d=

Md
d .b. f cd

or:

One obtains the minimum depth d for the maximum value of the reduced moment d.
Yet, attention must be paid here to the fulfilment of the ductility condition, with respect
for the imposed limit values lim or (d)max prescribed by the Belgian ANB.
d (d)max leads to:
d d min =

Md
( d ) max .b. f cd

In addition to this, one has also: lim (see table A4.2.3.7), ; taking into account
expression (4.2.4-6), this leads to:
As As ,max =

Md
min .d . f yd

4.2.6.5 Discussion
Comparison of the formulas for the optimum and minimum depth, shows that the
reduction of the depth to 2/3 of the optimum depth leads to an increase of the
reinforcement area As with 100%. This is not an economic solution: reduction of
depth leads to a higher cost in steel reinforcement.
Choosing minimum dimensions may lead to serious problems to respect the
serviceability conditions: limitations of stress levels, crack opening and
deformations. Simple rules will be presented in chapter 5 in these course notes,
which will allow making better choices of the depth, in order to avoid future
problems with the SLS conditions.

4-25

4.2.7

Doubly reinforced cross-section

4.2.7.1 When?
See arguments in paragraph 4.2.5.2; compression reinforcement is needed when d >
lim or d > max,ANB; the aim is to reduce the area of compressed concrete and thus to
lift the position of the NA in order to get the tensile steel in the plastic behaviour
domain before the ULS is reached.
4.2.7.2 Compatibility equations
The strain diagram to be considered for the analysis of the doubly reinforced section,
after the lifting of the NA, is situated in domain 2a: see figure 4.2-11. The further lifting
of the NA into domain 1b (by putting in more compressed steel) is not necessary and is
not an economical solution.
b

cu2

fcd

s2

s2

d2
As2

x
d

d-d2
As1

s1

s1

d1

s
s

lim

s1 > lim

Figure 4.2.7-1
The strain diagram to be considered for the analysis of the doubly reinforced section,
after the lifting of the NA, is situated in domain 2a
As before (for the singly reinforced cross-section), one may write:

s1 d x
1
s1 = cu 2 .
=
x

cu 2

(4.2.7-1)

Now can be added to that:

s2 x d2
=
cu 2
x

s 2 = cu 2 .

introducing = x/d and 2 = d2/d

4-26

(4.2.7-2)

4.2.7.3 Study of s2
The aim of this paragraph is to determine the smallest possible value of s2, and thus
also of s2. Figure 4.2.7-2 shows that the lifted strain diagrams are situated in domain
2a; the extreme position is the limit between the domains 1b and 2a, which gives the
minimum value of s2.
cu2
O'

s2 is minimum
1b
=0,259 (concrete<C50/60)

s2,min
h

2a

s
A

ud

Figure 4.2.7-2
Extreme minimum value for s2
For the concrete classes up to C50/60 and assuming that tensile steel strain is limited to
1%, s2 is minimum for = 0,259
Because:

s 2 = cu 2 .

3,5 2
2
.
=
1000

and with in general 2 0,10, one finds s2 0,215 %. Figure 4.2.7-3 shows that for this
strain, one may adopt s2 = fyd for all steel grades up to S500 (and using concrete up to
C50/60). For larger values of 2 and with steel S600, a more elaborated estimation of s2
is necessary.
s (MPa)

0,215 %

522

S600
S500

435

S400

348

S220

191

Es = 200 GPa

0,10

0,17 0,22 0,26

s (%)

Figure 4.2.7-3
Value of s2 corresponding with s2 0,215 %

4-27

4.2.7.4 Basic equations

Figure 4.2.7-4 shows the principle idea behind the development of the equations
necessary for the calculation of the doubly reinforced cross-section loaded by the
imposed design bending moment Md: the doubly reinforced cross-section is considered
as the superposition of two virtual cross-sections:
a simply reinforced cross-section which resists to the design moment Mn; the
maximum value of Mn is Mlim = lim.b.d2.fcd (otherwise this would not be a simply
reinforced cross-section!); and
a cross-section composed of two reinforcements: the compressive reinforcement
As2 and the complementary reinforcement Asc which is in tension. This crosssection has to resist the complementary design moment Mc, so that
Md = Mn + Mc.
The three cross-sections are thus subjected to the same strain diagram.
cu2
s2

d2
As2

As1

As2

Asc

Asn

d1

Md

Mn

s1

Mc

Figure 4.2.7-4
Decomposition of the doubly reinforced cross-section in two virtual cross-sections
The basic equations are thus:
Md = Mn + Mc

(4.2.7-3)

Mc = As2 . s2 . (d d2)
(rotation equilibrium, simple bending)

(4.2.7-4)

As1 = Asn + Asc

(4.2.7-5)

As2 . s2 = Asc . s1
(translation equilibrium, simple bending)

(4.2.7-6)

with

and

with

4-28

4.2.7.5 Scheme of the design calculation

1. Identification of the minimum necessary area of compressive steel As2,min
The minimum area As2 is necessary to equilibrate the complementary moment
equal to Mc = Md - Mlim , with: Mlim = lim.b.d2.fcd or Mlim = max ABN.b.d2.fcd.
In that way, expression (4.2.7-4) gives:
As 2,min = As 2,lim

M d lim .b.d 2 . f cd
=
s 2 .(d d 2 )

Temporarily, one may assume s2 = fyd ; later on, it may be necessary to calculate
a better estimation of As2,min, once a better value of s2 is known.
2. Choice of As2 > As2,min
The real area of compressed reinforcement As2 has now to be chosen, slightly
bigger than As2,min, in such a way that:
Mn = Md Mc = Md As2.s2.(d - d2) < Mlim
and thus in such a way that:
n < lim ou max,ANB
3. Calculation of Asn

n =

Mn
M As 2 . s 2 .(d d 2 )
= d
2
b.d . f cd
b.d 2 . f cd

with As2 given (= chosen) and s2 = fyd (in first approximation).

With this first value of n, one may use table A4.2.3.7 in order to determine
Asn and also .
The compatibility equations (4.2.7-2) permit to determine s2 ; the design stressstrain curve for the compressive steel is then used to determine s2. If s2 < fyd, it
is necessary to repeat the step mentioned above in order to find the final values
for As2, s2, Mn and Asn. If s2 = fyd, supplementary iteration is not necessary.
4. Conclusion
As1 = Asn + Asc
Expression (4.2.7-6) leads to:

4-29

Asc = As 2 .

s2
s1

If s2 = fyd (and knowing that s1 = fyd ), one finds: Asc = As2

4.3 Simple bending T-section

4.3.1

Introduction

Cast-in-place floor slabs are often composed of slabs supported by a series of beams.
Slab and beams may be considered as a series of T-beams. Each T-beam is
characterized by its wide flange in compressed concrete on the upper side; this means
that in bending, the ultimate capacity if the compressed concrete is in general not
reached. The analysis is thus normally situated in domain 1b.
4.3.2

The notion of effective width of the compressed flange of a T-beam

4.3.2.1 Terminology and designation

The effective width is sometimes called the cooperating width; notation: beff
4.3.2.2 Significance
Figure 4.3.2-1 presents a T-beam loaded in bending. The active part of the T-beam is
the web; indeed, load is transmitted by the stiff parts of a structure! The reasoning starts
by considering only the rectangular part or web of the T-beam without the flanges on
both sides. The compression on the upper side causes the shortening of the upper fibres;
point A in figure 4.3.2-1 shifts to the right. Suppose that two slabs are now attached to
(both sides of) the web before bending. Due to the bending, both slabs are dragged
along by the web through the shear stresses that appear in the connection areas.

Figure 4.3.2-1
The notion of effective width of the flanges of T-beams;

4-30

shear stresses in the contact areas between flanges and web explain the shear lag
phenomenon
The next step in the reasoning consists of
the isolation of an elementary slice AA' of the T-beam, close to the support
(figure 4.3.2-2), as well as
a slice of one of the half-flanges.
It can be observed that the difference between the compressive stresses c applied on the
areas A'FGD' and AIHD should be equilibrated by the shear stresses applied on the
area IFGH. The shear stresses which act in all similar parallel areas, are responsible for
the deformation in the horizontal plane of the beam AA'D'D BB'C'C. The fibres which
are the most distant from the web, such as AA' and BB', lag somewhat behind the fibres
close to the web; they appear to be less efficient in cooperating to resist to the bending.
This phenomenon is designated by the term shear lag. One also observes that the
assumption of BERNOULLI is not applicable because A'B'C'D' does not remain
straight. In order to be able to accept the assumption anyway, one has to narrow the
cooperative width of the flange towards the so-called effective width beff, within which
it may be assumed that strains are uniformly distributed.

hf

bw
Figure 4.3.2-2
Effective width of a T-beam; figure explaining the shear lag effect;
bw = width of the web (w < web) and
hf = depth (or heigth) of the compression flange (f < flange)

The effective width is not the same in all cross-sections along the longitudinal axis of
the beam; indeed, the shear lag effect is proportional to the shear force and is thus
maximal at the supports (larger shear gradient). In sections further away from the
supports, the compression force is better distributed over the whole width of the
compressive flange; see figure 4.3.2-3.

4-31

beff

Figure 4.3.2-3
The effective width of a T-beam is variable along the axis of the beam
4.3.2.3 Effective width: rules and prescriptions
The effective width of the compressive flange, which is taken into account in the design
calculation of a T-beam, should not be larger than the following limits:
the real width of the flange;
the same concrete cannot be used for two neighbouring beams; that means: beff e
(see figure 4.3.2-4);

e
Figure 4.3.2-4
Neighbouring T-beams

EN 1992-1-1:2004; 5.3.2.1 (part of paragraph 5.3, which presents acceptable

simplifications for identifying models for structural analysis) gives the following
prescriptions regarding the effective width of flanges:
article (1) : in T-beams, the effective flange width, over which uniform conditions
of stress can be assumed, depends on the web and flange dimensions, the type of
loading, the span, the support conditions and the transverse reinforcement;
article (4) : for structural analysis, where a great accuracy is not required, a
constant width may be assumed over the whole span;
article (3) : the effective flange width beff for a T-beam or L-beam may be derived
as (notations: see figures 4.3.2-5 and 4.3.2-6):
beff = beff,i + bw b

4-32

with beff,i = 0,2 . bi + 0,1 . l0 0,2 . l0 and beff,i bi

Notes:
- w < web;
- in figure 4.3.2-6, l0 represents the distance between the points of zero
moment.

Figure 4.3.2-5
Parameters for the determination of the effective flange width
(figure 5.3 in EN 1992-1-1:2004)

Figure 4.3.2-6
Span length l0, to be used for the determination of the effective flange width
(figure 5.2 in EN 1992-1-1:2004).

Note:
Figure 4.3.2-6 is valid with the following assumptions:
- the length of the cantilever (l3) should be less than half the adjacent span;
- the ratio of adjacent spans should lie between 2/3 and 1,5.

4-33

4.3.3

Design of T cross-sections at ULS: 1st method

4.3.3.1 Basic idea of the 1st method

The real T cross-section can be considered as the result of the subtraction of the
rectangular section (beff bw).(h hf) from the rectangular section beff.h: see figure 4.3.31. The properties for the section beff.h are indicated with (); the properties for the
section (beff bw).(h hf) are indicated with (). Both rectangular sections, as well as the
original T cross-section, are subjected to the same strain diagram. In this way, the basic
equations necessary to solve the design problem, are:

b = beff

Md = M'd M"d

(4.3.3-1)

As = A's A"s

(4.3.3-2)

b' = b

'c
x'='.d

hf

'.d

hf

"c

x"=".d"
axe neutre

d'=d

d'
d"=d-hf

As

bw

A's

A"s

d"

Figure 4.3.3-1
Basic figure for the elaboration of the equations necessary to solve the design problem
of T cross-sections in bending

4.3.3.2 Compatibility equations

Starting from figure 4.3.3-1, one may write:

'c
'
=
s 1 '

"c
"
=
s 1 "

and

and also;

"c
=
'c
with

x" = x' hf

and

'

hf

'

d" = d' - hf

4-34

The compatibility equations can be elaborated more in detail for the domains 2a and 1b:
see figure 4.3.3-2 and table 4.3.3-1.

hf

(a)
domain 2a

(b)
domain 1b

'c = cu2

'c
"c

"c
x

As

s = ud (or 1%)

0 < s < ud (or 1%)

Figure 4.3.3-2
Strain diagram in domain 2a (a) and in domain 1b (b)

Table 4.3.3-1
Compatibility equations for the T-cross-section
domain 2a
Domain 1b

'c = cu 2
"c = cu 2 .

'

s = cu 2 .

s = ud (or 1%)
hf

'

1 '
'

'c = ud .

'
1 '

"c = ud .

"
1 "

et "c = ud .

'

hf

d
1 '

4.3.3.3 Equilibrium equations

1. As = A's A"s
where A's and A"s can be detailed by means of the translation equilibrium:
As . s = ' . b . x' . fcd " . (b bw) . x" . fcd
2. Md = M'd M"d
where M'd and M"d can be detailed by means of the rotation equilibrium:
Md = ' . b . x' . fcd . (d 'G . x') " . (b bw) . x" . fcd . (d" "G . x")

4-35

4.3.3.4 Calculation scheme

Given: the T cross-section, with:
- dimensions: b (= beff), bw, h, hf
- materials: fyd, fcd
- loads: Md
Question: As = ?
Solution :
The solution scheme comprises the following steps.
1. If the cross-section were rectangular (b . h), the theory of the simply reinforced
rectangular cross-section would apply:
f
Md
d =
with
f cd = 0,85. ck
2
b.d . f cd
1,5
2. d table A4.2.3.7 identification of domain 2a or 1b = x/d
3. There are 2 possibilities:
if hf / d, the NA is situated in the flange and the T cross-section can be
calculated in the same way as a rectangular section (concrete in tension is
not considered); see figure 4.3.3-3.
NA

=
Figure 4.3.3-3
The NA is situated in the flange
if > hf / d, the NA is situated in the web; the model presented in figure
4.3.3-1 has now to be used.
4. A value is now chosen for ' which is somewhat larger than the obtained so far.
Indeed, has been obtained by considering a rectangular section and thus by
considering an area of compressed concrete that does not exist in reality: see
figure 4.3.3-4; for that reason, the NA should be put on a lower level.

4-36

Figure 4.3.3-4
The NA falls in the web; one takes into account too much compressed concrete
when working with a rectangular section
5. ' chosen x' is known (because d' = d) as well as ":

"=

x" x' h f
=
d " d ' h f

The compatibility equations in table 4.3.3-1 allow to determine 'c et "c.

On the basis of 'c, one can determine ' and 'G by numerical integration or by
using table A4.2.3.7. On the basis of "c, one can determine " and "G by
numerical integration. In order to perform the calculation of the T-section by
hand, one can determine " and "G by means of the auxiliary tables in Annex
A4.3.3.4, which are elaborated for the different concrete classes; table A4.3.3.4-a
is elaborated for the commonly used concrete classes up to C50/60, and is
represented below.

Note:
the choice of ud = (either 1% or 0,8 uk ) does not influence the values of " and
"G for a rectangular section.

4-37

Table A4.3.3.4-a
Determination of " and "G in function of "c for concrete classes up to C50/60 (this is
the first of the auxiliary tables in Annex A4.3.3.4)
"c

"

"G

0,000%

0,000

0,334

0,010%

0,049

0,336

0,020%

0,096

0,337

0,030%

0,142

0,339

0,040%

0,186

0,340

0,050%

0,228

0,342

0,060%

0,269

0,344

0,070%

0,308

0,345

0,080%

0,346

0,347

0,090%

0,381

0,349

0,100%

0,415

0,351

0,110%

0,448

0,353

0,120%

0,479

0,355

0,130%

0,508

0,357

0,140%

0,535

0,360

0,150%

0,561

0,362

0,160%

0,585

0,365

0,170%

0,608

0,367

0,180%

0,628

0,370

0,190%

0,648

0,373

0,200%

0,665

0,376

0,200%

0,665

0,376

0,208%

0,677

0,378

0,215%

0,688

0,381

0,223%

0,699

0,383

0,230%

0,709

0,386

0,238%

0,718

0,388

0,245%

0,726

0,390

0,253%

0,734

0,393

0,260%

0,742

0,395

0,268%

0,749

0,397

0,275%

0,756

0,399

0,283%

0,762

0,401

0,290%

0,768

0,403

0,298%

0,774

0,405

0,305%

0,780

0,407

0,313%

0,785

0,409

0,320%

0,790

0,410

0,328%

0,795

0,412

0,335%

0,799

0,414

0,343%

0,804

0,415

0,350%

0,808

0,417

C12/15 C50/60

0,9

0,42

0,8

0,41

0,7

0,4

0,6

0,39

0,5

0,38

0,4

0,37

G
0,3

0,36

0,2

0,35

0,1

0,34

0
0,000%

0,050%

0,100%

4-38

0,150%

0,200%

0,250%

0,300%

0,33
0,350%

One can now calculate A's, M'd and A"s, M"d, by means of the equilibrium
equations for each rectangular cross-section.
A's . s = ' . b . x' . fcd
with s = fyd (because lim !). Consequently:
A's = ' . ' . b . d . fcd / fyd
M'd = ' . b . x' . fcd . (d' 'G . x')
of M'd = ' . ' . b . d2 . fcd . (1 'G . ')

A"s . s = " . (b bw) . x" . fcd

A"s = " . " . (b - bw) . (d hf) . fcd / fyd
M"d = " . (b bw) . x" . fcd . (d" "G . x")
or

M"d = " . " . (b bw) . fcd . d"2 . (1 "G . ")

6. On the basis of the first choice of ', one thus obtains a first solution for As
(As = A's - A"s). Yet, it would be very surprising that on the basis of the first
choice of ', the condition Md = M'd - M"d would immediately be respected. Two
alternatives are possible:
the first solution of As (As = A's - A"s). may be adapted as follows:
As = ( A' A" ).

Md
M ' d M "d

if the difference between M'd - M"d and Md is too big, calculations have to
be started again with a new chosen value of ':
- smaller than the first choice of ', when M'd M"d > Md
- larger than the first choice of ', when M'd M"d < Md

4.3.4

Design of T cross-sections at ULS: 2nd method

4.3.4.1 Principle of the 2nd method

The equilibrium equations are applied to an equivalent virtual rectangular cross-section,
4-39

taking into account a stress diagram for the compressed concrete which is adapted to the
width of the real cross-section. This way of reasoning may be applied for the calculation
of all possible shapes of cross-sections. The method is explained in a schematic way in
the following text.
4.3.4.2 Systematic elaboration of the 2nd method
(1)

Characteristics of the cross-section

The T cross-section to be considered is represented in figure 4.3.4-1.
b = beff
hf

h
d
As

bw

Figure 4.3.4-1
Geometrical characteristics of the T cross-section
As before, the materials are characterized by fyd (steel) and fcd (concrete).
As = the area of the tension reinforcement.
The cross-section has to be calculated for the imposed design bending
moment Md.
(2)

Strain diagrams
The following symbols are used:
- c = maximum compressive strain in the extreme concrete fibre
(with c cu2 );
- s = tensile strain in the steel reinforcement (with s ud );
- x = the depth of the compression stress bloc, defined by the
position of the NA;
x
-=
d
Two different cases are possible, depending on the position of the NA.
Case 1: x h f

hf
d

4-40

The NA is situated in the flange. This case has been discussed before (in
4.3.3.4); the T cross-section can be calculated as a rectangular section.
( >

Case 2: x > h f

hf

)
d
The NA lies within the web of the T cross-section (figure 4.3.4-2).
c
hf

As

Figure 4.3.4-2
Strain diagram in the T cross-section: the NA lies within the web
The compressed zone has a variable width: the width of the flange is b (as a
matter of fact beff), the width of the web is bw (see figure 4.3.4-3).
Compression
zone
NA

As

Figure 4.3.4-3
Compression zone in the T cross-section with x > h f
The corresponding stress distribution diagram is presented in figure 4.3.4-4,
which is valid for domain 2a (c = cu2 and 0 s ud ).

4-41

hf

fcd

NA

As

Figure 4.3.4-4
Strain and stress distribution diagram for the T cross-section: the NA lies
within the web (domain 2a)

The equivalent virtual rectangular section

(3)

The T cross-section is replaced by a virtual rectangular cross-section with

width b and depth h, equivalent with the T-section from the statics point of
view (see figure 4.3.4-5).
b

hf

1
NA

'

c
x

fcd

G.x
Fc =

.x.fcd.b

As

reduction of stress by
factor bw/b

Figure 4.3.4-5
Strain and stress diagram for the equivalent virtual rectangular cross-section
The flange is composed of the same material (fcd) as in the real T-section.
The lower part of the compressive bloc, within the web, is replaced by a
fictitious material with adapted stresses: the ratio between the new stress and
the stress in the real section is given by the factor bw/b. The stress ' in the
rectangular section is thus:

'=
b
'= w
b

within the flange

within the web

4-42

The virtual rectangular cross-section is statically equivalent to the real

section: the stress acting on the width bw is statically equivalent to the
b
stress ' = w acting on the width b.
b
The earlier presented developments for rectangular sections can now be
applied to the virtual section, taking into account the following
modifications:
- - the filling coefficient is now:

=
-

1
x. f cd

x
0

' ( s).ds

the coefficient of the centre of gravity G is now:

x

1
G =
x

s. ' (s).ds = 1
' (s).ds x . f
0

cd

x
0

s. ' ( s ).ds

In which the distance s is defined in figure 4.3.4-6.

'
x

Figure 4.3.4-6
Definition of the distance s which is necessary for the calculation of and

In this way, the resultant compression force in the concrete is applied at the
distance G.x with respect to the extreme concrete fibre:

Fc = .x. f cd .b
This formula is, from a formal point of view, identical to the formula
developed for a rectangular section. Consequently, all other formulas and
reasoning that are characteristic for the rectangular cross-section, are
applicable.
4.3.5

Practical design table for T cross-sections in bending

Table 4.3.5-1 allows quick design at ULS of simply reinforced T cross-sections, loaded
in bending. The table is valid for concrete classes up to C50/60 and assuming ud = 1%.

4-43

The table shows in function of (hf/d, b/bw and d). It can be seen as the extension of
table A4.2.3.7 for rectangular sections.

Note:
- The upper part of the table shows values of which are printed in grey; these
values of are the same as those obtained for the same value of d in
rectangular cross-sections (this is because the NA falls within the flange). The
values of obtained for a rectangular section are presented in the second
column of the table, in order to show the difference with the solution for a
real T-section.
- In the lower part of the table, results are limited to lim.

Table 4.3.5-1
Design at ULS of simply reinforced T cross-sections, loaded in bending;
concrete classes up to C50/60 and ud = 1%.
(the table shows in function of and hf/d)

Important note concerning the values of and d in table 4.3.5-1:

A f yd
- definition of :
=
.
b.d f cd
- d is calculated with Md and taking into account a rectangular section (b.d)

Table: see next pages

4-44

0,000
0,010
0,020
0,030
0,040
0,050
0,060
0,070
0,080
0,090
0,100
0,110
0,120
0,130
0,140
0,150
0,160
0,170
0,180
0,190
0,200
0,210
0,220
0,230
0,240
0,250
0,260
0,270
0,280
0,290
0,300
0,310
0,320
0,330
0,340
0,350
0,360
0,370
0,380
0,390
0,400
0,410
0,420
0,430
0,440
0,450
0,460
0,470
0,472
S220
S400
S500
S600

for
rectangle
0,000
0,010
0,020
0,031
0,041
0,052
0,063
0,073
0,084
0,095
0,106
0,118
0,129
0,141
0,152
0,164
0,176
0,188
0,201
0,213
0,226
0,240
0,253
0,267
0,281
0,295
0,309
0,324
0,339
0,355
0,371
0,387
0,404
0,422
0,439
0,458
0,477
0,498
0,518
0,541
0,563
0,589
0,615
0,644
0,676
0,708
0,754
0,799
0,809
lim =
lim =
lim =
lim =
lim =
lim =
lim =
lim =

in function of and hf/d )

hf/d = 0,05

hf/d = 0,10

b/bw=10

b/bw=5

b/bw=3

b/bw=2

b/bw=10

b/bw=5

b/bw=3

b/bw=2

0,000
0,010
0,020
0,031
0,041
0,052
0,063
0,076
0,094

0,000
0,010
0,020
0,031
0,041
0,052
0,062
0,074
0,087
0,101
0,116
0,134
0,156
0,156
0,187

0,000
0,010
0,020
0,031
0,041
0,052
0,063
0,074
0,085
0,097
0,110
0,124
0,138
0,138
0,153
0,170
0,188
0,208
0,232
0,261

0,000
0,010
0,020
0,031
0,041
0,052
0,063
0,073
0,085
0,096
0,108
0,120
0,133
0,146
0,159
0,173
0,188
0,203
0,220
0,237
0,256
0,275
0,297
0,321
0,349
0,381

0,000
0,010
0,020
0,031
0,041
0,052
0,063
0,073
0,084
0,095
0,106
0,119
0,135
0,159

0,000
0,010
0,020
0,031
0,041
0,052
0,063
0,073
0,084
0,095
0,106
0,118
0,131
0,145
0,161
0,180
0,204
0,240

0,000
0,010
0,020
0,031
0,041
0,052
0,063
0,073
0,084
0,095
0,106
0,118
0,130
0,142
0,156
0,170
0,185
0,202
0,220
0,239
0,264
0,292

0,000
0,010
0,020
0,031
0,041
0,052
0,063
0,073
0,084
0,095
0,106
0,118
0,129
0,141
0,154
0,167
0,180
0,194
0,208
0,224
0,239
0,257
0,275
0,294
0,315
0,339
0,365
0,396
0,441

0,087
0,109
0,083
0,099
0,081
0,095
0,079
0,092

0,125
0,167
0,117
0,148
0,113
0,140
0,110
0,133

0,175
0,245
0,162
0,213
0,156
0,200
0,150
0,188

0,238
0,343
0,219
0,295
0,210
0,274
0,201
0,257

0,128
0,154
0,123
0,143
0,122
0,139
0,121
0,136

0,162
0,207
0,153
0,187
0,150
0,179
0,146
0,173

0,206
0,279
0,193
0,246
0,186
0,233
0,181
0,221

0,261
0,368
0,242
0,320
0,233
0,299
0,224
0,282

4-45

0,000
0,010
0,020
0,030
0,040
0,050
0,060
0,070
0,080
0,090
0,100
0,110
0,120
0,130
0,140
0,150
0,160
0,170
0,180
0,190
0,200
0,210
0,220
0,230
0,240
0,250
0,260
0,270
0,280
0,290
0,300
0,310
0,320
0,330
0,340
0,350
0,360
0,370
0,380
0,390
0,400
0,410
0,420
0,430
0,440
0,450
0,460
0,470
0,472
S220
S400
S500
S600

for
rectangle
0,000
0,010
0,020
0,031
0,041
0,052
0,063
0,073
0,084
0,095
0,106
0,118
0,129
0,141
0,152
0,164
0,176
0,188
0,201
0,213
0,226
0,240
0,253
0,267
0,281
0,295
0,309
0,324
0,339
0,355
0,371
0,387
0,404
0,422
0,439
0,458
0,477
0,498
0,518
0,541
0,563
0,589
0,615
0,644
0,676
0,708
0,754
0,799
0,809
lim =
lim =
lim =
lim =
lim =
lim =
lim =
lim =

b/bw=5

in function of and hf/d)

hf/d = 0,15
b/bw=3 b/bw=2

b/bw=5

hf/d = 0,2
b/bw=3 b/bw=2

hf/d = 0,3
b/bw=3 b/bw=2

0,000
0,010
0,020
0,031
0,041
0,052
0,063
0,073
0,084
0,095
0,106
0,118
0,129
0,140
0,152
0,164
0,178
0,193
0,210
0,231
0,259

0,000
0,010
0,020
0,031
0,041
0,052
0,063
0,073
0,084
0,095
0,106
0,118
0,129
0,140
0,152
0,164
0,177
0,190
0,204
0,220
0,237
0,254
0,275
0,299
0,329

0,000
0,010
0,020
0,031
0,041
0,052
0,063
0,073
0,084
0,095
0,106
0,118
0,129
0,140
0,152
0,164
0,176
0,189
0,202
0,216
0,231
0,246
0,262
0,278
0,296
0,315
0,336
0,358
0,385
0,416
0,457

0,000
0,010
0,020
0,031
0,041
0,052
0,063
0,073
0,084
0,095
0,106
0,118
0,129
0,141
0,152
0,164
0,176
0,188
0,200
0,214
0,228
0,245
0,264
0,289

0,000
0,010
0,020
0,031
0,041
0,052
0,063
0,073
0,084
0,095
0,106
0,118
0,129
0,141
0,152
0,164
0,176
0,188
0,201
0,213
0,227
0,241
0,257
0,274
0,293
0,314
0,338
0,371

0,000
0,010
0,020
0,031
0,041
0,052
0,063
0,073
0,084
0,095
0,106
0,118
0,129
0,141
0,152
0,164
0,176
0,188
0,201
0,213
0,227
0,240
0,255
0,270
0,286
0,302
0,320
0,339
0,359
0,382
0,408
0,438
0,479

0,000
0,010
0,020
0,031
0,041
0,052
0,063
0,073
0,084
0,095
0,106
0,118
0,129
0,141
0,152
0,164
0,176
0,188
0,201
0,213
0,226
0,240
0,253
0,266
0,280
0,294
0,309
0,324
0,341
0,359
0,381
0,406
0,439

0,000
0,010
0,020
0,031
0,041
0,052
0,063
0,073
0,084
0,095
0,106
0,118
0,129
0,141
0,152
0,164
0,176
0,188
0,201
0,213
0,226
0,240
0,253
0,266
0,280
0,294
0,309
0,324
0,340
0,357
0,375
0,394
0,415
0,439
0,465
0,497
0,541

0,196
0,247
0,189
0,228
0,184
0,219
0,181
0,212

0,235
0,312
0,222
0,280
0,215
0,265
0,210
0,254

0,283
0,393
0,264
0,345
0,254
0,324
0,246
0,306

0,229
0,287
0,221
0,267
0,218
0,260
0,214
0,252

0,262
0,345
0,249
0,313
0,244
0,300
0,237
0,287

0,304
0,417
0,285
0,370
0,275
0,350
0,266
0,331

0,312
0,411
0,299
0,379
0,293
0,365
0,288
0,354

0,341
0,467
0,322
0,420
0,312
0,399
0,304
0,382

4-46

4.3.6

Doubly reinforced T cross-section

Both methods, presented before, can also be applied for the design calculation of doubly
reinforced T-sections: see figure 4.3.6-1.

Note:
As mentioned before, the need for reinforcement in the compressed side of Tsections is exceptional!

Figure 4.3.6-1
Basic figure showing the philosophy to apply for the design calculation of doubly
reinforced T cross-sections by means of the 1st method presented in paragraph 4.3.3
before (subtraction of rectangular sections)

4.3.7

Generalization of the 2nd method (developed for T-sections) towards crosssections with variable shapes

4.3.7.1 Principle
The theory of the 2nd method developed for T cross-sections in paragraph 4.3.4 before,
can easily be generalized towards sections with variable shapes. The real section, with
variable width, has to be replaced by an equivalent virtual section in which the stresses
are reduced by application of a factor that takes into account the difference between the
real width and the width of the virtual rectangular section. This principle is illustrated
by means of two examples, presented in figures 4.3.7-1 and 4.3.7-2.

4-47

Example 1:
A first cross-section is composed of parts with constant width: see figure 4.3.7-1. The
stresses ' to be applied on the equivalent rectangular section are obtained by reducing
the real stresses by application of a factor that takes into account the difference between
the real width and the width of the virtual rectangular section.
bmax

c
x

NA

s
As

statically
equivalent

bmax

'

c
x

NA
reduction of
stress in function
of width

s
As

Figure 4.3.7-1
Equivalent rectangular section: example 1

4-48

Example 2:
The real cross-section presents a variable width.
bmax

c
x

b(s)

NA

As

Statically
equivalent

bmax

'

c
x

NA
reduction of
stress in function
of width
As

Figure 4.3.7-2
Equivalent rectangular section: example 2

4.3.7.2 Determination of and G for an arbitrary section

As explained before, the real section is replaced by an equivalent virtual rectangular
section the width of which is bmax, which is the maximum width of the real section.
With:
s:
defines the depth in the section (see figure 4.3.4-6),
(s) : the stress in the real section, at the level s
b(s) : the width of the real section, at the level s
4-49

'(s) : the virtual stress in the equivalent virtual rectangular section, at the
level s

and:

b( s )
( s)
bmax

' ( s) =

the integrals for the calculation of and G are thus:

G =

1
x. f cd

1
2
x . f cd .

x
0

b( s )
( s).ds
bmax
x
0

s.

b( s )
( s).ds
bmax

Once these parameters are determined, the formulas for the rectangular sections may be
applied.

4-50

4.4 Bending combined with compression

Preliminary note: the developments in this paragraph are limited to rectangular crosssections, but extrapolation of the method towards sections with variable shapes is
straightforward.
4.4.1

Domains to be analyzed

Table 4.1.4-1 shows that ULS strain diagrams corresponding to bending in combination
with axial compression, are situated in the domains 1b, 2a, 2b and 3.
4.4.2

Definition of the design problem

Given:
the quality of the selected materials: fyd et fcd;
the dimensions of the concrete cross-section: b, h (d), d1 and d2; see figure 4.4.21;
the design values of the imposed internal forces Md and Nd; the imposed loads
may also be defined as a axial compression force Nd that is applied with an
eccentricity e0 to the geometric centre (or centroid) of the cross-section. The fact
that the geometric centre is used instead of the centre of gravity is justified
because the last one is not known yet; in this stage of the project, only
architectural plans are available.
Question: the areas of reinforcement As1 and As2 = ?
d2
As2
h

Nd
d

Md

eo

Nd
As1
d1
b

Figure 4.4.2-1
Bending combined with axial compression: designation of the imposed internal forces

4.4.3

Analysis of domains 1b and 2a

4.4.3.1 When?
A strain diagram in these two domains can only be obtained if the bending is much
more important than the compression; Nd is applied with a large eccentricity eo.

4-51

4.4.3.2 Basic equations

The formulas in this paragraph are developed for a strain diagram in domain 2a; the
developments in domain 1b are completely similar.
Basic figure: see figure 4.4.3-1.
Application point of
axial force

Nd

C
d2

cu2
e0

As2
d

s2

c2

e1

fcd
Ns2=As2.s2
Nc=

.b.x.fcd

d-G.x
As1

s1

Ns1=As1.s1

d1

Figure 4.4.3-1
Basic figure for the analysis of bending combined with compression in domain 2a
Horizontal translation equilibrium:

Nd = As2 . s2 + . b . x . fcd As1 . s1

and thus

Nd + As1 . s1 = As2 . s2 + . b . x . fcd

or

s1 .( As1 +

Nd

s1

) = As 2 . s 2 + .b.x. f cd

(4.4.3-1)

Rotation equilibrium around the tensile reinforcement As1:

Nd . e1 = . b . x . fcd . (d G . x) + As2 . s2
with e1 = eccentricity of Nd with respect to As1.

4.4.3.3 First case: As2 = 0

The equations (4.4.3-1) and (4.4.3-2) are now written as:

4-52

(4.4.3-2)

 s1.( As1 +

Nd

s1

) = .b.x. f cd

(4.4.3-3)

and

Nd . e1 = . b . x . fcd . (d G . x)

(4.4.3-4)

These equations are completely similar to those for simple bending: see the equations
(4.2.3-6) and (4.2.3-7). The solution scheme is thus also similar to the one for simple
bending:
Given: Nd and e0 deduction of e1 ;
Calculation of:

d =

N d .e1
b.d 2 . f cd

with

f cd = 0,85.

f ck
1,5

Attention: the reduced moment is not calculated with Md but with Nd . e1!
By means of table A4.2.3.7, one finds:
A f
= s . yd
b.d f cd

and

s1

Attention: the value of As1 should now be derived from As that was determined
above, because the comparison of expressions (4.4.3-3) and (4.2.3-6) shows that

As1 = As

Nd

s1

(4.4.3-5)

4.4.3.4 Second case: As2 0

The equations (4.4.3-1) and (4.4.3-2) are completely similar to the ones for the doubly
reinforced concrete section loaded in simple bending, because the system of equations:
N
.b.x. f cd As 2 . s 2

As1 + d =
+

s1
s
1
s
1

N .e = .b.x. f .(d .x) + A . .(d d )

cd
G
s2
s2
2
d 1

is equivalent to the following system:

As1 = Asn + Asc

M d = M n + M c

4-53

The last system of equations has been exploited for the calculation of the doubly
reinforced section loaded in simple bending. The solution scheme is thus as follows:
First, it is assumed that As2 = 0; calculation of:
f
N d .e1
d =
with f cd = 0,85. ck
2
1,5
b.d . f cd
If d > lim in table A4.2.3.7, a compression reinforcement is necessary in order
to reduce d with a term which is function of As2, so that d falls into the useful
part of table A4.2.3.7.
As2 minimum is then determined by:

As 2 min = As 2 lim =

M d lim .b.d 2 . f cd
s 2 .(d d 2 )

in which Nd . e1 is used instead of Md !

Choice of As2.
Calculation of:

n =

N d .e1 As 2 . s 2 .(d d 2 )
b.d 2 . f cd

This n thus corresponds to the moment Mn which is resisted by the simply

reinforced section (see decomposition of the doubly reinforced section, presented
in figure 4.2.7-4). For this n, one finds in table A4.2.3.7 a corresponding value of
, which then permits to determine Asn;
Calculation of:

Asc =

As 2 . s 2

s1

Finally, one finds:

As1 +

Nd

s1

= Asn + Asc

Asn + Asc corresponds indeed to:

As1 +

Nd

s1

Consequently:

As1 = Asn + Asc

4-54

Nd

s1

(4.4.3-6)

4.4.4

Analysis of domain 2b

4.4.4.1 When?
A strain diagram in this domain can only be obtained if the compression is more
important than the bending; Nd is applied with a small eccentricity eo.

4.4.4.2 The limites of domaine 2b

See figure 4.1.3-2
Left limit:
x = d;
x = h;
Right limit:

= x/d = 1;
= x/d = h/d;

t = x/h = (h d1) / h = 1 t1
t = x/h = h/h = 1

Note:
- Attention should be paid to designations and symbols used: for the analysis of
domains 2b and 3 (in which practically the whole cross-section is in
compression), it is preferred to use the total depth of the section as reference
depth instead of the effective depth; the parameter t = x/h is used instead of
= x/d. In the same way, t1 = d1/h is used instead of 1 = d1/d.
- In practice, the assumption d1 = d2 is adopted (and thus t1 = t2).

4.4.4.3 Compatibility equations

See figure 4.4.4-1
d2

cu2
s2

As2

As1

s1
d1

Figure 4.4.4-1
Domain 2b; auxiliary figure for the elaboration of the compatibility equations
The equations are:

s1 = cu 2 .

+ t1 1
xd
x h + d1
= cu 2 .
= cu 2 . t
x
x
t

4-55

(4.4.4-1)

 s 2 = cu 2 .

t2
x d2
= cu 2 . t
x
t

(4.4.4-2)

4.4.4.4 Equilibrium equations

See figure 4.4.4-2.
d2

cu2

fcd

s2

s2

As2

c2
d

s1

s1

e2

G.x

Nd
N c=
.b.x.fcd

As1

Ns2=As2.s2

e0
O

Ns1=As1.s1

d1

Figure 4.4.4-2
Principle figure for the analysis of bending combined with compression in domain 2b
Translation equilibrium:

Nd = . b . x . fcd + As2 . s2 + As1 . s1

(4.4.4-3)

Rotation equilibrium around the most compressed reinforcement As2 :

Nd . e2 = . b . x . fcd . ( G . x d2) + As1 . s1 . (d - d2)

(4.4.4-4)

with e2 = the eccentricity of Nd with respect to As2.

Note:
The rotation centre to be used for the expression of the rotation equilibrium is
always the most loaded (= most important) reinforcement, which is:
- the tensile reinforcement, if there is one;
- the most compressive reinforcement if both reinforcements are compressed.
Assuming that:

d =

N d .e2 As1. s1.(d d 2 )

b.h 2 . f cd

with the following symbols:

x
d
d
t = ; t1 = 1 ; t 2 = 2
h
h
h
the rotation equilibrium equation can be written as:

4-56

(4.4.4-5)

d = .t . (G . t t2)

(4.4.4-6)

Inversely, t can also be expressed in function of d, by means of the following

quadratic equation:

. G . t2 - . t2 . t - d = 0
which has the following solution:

t2 + t22 +
t =

4. G

. d

2. G

(4.4.4-7)

t = f (d , t2 , et G )
in which and G correspond to the entire parabola-rectangle diagram, with
c = cu2. One thus finds and G in table A4.2.3.7, for domain 2a.

4.4.4.5 Solution scheme

Exploitation of the rotation equilibrium via equation (4.4.4-5).
A first choice can be made: s1 = 0; notation: (s1)1 = 0, in which the index 1
indicates that this is the first choice. Anyway, it is clear that in domain 2b, this
choice is not far removed from reality. Equation (4.4.4-5) is now written as:
( d )1 =

N d .e2
b.h 2 . f cd

Consequently:

t = f ((d )1 , t2)
via expression (4.4.4-7), in which and G correspond to the complete parabolarectangle diagram, with c = cu2 (values in table A4.2.3.7 for domain 2a).
Exploitation of the compatibility equations.
As t1 = t2, expressions (4.4.4-1) and (4.4.4-2) lead to:
( s1 )1 = cu 2 .

t + t1 1
t

( s 2 )1 = cu 2 .

4-57

t t1
t

 Exploitation of the design stress-strain diagram of the steel.

(s1)1 and (s2)1 lead to (s1)1 and (s2)1
Exploitation of the translation equilibrium
There are still two unknown parameters in equation (4.4.4-3): As1 and As2, while
all available equations have been exploited. In order to overcome this difficulty,
the following trick is proposed, by introducing the ratio of the reinforcement
areas:

As1
As 2

with = 0, , , 1, 2, 4 etc. = 1 corresponds to the case of "symmetric"

reinforcement (As1 = As2), which is often the case in columns for example.
With As1 = . As2, the translation equilibrium equation is rewritten:

Nd = . b . h . fcd . t + As2 . (s2 + . s1)

or

As 2 =

N d .b.h. f cd . t
s 2 + . s1

(4.4.4-8)

Taking into account the selected value of , one finds via (4.4.4-8): (As2)1 and
(As1)1 = . (As2)1
In domain 2b, the first approximation is in general conclusive because, as already
pointed out before, the first choice (s1)1 = 0 is close to reality. However, one can
improve the results by repeating the calculation sequence on the basis of the
results obtained after the first sequence:

d =

N d .e2 ( As1 )1.( s1 )1.(d d 2 )

b.h 2 . f cd

Consequently
(t)2 = f((d)2 , t2)
where and G correspond to the complete parabola-rectangle diagram, with c =
cu2 (values in table A4.2.3.7 for domain 2a).
This allows to determine (s1)2 and (s2)2, and via the stress-strain diagram of steel:
(s1)2 and (s2)2.
Finally:

4-58

( As 2 ) 2 =

N d .b.h. f cd .( t ) 2
( s 2 ) 2 + .( s1 ) 2

(As1)2 = . (As2)2
4.4.4.6 Auxiliary table for shortening hand calculations
In order to shorten hand calculations, annex A4.4.4.6 presents practical tables which are
composed on the basis of the following scheme:
S220

t1 = t2 = . . . .

s1

s2

s1

MPa

S400

s2

s1

MPa

MPa

S500

s2

MPa

s1

MPa

s2

MPa

S600

s1

MPa

s2

MPa

with t2 chosen and t1 = t2, and the following formulas:

d =

N d .e2 As1. s1.(d d 2 )

b.h 2 . f cd

t2 + t22 +
t =

. d

2. G

s1 = cu 2 .

t + t1 1
t

s 2 = cu 2 .
191 MPa

348 MPa
s1 = Es . s1
435 MPa

522 MPa

4. G

t t1
t

(S220)
(S400)

s 2 = Es . s 2

(S500)
(S600)

191 MPa

348 MPa

435 MPa

522 MPa

(S220)
(S400)
(S500)
(S600)

with

Es = 200 GPa = 200000 MPa

The table thus only allows making a faster transition from d towards the stresses s1
and s2 for different values of t2. Table A4.4.4.6 is elaborated for commonly

4-59

encountered values of t2: 0,06 ; 0,08 ; 0,10 ; 0,12 ; 0,14 and 0,16. The table is function
of the selected concrete class (the value of cu2 and the equation of the parabolarectangle diagram), but is independent of the choice of ud for steel.
Notes concerning table A4.4.4.6
The link between d and t is given by (4.4.4-5) and (4.4.4-6):

d =

N d .e2 As1. s1.(d d 2 )

= . t .( G . t t 2 )
b.h 2 . f cd

Limit on the left side of domain 2b: values of t

x=d

t =

x d h d1
= =
= 1 t1
h h
h

table 4.4.4-1

Table 4.4.4-1
Domain 2b: left limit for different values of t1
t1
0,06 0,08
0,10
0,12
0,94 0,92
0,90
0,88
t
Concrete classes up to C50/60
= 0,805
d 0,252 0,226 0,200 0,176
G = 0,418
Concrete class C55/67
= 0,742
d 0,216 0,193 0,170 0,148
G = 0,394
Concrete class C60/75
= 0,692
d 0,192 0,170 0,149 0,129
G = 0,378
Concrete class C70/85
= 0,625
d 0,164 0,145 0,126 0,109
G = 0,361
Concrete class C80/95
= 0,596
d 0,154 0,136 0,118 0,101
G = 0,356
Concrete class C90/105
= 0,581
d 0,149 0,131 0,114 0,098
G = 0,354

4-60

0,14
0,86

0,16
0,84

0,152

0,129

0,127

0,106

0,110

0,091

0,092

0,075

0,085

0,070

0,082

0,067

 Limit on the right side of domain 2b

x=h

t =

x d
= =1
h h

d = . (G t2) table 4.4.4-2

Table 4.4.4-2
Domain 2b: right limit for different values of t2
t2
0,06 0,08
0,10
0,12
1
1
1
1
t
Concrete classes up to C50/60
= 0,805
d 0,288 0,272 0,256 0,240
G = 0,418
Concrete class C55/67
= 0,742
d 0,248 0,233 0,218 0,203
G = 0,394
Concrete class C60/75
= 0,692
d 0,220 0,206 0,192 0,178
G = 0,378
Concrete class C70/85
= 0,625
d 0,188 0,176 0,163 0,151
G = 0,361
Concrete class C80/95
= 0,596
d 0,176 0,164 0,152 0,141
G = 0,356
Concrete class C90/105
= 0,581
d 0,171 0,159 0,148 0,136
G = 0,354

4.4.5

0,14
1

0,16
1

0,224

0,208

0,188

0,173

0,164

0,151

0,138

0,126

0,129

0,117

0,124

0,113

Analysis of domain 3

4.4.5.1 When?
A strain diagram in this domain can only be obtained if the compression becomes far
more important than the bending; Nd is applied with a very small eccentricity eo.

4.4.5.2 Compatibility equations

The whole concrete cross-section is in compression. The NA is situated between the
lower side of the section and the infinity. The strain diagram turns around the third
hinge: point C situated at a distance

4-61



h1 c 2
cu 2

from the most compressed fibre at the upper side: see figure 4.4.5-1.
d2

c_top
s2

As2
d

c2

As1

h1 c 2
cu 2

s1
c_bottom

d1

NA

Figure 4.4.5-1
Domain 3; auxiliary figure for the elaboration of the compatibility equations
The equations are the following:

c _ top = c 2 .

s2 = c2 .

x h1 c 2
cu 2

(4.4.5-1)

t 1 c 2
cu 2

t t2
x d2
= c2 .

t 1 c 2
x h1 c 2
cu 2
cu 2

c _ bottom = c 2 .

s1 = c 2 .

= c2 .

t 1
xh
= c2 .

t 1 c 2
x h1 c 2
cu 2
cu 2

(4.4.5-2)

+ t1 1
x (h d1 )
xd
= c2 .
= c2 . t

t 1 c 2
x h1 c 2
x h1 c 2
cu 2
cu 2
cu 2

4-62

(4.4.5-3)

(4.4.5-4)

Used notations for the steel reinforcement: index 1 is associated with the lower side (the
least compressed side); index 2 is associated with the upper side (the most compressed
side). In order to distinguish from c2 (see the design stress-strain diagram for concrete),
the designations c_top and c_bottom are used for the concrete strains instead of c1 and c2.

4.4.5.3 Use of the parabola-rectangle diagram

The determination of the area of the concrete stress diagram and the position of the
centre of gravity, is a particular problem, because a part of the parabola is not taken into
account: see figure 4.4.5-2. Attention: the filling coefficient and the coefficient of the
centre of gravity G are now defined with respect to the total depth h (instead of d).
fcd
xG = G.h

h1 c 2
cu 2

h
x

x-h

Figure 4.4.5-2
Domain 3; auxiliary figure for the determination of the area of the stress diagram and its
centre of gravity
The results, obtained by numerical integration of the equation of the stress diagram for
the different classes of concrete, are shown in table 4.4.5-1.

4-63

Table 4.4.5-1
Domain 3; filling coefficient and coefficient of the centre of gravity G in function of
t = x/h, for different concrete classes
C50/60

C55/67

C60/75

1
1,2
1,4
1,6
1,8
2
2,5
3
4
5

0,8095
0,8955
0,9341
0,9547
0,9669
0,9748
0,9855
0,9906
0,9951
0,9970
1

0,4160
0,4583
0,4748
0,4831
0,4878
0,4908
0,4948
0,4967
0,4983
0,4990
0,5

0,7440
0,8347
0,8834
0,9129
0,9321
0,9454
0,9651
0,9755
0,9858
0,9906
1

0,3927
0,4384
0,4591
0,4705
0,4776
0,4823
0,4892
0,4927
0,4963
0,4980
0,5

0,6934
0,7854
0,8396
0,8746
0,8986
0,9159
0,9430
0,9583
0,9743
0,9822
1

0,3769
0,4240
0,4470
0,4603
0,4689
0,4748
0,4836
0,4884
0,4934
0,4959
0,5

C70/85

C80/95

C90/105

1
1,2
1,4
1,6
1,8
2
2,5
3
4
5

0,6266
0,7193
0,7788
0,8196
0,8491
0,8712
0,9077
0,9296
0,9538
0,9667
1

0,3600
0,4084
0,4334
0,4485
0,4585
0,4656
0,4765
0,4827
0,4894
0,4928
0,5

0,5976
0,6908
0,7523
0,7954
0,8271
0,8512
0,8916
0,9163
0,9442
0,9592
1

0,3548
0,4036
0,4292
0,4448
0,4552
0,4626
0,4742
0,4808
0,4880
0,4918
0,5

0,5831
0,6769
0,7395
0,7839
0,8167
0,8418
0,8842
0,9102
0,9399
0,9559
1

0,3531
0,4020
0,4278
0,4436
0,4541
0,4617
0,4735
0,4803
0,4876
0,4915
0,5

Note regarding table 4.4.5-1:

it is observed that tends rapidly to 1 and that G tends rapidly to 0,5; for these values,
the parabola-rectangle diagram is transformed into a rectangular diagram with depth h
and width fcd.

4.4.5.4 Equilibrium equations

See figure 4.4.5-3.

4-64

d2

c_top
s2

As2

h1 c 2

cu 2

As2.s2

s2

Nc=

.b.h.fcd

x
As1

s1

s1
d1

e2

xG = G.h

Nd

c2

fcd

c_bottom

e0
O

As1.s1

Figure 4.4.5-3
Basic figure for the analysis of bending combined with compression in domain 3

Translation equilibrium:

N d = .b.h. f cd + As1. s1 + As 2 . s 2

(4.4.5-5)

Rotation equilibrium around the most compressed reinforcement As2:

N d .e2 = .b.h. f cd .( G .h d 2 ) + As1. s1 (d d 2 )

(4.4.5-6)

or

N d .e2 = .b.h 2 . f cd .( G t 2 ) + As1. s1 (d d 2 )

or

N d .e2 As1. s1.(d d 2 )

= .( G t 2 )
b.h 2 . f cd

(4.4.5-7)

with e2 = eccentricity of Nd with respect to As2.

First, the reduced moment is defined as:

d =

N d .e2 As1. s1.(d d 2 )

b.h 2 . f cd

(4.4.5-8)

d = .( G t 2 )

(4.4.5-9)

Expression (4.4.5-7) then gives:

4-65

In summary:

d = f ( , G , t 2 )

with et G = f ( t )

Inversely, one may also express t in function of d.

4.4.5.5 Solution scheme

Exploitation of the rotation equilibrium
The first choice is to put s1= 0; the equation (4.4.5-8) then gives:
( d )1 =

N d .e2
b.h 2 . f cd

With a chosen value of t2 and on the basis of the equation of the parabolarectangle diagram of the selected concrete class, one finds ()1, ()1 and (t)1.
Exploitation of the compatibility equations
The equations (4.4.5-1 to 4.4.5-4) allow the determination of (c1)1, (c2)1, (s1)1
and (s2)1.
Exploitation of the design stress-strain diagram of steel
(s1)1 and (s2)1 give (s1)1 and (s2)1
Exploitation of the translation equilibrium
Again (such as in domain 2b), there are still two unknown parameters: As1 and
As2, while all available equations have been exploited. In order to overcome this
difficulty, the ratio of the reinforcement areas is used again:

As1
As 2

The translation equilibrium equation is then written as:

N d .b.h. f cd = As 2 .( . s1 + s 2 )
and thus:

As 2 =

N d .b.h. f cd
s 2 + . s1

The choice of a value for , leads via (4.4.5-10) to:

4-66

(4.4.5-10)

( As1 )1 =

N d ( )1.b.h. f cd
( s 2 )1 + .( s1 )1

and
( As1 )1 = .( As 2 )1
This first approximation is in general not conclusive in domain 3, because the first
choice (s1)1 = 0 is less close to reality (than in domain 2b). The calculation
sequence has to be started up again with a new approximate value of d:

d =

N d .e2 ( As1 )1.( s1 )1.(d d 2 )

b.h 2 . f cd

The iterative calculation has to be continued until the approximate solution is

sufficiently acceptable. In practice, it is observed that two iteration steps are in
general sufficient.

4.4.5.6 Auxiliary table for shortening hand calculations

In order to shorten hand calculations, annex A4.4.5.6 presents practical tables which are
composed on the basis of the following scheme:
S220

t1 = t2 = . . . .

s1

s2

s1

MPa

s2

MPa

S400

s1

MPa

s2

MPa

with t2 chosen and with t1 = t2, and the following formulas:

d =

N d .e2 As1. s1.(d d 2 )

b.h 2 . f cd

s1 = c 2 .

s2 = c2 .

t + t1 1

t (1 c 2 )
cu 2
t t2

t (1 c 2 )
cu 2

4-67

S500

s1

MPa

s2

MPa

S600

s1

MPa

s2

MPa

191 MPa

348 MPa
s1 = Es . s1
435 MPa

522 MPa

(S220)
(S400)

s 2 = Es . s 2

(S500)
(S600)

191 MPa

348 MPa

435 MPa

522 MPa

(S220)
(S400)
(S500)
(S600)

with

Es = 200 GPa = 200000 MPa

The table thus only allows making a fast transition from d towards the stresses s1 and
s2 via , t, t, s1 and s2 for different values of t2 = t1. The annex A4.4.5.6 contains
tables A4.4.5.6-1 to A4.4.5.6-6 which are elaborated for the different concrete classes;
the tables are also characterized by the designation "a" to "f" in function of 6 commonly
encountered values for t2 : 0,06; 0,08; 0,10; 0,12; 0,14 and 0,16.
Notes concerning tables A4.4.5.6
Left limit of domain 3

x=h

t =
d = .( G t 2 )

x h
= =1
h h

table 4.4.5-2

Table 4.4.5-2
Domain 3: left limit for different values of t2
t2
0,06 0,08
0,10
0,12
1
1
1
1
t
Concrete classes up to C50/60
= 0,805
d 0,288 0,272 0,256 0,240
G = 0,418
Concrete class C55/67
= 0,742
d 0,248 0,233 0,218 0,203
G = 0,394
Concrete class C60/75
= 0,692
d 0,220 0,206 0,192 0,178
G = 0,378
Concrete class C70/85
= 0,625
d 0,188 0,176 0,163 0,151
G = 0,361
Concrete class C80/95
= 0,596
d 0,176 0,164 0,152 0,141
4-68

0,14
1

0,16
1

0,224

0,208

0,188

0,173

0,164

0,151

0,138

0,126

0,129

0,117

G = 0,356
Concrete class C90/105
= 0,581
G = 0,354

0,171 0,159

0,148

0,136

0,124

0,113

0,14

0,16

0,36

0,34

0,36

0,34

0,36

0,34

0,36

0,34

0,36

0,34

0,36

0,34

Right limit of domain 3

x=

t =
=1
t = 0,5
d = 0,5 t 2

table 4.4.5-3

Table 4.4.5-3
Domain 3: right limit for different values of t2
t2
0,06 0,08
0,10
0,12

t
Concrete classes up to C50/60
0,40
0,38
=1
d 0,44 0,42
G = 0,5
Concrete class C55/67
=1
0,40
0,38
d 0,44 0,42
G = 0,5
Concrete class C60/75
=1
0,40
0,38
d 0,44 0,42
G = 0,5
Concrete class C70/85
=1
0,40
0,38
d 0,44 0,42
G = 0,5
Concrete class C80/95
=1
0,40
0,38
d 0,44 0,42
G = 0,5
Concrete class C90/105
=1
0,40
0,38
d 0,44 0,42
G = 0,5

4-69

4.4.6

Right limit of domain 3: pure compression

4.4.6.1 Principle of the calculation

Figure 4.4.6-1 presents a rectangular cross-section loaded in pure compression: the
strain for the whole section is c2.
d2

c2

fcd

As2

Ns2

sd

d
Nc

As1

sd

Nd

Ns1

d1
b

(a)

(b)

(c)

(d)

Figure 4.4.6-1
Basic figure for the analysis of pure compression; rectangular cross-section;
(a) strain diagram; (b) stress distribution diagram;
(c) resultant forces for the stress blocs; (d) imposed load
The design problem can be solved by means of the horizontal translation equilibrium
equation:

Nd = Nc + Ns1 + Ns2
Nd = Ac . fcd + (As1 + As2) . sd
with sd specified in figure 4.4.6-2 and in table 4.4.6-1.

4-70

(4.4.6-1)

C70/85
C60/75
C80/95
C55/67
C90/105
C50/60

s (MPa)
522

S600
S500

435

S400

348

S220

191

0,096

0,174

0,217

0,261

s (%)

Figure 4.4.6-2
Stress in the steel reinforcement for strain s = c2

Table 4.4.6-1
Stress in the steel reinforcement for strain s = c2
S220
S400
S500

Concrete
class

c2

(fyd= 191 MPa)

(fyd= 348 MPa)

(fyd= 434 MPa)

(fyd= 522 MPa)

C50/60
C55/67
C60/75
C70/85
C80/95
C90/105

2,00
2,20
2,29
2,42
2,52
2,60

fyd
fyd
fyd
fyd
fyd
fyd

fyd
fyd
fyd
fyd
fyd
fyd

400 MPa
fyd
fyd
fyd
fyd
fyd

400 MPa
440 MPa
458 MPa
484 MPa
504 MPa
520 MPa

S600

4.4.6.2 Second order effects

In principle, pure compression is a loading case which is encountered typically for the
calculation of columns. In the design of columns, one has to take care of the eventual
second order effects (buckling) in function of the slenderness of the column. Second
order effects are discussed in the chapter on columns, further in these course notes,
where particular attention will be paid to the conditions indicating whether or not
buckling analysis of a column is necessary.
However, at this stage of the course, it is important to note that even if it is not
necessary to perform a buckling analysis (because of small slenderness), several
uncertainties have to be taken into account:
cases of pure compression are very sensitive to possible second order effects due
to imperfections in materials and cross-sections;

4-71

 in practice, it is almost impossible to assure the application of an axial force

exactly in the centre of gravity of the cross-section of the column. This is for
example due to:
- tolerances related to the casting of the column or of the construction element
that is supported by the column;
- the distorted stress distribution in the neoprene support pad due to a (small)
rotation angle of the supported element (see figure 4.4.6-3).

(a)

(b)

Figure 4.4.6-3
Eccentric application of the compression force onto the column, due to a small
rotation angle; (a) support pad in neoprene;
(b) stress distribution in neoprene pad
EN 1992-1-1:2004; 5.8.2(5) proposes to cover this problem by the introduction of an
additional 1st order eccentricity defined on the basis of the geometrical imperfections
(which have already been discussed in paragraph 2.5.2.3 in these course notes). A
structural member subjected to pure compression, which is not sensitive to second order
effects because of its reduced slenderness, has thus to be calculated for an eccentrically
applied axial compression force, applied with the additional eccentricity ea. In that way,
the design problem of columns is translated into a problem of bending combined with
compression (with small eccentricity, and thus in domain 3).
As a reminder (on the basis of paragraph 2.5.2.3 in these course notes):

ea = i . l0 / 2

(4.4.6-2)

in which:

l0 =

i =

the effective length of the isolated element; indeed, the formula is elaborated for
an isolated column, fixed at one end, free at the other end, characterized by an
effective length = 2 (length of the column).
the rotation angle with respect to the vertical line, which characterizes the
inclination of the element.

Note:
For walls and isolated columns in braced systems, ea = l0 /400 may be assumed
for the purpose of simplification.
4-72

4.5 Bending combined with tension

Preliminary note: the developments in this paragraph are limited to rectangular crosssections, but extrapolation of the method towards sections with variable shapes is
straightforward.
4.5.1

Domains to be analysed

Table 4.1.4-1 shows that ULS strain diagrams corresponding to bending in combination
with axial tension, are situated in the domains 1a, 1b and 2a.
4.5.2

Definition of the design problem

Given:
the quality of the selected materials: fyd and fcd;
the dimensions of the concrete cross-section: b, h (d), d1 and d2; see figure 4.5.21;
the design values of the imposed internal forces Md and Nd; the imposed loads
may also be defined as a axial tensile force Nd that is applied with an eccentricity
e0 to the geometric centre of the cross-section. To be remembered: the geometric
centre (or centroid) is used instead of the centre of gravity because the latter is as
yet unknown; in this stage of the project, only architectural plans are available.
Question: the reinforcement areas As1 and As2 = ?
d2
As2

Md

Nd

e0
Nd

As1
d1
b

Figure 4.5.2-1
Basic figure for the analysis of bending combined with tension

4.5.3

Analysis of the domains 1b and 2a

4.5.3.1 When?
A strain diagram in these two domains can only be obtained if the bending is much
more important than the tension; Nd is applied with a large eccentricity e0.
4-73

4.5.3.2 Basic equations

The formulas in this paragraph are developed for a strain diagram in domain 2a; the
developments in domain 1b are completely similar.
Principle figure: see figure 4.5.3-1.
c=cu2
s2

d2
As2

s2

c2

x
d

fcd
Ns2=As2.s2
Nc=

.b.x.fcd

d-G.x
e0

As1
d1
C

s1

s1

Ns1=As1.s1

e1
Nd

Figure 4.5.3-1
Principle figure for the analysis of bending combined with tension in domain 2a
Horizontal translation equilibrium:

Nd = Ns1 - Nc - Ns2
or

Nd = As1 . s1 - . b . x . fcd As2 . s2

or

. b . x . fcd + As2 . s2 = As1 . s1 Nd

and thus

.b.x. f cd + As 2 . s 2 = ( As1

Nd

s1

). s1

(4.5.3-1)

Rotation equilibrium around the tensile reinforcement As1:

Nd . e1 = . b . x . fcd . (d G . x) + As2 . s2 . (d d2)

with e1 = eccentricity of Nd with respect to As1.

4-74

(4.5.3-2)

These equations are the same as the ones for bending combined with compression in the
domains 1b and 2a: see equations (4.4.3-1) and (4.4.3-2); the only difference in equation
(4.5.3-1) with respect to equation (4.4.3-1) is the minus sign (instead of a plus sign)
before the term Nd/s1. The solution scheme is thus quasi identical to the one for
bending combined with compression.

4.5.3.3 First case: As2 = 0

The solution scheme comprises the following steps:
Nd and e0 are given deduction of e1;
calculation of:

d =

N d .e1
b.d 2 . f cd

by means of table A4.2.3.7, one finds:

As f yd
.
b.d f cd

and

s1

subtraction of As1 from the value of As that has been obtained just before:
As1 = As +

Nd

s1

(4.5.3-3)

Attention to the plus sign instead of the minus sign in expression (4.4.3-5)

4.5.3.4 Second case: As2 0

The solution scheme comprises the following steps:
assumption that As2 = 0 and calculation of:

d =

N d .e1
b.d 2 . f cd

If d > lim in table A4.2.3.7, a compression reinforcement is necessary in order

to reduce d by a term which is function of As2, causing d to fall into the useful
part of table A4.2.3.7
Calculation of As2 minimum via:

As 2 min = As 2 lim =

M d lim .b.d 2 . f cd
s 2 .(d d 2 )

4-75

Attention: Md is replaced by Nd . e1!

Choice of As2;
calculation of:

n =

N d .e1 As 2 . s 2 .(d d 2 )
b.d 2 . f cd

This n thus corresponds to the moment Mn which is resisted by the simply

reinforced section (see decomposition of the doubly reinforced section, presented
in figure 4.2.7-4). For this n, one finds a corresponding value of in table
A4.2.3.7, which then permits to determine Asn;
calculation of:
Asc =

As 2 . s 2

s1

and finally:
As1 = Asn + Asc +

4.5.4

Nd

(4.5.3-4)

s1

Analysis of domain 1a

4.5.4.1 When?
A strain diagram in this domain can only be obtained if the tension is far more important
than the bending; Nd is applied with a small eccentricity eo.

4.5.4.2 Equilibrium equations

See figure 4.5.4-1; the whole cross-section is in tension; concrete does not appear in the
design equations.

d2

s2

s2

As2

Ns2=As2.s2
e2

Nd
As1

s1=ud

s1

Ns1=As1.s1

e1

d1

Figure 4.5.4-1
Basic figure for the analysis of bending combined with tension in domain 1a

4-76

The rotation equilibrium equation is written two times, one time around the lower
reinforcement, one time around the upper reinforcement:
external moment = internal moment
Thus:

Nd . e1 = As2 . s2 . (d d2)

(4.5.4-1)

Nd . e2 = As1 . s1 . (d d2)

(4.5.4-2)

and

in which s1 = s1 (s1 = ud) = fyd for all steel grades S220, S400, S500 and S600.
Expression (4.5.4-2) leads to:
As1 =

N d .e2
s1 .(d d 2 )

with

s1 = fyd

The area As2 cannot be deduced from expression (4.5.4-1) because s2 is not known.
This problem is overcome by the introduction of :
As1
As 2

=
And thus:

As 2 =

As1

(4.5.4-3)

An optimal solution can be realised by imposing that the stresses in both reinforcements
are equal to the yield value fyd.
Division of (4.5.4-1) by (4.5.4-2) leads to:
N d .e1 As 2 . s 2 .(d d 2 )
=
N d .e2
As1 . s1 .(d d 2 )

and consequently:

s 2 = s1 .

e1 As1
e
.
= s1 . 1 .
e2 As 2
e2

4-77

(4.5.4-4)

The most economical solution is found for:

s2 = s1 = fyd
and thus for

e1
. = 1
e2

optimal =

or

e2
e1

or finally:
As1 e2
=
As 2 e1

The last expression means that the tensile load Nd should be applied in the centre of
gravity of both reinforcements, which is an obvious result (principle of the lever arms).

4.5.5

Left limit of domain 1a: pure tension

The tensile force Nd is applied in the geometric centre 0 of the cross-section (see figure
4.5.4-1):

e1 = e2 =

d d2
2

optimal =

e2
=1
e1

The design formula is thus:

As1 = As 2 =

Nd / 2
Nd
=
( s ) = ud 2. f yd

(4.5.5-1)

4.6 Interaction diagrams

4.6.1

Introduction

The reasoning aiming at the introduction and definition of the interaction diagrams
starts with the discussion of the design equations associated with one particular ULS
strain diagram. The following developments are based on the choice of an ULS strain
diagram in domain 3: see figure 4.6.1-1.

4-78

d2

s2

As2

c_top

fcd

s2

h1 c 2

cu 2

xG = G.h

s1
d1

Nd

c2

e0

Nc=

.b.h.fcd

x
As1

As2.s2

s1
c_bottom

As1.s1

Figure 4.6.1-1
Basic figure for the analysis of bending combined with compression in domain 3
Translation equilibrium:
Nd = . b . h . fcd + As1 . s1 + As2 . s2
or
Nd = . b . h . fcd + As2 . (s2 + . s1) with = As1 / As2
which gives
As 2 =

N d .b.h. f cd
. s1 + s 2

(4.6.1-1)

Rotation equilibrium, written around the geometric centre 0:

h
h
h
N d .e0 = . f cd .b.h.( G .h) + As 2 . s 2 .( d 2 ) As1 . s1 .( d1 )
2
2
2
or
N d .e0 = . f cd .b.h 2 .( 1 G ) + As 2 . s 2 .h.( 1 t 2 ) . As 2 . s1 .h.( 1 t1 )
2
2
2
or
N d .e0 = . f cd .b.h 2 .( 1 G ) + As 2 .h.( 1 t ).( s 2 . s1 )
2
2
As2 in (4.6.1-2) is replaced by (4.6.1-1):

4-79

(4.6.1-2)

N .b.h. f cd
N d .e0 = . f cd .b.h 2 .( 1 G ) + h.( 1 t ).( s 2 . s1 ). d
2
2
. s1 + s 2

or, by rearranging terms:

( 1 t ).( s 2 . s1 ) h.( 1 t ).( s 2 . s1 )

2
1
.N d
+

N d .e0 = . f cd .b.h . ( G ) 2
s 2 + . s1
s 2 + . s1

(4.6.1-3)
Equation (4.6.1-3) presents the relationship between the bending moment Md = Nd . e0
and the axial compression force Nd. One observes that:
for a given set of values for , t, b, h, fyd, fcd, and
for a chosen ULS strain diagram (for which G, s1 and s2 are thus known),
the relationship (4.6.1-3) is a linear function Md = f(Nd), represented by a straight line in
figure 4.6.1-2.

Md = f(Md)

Md
Nd

Figure 4.6.1-2
Md = f(Nd), associated with one particular ULS strain diagram in domain 3, and taking
into account a set of chosen values for , t, b, h, fyd and fcd, appears to be a linear
function

Note :
By applying the procedure demonstrated above (exploitation of translation and
rotation equilibrium equations), one obtains a straight line Md = f(Nd) associated
with each ULS strain diagram (also for those strain diagrams situated in other
domains than domain 3).

Each point, represented by the coordinates (Md,Nd), of the straight line corresponds to a
certain amount of reinforcement; for the case of domain 3, it may be reminded that the
reinforcement areas are given by:
As 2 =

N d .b.h. f cd
s 2 + . s1

4-80

As1 = . As2
and thus
As tot = As1 + As2
which leads to a mechanical reinforcement ratio:

As tot f yd
.
b.h f cd

Each point of the line in figure 4.6.1-2 corresponds to a specific value of .

4.6.2

The interaction diagram

4.6.2.1 Elaboration and significance

An interaction diagram is elaborated for the given cross-section, by considering all
possible ULS strain diagrams in all domains 1a to 3, and thus the whole range of values
of = x/d varying between - and +: see figure 4.6.2-1.

c2

ud

cu

cu2

cu2

c2
d2

su

cu2

h
d1

ud

ud

ud

ud

0,167

0,259

c2
1,0

1+

d1
d

Figure 4.6.2-1
Schematic overview of the different positions of the ULS strain diagram for a given
cross-section, to be considered for the elaboration of an interaction diagram. The values
mentioned for are those corresponding to c2 = 0,002; cu = 0,0035 and ud = 0,01
The resisting internal forces Nd and Md are calculated for each position of the ULS
strain diagram, on the basis of the stress-strain diagrams for steel and concrete, for
different values of As1 and for the chosen parameter = As1 / As2.
The resulting diagram comprises all combinations of Nd and Md that are possible for the
given cross-section. Figure 4.6.2-2 shows such an interaction diagram in a schematic
way. The diagram is composed of:
4-81

- a tree of straight lines which correspond to the ULS strain diagrams, and
- a series of iso-reinforcement curves.
It is important to notice that there is only 1 interaction diagram for the following set of
chosen parameters:
the shape and dimensions of the cross-section (b and h for a rectangular section);
the materials properties: fyd et fcd;
the ratio = As1 / As2;
t (= t1 = t2)

compression
iso-reinforcement
curve

1 interaction diagram for:

= As1/As2
d1/h = d2/h
fyd
fcd
b, h (section)

=Nd.e0

tension

Figure 4.6.2-2
Schematic representation of an interaction diagram

The interaction diagrams are useful to solve the following problems:

determination of the maximum axial force that can be resisted by a given crosssection (geometry, As1 and As2) which is loaded by an imposed bending moment ;
determination of the maximum bending moment that can be resisted by a given
cross-section (geometry, As1 and As2) which is loaded by an imposed axial force;
determination of the reinforcement areas As1 and As2 of a given cross-section with
known geometry, subjected to bending combined with an axial force. The
comparison of different interaction diagrams, elaborated for different values of
As2/As1, allows the determination of the minimum value of the total reinforcement
area As1 + As2.

4-82

The interaction diagrams can also be used for the case of pure bending (Nd = 0) by
considering the values on the Md axis.
4.6.2.2 Further generalisation
In order to reduce the number of parameters associated with each interaction diagram,
the last ones are represented in a system of axis corresponding to reduced moments
and to reduced axial forces ; moreover, the iso-reinforcement curves correspond to
the mechanical reinforcement ratio : see figure 4.6.2-3. This permits a more rational
representation of interaction diagrams, which are now independent from the concrete
class and the dimensions b and h. With this way of representation, there is only one
interaction diagram for each chosen set of the following parameters:
the shape of the cross-section (rectangular, circular, etc.);
the steel grade fyd;
the reinforcement ratio = As1 / As2;
t (= t1 = t2).

As tot f yd
.
b.h f cd

with

As tot = As1 + As 2

Md
b.h 2 . f cd

1 diagram for:

Nd
b.h. f cd

shape of cross-section
= As1/As2

fyd

Figure 4.6.2-3
Schematic representation of a practical interaction diagram by using reduced loads

4.6.3

Examples of interaction diagrams

Examples of interaction diagrams can be found in typical vade-mecums related to

concrete design. Some of these examples are shown in annex A4.6.3.
Figures A4.6.3-1 to A4.6.3-5 (LAMBOTTE, 1988) are valid for the following
parameters:
rectangular section;
= 1 (As1 = As2);
t = 0,10;
steel grades S220, S400 or S500:
4-83

figure A4.6.3-1: S220;

figure A4.6.3-2: S400;
figure A4.6.3-3: S400; enlargement of figure A4.6.3-2;
figure A4.6.3-4: S500;
figure A4.6.3-5: S500; enlargement of figure A4.6.3-4.
Figures A4.6.3-6 to A4.6.3-9 (BETONVERENIGING, 1997) show an alternative
representation and are valid for the following parameters:
figure A4.6.3-6: rectangular section, = 1, t = 0,10 ;
figure A4.6.3-7: rectangular section, = 1, t = 0,15 ;
figure A4.6.3-8: rectangular section, = 1, t = 0,10 ;
figure A4.6.3-9: rectangular section, = 1, t = 0,15 ;

4-84

5 Chapter 5
Serviceability limit states (SLS)
5.1 Introduction
EN 1992-1-1:2004; 7.1(1) identifies the common serviceability limit states (SLS):
- stress limitation;
- crack control;
- deflection control.
Other limit states (such as vibrations) may be of importance in particular structures, but
are not covered in Eurocode 2.
The observation should be made here that, with respect to the earlier versions of
Eurocode 2, there is an important evolution in prescriptions dealing with durability. A
better knowledge of degradation mechanisms in concrete and damage mechanisms such
as corrosion of steel reinforcement, has lead to the introduction of new prescriptions
aiming to help in avoiding the damage mechanisms. This is why today, it is asked to
apply specific rules regarding stress control, crack limitation etc. in function of the
environmental conditions, already from the very beginning of the design project. That is
the reason why this chapter on SLS starts with an overview of the exposure classes
and environment classes.

5.2 Exposure classes and environment classes

Exposure classes define chemical and physical conditions to which a structure is
exposed in addition to the mechanical actions. Table 5.2-1 in these course notes shows
an overview of the 18 exposure classes, defined in EN 1992-1-1:2004; 4.2(2) table
4.1, in full accordance with EN 206-1. The 18 exposure classes are grouped in 6
principal classes which depend on one or another aggressive action. The designation is
as follows: character X followed by a second character which indicates the aggressive
action:
C for carbonation;
D for deicing salt (attack by chlorides);
S for sea water (attack by chlorides);
F for frost/thaw attack ;
A for aggressive environment (chemical attack).
This character is followed by a number which corresponds to a particular characteristic
of the attack conditions. The table also gives informative examples where the exposure
classes may occur.
NBN EN 1992-1-1 ANB:2009 stipulates that table 4.1 in EN 1992-1-1:2004 must be
completed by table 1 in NBN B15-001; this last table is reproduced in table 5.2-2 in
these course notes, which shows an overview of the 13 so-called environment classes
(terminology used in NBN B15-001). The table also presents the agreement between the
the environment classes (NBN-ANB) and the exposure classes (EN).

5-1

Note: The composition of the concrete affects both the protection of the
reinforcement and the resistance of the concrete to possible environmental
attacks. This may lead in certain cases, to the choice of higher strength classes
than required for the structural design. Table 5.2-3 in these course notes
reproduces the prescriptions on this topic mentioned in EN 1992-1-1:2004;
Annex E - table E.1N.
Regarding this topic, NBN EN 1992-1-1 ANB:2009 refers to NBN EN 206-1
and NBN B15-001.

5-2

Table 5.2-1
Exposure classes related to environmental conditions, according to EN 1992-1-1:2004, based
on EN 206-1 (tableau 4.1 de l' EN 1992-1-1:2004)
Exposure classes
Classes dexposition
Blootstellingsklassen

NBN EN 206-1

The European norm NBN EN 206-1 defines 18 exposure classes for concrete, related to environmental effects.
La norme europenne NBN EN 206-1 dfinit 18 classes dexposition pour le bton, en fonction des effets de
lenvironnement.
De Europe norm NBN EN 206-1 bepaalt 18 blootstellingsklassen voor beton, in functie van de invloed van de omgeving.
Class
Classe
Klasse

Description
Description
Omschrijving

Examples
Exemples
Voorbeelden

X0 : no risk of corrosion or attack

X0 : pas de risqu de corrosion ou dattaque
X0 : geen risico op corrosie of aantasting

X0

Concrete without reinforcement : all exposures

except where there is freeze/thaw, abrasion or
chemical attack
Concrete with reinforcement : very dry

Concrete inside buildings with very low air humidity

Bton non arm : tous les environnements, sauf avec

gel/dgel, abrasion ou attaques chimiques
Bton arm : trs sec
Beton zonder wapening : alle milieus behalve bij
vorst/dooi, afslijting of chemische aantasting
Beton met wapening : zeer droog

Bton lintrieur avec humidit de lair trs faible

Beton binnen gebouwen met zeer lage luchtvochtigheid

XC : Corrosion induced by carbonation (concrete with reinforcement exposed to air and moisture)
XC : Corrosion provoque par la carbonatation (bton arm expos lair et lhumidit)
XC : Corrosie genitieerd door carbonatatie (beton met wapening blootgesteld aan lucht en vocht)

XC1

Dry or permanently wet

- Concrete inside buildings with low air humidity

- Concrete permanently submerged in water

Sec ou humide en permanence

- Bton lintrieur de btiments o le taux dhumidit

de lair est faible
- Bton constamment immerg

Droog of blijvend nat

- Beton binnen gebouwen met lage luchtvochtigheid

- Beton blijvend onder water

5-3

XC2

Wet, rarely dry

- Concrete surfaces subject to long-term water contact

- Many foundations

Humide, rarement sec

- Surfaces de bton en contact long terme avec leau

- Nombreuses fondations

Nat, zelden droog

- Betonoppervlakken langdurig in contact met water

- Veel funderingen

Moderate humidity

- Concrete inside buildings with moderate or high air

humidity
- External concrete sheltered from rain

Humidit modre

- Bton lintrieur de btiments o le taux dhumidit

de lair est modr ou lev
- Bton lextrieur, abrit de la pluie

XC3

- Beton binnen gebouwen met matige of hoge

luchtvochtigheid
- Buiten beschut tegen regen

Matige vochtigheid

XC4

Cyclic wet and dry

Concrete surfaces subject to water contact, not within

exposure class XC2

Alternance dhumidit et de schage

Surfaces de bton en contact avec leau, ne faisant pas

partie de la classe XC2

Wisselend nat en droog

Betonoppervlakken blootgesteld aan contact met water,

niet onder klasse XC2

XD : Corrosion induced by chlorides other than from sea water (concrete with reinforcement subject to contact with
water containing chlorides, including de-icing salts, from sources other than from sea water)
XD : Corrosion provoque par des chlorures, de provenance autre que leau de mer (bton arm en contact avec de leau
contenant des chlorures, y compris les sels de dneigement, hormis leau de mer)
XD : Corrosie genitieerd door chloriden uit andere bronnen dan zeewater (beton met wapening in contact met water dat
chloriden, inclusief dooizouten, bevat die komen uit andere bronnen dan zeewater)

XD1

Moderate humidity

Concrete surfaces exposed to airborne chlorides

Humidit modre

Surfaces de bton exposes des chlorures dans lair

Matige vochtigheid

Betonoppervlakken blootgesteld aan chloriden uit de

lucht

Wet, rarely dry

- Swimming pools
- Concrete exposed to industrial waters containing
chlorides

Humide, rarement sec

- Piscines
- Bton expos des eaux industriels contenant des
chlorures

XD2

- Zwembaden
- Beton blootgesteld aan chloridehoudend industriewater

Nat, zelden droog

5-4

Cyclic wet and dry

- Parts of bridges exposed to spray containing

chlorides
- Pavements
- Car park slabs

Alternance dhumidit et de schage

- Parties de ponts exposes des projections contenant

des chlorures
- Pavages
- Stationnements de vhicules

XD3

- Brugdelen blootgesteld aan chloridehoudend spatwater

- Verhardingen
- Parkeerplaatsen voor voertuigen

Wisselend nat en droog

XS : Corrosion induced by chlorides from sea water (concrete with reinforcement subject to contact with chlorides
from sea water or air carrying salt originating from sea water)
XS : Corrosion provoque par des chlorures deau de mer (bton arm en contact avec des chlorures deau de mer ou de
lair contenant des sels dorigine marine)
XS : Corrosie genitieerd door chloriden uit zeewater (beton met wapening blootgesteld aan chloriden afkomstig uit
zeewater of aan lucht die zouten bevat uit de zee)

XS1

XS2

XS3

Exposed to airborne salt but not in direct contact

with sea water

Structures near to or on the coast

Expos lair vhiculant du sel marin, mais pas en

contact direct avec leau de mer

Construction prs de ou la cte

Blootgesteld aan zouten uit de lucht, maar niet in

direct contact met zeewater

Constructies bij of aan de kust

Permanently submerged

Parts of marine structures

Immerg en permanence

Parties de structures marines

blijvend ondergedompeld

Delen van constructies in zee

Tidal, splash and spray zones

Parts of marine structures

Zones de marnage, zones soumises des projections

ou des embruns

Parties de structures marines

Getijde-, spat- en nevelzone

Delen van constructies in zee

XF : Freeze/thaw attack with or without de-icing agents

XF : Bton expos des cycles de gel/dgel, avec ou sans sels de dverglaage
XF : Beton blootgesteld aan vorst-dooicycli, met of zonder dooizouten

XF1

Moderate water saturation, without de-icing

agent

Vertical concrete surfaces exposed to rain and

freezing

Saturation en eau modre, sans sels de dverglaage

Surfaces de bton verticales, exposes la pluie et au gel

Matige waterverzadiging zonder dooizouten

Verticale betonoppervlakken blootgesteld aan regen en

vorst

5-5

XF2

Moderate water saturation, with de-icing agent

Vertical concrete surfaces of road structures exposed

to freezing and airborne de-icing agents

Saturation en eau modre, avec sels de

dverglaage

Surfaces de bton verticales, de structures routires,

exposes au gel et des sels de dneigement vhiculs
par lair
Verticale betonoppervlakken van wegconstructies
blootgesteld aan vorst en met de lucht meegevoede
dooizouten

Matige waterverzadiging met dooizouten

XF3

XF4

High water saturation, without de-icing agent

Horizontal concrete surfaces exposed to rain and

freezing

Saturation en eau leve, sans sels de dverglaage

Surfaces de bton horizontales, exposes la pluie et au

gel

Hoge waterverzadiging zonder dooizouten

Horizontale betonoppervlakken blootgesteld aan regen

en vorst

High water saturation, with de-icing agent or sea

water

- Road and bridge decks exposed to de-icing agents

- Concrete surfaces exposed to direct spray containing
de-icing agents and freezing
- Splash zones of marine structures exposed to
freezing

Saturation en eau leve, avec sels de dverglaage

- Routes et tabliers de ponts exposs aux agents de

dverglaage
- Surfaces de bton exposes des projections directes
contenant des sels de dverglaage, et au gel
- Constructions marines soumises des claboussures et
au gel

Hoge waterverzadiging met dooizouten of zeewater

- Wegen en brugdekken blootgesteld aan dooizouten

- Betonoppervlakken blootgesteld aan direct gesproiede
dooizouten en aan vorst
- Spatzone van constructies in zee blootgesteld aan vorst

XA : Chemical attack (concrete exposed to chemical attack, e.g. from natural soils and ground water)
XA : Attaques chimiques (bton expos des attaques chimiques, p.ex. du sol naturel ou des eaux souterraines)
XA : Chemische aantasting (beton blootgesteld aan chemische aantasting, o.a. natuurlijke grond en grondwater)
Chemical compositions are detailed in NBN EN 206-1,
(table 2)

Slightly aggressive chemical environment

XA1

XA2

Environnement faible agressivit chimique

Zwak agressieve chemische omgeving

Les compositions chimiques sont dtailles dans la NBN

EN 206-1 (table 2)

Moderately aggressive chemical environment

Chemische samenstellingen worden in NBN EN 200-1

(tabel 2) gedetailleerd

Environnement agressivit chimique modre

Matig agressieve chemische omgeving
Highly aggressive chemical environment

XA3

Environnement forte agressivit chimique

Sterk agressieve chemische omgeving

5-6

Table 5.2-2
Environment classes according to NBN EN 1992-1-1 ANB:2009 (table 1 in NBN B15-001) and
agreement with the exposure classes of EN 1992-1-1:2004
Environment classes
Classes denvironnement
Milieuklassen

NBN B15-001

The Belgian norm NBN B15-001 defines 13 environment classes for concrete. The correspondence with the exposure
classes from NBN EN 206-1 is given as follows.
La norme belge NBN B15-001 dfinit 13 classes denvironnement pour le bton. La correspondance avec les classes
dexposition de la NBN 206-1 est donne ci-dessous.
De Belgische norm NBN B15-001 bepaalt 13 milieuklassen voor beton. De overeenkomst met de blootstellingsklassen van
NBN EN 206-1 wordt hieronder gegeven.
Class
Classe
Klasse

Corresponding exposure classes (NBN EN 206-1)

Classes dexposition correspondantes (NBN EN 206-1)
Overeenkomstige blootstellingsklassen (NBN EN 206-1)

Description
Description
Omschrijving

Concrete without
reinforcement
Bton non arm
Ongewapend beton
Non aggressive environment (only for concrete
without reinforcement)

E0

Environnement non agressif (uniquement pour bton

non arm)

Reinforced and
prestressed concrete
Bton arm et prcontraint
Gewapend en
voorgespannen beton

X0

X0

XC1

X0

XC2

XF1

XC3, XF1

XF1

XC4, XF1

Niet schadelijke omgeving (enkel voor ongewapend

beton)
Inside buildings
Environnement intrieur

EI

Binnenomgeving
EE

(outside buildings, environnement extrieur, buitenomgeving)

Outside buildings, no freeze attack

EE1

Environnement extrieur, pas de gel

Buitenomgeving, geen vorst
Outside buildings, freeze but no contact with rain

EE2

Environnement extrieur, gel mais pas de contact

avec la pluie
Buitenomgeving, vorst maar geen contact met regen
Outside buildings, freeze and contact with rain

EE3

Environnement extrieur, gel et contact avec la pluie

Buitenomgeving, vorst en contact met regen

5-7

EE4

Outside buildings, freeze and de-icing agents

(parts of road infrastructures)

XF4

XC4, XD3, XF4

XA1

XC2, XS2, XA1

XF1

XC4, XS1, XF1

XA1

XC1, XS2, XA1

XF4, XA1

XC4, XS3, XF4,

XA1

Environnement extrieur, gel et sels de dverglaage

(lments dinfrastructure routire)
Buitenomgeving, vorst en dooizouten (delen van
verkeersinfrastructuur)

ES (sea environment, environnement marin, zeeomgeving)

ES1

No contact with sea water, but with air carrying

salt (up to 3 km from the coast) and/or with
brackish water
No freeze
Pas de contact avec leau de mer, mais bien avec
lair marin (jusqu 3 km de la cte) et/ou avec de
leau saumtre
Pas de gel
Geen contact met zeewater, wel met zeelucht (tot 3
km van de kust) en/of met brakwater
Geen vorst

ES2

No contact with sea water, but with air carrying

salt (up to 3 km from the coast) and/or with
brackish water
Freeze
Pas de contact avec leau de mer, mais bien avec
lair marin (jusqu 3 km de la cte) et/ou avec de
leau saumtre
Gel
Geen contact met zeewater, wel met zeelucht (tot 3
km van de kust) en/of met brakwater
Vorst
Contact with sea water submerged elements

ES3

Contact avec leau de mer lments immergs

Contact met zeewater ondergedompeld elementen
Contact with sea water tidal and splash zones

ES4

Contact avec leau de mer lments exposs aux

mares et aux claboussures
Contact met zeewater getijden- en spatzone

EA (agressive environment, environnement agressif, agressieve omgeving)

Slightly aggressive chemical environment

EA1

Environnement faible agressivit chimique

XA1

XA1

XA2

XA2

Zwak agressieve chemische omgeving

Moderately aggressive chemical environment

EA2

Environnement agressivit chimique modre

Matig agressieve chemische omgeving

5-8

Highly aggressive chemical environment

EA3

Environnement forte agressivit chimique

XA3

XA3

Sterk agressieve chemische omgeving

Table 5.2-3
Concrete strength classes to be used for particular environment exposure classes, prescribed by EN
1992-1-1:2004 (Table E.1N in Annex E in EN 1992-1-1:2004)
Exposure classes according to table 5.2-1 (table 4.1 in EN1992-1-1:2004)

Corrosion
Carbonation-induced corrosion

Chloride-induced corrosion

Chloride-induced corrosion
from sea-water

Indicative strength class

XC1

XC2

C20/25

C25/30

XC3

XC4

XD1

C30/37

XD2

C30/37

XD3

XS1

C35/45

C30/37

XS2

XS3

C35/45

Damage to concrete
No risk

Indicative strength class

Freeze/thaw attack

Chemical attack

X0

XF1

XF2

XF3

C12/15

C30/37

C25/30

C30/37

5-9

XA1

XA2
C30/37

XA3
C35/45

Table 5.2-4
Concrete strength classes to be used for particular environment classes, prescribed by
NBN EN 1992-1-1 ANB:2009, in accordance with NBN B15-001

Bton non
arm
C12/15

Bton arm et
prcontraint

EI
EE (Outside buildings)
Outside buildings, no freeze attack
EE1
Outside buildings, freeze but no
EE2

C12/15

C16/20

C12/15
C25/30

C20/25
C25/30

EE3

C25/30

C30/37

C35/45
C25/30 A (*)

C35/45
C30/37 A (*)

ES (Sea environment)
No contact with sea water, but with
ES1

C20/25

C30/37

ES2

C25/30

C30/37

C25/30

C35/45

C35/45
C25/30 A (*)

C35/45
C30/37 A (*)

C25/30

C25/30

C30/37

C30/37

C35/45

C35/45

Classe
E0

EE4

ES3
ES4

Description
Non aggressive environment (only
for concrete without reinforcement)
Inside buildings

contact with rain

Outside buildings, freeze and
contact with rain
Outside buildings, freeze and deicing agents (parts of road
infrastructures)

air carrying salt (up to 3 Km from

the coast) and/or with brackish
water No freeze
No contact with sea water, but with
air carrying salt (up to 3 Km from
the coast) and/or with brackish
water - Freeze
Contact with sea water
submerged elements
Contact with sea water tidal and
splash zones

EA (Aggressive environment)
Slightly aggressive chemical
EA1

environment
Moderately aggressive chemical
EA2
environment
Highly aggressive chemical
EA3
environment
(*) : ( A) concrete with air entraining agents

5-10

---

5.3 Stress limitation

5.3.1

Prescriptions

5.3.1.1 Compression stresses in concrete

Reference: EN 1992-1-1:2004; 7.2(1), (2) and (3)
The compressive stress in concrete should be limited in order to avoid
longitudinal cracks, micro-cracks or high levels of creep.
Longitudinal cracks may occur if the stress level under the characteristic
combination of loads exceeds a critical value. Such cracking may lead to a
reduction of durability. In the absence of other measures, such as an increase in
the cover to reinforcement in the compressive zone or confinement by transverse
reinforcement, it is appropriate to limit the compressive stress. EN 1992-11:2004; 7.2(2) proposes to limit c to 0,6.fck for the exposure classes XD, XF and
XS.
NBN EN 1992-1-1 ANB:2009 is more restrictive prescribing: c 0,5.fck for the
exposure classes XD, XF and XS and c 0,6.fck for all other exposure classes.
If creep could result in unacceptable effects on the functioning of the structure, it
is necessary to limit the stress level under the quasi-permanent combination of
loads to 0,45.fck ; if stress is less than 0,45.fck, linear creep may be assumed ; if the
stress in the concrete exceeds this limit, non-linear creep should be considered.
5.3.1.2 Tensile stress in steel
Reference: EN 1992-1-1:2004; 7.2(4) and (5)
Tensile stresses in the reinforcement should be limited in order to avoid inelastic
strain, unacceptable craking or deformation.
Unacceptable cracking or deformations may be assumed to be avoided, if under
the characteristic combination of loads, the tensile stress in the reinforcement
does not exceed 0,8.fyk. Where the stress is caused by an imposed deformation,
the tensile stress should not exceed fyk.
5.3.1.3 The cross-section to be taken into account for stress calculations
Reference : EN 1992-1-1:2004; 7.1(2)
For the calculation of stresses, cross-sections may be assumed to be uncracked provided
that the flexural tensile stress does not exceed fctm.
Discussion: stress calculations may be performed on the basis of two alternatives
regarding the cross-section to be used:
the section is uncracked: it is assumed that the whole area of concrete is
active and that concrete and steel behave elastically in tension and in
compression;
the section is cracked: the concrete behaves elastically in compression, but
does not transmit any tensile stress.

5-11

5.3.2

Formulas for the calculation of the normal stresses in a reinforced concrete

cross-section in SLS

5.3.2.1 Basic assumptions

Elastic behaviour for concrete in compression and for steel in tension and
compression.
Es tension = Es compression = 200 GPa Ec
The concrete in the tension part of the cross-section is considered to be cracked
and is not taken into account for the stress calculations.
The assumption of BERNOULLI is accepted.
The equations are elaborated for cross-sections with arbitrary shape but with a
vertical axis of symmetry. The loads are applied in the plane of symmetry.
5.3.2.2 Simple bending
1. DEFINITION OF THE PROBLEM: see figure 5.3.2-1

z'

O'

d2

c=Ec.c

c
s2

As2
x
h

y'

s2=Es.s2

NA

As1

s1

(c)y'

s1=Es.s1

d1
y'

Figure 5.3.2-1
Simple bending in SLS. Questions: x = ? c = ? s = ?
Discussion:
Simple bending the NA is situated somewhere inside the cross-section,
because bending necessitates equilibrium between the tension and
compression forces.
A perfect bond is assumed between steel and concrete; thus s = c at the
interface. Consequently:

s
Es

c
Ec

s = Es/Ec . c = . c

5-12

(5.3.2-1)

- for loads characterized by a short duration time: 6 7 ;

- for long duration loads (which may lead to creep): = 15 is a commonly
used value for preliminary design.

Note: the previous version of EC2 (1998) prescribed effectively

the value = 15 for long term analysis and in the case where
more than 50% of the stresses is due to quasi-permanent loads.
This prescription has been dropped in the present standard
(version 2004).

The concrete in the tensile zone does not take up any stress.
The elaboration below necessitates the introduction of an auxiliary system
of axes (x y z) with the origin situated at the extreme (most compressed)
top fibre; the positive y'-axis is chosen in downwards direction.

2. SOLUTION - Elaboration of the basic formulas

(a) Linearity of the stress diagram
All stresses can be expressed in function of the maximum compressive
stress in the concrete cmax , which appears at the top fibre of the crosssection (and which is designated in a simplified way by c in the following
text):
- stress in the concrete at an arbitrary level y':

( c ) y ' = c .
-

x y'
x

(5.3.2-2)

steel:

s1 = . c .

x d2
dx
and s 2 = . c .
x
x

(5.3.2-3)

(b) Translation equilibrium

N compression = N tension

N c + N s 2 = N s1

Ac compressed

( c ) y ' .dAc + As 2 . s 2 = As1 . s1

Replacement of the stresses by their expressions in function of c :

5-13

Ac compressed

Division by

x d2
x y'
dx
.dAc + . c
. As 2 = . c .
. As1
x
x
x

Ac compressed

x. Ac compressed

( x y ' ).dAc + .( x d 2 ). As 2 = .(d x). As1

Ac compressed

y '.dAc + .x. As 2 .d 2 . As 2 .d . As1 + .x. As1 = 0

x. Ac compressed + . As 2 + . As1 =

Ac compressed

y '.dAc + .d 2 .As 2 + .d . As1

in which:
- ( Ac compressed + . As 2 + . As1 ) may be called the fictive section

Af ;
- the second member is the static moment of the fictive section with
respect to the top fibre (thus with respect to the z-axis).
Finally:

x=

( S Af ) z '

(5.3.2-4)

Af

Note: Significance of expression (5.3.2-4): the NA passes by the

centre of gravity of the fictive section!

(c) Rotation equilibrium around the NA

M =

Ac compressed

( c ) y ' .dAc .( x y ' ) + As 2 . s 2 .( x d 2 ) + As1. s1.(d x)

Replacement of the stresses by their expressions in function of c :

M=

c
x

Ac compressed

( x y ' ) 2 .dAc + .

5-14

c
x

.( x d 2 ) 2 . As 2 +

. c
x

.(d x) 2 . As1

M=

2
2
2
Ac compressed ( x y' ) .dAc + ( x d 2 ) . . As 2 + (d x) . . As1
x

moment of inertia of the fictive section A f with respect to the NA

Finally:

c =

M
.x
( I A f ) NA

(5.3.2-5)

3. PARTICULAR CASE: simply reinforced rectangular section

See figure 5.3.2-2.
c

Nc

z/3

x
h

N.L.-A.N.
d

As1
b

s1

s1

Ns1

d1

s1/

Figure 5.3.2-2
Simple bending; SLS; simply reinforced rectangular section

One finds
A f = b.x + . As1

(S )

Af z '

x
= b.x. + . As1.d
2

which leads to
b.x 2
+ . As1.d
( S Af ) z '
2
x=
=
Af
b.x + . As1

or
b.x 2
+ . As1.x . As1.d = 0
2
Solution of the quadratic equation:

5-15

x=

. As1
d
. As1
+
+ 2. . As1.
b
b
b

(5.3.2-6)

1
Moreover: M = N c .z = .b.x. c .z , which permits to write:
2
2M
c =
b.x.z
In an analogous way: M = N s1.z = As1. s1.z , which permits to write:

s1 =

M
As1.z

with
z=d

x
3

4. PARTICULAR CASE: doubly reinforced rectangular section

See figure 5.3.2-3.
d2

c
s2

s2

As2
x
h

s2/

N.L.-A.N.

Ns2
Nc

As1
b

Ns2+Nc

s1

s1
Ns1

d1

s1/

Figure 5.3.2-3
Simple bending; SLS; doubly reinforced rectangular section
One finds:
A f = b.x + .( As1 + As 2 )

( S Af ) z ' =

b.x 2
+ .( As1.d + As 2 .d 2 )
2

which gives:
b.x 2
+ .( As1.d + As 2 .d 2 )
( S Af ) z '
x=
= 2
Af
b.x + .( As1 + As 2 )

5-16

Solution of the quadratic equation:

.( As1 + As 2 )
.( As1 + As 2 ) 2. .( As1.d + As 2 .d 2 )
+
x=
+
b
b
b

(5.3.2-7)

Finally:

c =

M .x
( I A f ) NA

with (see figure 5.3.2-4)

2

( I A f ) NA

b.x 3
x
2
2
=
+ b.x. + . As1 .(d x ) + As 2 .( x d 2 )
3
2

b.x 3
2
2
=
+ . As1 .(d x ) + As 2 .( x d 2 )
3

d2
As2
x

x-d2

NA
d
d-x
As1
b
Figure 5.3.2-4
Auxiliary figure for the determination of the moment of inertia ( I A f ) AN

Moreover:

s1 = . c .

dx
x

s 2 = . c .

x d2
x

and

5-17

Note: z = ?

z=

( I A f ) NA . c
M
dx
=
with s1 = . c .
N s1
x. As1 . s1
x
=

( I A f ) NA

. As1 .(d x)

( I A f ) NA
1

( S A f2 ) NA

b.x 3
+ . As1 .(d x) 2 + As 2 .( x d 2 ) 2
= 3
. As1 .(d x)

( S A f ) NA = 0 because the NA passes through G Af ; and thus:

x
b.x. + . As 2 .( x d 2 ) . As1.(d x) = 0
2
2
b.x
= .[ As1.(d x) As 2 .( x d 2 )]
2
b.x 3 2
= . .[As1.(d x) As 2 .( x d 2 )].x
3
3

to be introduced in the expression of z:

z=

2. .x.[As1.(d x) As 2 .( x d 2 )] + 3. . As1.(d x) 2 + As 2 .( x d 2 ) 2
3. . As1.(d x)

Rearranging terms leads to:

=

2
2 A x d 2 As 2 ( x d 2 )
.x .x. s 2 .
+
.
+d x
3
3 As1 d x As1 d x

x As 2 x d 2
2
+
.
. x d 2 .x
3 As1 d x
3

Or finally
z=d

x As 2 x d 2 x

.
. d 2
+
3 As1 d x 3

x
Conclusion: z=d- 3 when
As2=0

x
d2= 3(Ns2 and Nc coincide)
d2=x

5-18

(5.3.2-8)

5. PARTICULAR CASE: simply reinforced T cross-section

1st case: the NA is situated in the flange. This case is solved in the same way as for a
rectangular cross-section.
2nd case: the NA is situated in the web: figure 5.3.2-5.
b

c
hf

Nc
x
d

As1

s1

s1

Ns1
bw

Figure 5.3.2-5
Simple bending; SLS; simply reinforced T cross-section
The part of the web that is in compression, resists to rather small stresses. The
following equations are elaborated on the basis of the assumption that the
compressed part of the web is neglected.
One finds:
A f = b.h f + . As1

(S )

Af z '

b.h 2f

+ . As1.d

which gives:
x=
And because:

(I )
Af

NA

(S )

Af z '

Af

b.h 2f

+ . As1.d
= 2
b.h f + . As1

b.x 3 b.(x h f
=

3
3

+ . As1 .(d x )

(5.3.2-9)

it is concluded that:

c =

M .x
M .x
=
3
3
( I A f ) NA b.x
b.(x h f )
2

+ . As1 .(d x )
3
3

and

5-19

 s1 = . c .

dx
x

Note: alternative formulas

M=Nc.z with Nc: see figure 5.3.2-6
c
O'

hf

Nc
x
(c)y'

y'

Figure 5.3.2-6
Auxiliary figure for the determination of Nc
Nc =

hf

( c ) y ' .b.dy'
Nc =

with ( c ) y ' = c .

hf

.b.h f . x
2
x

and thus
M =

hf

.b.h f . x
x
2

.z

Consequently:

c =

M .x
h

b.h f . x f
2

M = N s1.z = As1. s1.z

and thus

s1 =

z=?

5-20

M
As1.z

.z

x y'
x

z=

M
N s1

=
=

Af

NA

. c

x. As1 . s1

(I )
Af

NA

x. As1 .(d x)

s1 = . c .

with
=

(I )
Af

dx
x

NA

S 2
A f NA
1

b.x 3 b.(x h f )
2

+ . As1 .(d x )
3
3
1
2
= .b.h f . 3.x 2 3.x.h f + h 2f + . As1 .(d x )
3
hf

b.h f . x
. As1 .(d x ) =
=

part below the NA

( )

avec I A f

(I )

NA

and S A f2

NA
1

part above the NA

Result of the development:

h 3.x 2.h f
z =h f .
3 2.x h f

(5.3.2-10)

3rd case: beams with large height and thin flanges; the NA is situated in the web.
In this case, neglecting the compressed part of the web is too far from reality. See
figure 5.3.2-5.
One finds:

A f = b.h f + bw .(x h f ) + . As1

(S )

Af z '

b.h 2f

x hf

+ bw .(x h f ). h f +
2
2

and
x=

+ . As1.d

(S )

Af z '

Af

This leads to the following quadratic equation:

h 2f

bw 2
.x + ((bw b ).h f . As1 ).x + .(b bw ) + . As1.d = 0
2
2

and the solution:

5-21

x=

1
. (bw b ).h f . As1
bw

1
+
bw

[(b

b ).h f . As1

h2

+ 2.bw . f .(b bw ) + . As1.d

2

(5.3.2-11)

Finally:

c =

M .x
( I A f ) NA

with
3

( I A f ) NA

b.x 3 (b bw ).( x h f )
=

. As1 .(d x) 2
3
3

and

s1 = . c .

dx
x

6. PARTICULAR CASE: doubly reinforced T cross-section

1st case: the NA is situated in the flange. This case is solved such as for a
rectangular cross-section.
2nd case: the NA is situated in the web: see figure 5.3.2-7.
b
d2
hf

c
s2

s2

As2
x

Ns2
Nc

d
As1

s1

s1
Ns1

bw

Figure 5.3.2-7
Simple bending; SLS; doubly reinforced T cross-section

In a first stage, it is assumed to neglect the compressed part of the web.

One finds:

5-22

A f = b.h f + .( As1 + As 2 )
( S Af ) z ' =

b.h 2f
2

+ .( As1.d + As 2 .d 2 )

and thus:
x=

(S Af ) z '
Af

b.h 2f

+ .( As1.d + AS 2 .d 2 )
2
=
b.h f + .( As1 + As 2 )

(5.3.2-12)

Finally:
M .x
( I A f ) NA

c =
with
3

b.x 3 b.( x h f )
=

+ . As1 .(d x) 2 + As 2 .( x d 2 ) 2
3
3

( I A f ) NA

and

dx
d
x d2
= . c .
x

s1 = . c .
s2

3rd case: beams with large height and thin flanges; the NA is situated in the web.
In this case, neglecting the compressed part of the web is too far from reality. See
figure 5.3.2-7.
One finds:
A f = b.h f + bw .( x h f ) + .( As1 + As 2 )

(S Af ) z ' =

b.h 2f

x hf

+ bw .( x h f ). h f +
2
2

+ .( As1.d + As 2 .d 2 )

Expression (5.3.2-4) leads to a quadratic equation, which has the following

solution:
x=

1
. h f .(bw b) .( As1 + As 2 )
bw
+

1
bw

[h .(b
f

]
]

b) .( As1 + As 2 ) +

bw (b bw ) 2

.
.h f + .( As1.d + As 2 .d 2 )
2 2

(5.3.2-13)

5-23

Finally:
M .x
( I A f ) NA

c =

with
3

( I A f ) NA

b.x 3 (b bw ).( x h f )
=

+ . As1 .(d x) 2 + . As 2 .( x d 2 ) 2
3
3

and
dx
x
x d2
= . c .
x

s1 = . c .
s2

5.3.2.3 Bending combined with compression

1. Introduction: principal aspects
- simple bending: see figure 5.3.2-8.

z'
As2
M

G
O

O
V
As1

s1/

y'

Figure 5.3.2-8
Simple bending

The NA is situated at a distance equal to:

5-24

( S Af ) z '

x=

Af

from the z-axis, and passes through the centre of gravity G of the fictive section
Af..
- Pure compression: see figure 5.3.2-9.
d2

c=Ec.

s=Es.

As2
h

G
d

y'N

O
As1

s=Es.
d1

Figure 5.3.2-9
Pure compression
Question: what is the location of the resultant force N of the compressive
stresses?
In order to answer to this question, the rotation equilibrium around the z'-axis is
written:
h

N . y ' N = ( c ) y ' . y '.b.dy '+ As1. s1.d + As 2 . s 2 .d 2

0

with
h

N = ( c ) y ' .b.dy + As1. s1 + As 2 . s 2

0

and

s1 = s 2 = . c
which leads to:
h

y'N

( ) . y'.b.dy' + ( . A .d + . A .d ).
=
( ) . y'.b.dy' + ( . A + . A ).
c y'

s1

s2

c y'

s1

with
( c ) y ' = c

5-25

s2

After division by c, on gets:

y' N

y '.b.dy '+ . As1 .d + . As 2 .d 2

b.dy '+ . As1 + . As 2

(S Af ) z '

Af

(= x in simple bending !)

Conclusion: in the case of pure compression, N is applied in the centre of gravity

GPC of the fictive section Af..
Moreover, one finds:

c =

N
Af

because
h

N = ( c ) y ' .b.dy ' + As1. s1 + As 2 . s 2

0

with
( c ) y ' = c en s1 = s 2 = . c
And thus:
h
N = c . b.dy '+ .( As1 + As 2 )
0

Af

N = c .Af
- Bending combined with compression
N is applied with an eccentricity with respect to GPC (the centre of gravity of the
fictive section for pure compression).
The two situations shown in figure 5.3.2-10 are equivalent.
N
GPC

eGPC

GPC

N
O

5-26

N.eG PC

Figure 5.3.2-10
Equivalent representations for combined bending

The stress diagram is thus composed of two parts (superposition of):

the effect of the axial force N (applied in GPC), with the stress c =

N
all
Af

over the cross-section;

the effect of the bending moment N.eG, which has its own NA passing by
GSB. (SB = simple bending).
See figure 5.3.2-11.

GSB
O

NA of the combined bending

Figure 5.3.2-11
Combined bending in SLS
2. Formulas for the calculation of stresses
See figure 5.3.2-12.

N
u
d2

c
eG

As2
GEB
h

s2

eo

v'

O
v
As1
b

A.N.

s1
d1

N/Af

5-27

Figure 5.3.2-12
Bending combined with compression; SLS

The basic formula is:

c =

N
N .eG
N
e . '
+
. ' =
.(1 + G 2 )
A f ( I A f )G
Af
if

Auxiliary development with the aim to determine eG:

AN = 0 =

N .eG
e .
N
N

. =
.(1 + G2 ) = 0
A f ( I Af )G
Af
if

Which leads to:

eG =

i 2f

The basic formula is now written as follows:

c =

N
N x
N .x
'
.(1 + ) =
. =
Af

A f ( S A f ) AN

And thus:

c =

N .x
N .x
=
2
( S Af ) NL b.x
+ . As 2 .( x d 2 ) . As1.(d x)
2

Moreover:

s1 = . c .

dx
x

s 2 = . c .

x d2
x

and

Calculation of x

5-28

i 2f

+ ' = (eG + ) u = + u

i 2f + 2
A f .(i 2f + 2 )
u =
u
A f .

with
A f (i 2f + 2 )

= A f .i 2f + A f . 2

= ( I A f ) G + A f . 2
= ( I A f ) NA

and
( I A f ) NA =

b.x 3
+ . As 2 .( x d 2 ) 2 + . As1 .(d x) 2
3

x=

( I A f ) NA
( S A f ) NA

b.x 3
+ . As 2 .( x d 2 ) 2 + . As1.(d x) 2

x= 3 2
b.x
+ . As 2 .( x d 2 ) . As1.(d x)
2
This leads to a 3rd degree equation in x:
x.
.[( As1.d + As 2 .d 2 ) + .( As1 + As 2 )]
x 3 + 3. .x 2 + 6.
b
6.

. ( As1.d + As 2 .d 2 ). + ( As1.d 2 + As 2 .d 22 ) = 0
b

(5.3.2-14)

5.3.2.4 Bending combined with tension

1. Case with small eccentricity
The tensile force is applied within the central core; the NA falls outside the crosssection; tension is dominant; see figure 5.3.2-13.

5-29

As2

Ns2

h/2-d2

h/2

O
eo

h/2-d1

N
As1

Ns1

Figure 5.3.2-13
Bending combined with tension small eccentricity - SLS
The whole section is in tension:
h
d 2 + e0
N
N s1 = N . 2
s1 = s1
d d2
As1
N s2

h
d1 e0
N
2
= N.
s2 = s2
d d2
As 2

2. Case with large eccentricity

Figure 5.3.2-14.

d2
As2

v'

GEB
h

s2

NA

O
u
eG
eo

As1
b

s1

d1

N/Af
N

Figure 5.3.2-14
Bending combined with tension large eccentricity - SLS

5-30

The basic formula is:

c =

N .eG
N
. '
Af
( I Af )G

Auxiliary development with the aim to determine eG:

AN = 0 =

N .eG
e .
N
N

. =
.(1 G2 ) = 0
A f ( I Af )G
Af
if

which leads to:

eG =

i 2f

The basic equation can now be rewritten:

c =

N
Af

N .x
' N ' N x
. 1 =
.
. =
=

A f A f ( S A f ) NA

Moreover:

s1 = . c .

dx
x

Calculation of x
x

(eG + ) =

(i 2f + 2 )

A f .(i 2f + 2 )
A f .

( I A f ) NA
( S A f ) NA

The development leads to a 3rd degree equation:

x
x 3 3. .x 2 + 6. . .[( AS1.d + As 2 d 2 ) .( As1 + AS 2 )]
b

(5.3.2-15)

+ 6. . ( As1.d + As 2 .d 2 ). ( As1.d + As 2 .d ) = 0
b
5.3.3

2
2

Method for the determination of the position of the NA

The method presented below helps to determine the position of the NA, in particular for
cross-sections with complex shapes (example: the cross-section of a chimney). The
method is explained by means of the example of an I-shaped section in reinforced
5-31

concrete, loaded by an eccentrically applied axial compression force N: see figure 5.3.31.
The method starts with the subdivision of the cross-section into elementary fictive areas
f (fictive = compressed concrete and As). The position of each elementary area f is
determined by:
- the distance t towards the line of action of the force N, and
- the distance s towards the NA.
The positive direction of t and s is indicated in figure 5.3.3-1.
Figure 5.3.3-1 also shows the stress distribution. Each area f is associated with the
stress f.
One may now write the three following equations:
Translation equilibrium: N = f . f
Rotation equilibrium around the line of action of the force N: 0 = f . f .t
Description of the linearity of the stress distribution: f = f . s/x

Action line of the

force N

N
t

NA
As1

s1/

Figure 5.3.3-1
The I-shaped section in reinforced concrete is subdivided in elementary fictive areas f
The rotation equilibrium equation may be rewritten in the following way:
0 = f . f .t
0 = f . c . s/x . t
0 = c /x . f . s . t
And finally:
f . t . s = 0

(5.3.3-1)

This relationship can now be exploited in order to determine the position of the NA; see
figure 5.3.3-2. Indeed:
5-32

( f . t . s) for f above the NA = ( f . t . s) for f below the NA

<0

>0

f . t . s < 0
f
>0

f . t . s

>0

f . t . s > 0

<0

NA
As1

<0

f . t . s = . As1 . t . s < 0
>0 <0

Figure 5.3.3-2
Method for the determination of the position of the NA
Starting at the upper side of the cross-section, the values of (f . t . s) are accumulated
(see right part in figure 5.3.3-2). These values (f . t . s) are negative until the line of
action of the force N. From that level on, the added values of (f . t . s) are positive.
Below the NA, only the reinforcement As1 leads to a contribution: ( . As1 . t . s), which
is also negative.
One finds the position of the NA on the level where the cumulative value ( f . t . s)
for the compressed concrete is equal to (f . t . s) associated with the steel in tension.
Note: this reasoning can be worked out graphically, but is very much appropriate for
numerical elaboration.

5.4 Crack control

5.4.1

General considerations

Reference : EN 1992-1-1:2004; 7.3

Cracking is normal in reinforced concrete structures subjected to bending, shear,
torsion or tension resulting from either direct loading or restraint or imposed
deformations. Yet, cracking should be limited to an extent that will not impair the
proper functioning or durability of the structure or cause its appearance to be
unacceptable.
Cracks may also arise from other causes such as plastic shrinkage or expansive
chemical reactions witin the hardened concrete. Such cracks may be unacceptably
large but their avoidance and control lie outside the scope of EN 1992-1-1:2004.
In the absence of specific requirements (such as for example water-tightness), it
may be assumed that limiting the crack width to a maximum acceptable value
wmax under the quasi-permanent combination of loads, is satisfactory for
5-33

reinforced concrete members in buildings with respect to appearance and

durability. EN 1992-1-1:2004 recommends some values of wmax, in function of
the exposure class that has to be considered (or the environment class in ANB):
see table 5.4.1-1. It appears from this table that a limitation of the crack width to
0,3mm is acceptable for most exposure classes. For exposure classes X0 and XC1
(ANB: environment classes E0 and EI), the crack width has no influence on
durability and higher values for wmax can be accepted, but limitation is necessary
to guarantee acceptable appearance. Note: the conditions for prestressed concrete
are more severe because of the higher vulnerability of prestressing steel.
Special measures may be necessary for members subjected to exposure class XD3
(ANB: environment class EE4); the choice of appropriate measures depends upon
the nature of the aggressive agent involved.
Table 5.4.1-1
Recommended values of maximum crack width wmax (mm) in function of exposure
classes (Table 7.1N in EN 1992-1-1:2004) and environment classes (Table 7.1NANB in NBN EN 1992-1-1 ANB)
wmax (mm)
Exposure class
Environment class
Reinforced concrete / quasipermanent combination of loads
X0, XC1
EI
0,4
EE1, EE2, EE3, EE4,
XC2, XC3, XC4,
ES1, ES2, ES3, ES4
XD1, XD2, XD3,
0,3
XS1, XS2, XS3
In order to limit the crack width to acceptable values, EN 1992-1-1:2004; 7.3, stipulates
that it is not necessary to calculate explicitly the crack widths explicitly, on the
condition that the following rules (which are detailed in the following paragraphs) are
respected:
-

in all sections where tension is expected, a minimum amount of bonded

reinforcement is required; the amount must be high enough to resist, without
yielding, to the tensile force in the concrete just before cracking. This means that,
when a crack appears in the tensile concrete, there is enough reinforcement to
resist the tensile force that is liberated by the cracking;

the bar diameter and the distance between bars (= spacing) are restricted.
Although full justification will become clear in chapter 6, with the discussion of
the anchorage phenomenon, an intuitive reasoning helps already to understand the
rule :
- bars with a large diameter transmit large forces which, in spite of the
increase in perimeter surface, leads to a more severe loading of the
surrounding concrete;
- for highly loaded bars (s large), the distance should not be too large
between bars, in order to avoid that parts of the concrete area stay
behind and do not effectively cooperate with the heavily loaded
concrete around the bars.
5-34

 Note: EN 1992-1-1:2004; 7.3.3(1) specifies that specific measures to control

cracking are not necessary for reinforced concrete slabs in buildings, when the
following conditions are fulfilled:
- the slabs are subjected to bending without significant axial tension;
- the height (or overall depth) of the slab does not exceed 200mm;
- the reinforcement is calculated and provided according to the prescriptions
of the EN (see chapter on slabs).

5.4.2

Minimum reinforcement areas

Reference: EN 1992-1-1:2004; 7.3.2

Unless a more rigorous calculation shows lesser areas to be adequate, the required
minimum area of reinforcement is calculated from the equilibrium between the tensile
force in concrete just before cracking and the tensile force in the reinforcement at
yielding or at a lower stress if necessary to limit the crack width. The principle is thus
that when concrete cracks, enough reinforcement must be in place to resist the liberated
force; this is the fundamental meaning of the following expression:
As,min . s = kc . k . fct,eff . Act

(5.4.2-1)

with:
As,min = minimum area of reinforcing steel within the tensile zone;
= area of concrete within the tensile zone. The tensile zone is that part of
Act
the section which is calculated to be in tension just before formation of the first
crack. For a rectangular section in bending, it is common practice to take half of
the full section as tensile zone (instead of determining the exact position of the
NA via the position of the centre of gravity of the fictive section);
s
= maximum stress permitted in the reinforcement immediately after
formation of the crack. This may be taken as the yield strength fyk of the
reinforcement. However, a lower value may be needed in order to satisfy the
crack width limits according to the maximum bar size or spacing (see further in
table 5.4.3-1);
fct,eff
= the tensile strength of the concrete effective at the time when the
cracks may first be expected to occur. A suitable choice may be to adopt the mean
tensile strength fctm (table 3.4.3-1), or a lower value fctm(t) if cracking is expected
earlier than 28 days (see paragraph 3.4.3.7); this might occur even within 3 to 5
days after casting, if the imposed deformation is mainly caused by the evacuation
of hydration heat (of course in function of the environmental conditions, the
shape of the construction element, the nature of the formwork material, etc).
k
= coefficient which allows for the effect of non-uniform selfequilibrating stresses, which lead to a reduction of restraint forces. This
coefficient permits to reduce the minimum reinforcement thanks to the reduced
influence of imposed deformations after cracking; moreover, tensile stresses are

5-35

eliminated in the neighbourhood of a crack which causes a reduction of the risk to

have new cracks created). Recommended values for k:
- 1,0 for webs with h 300 mm or flanges with widths 300 mm ;
- 0,65 for webs with h 800 mm or flanges with widths 800 mm ;
- intermediate values to be determined by interpolation.
kc = coefficient which takes account of the stress distribution within the section
immediately prior to cracking. The stress distribution is function of the load
combination effects and the restraint of imposed deformations.
For pure tension: kc = 1,0.
For bending or bending combined with axial forces, the following values are
recommended:
- for rectangular section and webs of box sections and T-sections:

c
k c = 0,4.1
*
k1 (h / h ) f ct ,eff

(5.4.2-2)

- for flanges of box sections and T-sections:

k c = 0,9.

with
c =

Fcr
0,5
Act . f ct ,eff

the mean stress of the concrete acting on the part of the section under
consideration:

c =
NEd =

h*

(5.4.2-3)

N Ed
b.h

(5.4.2-4)

the axial force acting on the part of the cross-section under

consideration (compressive force positive!). This force should be
determined considering the characteristic values of prestress and axial
forces under the relevant combination (SLS) of actions;
h* = h
for h < 1,0 m
*
h = 1,0 m for h 1,0 m

k1 =

coefficient considering the effects of axial forces on the stress

distribution:
- k1 = 1,5 if NEd is a compressive force;
- k1 = (2 . h*)/ (3 . h) if NEd is a tensile force.

Fcr =

absolute value of the tensile force within the flange immediately prior
to cracking, due to the cracking moment calculated with fct,eff.

5-36

Note 1: The expression (5.4.2-1) shows that the use of a higher concrete quality
than the one that has been used in the design project, may lead to
unexpected consequences; indeed, a better concrete leads to a larger
area of minimum reinforcement!

Note 2: In profiled cross-sections like T-beams and box girders, the minimum
reinforcement should be determined for the individual parts of the
section: webs and flanges!

5.4.3

Restrictions of bar diameter and spacing

In addition to the provision of the minimum amount of bonded reinforcement, EN 19921-1 :2004 ; 7.3.3 proposes limitations for bar diameters and spacing between bars, in
order to avoid (without detailed calculations) excessive crack widths:
- for cracking caused dominantly by restraint, the bar size has to be limited;
- for cracking caused by loading, either the bar diameter should be limited or the
spacing.
The prescriptions regarding bar size and spacing are presented in the form of tables,
which have been elaborated for a representative cross-section.
Table 5.4.3-1
Crack control: maximum bar diameter *s in function of steel stress and the maximum
crack width wk (Table 7.2N in EN 1992-1-1:2004)
Steel stress (MPa)
160
200
240
280
320
360
400
450

Maximum bar diameter *s (mm)

wk = 0,4 mm
wk = 0,3 mm
wk = 0,2 mm
40
32
25
32
25
16
20
16
12
16
12
8
12
10
6
10
8
5
8
6
4
6
5
-

Table 5.4.3-2
Crack control: maximum distance between bars (spacing) in function of steel stress and
the maximum crack width wk (Table 7.3N in EN 1992-1-1:2004)
Maximum bar spacing (mm)
Steel stress (MPa)
wk = 0,4 mm
wk = 0,3 mm
wk = 0,2 mm
160
300
300
200
200
300
250
150

5-37

240
280
320
360

250
200
150
100

200
150
100
50

100
50
-

The values shown in both tables are obtained on the basis of the following assumptions:
- h = 400 mm;
- cross-section with only 1 layer of reinforcement;
- concrete cover c = 25 mm;
- fct,eff = 2,9 MPa (concrete class C30/37);
- hcr = 0,5.h with hcr (cracked) = depth of the tensile zone immediately prior to
cracking, considering the characteristic values of axial forces under the quasipermanent combination of actions;
- h d = 0,1.h;
- k1 = 0,8; kc = 0,4; k = 1 (coefficients: see formula (5.4.2-1));
- maximum concrete stress that takes account of creep = 0,5.fck (this is in
accordance with the selection of the quasi-permanent combination of actions!);
- long term loading;
- the steel stress is calculated on the basis of a cracked section under the quasipermanent combination of actions.
Consequently, the values of the maximum diameter, presented in table 5.4.3-1 have to
be adjusted if conditions differ from the assumptions. The modified maximum diameter
is found by means of the following expression:
- for bending

s = *s .( f ct ,eff / 2,9).

k c .hcr
2(h d )

(5.4.2-5)

s = *s .( f ct ,eff / 2,9).

hcr
8(h d )

(5.4.2-6)

for pure tension

with:
- s = the adjusted maximum diameter;
- *s = the maximum bar size given in table 5.4.3-1;
- hcr = hcr (cracked) = depth of the tensile zone immediately prior to cracking,
considering the characteristic values of axial forces under the quasipermanent combination of actions.

5.4.4

Skin reinforcement

Reference: EN 1992-1-1:2004; 7.3.3(3)

5-38

Beams with a total depth of 1,0 m or more, where the main reinforcement is
concentrated in only a small proportion of the depth, should be provided with additional
skin reinforcement to control cracking on the side faces of the beam. This reinforcement
should be evenly distributed between the level of the tension steel and the NA and
should be located within the links; see figure 5.4.4-1. The area of the skin reinforcement
should not be less than the amount obtained from expression (5.4.2-1), taking k = 0,5
and s = fyk. The spacing and size of suitable bars may be obtained from tables 5.4.3-1
and 5.4.3-2, assuming pure tension and a steel stress of half the value assessed for the
main tension reinforcement.

x
NA

d-x

Figure 5.4.4-1
Skin reinforcement to control cracking in beams with large depth, is placed within the
links

Note: EN 1992-1-1:2004; 9.2.4 (1) stipulates that, in certain cases, it may be

necessary to provide surface reinforcement either to control cracking or to
ensure adequate resistance to spalling of the cover. Detailing rules for this
surface reinforcement are given in the informative annex J of EN 1992-11:2004:
- surface reinforcement to resist spalling should be used where the main
reinforcement is made up of bars with a diameter greater than 32 mm, or
bundled bars with an equivalent diameter greater than 32 mm;
- the surface reinforcement consists of wire mesh or small diameter bars, and is
placed outside the links: see figure 5.4.4-2;
- the area of the surface reinforcement As,surf should be not less than 0,01.Act,ext
in the two directions parallel and orthogonal to the tension reinforcement in
the beam. Act,ext is the area of the tensile concrete external to the links as
shown in figure 5.4.4-2. If the cover to reinforcement is greater than 70 mm
(for example for particular durability requirements), then As,surf,min may be
equal to 0,005.Act,ext in each direction;
- the prescriptions regarding minimum cover to reinforcements are also
applicable to the surface reinforcement;
- the longitudinal bars of the surface reinforcement may be taken into account
as longitudinal bending reinforcement and the transverse bars as shear

5-39

reinforcement provided that they meet the requirements for the arrangement
and anchorage of these types of reinforcement.

Figure 5.4.4-2
Surface reinforcement to resist spalling, is placed outside the links
(Figure J.1 in annex J in EN 1992-1-1:2004)

5.4.5

Calculation of crack width

See chapter on cracking and deflections

5.5 Deflection control

5.5.1

General considerations

Excessive deformations of construction elements may lead to reduction of strength and

stability, but may also affect the proper functioning of the structure. The deflection
control has gained in importance in the recent versions of the standard because of:
- the use of more refined calculation methods,
- the continuous improvement of materials properties which allows to build more
slender and more economical structures which are also more sensitive to
deformations.
The materials that are used for non-bearing partition walls in buildings, are
characterized by a large variety of deformation properties; this also leads to difficulties
in determining appropriate limit values for deflections of construction elements.
EN 1992-1-1:2004 ; 7.4 contains the following recommendations:
the deformation of a member or structure should not adversely affect its proper
functioning or appearance;
together with the client, appropriate limiting values of deflection should be
established, which take account of
- the nature of the structure, of the finishes, partitions and fixings,
5-40

- the function of the structure,

- connected elements such as partitions, glazing, cladding, services or
finishes,
- the proper functioning of machinery or apparatus supported by the
structure,
- the risk og having water accumulation on flat roofs, etc.
The NBN EN 1992-1-1 ANB:2009 also refers to NBN B03-003 for specific limits
of deflections in function of the use of the structure;
in order to preserve appearance and general utility of the structure, the calculated
sag of a beam, slab or cantilever subjected to quasi-permanent loads, should be
limited to span/250; the sag is assessed relative to the supports. Deflections that
could damage adjacent parts of the structure should be limited to span/500 for
quasi-permanent loads;
the sag is assessed relative to the supports. Pre-camber may be used to
compensate for some or all of the deflection but any upward deflection
incorporated in the formwork should not exceed span/250.
5.5.2

Cases where calculations may be omitted

The calculation of deflections is not an easy task, because deflections are determined by
the influence of many parameters. Sometimes, major differences appear between real
deformations and previsions, especially when the bending moment is close to the
cracking moment (= bending moment that causes the first crack).
Generally, and in particular for project design, it is not necessary to calculate the
deflections explicitly; for deflection problems in normal circumstances, it is sufficient to
apply the simple rule of the limitation of the span/depth ratio. On the basis of a large
number of representative calculations and tests on a wide variety of beams, formulas
have been developed defining limits of the span/depth ratio; provided that beams or
slabs comply with these limits, their deflections may be considered as not exceeding the
limits span/250 and span500. The span/depth limits are:
3/ 2

0

0
l
= K 11 + 1,5 f ck
+ 3,2 f ck
1

0
l
1
= K 11 + 1,5 f ck
+
' 12
d

with
l/d =
K=
0 =
=

f ck

'

for 0

for > 0

(5.5.2-1)

(5.5.2-2)

the limit span/depth;

factor to take into account the different structural systems: see table 5.5.2-1;
reference reinforcement ratio = fck . 10-3;
required tension reinforcement ratio to resist the moment due to the design
loads (at mid-span for simply supported beams, at support for contilevers);

5-41

fck =

required compression reinforcement ratio to resist the moment due to the

design loads (at mid-span for simply supported beams, at support for
contilevers);
is in MPa units.

Table 5.5.2-1
Values of the factor K in the formulas of l/d for deflection control (part of table 7.4N in
EN 1992-1-1:2004)
Structural system
Value of K to
be used in
(5.5.2-1) and
(5.5.2-2)
- simply supported beam
1,0
- one- or two-way spanning simply supported slab
- end span of continuous beam
1,3
- end span of one-way continuous slab
- end span of two-way spanning slab continuous over one long
side
- interior span of beam or one way or two-way spanning slab
1,5
- flat slab = slab supported on columns without beams (based on
1,2
longer span)
- cantilever
0,4

Expressions (5.5.2-1) and (5.5.2-2) have been derived on the assumption that
- under the appropriate design load at SLS,
- at a cracked section at the mid-span of a beam or slab or at the support of a
contilever,
the steel stress = 310 MPa. This means that steel grade fyk = 500 MPa is used; it is
expected that the steel S500 works at SLS at a stress = 310 MPa. When other stress
levels are used (or other steel grades), the values obtained using expressions (5.5.2-1)
and (5.5.2-2) should be multiplied by 310/s. According to EC2, it is conservative to
assume that:
310
500
=
(5.5.2-3)
As ,req
s
f yk .
As , prov
with
s =

the tensile steel stress under the design load at SLS, at mid span for simply
supported beams or at support for cantilevers;
As,prov = area of steel provided at this section;
As,req = area of steel required at this section for ULS.
The expressions (5.5.2-1) and (5.5.2-2) must be completed by the following additional
prescriptions:

5-42

 for flanged sections (T-beams) where the ratio of the flange breadth to the rib
breadth exceeds 3, the values of l/d should be multiplied by 0,8;
for beams and slabs (other than flat slabs) with spans exceeding 7 m, which
support partitions liable to be damaged by excessive deflections, the values of l/d
should be multiplied by 7/leff (with leff in m);
for flat slabs where the greater span exceeds 8,5 m, and which support partitions
liable to be damaged by excessive deflections, the values of l/d should be
multiplied by 8,5/leff (with leff in m).
Table 5.5.2-2 shows the limit values (maximum!) for the span/depth ration l/d,
calculated by means of expressions (5.5.2-1) and (5.5.2-2), for several reinforced
concrete members, taking into account the following:
-

no axial compression ;
concrete class C30/37;
tensile steel stress s = 310 MPa;
geometrical reinforcement ratio = 0,5 %. This is a light reinforcement, typically
used in slabs. This case is designated in the table by lightly stressed concrete;
geometrical reinforcement ratio = 1,5 %. This is an important reinforcement,
typically used in beams. This case is designated in the table by highly stressed
concrete.

5-43

Table 5.5.2-2
Maximum values of the span/depth ratio l/d for reinforced concrete members without
axial compression, C30/37; s = 310 MPa
(Table 7.4N in EN 1992-1-1:2004)
Structural system
= 1,5 % = 0,5 %
- simply supported beam
14
20
- one- or two-way spanning simply supported slab
- end span of continuous beam
18
26
- end span of one-way continuous slab
- end span of two-way spanning slab continuous over one
long side
- interior span of beam or one way or two-way spanning
20
30
slab
- flat slab = slab supported on columns without beams
17
24
(based on longer span)
- cantilever
6
8
Notes:
- the values shown in the table have been chosen to be conservative in general;
calculation may show that thinner members are possible;
- for two-way spanning slabs, the control of the span/depth ration should be carried
out on the basis of the shorter span. For flat slabs the longer span should be
taken.

Note: The maximum values of the span/depth ratio, obtained with expressions
(5.5.2-1) and (5.5.2-2) and table 5.5.2-2 are the result of a parametric analysis
performed on a wide variety of beams and slabs, for which the deflections have
been calculated for different concrete classes and different reinforcement areas.
For a given reinforcement, the bending moment at failure has been determined
and the quasi-permanent load has been estimated at 50% of the failure load. The
span/depth ratio has then been deduced from the requirements regarding the
maximum deflection.

5.5.3

Deflection control by calculation

See chapter 9 in these course notes.

5.6 Estimation of the depth of cross-sections for preliminar design

5.6.1

Introduction

At the start of a design project, the structural actions have to be determined. One of the
loads to be estimated, is the self-weight of the structural elements to design. However,
the self-weight is function of the dimensions which still have to be determined (because
the dimensions are determined by the loads!). A critical point in the start-up of a design
calculation is thus the first estimation of the depth h or the effective depth d of the
beam.
At this stage, it is important to notice that when a design is performed by the limit state
method, the dimensions of a structural element are sometimes not determined by the
5-44

ULS conditions (the strength), but rather by the SLS conditions, and in particular by the
deflection limitation requirements (problem of rigidity).
The following paragraphs present indications and rules that help in determining the
depth of a cross-section.
5.6.2

Rules with respect to the SLS deflection control

The condition l/d constant, with the constant presented in table 5.5.2-2, permits to find
an initial estimation (conservative) of the effective depth d.

Note: the determination of the span of beams and slabs requires the application
of the simplification rules of structural systems, presented in figure 2.5.2-1.

5.6.3

Rules with respect to the cover of the reinforcement

5.6.3.1 Introduction
The following rules help to pass on from the effective depth d to the total depth h, which
is necessary for the calculation of the self-weight.
5.6.3.2 General considerations
References: EN 1992-1-1:2004; 4.4.1 and NBN EN 1992-1-1 ANB:2009; 4.4.1.2(5)
The concrete cover is the distance between the surface of the reinforcement closest to
the nearest concrete surface (including links and surface reinforcement) and the nearest
concrete surface. The nominal cover, which is specified on the drawings, is defined as
follows:
cnom = cmin + cdev

(5.6.3-1)

with
cnom = the nominal cover;
cmin = the minimum cover;
cdev= supplementary deviation term for the cover.
5.6.3.3 Minimum cover
The minimum cover cmin is provided in order to ensure:
the transmission of bond forces between steel and adjacent concrete;
the protection of the steel against corrosion (durability);
an adequate fire resistance (see EN 1992-1-2).
The minimum cover cmin is determined by the formula:
cmin = max (cmin,b ; cmin,dur ; 10 mm)
with
5-45

(5.6.3-2)

cmin,b =
cmin,dur =

minimum cover due to bond requirement (b < bond);

minimum cover due to environmental conditions (dur < durability).

Note: EN 1992-1-1:2004 allows the reduction of minimum cover for use of

stainless steel or additional protection such as polymer coatings.

5.6.3.4 Minimum cover requirements with regard to bond: cmin,b

In order to assure the transmission of the forces in the steel reinforcement to the
adjacent concrete without the reinforcement slipping, a minimum cover thickness is
necessary. This is associated with the way how forces are introduced into the concrete
via the reinforcement. Figure 5.6.3-1 shows one bar, anchored in a concrete mass,
loaded in tension; the force in the steel is transmitted to the concrete via the ribs on the
bars surface: figure 5.6.3-1(a). This leads to :
- the radial spread-out of compression stresses in the concrete in transverse
direction with respect to the bar: figure 5.6.3-1(b) ;
- circumferential tensile stresses in the concrete. These stresses are more important
for a thin cover layer and for too small distances between adjacent bars: figure
5.6.3-1(c). For small cover thickness and too small spacing between bars, a
typical crack pattern may be observed: figure 5.6.3-1(d). The cracking risk is also
more important for large bar diameters (a phenomenon which will be discussed
in chapter 6 dealing with the anchorage of bars).

Figure 5.6.3-1
Transmission of bond forces: (a) introduction of the forces into the concrete by the ribs
on the bars surface; (b) radial compression forces in the concrete; (c) tensile stresses in
the concrete; (d) typical crack pattern for insufficient cover
EN 1992-1-1:2004; 4.4.1.2(3) stipulates that in order to transmit bond forces safely and
to ensure adequate compaction of the concrete, the minimum cover cmin,b should not be
less than:
or n

5-46

with
=
n =
dg =

( + 5 mm) or ( n + 5 mm) when dg > 32 mm

diameter of one bar;
equivalent diameter of a group of bundled bars;
nominal maximum aggregate size.
Note: bundles bars
EN 1992-1-1:2004; 8.9.1: bars with the same characteristics (type and grade) can
be bundled; bars of different sizes may be bundled provided that the ratio of
diameters does not exceed 1,7. The advantage of bundled bars is that only one
diameter should be considered in design calculations. The bundle is replaced by
a notional bar having the same sectional area and the same centre of gracity as
the bundle. The equivalent diameter n of the bundle is defined by:

n = nb 55mm
with
nb =

(5.6.3-3)

number of bars in the bundle, where nb is limited to:

- 4 for vertical bars in compression (columns) and for bars in lapped
joints;
- 3 for all other cases.

5.6.3.5 Minimum cover requirements with regard to durability: cmin,dur

Note:
In the following text, the prescriptions are taken from NBN EN 1992-1-1
ANB:2009; they differ from the ones in EN 1992-1-1:2004.

The minimum cover necessary to ensure durability (see table 5.6.3-1) is determined by:
the exposure class to be considered (environment class in ANB), and
the "structural class" in which the structure has to be situated.
EN 1992-1-1 recommends adopting the structural class S4 for a design working life of
50 years (also accepted by ANB!). Table 5.6.3-2 presents possible modifications to the
structural class in function of several criteria. In practice, the procedure is as follows:
one examines in table 5.6.3-2 if class S4 can be adopted or if the requirements
should be more or less severe;
table 5.6.3-1 is then used, on the basis of the selected class, to determine the value
of cmin,dur.

Note:
When freez/thaw or chemical attack on concrete (classes XF and XA) are
expected, special attention should be given to the concrete composition

5-47

(see EN 206-1). The values in table 5.6.3-1 are normally sufficient for
such situations.
Table 5.6.3-1
Minimum cover cmin,dur with respect to durability of reinforcement
(Table 4.4N-ANB in NBN EN 1992-1-1 ANB:2009)
Exposure classes and environment classes
XC1
XC2,
XC4
XD1,
XD2,
XD3,
XC3
XS1
XS2
XS3
Structural class
EI
EE1,
EE3
ES2
ES1,
EE4,
EE2
ES3
ES4
S1
10
10
15
20
25
30
S2
10
15
20
25
30
35
S3
10
20
25
30
35
40
S4
15
25
30
35
40
45
S5
20
30
35
40
45
50
S6
25
35
40
45
50
55
For concrete subjected to a chemical aggressive environment (XA and EA), the biggest
of all values cmin,dur which are needed for the other exposure and environment classes,
has to be applied

5-48

Table 5.6.3-2
Recommended structural classification: possible modifications of the structural class in
function of several criteria
(Table 4.3N-ANB in NBN EN 1992-1-1 ANB:2009)
Exposure and environment class
X0, XC1 XC2, XC3
XC4
XD1
XD2, XS1 XD3, XS2,
XS3
E0, EI
EE1, EE2
EE3
ES1, ES2 ES3, EE4,
Criterium
ES4
Design working life of
100 years
Strength class

increase class
by 2

increase class
by 2

increase class
by 2

increase class
by 2

increase class
by 2

increase class
by 2

C30/37
reduce class
by 1

C35/45
reduce class
by 1

C40/50
reduce class
by 1

C40/50
reduce class
by 1

C40/50
reduce class
by 1

C45/55
reduce class
by 1

reduce class
by 1

reduce class
by 1

reduce class
by 1

reduce class
by 1

reduce class
by 1

reduce class
by 1

reduce class
by 1

reduce class
by 1

reduce class
by 1

reduce class
by 1

Member with slab

reduce class
geometry (position of
by 1
reinforcement not
affected by
construction process)(*)
Special quality control
reduce class
of the concrete
( )
by 1
production ensured *
(*) : = quality control system is present

5.6.3.6 Allowance in design for deviation cdev

In order to take account of possible deviations during the construction phase, the
minimum cover has to be increased by a supplement of 10 mm (this is prescribed by EN
and by ANB!).
In certain cases, the supplementary cover may be reduced:
where production is subjected to a quality assurance system in which the
monitoring includes measurements of the concrete cover, the supplement cdev
may be reduced: 5 mm cdev 10 mm;
where a very accurate quality control system is used for monitoring (see
conditions in ANB) with in particular accurate monitoring of temperature, the
supplement cdev may be reduced: 0 mm cdev 10 mm;
In certain cases, the supplementary cover should be increased:
for concrete cast against uneven surfaces, the nominal (in ANB minimum in
EN) cover should be increased with a larger supplementary value, determined in
function of the unevenness. EN and ANB also present rules for concrete cast
directly against soil.

5-49

6 Chapter 6
Detailing of reinforcement

Preliminary note:
The following prescriptions are taken from EN 1992-1-1:2004 Sections 8 and
9. The prescriptions apply to ribbed reinforcement (bar and mesh) subjected to
static loading. They may not be sufficient for:
- dynamic loading caused by seismic effects, machine vibration, impact
loading, etc.;
- painted and epoxy or zinc coated bars.

6.1 Reinforcement in one cross-section

6.1.1

Minimum and maximum reinforcement areas

Reference: EN 1992-1-1:2004; 9.2.1.1

The area of longitudinal tension reinforcement should not be less than:
the minimum reinforcement to control cracking (see chapter 5 in these
course notes - SLS);
the value of As,min, defined by the formula:
As , min = 0,26.

with
bt =

fctm =

f ctm
.bt .d
f yk

but not less than 0,0013.bt.d

(6.1.1-1)

the mean width of the tension zone; for a T-beam with the flange in
compression, only the width of the web is taken into account in
calculating the value of bt;
should be determined with respect to the relevant strength class of the
concrete.
Note:
Sections containing less reinforcement than As,min should be
considered as unreinforced.

The cross-sectional area of tension or compression reinforcement should not

exceed As,max = 0,04.Ac, outside lap locations.

6-1

6.1.2

Choice of bar diameter and number of bars

The calculated reinforcement area As (or As1 and As2) in mm2, has to be translated into a
number of bars. To this end, one can use table 4.2.5-1 (As in function of diameter and
number of bars, presented in chapter 4). In general, it is recommended to limit the
number of different diameters in one project!
Yet, other factors have also to be taken into account when bar diameters have to be
selected; indeed, the design of the reinforcement should also take account of factors
such as optimization of the production and assembly and cost. The cost of the
reinforcement is mainly determined by:
the cost of the steel (50 %);
the wages of the people necessary for the design calculations and the realization
of the construction drawings (10 %);
the handling and treatment of the steel in precast factory or on site: cutting,
bending, bundling, transport and assembly (40 %).
The cost for handling and treatment is also determined by the bar diameter.
The choice of a large diameter has some advantages:
reduction of the number of bars to cut, to bend and to put in the formwork;
easy casting and compaction process.
However, disadvantages become important with diameters larger than 25 mm:
more difficult transport;
more heavy equipment is needed for the bending process;
larger mandrel diameters have to be adopted in order to avoid bending cracks
in the bar. But on the other hand, this may be limited by the available
dimensions of the reinforcement cage;
large diameters ask for large cover thickness and may lead to problems for
crack width control.
It is clear that rational design of reinforcement leads to economic solutions.
WALRAVEN (1995) formulated the problem in this way: it is recommended to choose
large diameters, but not too large.
6.1.3

Spacing of bars

Reference: EN 1992-1-1:2004; 8.2

For technological reasons, it is necessary to adopt minimum distances between bars; this
helps avoiding the reinforcement to act as a sieve when casting takes place, with bad
quality concrete in the neighbourhood of the bars as a consequence. Moreover, a
sufficient quantity of concrete surrounding each bar, is needed to ensure bond quality.
The following prescriptions are recommended:
the clear distance (horizontal and vertical) between individual parallel bars or
horizontal layers of parallel bars should not be less than (see figure 6.1.3-1):
The maximum bar diameter;
dg + 5mm, with dg = the maximum size of aggregate;
20 mm.
where bars are positioned in separate horizontal layers, the bars in each layer
should be located vertically above each other. There should be sufficient space
6-2

between the resulting vertical columns of bars to allow access for vibrators and
good compaction of the concrete: see figure 6.1.3-2.
lapped bars may touch one another within the lap length (lapped joints: see
further in this chapter).

dg + 5 mm
20 mm

Figure 6.1.3-1
Detailing concerning the spacing between individual bars

Figure 6.1.3-2
Sufficient space between the vertical columns of bars to allow access for vibrators

Note:
One may observe that the requirements regarding spacing between bars
are not very severe. For practical reasons, larger spacing is recommended
on construction site: in general 50 mm is adopted as a minimum value.

6.2 Longitudinal reinforcement in beams: identification of the

principal problems
6.2.1

Statically determined beams

The main reinforcement in beams (definition: span l 3.h) is positioned in function of

the evolution of the bending moment: see figure 6.2.1-1. The most unfavourable
moment distribution is considered. The reinforcement area is first determined in the

6-3

cross-section with the highest load; yet, is not necessary to have the same reinforcement
area everywhere along the beam, which means that reinforcement has to be curtailed. A
complementary problem to the cutting of bars, is the determination of the anchorage
length at the ends of the bar.

Figure 6.2.1-1
Schematic representation of the main reinforcement in a simply supported beam loaded
in simple bending by uniformly distributed load

Note: technological reinforcement

Technological reinforcement is provided in the zones where reinforcement is
not needed for strength reasons (for example in the compression zone of a
symply supported beam). The area of the technological reinforcement is in
principle not calculated, but is determined by technological rules . This
reinforcement, which is in general rather light, is provided for several reasons:
- limitation of crack development due to shrinkage ;
- improvement of the resistance against unforeseen small impact loads ;
- easy fixation of links (or stirrups shear reinforcement !) and practical
usefulness for the realization of a reinforcement cage.

6.2.2

Continuous beams

The position of the reinforcement is function of the sign of the bending moment: see
figure 6.2.2-1.

6-4

Figure 6.2.2-1
Schematic representation of the main reinforcement in a continuous beam subjected to
bending loads

6.2.3

Long beams

Beams with large spans need long reinforcement bars. Yet, the practical length of bars is
limited in general to 12 m, because of the increased weight, difficulties in manipulation
of the bars (picking up with crane), length of lorry platforms, etc. Bars have thus to be
connected. Special connectors (in high quality steel) can be used, but this solution has
disadvantages:
- the presence of these thick pieces of steel may hamper the easy realization of the
reinforcement cage (bars can not slide easily, etc.) ;
- high cost, because of the high quality of the connectors with machined
cylindrical holes in which the ends of the bars can be screwed; the ends of the
bars have thus to be machined in order to create the necessary screw thread.
In general, the solution of lapping is used: the connection between two individual bars is
realized by means of a lap with a certain length (very similar to an adhesive bond
between two metal substrates). Particular rules are available to determine the
appropriate lap length; see figure 6.2.3-1.

6-5

Figure 6.2.3-1
Long bars are realized by means of lapping of bars

6.3 Anchorage of longitudinal reinforcement

6.3.1

Parametric analysis of the anchorage problem

6.3.1.1 Introduction
At first sight, a simple solution for the anchorage of longitudinal reinforcement might be
the provision of anchor blocs at the ends of the bars: see figure 6.3.1-1. Such a solution
is however costly and is not used in practice. In general, bars are anchored by means of
an anchorage length in the in the anchor zone.

Figure 6.3.1-1
A theoretical solution for the anchorage of bars by means of anchor blocs at the ends of
the bars
6.3.1.2 Working principle of an anchorage
The working principle of an anchorage at the end of a bar, is shown in figure 6.3.1-2.
The force in the steel bar is transferred to the concrete by means of radial struts of

6-6

compressed concrete, which are inclined with respect to the axis of the steel bar. Figure
6.3.1-2(b) shows that in order to realize equilibrium of forces in each node at the surface
of the steel bar, tensile forces have to be introduced, perpendicular to the steel bar.
These tensile forces can be resisted by the concrete tensile strength when a sufficiently
thick cover layer is present.
compression

tension
compression in
concrete strut

Tensile
force in
concrete

Tensile force
in steel

(a)

(b)

Figure 6.3.1-2
Anchorage of the end of a straight bar;
(a) scheme of the force transmission; (b) equilibrium of forces
The calculation of the required anchorage length takes into consideration the notion of
bond stress, which is in fact due to the horizontal component of the diagonal
compression forces shown in figure 6.3.1-2. The distribution of the bond stress along
the steel reinforcement depends on strain situation in the concrete cover:
- when concrete strains are in the elastic domain, the bond stress is most important
in the location where the force is introduced (see theory of bonded joints): figure
6.3.1-3;
- when the stiffness of the material is gradually reduced due to the accumulation of
microscopic cracks, a redistribution of stresses takes place; this leads to the
appearance of the important bond stresses at the end of the bar.
In practice, the evolution of the bond stress is also influenced by creep and the eventual
cyclic nature of the loads.
Because of the many uncertainty factors, and in order to avoid too elaborated
calculations, a constant bond stress is adopted along the whole anchorage length.

Figure 6.3.1-3
Evolution of the bond stress along the anchorage length; solid line = bond stress
distribution for elastic strains in the concrete cover; dashed line = bond stress

6-7

distribution at failure; dash-dot line = simplified model adopted for the bond stress
distribution
6.3.1.3 Principle determination of the anchorage length
A formula for the calculation of the anchorage length lb can be developed on the basis of
figure 6.3.1-4. An equilibrium of forces equation is written: the anchorage force (which
is distributed over the whole length of the anchorage lb) = the force As.s which is
transferred by the steel bar to the concrete:

. . .lb =

. 2
4

. s

And thus:

lb =

s
.
4

(6.3.1-1)

lb

Ns =

. 2 . s
4

Figure 6.3.1-4
Auxiliary figure for the determination of the anchorage length lb
6.3.1.4 Parameters affecting the bond strength
Experimental analysis by means of pull out" tests (WALRAVEN, 1995) has shown
that the bond strength
- increases with better concrete quality and with higher concrete cover thickness,
but
- decreases with large diameter bars or bundles (WALRAVEN, 1995).
Another important parameter is the position of the reinforcement in the structural
element: indeed, the quality of the concrete as well as the bond strength varies in
function of the considered position. This is related to the compaction process.
Compaction (due to vibration) causes a volumetric contraction and leads to the setting
of the concrete: figure 6.3.1-5. The plastic concrete paste flows around the bar, but the
bar itself does not move (because it is fixed in the formwork); this may lead to
weakened zones under the bar, which are filled with water and fine particles. The
quality of the concrete just under the bars is low and this affects the bond strength. The
setting phenomenon does not appear

6-8

at the level of the lower bars in the cross-section, close to the bottom of the
formwork;
- along inclined bars, because the liquid fraction will move upwards along the
inclined bar during compaction.
The standard takes account of this phenomenon by considering zones in a reinforced
concrete element with good or poor bond conditions (see further).

Level after casting

Level after compaction
bar A

bar B

Poor quality
concrete

Good quality
concrete

Figure 6.3.1-5
Influence of the position of bars in the cross-section, on the bond quality
6.3.1.5 Other methods of anchorage
Figure 6.3.1-6 shows the support area of a simply supported beam. An inclined shear
crack (see chapter 7 in these course notes) starts from the lower side of the beam at the
edge of the support. In principle, the main reinforcement, which is needed to help in
supporting the shear load (see chapter 7), should be well anchored to the left. It is clear
that it is not possible to have the whole anchorage length for a straight bar on the left of
the support, unless the beam gets a cantilever part, which is sometimes not possible or
not practical anyway. This example shows that it may be necessary to shorten the
anchorage length.

Figure 6.3.1-6
The anchorage length of the straight bar should be positioned to the left of the shear
crack. This asks for a cantilever solution or for a method to reduce the anchorage length

6-9

Several methods are available to reduce the anchorage length:

by increasing the reinforcement area; formula (6.3.1-1) shows that a smaller value
of s leads to a smaller value of lb;

by bending the end of the bar into a standard bend, a hook or loop;

by welding transverse bars to the main reinforcement: figure 6.3.1-7.

Figure 6.3.1-7
Reduction of the anchorage length by means of transverse welded bars;
(a) top view; (b) side view

6.3.2

Prescriptions concerning the anchorage of longitudinal reinforcement

Reference: EN 1992-1-1:2004; 8.4 (confirmed by ANB)

6.3.2.1 Methods of anchorage
Next to the anchorage by straight bars, the EC2 distinguishes four alternative types for
the anchorage for bars in tension (see figure 6.3.2-1):
- the standard bend (NL: elleboog; FR: coude),
- the standard hook (NL: haak; FR: crochet),
- the standard loop (NL: lus; FR: boucle), and
- the anchorage by means of welded transverse bars.
Attention: bends and hooks do not contribute to compression anchorages.

Note:
The concrete at the inner side of a bend, hook or loop is vulnerable for crushing
or splitting. Moreover, as already pointed out earlier, bending cracks in the
reinforcement bars have to be avoided. That is the reason why EN 1992-1-

6-10

1:2004; 8.3 recommends values for permissible mandrel diameters for bent bars
(in function of the bar diameter).

Figure 6.3.2-1
Methods of anchorage other than by straight bar: (a) standard bend; (b) standard hook;
(c) standard loop; (d) anchorage by means of welded transverse bars
6.3.2.2 Ultimate bond stress
The design value of the ultimate bond stress fbd (NL: grenskleefspanning; FR: contrainte
limite d'adhrence) for ribbed bars, is given by the formula:
fbd = 2,25 . 1 . 2 . fctd
with
fctd =

1 =

(6.3.2-1)

the design value of concrete tensile strength, for which the value of fctk,0,05 (see
table 3.4.3-1) can be used. Due to the increasing brittleness of higher strength
concrete, fctd should be limited to the value of fctk,0,05 for concrete class C60/75.
Table 6.3.2-1 presents an overview of the values of fctd to be used for the
different concrete classes.
coefficient related to the quality of the bond condition and the position of the bar
in the cross-section: see figure 6.3.2-2. The bond quality is indeed not only
dependent on the nature of the surface of the bar, but also on the dimensions of
the structural element, the position of the bar in the cross-section and the
inclination of the bar during the casting phase. Two values are recommended for
the coefficient 1:
1 = 1 for good bond conditions;
1 = 0,7 for all other cases (= called "poor" bond conditions).
6-11

2 =

coefficient related to the bar diameter. Large bar diameters (which may transfer
large tensile forces) may lead to increased loading of the bond and thus to an
increased risk for the quality of the bond. The following values are
recommended for the coefficient 2:
2 = 1 for bars with 32 mm;
1 = (132 - )/100 for bars with > 32 mm.

Table 6.3.2-1
Recommended values for the concrete tensile stress to be used in formula (6.3.2-1) for
the calculation of the bond stress fbd
Concrete classes
fck
(MPa)
fck,cube
(MPa)
fctk,0.05
(MPa)
fctd
(MPa)

12

16

20

25

30

35

40

45

50

55

60

70

80

90

15

20

25

30

37

45

50

55

60

67

75

85

95

105

1,1

1,3

1,5

1,8

2,0

2,2

2,5

2,7

2,9

3,0

3,0

3,2

3,4

3,5

1,1

1,3

1,5

1,8

2,0

2,2

2,5

2,7

2,9

3,0

3,0

3,0

3,0

3,0

Figure 6.3.2-2
Overview of bond conditions: unhatched zone = good bond conditions; hatched zone =
poor bond conditions (figure 8.2 in EN 1992-1-1:2004)
6.3.2.3 The basic required anchorage length lb,rqd
The basic required anchorage length lb,rqd (NL: vereiste basis verankeringslengte; FR:
longueur dancrage de base requise) is determined on the basis of expression (6.3.1-1);
lb,rqd is necessary for anchoring the force As.sd in a straight bar, assuming constant bond
6-12

stress equal to fbd. Expression (6.3.1-1) then leads to:

lb ,rqd = . sd
4 f bd
with
sd =
fbd =

(6.3.2-2)

design stress of the bar at the position from where the anchorage is measured;
ultimate bond stress.

6.3.2.4 Design anchorage length lbd

The design anchorage length lbd is determined from the following expression:
lbd = 1 . 2 . 3 . 4 . 5 . lb,rqd lb,min

(6.3.2-3)

with
i a series of coefficients which take account of various factors that influence the
bond characteristics (discussion of these coefficients: see further):
1
effect of the form or shape of the bar;
2
effect of the thickness of the concrete cover;
3
effect of confinement by the presence of non-welded transverse
reinforcement;
4
influence of welded transverse bars along the design anchorage length;
5
effect of the pressure transverse to the plane of splitting along the design
anchorage length.
with the product (2 . 3 . 5) 0,7
lb,min is the minimum anchorage length, determined by:
for anchorages in tension: lb,min > max (0,3.lb,rqd ; 10. ; 100 mm);
for anchorages in compression: lb,min > max (0,6.lb,rqd ; 10. ; 100 mm);
Important: the design anchorage length lbd is measured along the centreline of the bar:
see figure 6.3.2-3.

Figure 6.3.2-3
The design anchorage length lbd is measured along the centreline of the bar

6-13

6.3.2.5 Discussion of the coefficients i in the formula (6.3.2-3) for the calculation of
the design anchorage length lbd
1

coefficient to take account of the influencing factor: shape of the bar;

for a straight anchorage:
- in compression: 1 = 1,0
- in tension: 1 = 1,0
for anchorages other than straight (bend, hook, loop):
- in compression: 1 = 1,0
- in tension: 1 = 0,7 if cd > 3., otherwise 1 = 1,0. The value of cd is taken
from figure 6.3.2-4; cd gives an indication if the bar is sufficiently covered
by concrete (in order to make a good bond possible).

Figure 6.3.2-4
Auxiliary figure for the determination of the factor cd, which indicates if the
bar is sufficiently covered by concrete (figure 8.3 in EN 1992-1-1:2004)
2

coefficient to take account of the influencing factor: thickness of the concrete

cover;
for a straight anchorage:
- in compression: 2 = 1,0
- in tension: 2 = 1 - 0,15.(cd ) / with the conditions 2 0,7 and
2 1,0 ; the value of cd is taken from figure 6.3.2-4.
for anchorages other than straight (bend, hook, loop):
- in compression: 2 = 1,0
- in tension: 2 = 1 - 0,15.(cd 3) / with the conditions 2 0,7 and
2 1,0; the value of cd is taken from figure 6.3.2-4.

coefficient to take account of the influencing factor: confinement by transverse

reinforcement not welded to main reinforcement;
for all types of anchorages (straight and non-straight):
- in compression: 3 = 1,0
- in tension: 3 = 1 - K. with the conditions 3 0,7 and 3 1,0;
the factor is defined by = (Ast - Ast,min) / As
with
Ast
cross-sectional area of the transverse reinforcement along
the design anchorage length lbd;
6-14

Ast,min

cross-sectional area of the minimum transverse

reinforcement, which always has to be provided with
anchorages (in tension and in compression). Ast,min =
0,25.As for beams and 0 for slabs;
As
area of a single anchored bar with maximum bar diameter.
The factor K is taken from figure 6.3.2-5.

Figure 6.3.2-5
Auxiliary figure for the determination of factor K, necessary for the
calculation of the coefficient 3, for beams and slabs
(figure 8.4 in EN 1992-1-1:2004)
4

coefficient to take account of the influencing factor: confinement by welded

transverse reinforcement along the anchorage length (see scheme in figure 6.3.21);
for all types of anchorages (straight and non-straight):
- in compression: 4 = 0,7;
- in tension: 4 = 0,7.

coefficient to take account of the influencing factor: confinement by transverse

pressure to the plane of splitting along the anchorage length;
for all types of anchorages (straight and non-straight):
- in compression : 5 = 1,0;
- in tension: 5 = 1 - 0,04.p with the conditions 5 0,7 and 5 1,0 ; p is the
value of the transverse pressure (in MPa) at ULS along lbd.

6.3.2.6 Simplified alternative to the prescriptions in paragraph 6.3.2.4

Paragraph 6.3.2.4 presents recommendations to determine the design anchorage length
lbd which is measured along the centreline of the bar: see figure 6.3.2-3. A simplified
alternative may be applied to the tension anchorage, by means of the equivalent
anchorage length" lb,eq which is defined in figure 6.3.2-6. lb,eq may be taken as:
bend, hook and loop: lb,eq = 1 . lb,rqd;
welded transverse bars: lb,eq = 4 . lb,rqd;

6-15

Figure 6.3.2-6
Simplified alternative for tension anchorage, by application of the equivalent
anchorage length" lb,eq, for different types of anchorages:
(a) bend; (b) hook; (c) loop; (d) anchorage by means of welded transverse bars
(figure 8.1 in EN 1992-1-1:2004)
6.3.2.7 Anchorage by means of welded bars
It appears from the paragraphs discussed before that the use of welded transverse bars
leads to an increase of the anchorage capacity.
EN 1992-1-1:2004; 8.6 proposes a model for the calculation of the positive effect of a
welded bar and the translation of this effect into a reduction of the anchorage length lbd,
which may even become smaller than lb,min.

6.3.3

Complementary prescriptions for the anchorage of the main reinforcement

particular rules for beams

6.3.3.1 Anchorage of bottom reinforcement at an end support

Reference: EN 1992-1-1:2004; 9.2.1.4 (confirmed by ANB)
-

At simple supports (supports with little or no end fixity assumed in design), the
area of bottom reinforcement should be at least 25 % of the area of steel provided
in the span;
the anchorage of the main reinforcement at the end support should be able to resist
the following tensile force:
FE = VEd .

ai
+ N Ed
z

6-16

(6.3.3-1)

with:
NEd = design value of eventual axial force;
VEd = design value of the shear force at the end support;
ai =
z (cotg cotg )/2
The justification of this formula is presented in chapter 7 of these course notes,
with the discussion of the shift rule .
the anchorage length lbd (defined in paragraph 6.3.2.4) is measured from the line
of contact between beam and support: see figure 6.3.3-1.

Figure 6.3.3-1
Anchorage of bottom reinforcement at an end support:
(a) direct support: beam supported by wall or column;
(b) indirect support: beam intersection another supporting beam
(figure 9.3 in EN 1992-1-1:2004)
6.3.3.2 Anchorage of bottom reinforcement at intermediate supports
Reference: EN 1992-1-1:2004; 9.2.1.5 (confirmed by ANB)
-

At intermediate supports, the area of bottom reinforcement should be at least 25 %

of the area of steel provided in the span;
the anchorage length should not be less than 10. (for straight bars) or not less
than (see figure 6.3.3-2):
- the diameter of the mandrel for hooks and bends with bar diameters at least
equal to 16 mm;
- twice the diameter of the mandrel for hooks and bends where the bar
diameter is less than 16 mm.
the reinforcement required to resist possible positive moments at the intermediate
supports (tension at the bottom side) due to settlement of the support, explosion,
etc., should be continuous ; this may be achieved by means of lapped bars; see
figure 6.3.3-2.

6-17

Figure 6.3.3-2
Anchorage of bottom reinforcement at intermediate supports; dm = diameter of the
mandrel (figure 9.4 in EN 1992-1-1:2004)

6.4 Connection of bars

6.4.1

Introduction

The commercial length of reinforcement bars is limited in general to 12 m; longer bars

are realized by connecting individual bars. Forces are transmitted from one bar to
another by:
welding;
mechanical devices, such as couplers with threaded cylindrical holes in which the
ends of bars can be screwed;
lapping of bars, with or without bends or hooks (lapping = NL: overlapping; FR:
recouvrement).
The last method is discussed in the following.

6.4.2

Parametric analysis of laps

Figure 6.4.2-1 shows a connection loaded in tension. When the distance between the
two bar ends is small (similarity with adhesive bonded joints), a series of compressed
diagonal struts appear, which show an inclination of about 45 (in a simplified way)
with respect to the bar axis. The compressed struts may be separated by cracks. In order
to get equilibrium of forces, it is necessary to introduce tensile forces which are
perpendicular to the bar axis and which equilibrate the longitudinal force in the bars and
the diagonal forces in the struts. When the tensile forces are not too large, the tensile
strength of the concrete surrounding the bars, may be sufficient to assure the
equilibrium.
Figure 6.4.2-1 also remembers that the transmission of forces in a lapped joint, takes
essentially place at the ends of the lap (see also theoretical aspects of adhesive bonded
joints).

6-18

tension

strut in
compression

l0
Figure 6.4.2-1
Various forces are present in a lapped joint
Figure 6.4.2-2 shows another damage mechanism which may appear with laps in
reinforce concrete. Because of the flexural rigidity of a bar, the concrete cover may be
pushed off (spalling effect) when the lap is situated in a heavily curved zone. This is
accompanied by tensile loading of the concrete in transverse direction at the ends of the
bars, which explains the necessity of transverse reinforcements at the ends of the
longitudinal bars.

Figure 6.4.2-2
Dislocation of the concrete cover (spalling effect) due to the presence of a lapped joint
in a heavily bended zone of a beam
The potential damage mechanisms discussed before, explain why it is not recommended
to have all the laps at the same location along the beam axis. Tensile stress
concentrations can be avoided by an adequate distribution of laps over the length of the
construction element; see figure 6.4.2-3.

6-19

Bottom side of section

Figure 6.4.2-3
Distribution of lapped joints along the beams axis
From the text before, it is concluded that lapped joints have to be used with precautions.
Yet, a lapped joint is a common and even useful tool in the practical realization of the
reinforcement of a structural element. Figure 6.4.2-4 shows a typical case where the use
of laps has a positive effect. During the manipulation of reinforcement bars (cutting,
bending), dimensional deviations are difficult to avoid. Consequently, a bar which is
slightly too short may present insufficient anchorage lengths, while a too long bar will
lead to insufficient concrete cover at the ends. It is preferred to avoid bars with exact
dimensions because it is difficult on the construction site to correct adequately errors
regarding lengths of bars. The use of lapped joints may help to solve this problem, as is
illustrated in figure 6.4.2-4.
c
bar with exact length

lapped joint

Figure 6.4.2-4
The use of lapped joints helps to solve problems with dimensional variations for bars
with exact length: the presence of the lap permits to ensure a sufficient concrete cover
at the ends of the beam

6.4.3

Prescriptions regarding lapped joints

Reference: EN 1992-1-1:2004; 8.7 (confirmed by ANB)

6.4.3.1 Arrangement of lapped joints
The following general principles have to be applied:

6-20

the detailing and arrangement of the laps of several bars in a structural element
should be such that the transmission of forces from one bar to another is assured;
it is recommended to avoid having all (tensile) laps in the same area;
tensile laps should not be located in areas of high bending moments;
laps should be arranged symmetrically in cross-sections, and laps should be
parallel to the concrete surface.

The standard recommends the following prescriptions (see figure 6.4.3-1):

the clear distance between lapped bars should not be greater than 4. or 50 mm,
otherwise the lap length l0 should be increased by the length equal to the clear
space where it exceeds 4. or 50 mm;
the longitudinal distance between two adjacent laps should not be less than 0,3.l0;
the clear distance between two adjacent lap joints should not be less than 2. or 20
mm;
when the conditions mentioned above are met, the permissible percentage of
lapped bars in tension may be 100 % where the bars are all in one layer. Where
the bars are in several layers, the percentage should be reduced to 50 %;
all bars in compression and secondary (technological) reinforcement may be
lapped in one section.

Figure 6.4.3-1
Practical arrangements for adjacent laps
(figure 8.7 in EN 1992-1-1:2004)
6.4.3.2 Design lap length l0
The design lap length l0 is determined by means of the following formula:
l0 = 1 . 2 . 3 . 5 . 6 . lb,rqd l0,min

(6.4.3-1)

with
lb,rqd calculated from expression (6.3.2-2)
l0,min > max (0,3.6.lb,rqd ; 15. ; 200 mm);
1
idem as for anchorage length;
2
idem as for anchorage length;
3
idem as for anchorage length, but with one difference:
Ast,min
= the area of the transverse reinforcement which has
6-21

always to be provided, is now equal to 1,0.As.(sd/fyd),

with:
As
= area of one lapped bar.
5
idem as for anchorage length;
6
= (1/25)0,5 with 1,0 6 1,5 and with 1 = percentage of reinforcement
lapped within 0,65.l0 from the centre of the lap length considered: see figure 6.4.32. Table 6.4.3-1 presents some values for 6.

Figure 6.4.3-2
Percentage of lapped bars in the neighbourhood of one lap
(figure 8.8 in EN 1992-1-1:2004)

Table 6.4.3-1
Values of the coefficient 6 for the calculation of the lap length
(table 8.3 in EN 1992-1-1:2004)
Percentage of lapped bars in the neighbourhood of
<
33 % 50 %
>
one laps (see figure 6.4.3-2)
25 %
50 %
6
1
1,15
1,4
1,5
Intermediate values may be determined by interpolation

6.4.3.3 Transverse reinforcement in the lap zone: bars in tension

Transverse reinforcement is required in the lap zone to resist transverse tension forces.
The following recommendations should be applied:
- where the diameter of the lapped bar < 20 mm, or the percentage of lapped bars
is less than 25 %, then any transverse reinforcement or links necessary for other
reasons (context of stirrups as shear reinforcement) may be assumed sufficient to
resist to the transverse tensile forces along the lap joint;
- where the diameter of the lapped bar 20 mm, the transverse reinforcement
should have an area Ast (sum of all legs parallel to the layer of the

6-22

reinforcement with the lap) not less than the area As of one lapped bar: Ast As.
The transverse bars should be placed:
- perpendicular to the direction of the lapped reinforcement;
- between the lapped bars and the concrete surface;
- concentrated at the ends of the lap, as shown in figure 6.4.3-3 (a);
if more than 50 % of the reinforcement is lapped at one point and the distance
between adjacent laps is 10., transverse reinforcement should be formed by
links or U bars anchored into the body of the section.

6.4.3.4 Transverse reinforcement in the lap zone: bars in compression

With a lap joint loaded in compression, there is a risk that a part of the load will be
transferred to the concrete by the end of the bar, which may lead to the splitting of the
concrete. That is the reason why, in addition to the rules for laps in tension, one bar of
the transverse reinforcement should be placed outside each end of the lap length and
within 4. of the ends of the lap length: see figure 6.4.3-3 (b).

Figure 6.4.3-3
Arrangement of transverse reinforcement along a lap joint
(figure 8.9 in EN 1992-1-1:2004)
6.4.3.5 Note: links may be used as transverse reinforcement along lap joints
It is already mentioned in paragraph 6.4.3.3 that in certain cases, the existing
reinforcement (parts of links) is sufficient to resist the transverse tensile forces along a
lapped joint.
Figure 6.4.3-4 (a) shows main reinforcement bars in overlap in the horizontal plane; the
horizontal parts of the links (which are present anyway to assure resistance to shear
6-23

see chapter 7) can be taken into account in the calculation of transverse reinforcement
along the lap joint.
Attention: figure 6.4.3-4 (a) shows that the horizontal leg of the link may be designed
for only one lap. When the main reinforcement bars are in overlap in the vertical plane,
the transverse tensile forces have to be added (figure 6.4.3-4 (b)): the vertical legs of the
link have to resist the total force Ns of the bars in overlap.

(a)

(b)
.Ns

Ns

.Ns

Ns

Figure 6.4.3-4
The legs of links can be used for the design of the transverse reinforcement along
lapped joints; (a) main reinforcement bars in overlap in the horizontal plane; (b)
main reinforcement bars in overlap in the vertical plane
6.4.3.6 Laps for welded mesh fabrics made of ribbed wires
See EN 1992-1-1:2004; 8.7.5

6.5 Anchorage of and lapping large diameter bars ( > 32 mm)

The loads that are transmitted may be large; transverse tension and splitting forces are
higher. Additional prescriptions are available for the transverse reinforcement, in
particular for the anchorages (where mechanical devices are recommended). In general,
lap joints are not used with large diameter bars; mechanical connectors should be used.
See EN 1992-1-1:2004; 8.8

6.6 Anchorage of and lapping bundles of bars

See EN 1992-1-1:2004; 8.9

6-24

7 Chapter 7
Design for shear loads
7.1 Introduction
The aim of chapter 4 was to show how to calculate the main reinforcement of a beam
subjected to bending moments and axial forces which give rise to distributions of
normal stresses in cross-sections. The main principle was that steel reinforcement is
provided in those areas where the concrete cracks due to large tensile stresses.
Figure 7.1-1 shows a beam in a four point bending disposition. The central part of the
beam is loaded in pure bending; the longitudinal reinforcement is calculated as
indicated in chapter 4. The two parts of the beam between the concentrated forces and
the supports are subjected to a more complex loading because of the combination of the
bending moment and the shear force. Yet, in the early days of reinforced concrete,
people tried out the behaviour of beams with only longitudinal reinforcement and
observed for increasing loads the appearance of inclined cracks in the zones with shear
loads. Without special reinforcement to bridge the inclined cracks, it is even observed
that failure of the beam is determined by shear: one crack is prolonged suddenly up to
the upper side of the beam which causes the total collapse of the structural element (as
shown in figure 7.1-2). This type of failure happens in a sudden (brittle) way and has
thus absolutely to be avoided. The logical solution is to provide inclined reinforcement,
perpendicular to the cracks (figure 7.1-3), but a valuable alternative is to use vertical
links (or stirrups) which bridge the crack at a certain angle.

Bending moment

Shear force

Figure 7.1-1
Four point bending test, applied on a beam with only longitudinal reinforcement

7-1

Figure 7.1-2
Typical failure mode for a beam with only longitudinal reinforcement: a shear crack
leads to total collapse in a sudden way

(a)

(b)

Figure 7.1-3
Two possible solutions to bridge shear cracks; (a) inclined shear reinforcement:
longitudinal bars (main reinforcement on the bottom side) may be bent up to the upper
side of the beam instead of being simply curtailed; (b) vertical shear reinforcement:
links or stirrups

7.2 Members not requiring shear reinforcement

7.2.1

Introduction

Providing shear reinforcement leads to a substantial cost; it is thus useful to analyse the
conditions which may allow omitting this type of reinforcement. This paragraph focuses
on the determination of the shear resistance of members without shear reinforcement.
7.2.2

A starting point: overview of results from theory of elasticity for beams with
continuous, homogeneous, isotropic and elastic materials

The following paragraph presents an overview of main notions and formulas concerning
shear forces and shear stresses in a beam loaded in bending, taken from theory of
elasticity and strength of materials courses. The formulas are valid for homogeneous,
isotropic, continuous and elastic materials. The setting is defined in figure 7.2.2-1.

7-2

P
x

NA

N'+dN'

N'

dx

+
x
Vy > 0

Mz > 0

Figure 7.2.2-1
Principle figure for the elaboration of the formulas for shear stresses in beams loaded in
bending

Main results are:

longitudinal shear force dN ' = b .b.dx
rotation equilibrium in a cross-section leads to (figure 7.2.2-2):
M = N '.z
dM = dN '.z

N'
z

NA

M
N

Figure 7.2.2-2
Rotation equilibrium in a cross-section of a beam loaded in bending
the relationship between Vy and Mz (in absolute values):
dM z
Vy =
dx
the shear stress on the level of the NA:
dN '
dM
V
NL =
=
=
b.dx b.z.dx b.z

(7.2.2-1)

the formula of JOURAWSKI for the shear stress in a certain point (or on a certain
level) of the cross-section:

7-3

 xy =

V .S z
b.I z

(7.2.2-2)

with
xy the y-component of the shear stress on an elementary surface perpendicular to
the x-axis;
V the shear load in the cross-section;
b the width of the cross-section at the level where the shear stress is
determined;
I z the moment of inertia of the full cross-section with respect to the z-axis (axis
passing through the centre of gravity G);
S z the static moment of the part of the cross-section situated above the level
where the stress is determined, with respect to the z-axis.

xy is characterized by a parabolic distribution for a rectangular cross-section,

with maximum value at the level of the NA equal to

3 V
.
.
2 b.h

Figure 7.2.2-3 presents a beam loaded by a uniformly distributed load. In uncracked

situation, and assuming continuous, homogeneous, isotropic and elastic material, one
obtains the set of trajectories of the principal stresses. The orientation and magnitude of
the stresses are determined in each point with a theory of elasticity approach; MOHRs
circle can be used for graphical representation. Figure 7.2.2-4 presents in a schematic
way the reasoning that permits to determine the principal orientations, the principal
elementary areas and principal stresses in a point A on the NA (x = 0; xy max) and in
point B in the cross-section. It is observed in point A that the principal tensile stress has
the same magnitude as the shear stress and is oriented with an angle of 45 with respect
to the axis of the beam. In punt B, the principal elementary area with the largest
principal tensile stress is much more horizontally oriented. These results help to
understand the crack pattern due to shear load in a beam in reinforced concrete with
only main reinforcement and in which, from a macroscopic point of view, the concrete
may be considered as a homogeneous material: see figure 7.2.2-5.

Figure 7.2.2-3
Principal stress trajectories in uncracked situation (continuous, homogeneous, isotropic
and elastic material)

7-4

V>0

x
A

V .S
b.I

XA
XB

1B
poleB

1A

YB
YA = poleA

B
A

Tensile zone cracks

Figure 7.2.2-4
Application of MOHRs circle for the identification of the principle tensile stress at the
NA (axis of the beam); deduction of the crack pattern influenced by the presence of
shear

7-5

Figure 7.2.2-5
Figure (a) presents the trajectories of the principle compression stresses in an uncracked
beam; figure (b) presents the experimentally observed crack pattern obtained by a four
points bending test on a beam in reinforced concrete without shear reinforcement
(WIGHT, 2009)
7.2.3

Effect of the cracking in reinforced concrete (beam without shear

reinforcement)

The appearance of cracks has an important influence on the further distribution of

internal forces. As cracks develop in the lower part of the beam, the NA is shifted
upwards which leads to the vertical elongation of the cracks; these cracks only deviate
towards the 45 orientation on the level of the new NA. This explains why the crack
pattern shown in figure 7.2.3-1 is characterized by much more vertical cracks than the
45 disposition in uncracked material. When load intensity increases, cracking
continues until one crack becomes instable: that means that the crack develops in a
brittle way over the whole depth of the beam. Internal equilibrium is not possible any
more and failure is reached.

Figure 7.2.3-1
7-6

Crack development with increasing load, in a beam in reinforced concrete with main
reinforcement and without shear reinforcement (WALRAVEN, 1995)
Another consequence of the cracking is that equations (7.2.2-1) and (7.2.2-2) are strictly
not valid anymore. Moreover, the stress distribution in the cross-section, in ULS, is nonlinear and is thus highly different from the distribution in uncracked state.

Note:
WALRAVEN (1995) assumes that the following formula still allows
determining a reasonable estimation of the mean" shear stress in a section in
reinforced concrete:
V
=
(7.2.3-1)
b.z
with
b = the width of the cross-section or the minimum width of the web of I- or Tbeams;
z = the lever arm, which in first approximation can be taken as 0,9.d.

7.2.4

Mechanisms of the transfer of shear loads in a cracked beam in reinforced

concrete

Figure 7.2.4-1 gives an overview of the different mechanisms which explain the transfer
of the shear load in a beam in reinforced concrete without shear reinforcement.

(a)
(b)

(c)

(d)

Figure 7.2.4-1
Mechanisms for shear load transfer in cracked reinforced concrete:
(a) uncracked concrete in compression; (b) tensile stresses at the tip of the crack;
(c) granulate interlocking; (d) dowel action
The following mechanisms are identified:
the uncracked compression concrete in the upper part of the beam (above the
shear crack) is able to transfer high shear loads;
7-7

 tensile contact stresses are present at the crack tip as long as both sides are not
separated more than w 0,15 mm (WALRAVEN, 1995). In order to further open
the crack tip, an additional tensile force has to be developed;
the shear displacement of one part of the beam with respect to the other part is
hindered by the mechanical friction resistance provided by the sliding of two
irregular crack surfaces. This is called the aggregate interlocking effect;
the shear displacement of one part of the beam with respect to the other part is
also hindered by the dowel action of the main reinforcement bars. On top of the
local shear resistance of the steel bars, one may also take account of the resistance
to local crushing of the concrete adjacent to the bars: figure 7.2.4-2.

Figure 7.2.4-2
Dowel action of the main reinforcement and resistance to local crushing of the
concrete adjacent to the bars

It can thus be concluded that the following factors determine the shear load bearing
capacity of beams without shear reinforcement:
the concrete class;
the main reinforcement ratio (a larger ratio also leads to smaller crack widths);
the width of the cross-section;
the depth of the cross-section. An important observation is that shear load bearing
capacity indeed increases with depth but less than proportional. This is a well
known phenomenon in the course on failure mechanics: a large crack is more
sensitive for instable elongation than a short crack (small sections are more
efficient to bear shear loads);
an eventual axial force, which may influence the crack width.

7-8

7.2.5

The shear resistance of a beam in reinforced concrete without shear

reinforcement

Reference: EN 1992-1-1:2004; 6.2.2

The design value for the shear resistance VRd,c in a beam without shear reinforcement is
determined by means of the following empirical formula:

VRd ,c

= C Rd ,c .k .(100. l . f ck ) 3 + k1. cp .bw .d

(7.2.5-1)

with a minimum of:

VRd ,c ,min imum = vmin + k1. cp .bw .d

(7.2.5-2)

where:
VRd,c is expressed in N;
fck is expressed in MPa;
200
with d in mm and k 2;
d
d = effective depth determining the distance between the centre of gravity of the
main reinforcement to the most compressed concrete fibres (top layer of beam);
bw = smallest width of the cross-section in the tensile area;

k = 1+

cp =

N Ed
< 0,2.fcd
Ac

N Ed (expressed in N): axial force in the cross-section due to loading or

prestressing (positive sign for compressive load);
Ac (expressed in mm2): area of concrete cross-section;

l =

Asl
0,02 ;
bw .d

with Asl = area of the tensile reinforcement which extends at the least over the
distance d+lbd beyond the section considered (see figure 7.2.5-1). Note: lbd is the
required anchorage length, discussed in chapter 6 in these course notes;
0,18
C Rd ,c =
; assuming c = 1,5 leads to CRd,c = 0,12;

k1 = 0,15;
vmin = 0,035.k3/2.fck1/2;
The introduction of the recommended values k1 = 0,15 , CRd,c = 0,12 and vmin in
equations (7.2.5-1) and (7.2.5-2) leads to the following equations for the design value of
the shear resistance of a beam without shear reinforcement:

7-9

V Rd ,c = 0,12.k .(100. l . f ck ) 3 + 0,15. cp .bw .d

(7.2.5-3)

with a mimimum of:

V Rd ,c ,min imum = 0,035.k 3 / 2 . f ck

1/ 2

+ 0,15. cp .bw .d

(7.2.5-4)

Figure 7.2.5-1
Definition of Asl in the formula for the calculation of the shear resistance of a beam
without shear reinforcement: one can only take account of those bars which are
adequately anchored; (a) end support; (b) intermediate support (Figure 6.3 in EN 19921-1:2004)

The verification of the shear load bearing capacity of a structural member without shear
reinforcement is thus performed by the comparison, in the cross-section to be
considered, of the design value of the imposed shear load VEd with VRdc.

7.3 Members requiring design shear reinforcement

7.3.1

Introduction

If preliminary calculation shows that the shear load bearing capacity of the member
without shear reinforcement, is not large enough to withstand the imposed shear force
(thus if VEd > VRd,c), an adequate shear reinforcement is necessary. The shear
reinforcement provides replacement of the shear load bearing capacity which disappears
gradually with growing cracks, the reduction of the thickness of the compressed
concrete arch and the increased crack width which reduces the granulate interlocking
resistance. The presence of shear reinforcement allows to further increase loads while
avoiding catastrophic beam shear failure before the full exploitation of the bending
capacity.
Throughout the years, it was not easy to find an international agreement on a shear
reinforcement calculation model. The models proposed in the CEB-FIP Model Code
(precursor of EC2) and later on in the EC2, have been reworked several times.
The shear reinforcement calculation model has been developed on the basis of
remarkable experimental results.

7-10

7.3.2

Remarquable experimental results

7.3.2.1 Result 1: beams in reinforced concrete may be analyzed by means of an

analoguous truss
Experiments on beams with shear reinforcement (links for example) reveal that the
crack pattern is determined by the presence of the links: figure 7.3.2-1. The cracks in the
zone loaded by shear, show a regular pattern and are even somewhat parallel in long
beams. In between the cracks, compressive concrete struts are identified. The struts
guide the loads applied on the upper side of the beam towards the lower side of the
beam; from there on, the loads are back again transferred towards the upper side by
means of the links; this is a regular process all along the length of the beam.
These experimental observations are the basis of the papers written independently by
the Swiss engineer RITTER in 1899 and the German engineer MRSCH in 1902, in
which they both proposed to describe the shear load transfer in reinforced concrete
beams by means of an analogous truss (WIGHT, 2009).

Asw
s

z.cotg

Figure 7.3.2-1
Schematic representation of the regular crack pattern in a beam in reinforced concrete
with shear reinforcement: identification of an analogous truss. Asw represents the crosssection of 1 link with two legs

The truss system is composed of fours types of members:

the non-cracked arch with compressed concrete at the upper side of the
beam, acts as the top compression member of the truss;
the horizontal tension steel (main reinforcement) acts as bottom chord of the
truss; the distance between top and bottom member is the lever arm z;

7-11

the diagonal compression members inclined at an angle , represent the

concrete compression struts between the (parallel) shear cracks;
the transverse tension members in the truss, characterized by the distance
z.cotg between them, represent the shear reinforcement (in this example
composed of vertical links).

7.3.2.2 Result 2: the relationship between imposed shear load and the necessary shear
reinforcement
The area Asv of a vertical member in the truss in figure 7.3.2-1, is equal to:
As = Asw .

z. cotg
s

(7.3.2-1)

with
Asw
the cross-section of 1 link (2 vertical legs);
s
the distance between adjacent links.
The force that has to be resisted by the vertical member is indeed the shear load V.
Consequently, the tensile stress sv in the vertical member is:

V
V
s
=
.
As Asw z. cotg

s =

(7.3.2-2)

The steel stress (in the links) has to be limited to the design strength fywd. This
reasoning, fully based on the truss analogy, leads to the value of the maximum shear
load that can be supported:
Vu =

Asw
.z. cotg . f ywd
s

(7.3.2-3)

However, experimental results (WALRAVEN, 1995) show that the real behaviour does
not fully coincide with the one suggested by the truss analogy. Figure 7.3.2-2 shows, in
a schematic way, the experimentally measured relationship between the steel stress sv
in the shear reinforcement and the applied shear load V; the solid line shows the
experimental relationship while the dashed line shows the relationship according to the
truss analogy via expression (7.3.2-2).
sv
fywd
stress in links
according to truss
analogy model

Vc

Effective stress
measured in the
vertical legs of the
links
Applied shear
load V

Vc

7-12

Figure 7.3.2-2
The steel stress sv in the vertical links in function of the imposed shear load V: solid
line = experimental measurement; dashed line = theoretical relationship according to the
truss analogy
Figure 7.3.2-2 shows that the shear reinforcement is practically not working when
small values of shear loads are applied. Shear reinforcement is only activated (increase
of steel stress sv in the links) from the moment on that a shear crack appears. It is
learned from the experiments that in cracked situation, the imposed shear load V is
transferred by two mechanisms:
partly by the truss mechanism;
partly by an extra bearing mechanism, which can be explained by:
the fact that the hinges in the idealized truss system are not
hinges at all in reality; the nodes of the truss transfer also
moments;
crack surfaces are not smooth and straight, but are very irregular
in shape;
a part of the load is transferred by the uncracked compression
arch to the supports and by the dowel action of the main
reinforcement.
The sum of all non-truss mechanisms can be called Vc; it is as if this part of the load
transfer is taken care off by the concrete (c < concrete). It is observed that
Vc is practically constant during loading, on the condition that the
mechanisms which explain the concrete part Vc are not too much destroyed
by too large crack widths;
Vc is practically equal to the shear load that causes inclined cracks to appear.
This leads to the assumption that this shear load is nothing else than the
shear load bearing capacity VRd,c of the same beam but without shear
reinforcement.
The experimental result mentioned above, has been confirmed for cross-sections with
various shapes and reinforcement ratios. It is an important result which has lead to the
rule in earlier versions of EC2 (1995, 1998) that shear reinforcement in beams could be
calculated for the shear load (V-VRd,c) only. The actual version of EC2 (2004) adopts
another point of vue (see further).
7.3.3

Analogous truss models

7-13

Fcd

As

Ftd

Figure 7.3.3-1
Basic model of analogous truss system for the development of the formulas for shear
load verification of a beam in reinforced concrete

Figure 7.3.3-1 presents the general truss model that is used for the development of the
formulas for shear load verification of a beam in reinforced concrete. The inclination
angle of the shear reinforcement with the beams axis is called . For inclined bars:
< 90 (typical 45); for vertical links: = 90. The inclination angle of the cracks, and
thus also the inclination angle of the concrete compression strut, is called . The limit
values for the angle are fixed in the standard:
EN 1992-1-1:2004; 6.2.3(2):
1 cotg 2,5

(7.3.3-1)

45,0 21,8

(7.3.3-2)

which corresponds to:

The Belgian ANB takes account of the effect of an eventual axial force or prestressing
force which lead to less inclined cracks; this is illustrated by the principle reasoning by
means of MOHRs circle in figure 7.3.3-2. The ANB defines the limit values for as
follows:
1,0 cotg cotg max

(7.3.3-3)

with
cotg max = 2 +

k1. cp .bw .d .s
Asw .z. f ywd

where:
k1, cp, bw, d : defined in paragraph 7.2.5;
7-14

(7.3.3-4)

 Asw
s
z
fywd

= cross-sectional area of one shear reinforcement: one inclined bar or two

vertical legs of one link;
= spacing of the adjacent shear reinforcement;
= lever arm between the compression and tensile members of the truss;
ANB accepts z = 0,9.d if cp = 0;
= design yield strength of shear reinforcement.

With cp non 0, one may even adopt cotg = 3; this assumption corresponds to very
slightly inclined cracks, with an inclination angle of only 18,4.
If cp = 0, application of ANB leads to the following limit values:
1 cotg 2

(7.3.3-5)

45,0 26,6

(7.3.3-6)

which corresponds to:

7-15

YA

XA

V .S
b.I

V>0

(druk)
x

A'

Y'A
X'A

V .S
b.I

X'A

XA

1A'

1A

Y'A = YA = poolA

poolA'

Figure 7.3.3-2
Auxiliary reasoning by means of MOHRs circle to show that the presence of
axial compression stresses leads to a less inclined crack angle

7-16

Note:
It is thus observed that the actual standard accepts the choice of rather small
values of the crack inclination angle and thus of the concrete compression
struts in the truss model. The justification for this choice and the discussion of its
consequences is presented further in this chapter.

7.3.4

Design of the truss members

7.3.4.1 Introduction
The truss model in figure 7.3.3-1 contains four components:
the vertical or inclined tension reinforcement which represent the shear
reinforcement (links or stirrups or inclined bars);
the concrete compression struts, with an inclination angle ;
the compression member on top;
the tension member at the bottom (the bottom chord member).
Design for shear means that each of all four members of the truss is designed strong
enough in order to make the beam able resisting the imposed shear load.
7.3.4.2 The shear reinforcement
1. The force in the truss member
The assumed truss model is once again presented in figure 7.3.4-1. The method of
sections (method of RITTER) may be applied to determine the force T in the
inclined truss member; vertical translation equilibrium leads to:
V
T=
(7.3.4-1)
sin

Figure 7.3.4-1
Application of the method of sections (RITTER) to determine the force in the
inclined truss member

7-17

2. The maximum shear load VRd,s that can be resisted by the shear reinforcement
A schematic representation of the truss is shown in figure 7.3.4-2; the figure
shows clearly that each single stirrup or inclined bar that is represented as
inclined truss member (associated with each inclined strut), represents in fact a
series of stirrups or bars distributed along each crack with spacing s.

s
dstrut

z.cotg

z.cotg

z.cotg

Figure 7.3.4-2
Auxiliary figure for the determination of the shear reinforcement

The maximum value of the shear load VRd,s that may be resisted by the inclined
tensile truss member is:
V Rd , s = (n. Asw . f ywd ). sin

with:
Asw

(7.3.4-2)

the cross-sectional area of 1 stirrup (2 legs!) or of 1 inclined bar;

f ywd

the design yield strength of the shear reinforcement;

the number of links or bars that is distributed along the distance

z (cotg + cotg ). The number is equal to:

n=

sin

z.(cotg + cotg )
s

with s = the spacing of the stirrups or bars;

for the vertical projection.

The formula for the maximum shear load that may be resisted by the shear
reinforcement is thus:

7-18

VRd ,s =

Asw
.z. f ywd .(cotg + cotg ). sin
s

(7.3.4-3)

For vertical links with = 90 (cotg = 0 ; sin = 1) , the formula is:

VRd ,s =

Asw
.z. f ywd . cotg
s

(7.3.4-4)

Note 1:
With cp = 0, z = 0,9.d may be assumed.

Note 2:
The earlier versions of the standard (1995, 1998) proposed to apply the
so called standard method in which the inclination angle of all
compression strut was = 45. With this assumption, the formulas are:
- with inclined shear reinforcement:
A
VRd ,s = sw .z. f ywd .(1 + cotg ). sin
(7.3.4-5)
s
- with vertical stirrups:
A
V Rd , s = sw .z. f ywd
(7.3.4-6)
s

3. The necessary shear reinforcement to resist the imposed shear load VEd
The necessary shear reinforcement per unit length (along the beams axis) can be
deduced from expression (7.3.4-3):
- for inclined shear reinforcement:
Asw
VEd
=
s
z. f ywd .(cotg + cotg ). sin

(7.3.4-7)

- for vertical stirrups:

Asw
VEd
=
s
z. f ywd . cotg

(7.3.4-8)

Note:
The last formula allows to conclude that the choice of a smaller value of
the angle leads to a smaller cross-sectional area of shear reinforcement
( smaller cotg larger). The adoption of less inclined cracks in the
truss model thus leads to savings in shear reinforcement.
7-19

This conclusion can also be explained in another way: if cracks are less
inclined, the principal tensile stress (which is perpendicular to the crack)
is oriented more vertically; this means that vertical stirrups are used
more efficiently which leads to the reduction of the number of stirrups
needed.
7.3.4.3 The concrete compression struts

1. The force in the truss member

The truss model is shown in figure 7.3.4-3. Vertical translation equilibrium leads
to the identification of the force D in the inclined compression member:
D=

V
sin

(7.3.4-9)

Figure 7.3.4-3
Application of the method of sections (RITTER) to determine the force in the
inclined concrete compression member

2. The maximum shear load VRd,max that can be resisted by the concrete compression
member
The maximum value of the compression force D that may be resisted by the
inclined concrete strut is equal to the product of the maximum concrete
compression strength with the cross-sectional area of the strut; the last one is
deduced from figure 7.3.4-2: cross-sectional area of the strut = b.d strut .
The maximum concrete compression strength to be used for the strut calculation,
is defined in EN 1992-1-1:2004; 6.2.3(3) and is limited to v.fcd
with fcd = fck / 1,5 (and not fcd = 0,85 . fck / 1,5 !)

Note:
It should be remembered here that EN 1992-1-1:2004; 3.1.6 defines fcd as
7-20

fcd = cc . fck / c
with cc a factor for which the value 1 is recommended.
In Belgium, the National Annex (NBN EN 1992-1-1 ANB) recommends
the use of the value cc = 0,85 for verification in ULS for axial loads,
bending and combined axial force with bending; for other loading types
(shear and torsion), cc = 1 should be used. This means in practice that
for calculations in accordance with NBN EN 1992-1-1 ANB, the
following design values have to be used for the compressive strength of
concrete:
for ULS design of the main reinforcement (thus for normal
stresses due to axial loads and bending moments): fcd = 0,85 . fck /
1,5
for ULS design of shear reinforcement (necessary to take up shear
loads and torsion): fcd = fck / 1,5

and with = strength reduction factor, defined by:

v = 0,6.( 1

f ck
250

) 0,5

(7.3.4-10)

in which fck is expressed in N/mm2.

The additional strength reduction factor has to be applied to the concrete design
strength for the calculation of the struts in order to take account of the complex,
two-dimensional stress situation in the struts. Indeed, the struts are intersected by
links or by inclined bars which are loaded in tension; due to the adherence
between steel and concrete, the transverse tensile stresses cause the weakening of
the compressive struts. Formula (7.3.4-10) is the result of experimental tests.

Note:
EN 1992-1-1:2004; 6.2.3(3) stipulates that when the design stress in the
shear reinforcement is less than 80% of fywk, the following values may be
adopted for the reduction factor :
- = 0,6 for fck 60 MPa;
- = 0,9 fck/200 > 0,5 for fck > 60 MPa

The maximum compression force D that can be resisted by the strut is thus:
v.fcd.b.dstrut. The vertical projection of this force is designated in EN 1992-11:2004 with the symbol VRd,max; this force has to be compared to the imposed
shear load VEd.
VRd,max may be further expressed as (see figure 7.3.4-2):
V Rd ,max = v. f cd .b.d strut . sin
with

7-21

(7.3.4-11)

sin =

d strut
z.(cotg + cotg )

and thus, with inclined (angle ) shear reinforcement:

VRd ,max = v. f cd .b.z. sin 2 .(cot g + cot g ) = . f cd .b.z.

cotg + cotg
1 + cotg 2

(7.3.4-12)

With vertical links = 90 (cotg = 0 ; sin = 1) , the formula is:

V Rd ,max = v. f cd .b. sin 2 .z. cotg = . f cd .b.z.

or also:

1
cotg + tg

V Rd ,max = v. f cd .b.z. sin . co s

(7.3.4-13)

Note 1:
If cp = 0, z = 0,9.d may be assumed.

Note 2:
The earlier versions of the standard (1995, 1998) proposed to apply the
so called standard method in which the inclination angle of all
compression strut was = 45.

= 45 (cot g = 1; sin =

2
)
2

With this assumption, the formulas are:

- with inclined shear reinforcement:
1
VRd ,max = v. f cd .b.z. .(1 + cotg )
2
- with vertical stirrups:
1
VRd ,max = .v. f cd .b.z
2

(7.3.4-14)

(7.3.4-15)

3. Stress control in the concrete compression strut

The stress is deduced from expression (7.3.4-12):
VEd
VEd
1
1 + cotg 2
c =
.
.
=
b.z sin 2 .(cotg + cotg ) b.z cotg + cotg

(7.3.4-16)

This formula allows to observe that the choice of a smaller value of inclination
angle leads to larger compression stresses in the concrete strut. This result is
obvious when looking at figure 7.3.4-4: a less inclined strut has to transfer a
7-22

larger compression force D in order to generate the same resisting shear force. It
is observed that this does not cause problems in most practical normal cases,
because the stress is in general quite smaller than the acceptable stress v.fcd (see
applications). Yet, problems may arise when small inclination angles are chosen.
Expression (7.3.4-16) also shows that the stress in the concrete compression strut
gets smaller with the use of inclined bars ( < 90).

Ved

Ved

Vstrut

Vstrut

Figure 7.3.4-4
A less inclined strut has to transfer a larger compression force D in order to
generate the same resisting shear force

Note:
EN 1992-1-1:2004; 6.2.3(3) stipulates that when prestressing is applied,
the value of VRd,max may be increased, in order to take account of the
fact that cracks are closed.

7.3.4.4 The upper chord and the bottom chord of the truss model
1. The forces in the truss members
The forces in the upper and bottom chord can be determined by expressing the
equilibrium of the forces applied to the part of the beam shown in figure 7.3.4-5.
At the level of the considered cross-section, a whole series of compression struts
are cut; all these compression forces have a resultant force which is D and which
is applied at half depth; the magnitude of its vertical component Dy is equal to V.
In the same cross-section, a whole series of tensile reinforcement bars (stirrups or
bars) are cut; the resultant force of all these tensile forces is T, which is applied at
half depth; the magnitude of its vertical component Ty has to be equal to V.
Consequently, the horizontal components are:
Dx = V . cotg
Tx = V . cotg

7-23

q
C Nc
D

Dx
Tx

z
z/2

Ns

Dy

Dx

z/2

Tx

Ty
Dx=V.cotg
Dy=V

Tx=V.cotg
Ty=V

Figure 7.3.4-5
Auxiliary figure for the determination of the member forces in the upper truss
member and in the bottom chord
The other forces that are applied to the isolated left part of the beam, are:
the imposed uniformly distributed load q;
the support reaction force R;
the force Nc in the arch of compressed non-cracked concrete;
the tensile force Ns in the main reinforcement.
Rotation equilibrium written around point S (figure 7.3.4-5) leads to:
R.x s

q.x s2
z
z
= N c .z + D x . Tx .
2
2
2

The first member of this equation is nothing else than the bending moment Mz in
the considered cross-section, and thus:
z
M z = N c .z + V .(cotg cotg ).
2
The force in the upper truss member is thus:

7-24

Nc =

Mz V
.(cotg cotg )
z
2

(7.3.4-17)

In an analogous way, the rotation equilibrium around point C (figure 7.3.4-5)

leads to:
Ns =

Mz V
+ .(cotg cotg )
z
2

(7.3.4-18)

2. Discussion
The result of expression (7.3.4-18) is important because this shows that, in zones
with shear loads, the force to be transmitted by the main reinforcement does not
M
only depend on the bending moment Mz; indeed: N s z . The main
z
reinforcement is loaded by an additional tensile force which increases with
decreasing value of (a disadvantage of choosing less inclined cracks and thus
less inclined concrete struts). The consequences of this observation are illustrated
in a visual way for the particular case with the choices: = 45 (cot g = 1) and
vertical stirrups with = 90 (cotg = 0); expression (7.3.4-18) is than written:
Ns =

Mz V
+
z
2

(7.3.4-19)

Expression (7.3.4-19) is represented in a schematic way in figure 7.3.4-6, for a

uniformly distributed load and for a concentrated load.

(a)

(b)

qL/2
Q/2

V/2

|V/2|

|V/2|
Ns

V/2

V/2

V/2

Ns

M/z

M/z

Figure 7.3.4-6

7-25

Ns
Ns

Schematic representation of the increase in tensile force in the main

reinforcement due to the shear load, for two beams (a) and (b), and with the
assumptions = 45 and = 90 (vertical stirrups)

In the case of the concentrated load (figure 7.3.4-6(b)), one notes Mz=V.x.
Substitution in expression (7.3.4-19) leads to:
Ns =

V .( x + 12 .z )
z

(7.3.4-20)

Expression (7.3.4-20) shows that in order to calculate the force Ns in the section x,
one may not use the bending moment at the distance x from the support, but
z
instead of that, has to use the bending moment at the distance x + from the
2
support. The bending moment diagram has thus to be shifted over the
1
distance .z , in unfavourable direction. In the more general case with arbitrary
2
values of and , the distance over which the bending moment diagram has to be
1
shifted is .z.(cotg cotg ) .
2

Note:
The additional tensile force in the main reinforcement disappears when
inclined bars with = 45 (cotg = 1) are used in combination with the
assumption = 45.

3. Prescriptions EN 1992-1-1:2004
The main reinforcement has to be designed for a supplementary force which is
due to the shear load; the problem is due to the fact that the orientation of the
reinforcement does not coincide with the orientation of the principal tensile stress.
EN 1992-1-1:2004; 6.2.3(7) stipulates that the main reinforcement should be
calculated for an additional tensile force Ftd which is due to the imposed shear
force VEd and which may be determined by means of expression (7.3.4-18):
Ftd = 0,5.VEd.(cotg cotg )

(7.3.4-21)

and with the condition that (see figure 7.3.4-7):

(MEd/z) + Ftd MEd,max/z
in which MEd,max is the maximum moment along the beam.

7-26

(7.3.4-22)

In practice, the rule just mentioned above is translated into the much more
practical alternative which is called the shift rule: the bending moment diagram
is shifted over the distance al, which also means that the length of the main
reinforcement bars is increased with al at each end. The shifting has thus
essentially consequences for the curtailment of the longitudinal tension
reinforcement. The prescriptions in EN 1992-1-1:2004; 9.2.1.3 can be
summarized as follows:
for structural members without shear reinforcement, the moment curve may
be shifted over the distance al = d;
for structural members with shear reinforcement, the moment curve may be
shifted over the distance al = z/2.(cotg - cotg ). As already said before,
in the absence of axial compression loads, z may be assumed equal to 0,9.d;
the curtailed bars should be anchored with lbd from the point on where the
bars are not useful anymore. The diagram of the resisting tensile forces
should engulf the envelope diagram of the imposed tensile forces, after
application of the shift rule: see figure 7.3.4-7;
the anchorage length of a bent-up bar which contributes to the resistance to
shear, should not be less than 1,3.lbd in the tension zone and 0,7.lbd in the
compression zone.

Figure 7.3.4-7
Envelope diagram for the calculation of structural members subjected to bending,
with indication of the anchorage lengths to be applied
(figure 9.2 in EN 1992-1-1:2004)

7-27

7.3.4.5 Discussion of the variable strut inclination method

The previous versions of the standard (1995,1998) recommended the so called standard
method, which is characterized by the choice of the inclination angle of all concrete
compression struts equal to 45. This assumption leads to a relatively simple
verification of successive cracks with a constant inclination angle, starting with the first
crack for the largest imposed shear load: see figure 7.3.4-8. However, the variable strut
inclination method was also mentioned in parallel to the standard method. The
adoption of smaller strut inclination angles leads to the following effects:
- less shear reinforcement per unit length along the beam;
- more important loading of the concrete compression struts, with higher stress
levels;
- larger shift length al and thus longer main reinforcement bars.
The 2004 version of EC2 does not mention the standard method anymore. Moreover, it
is observed that the standard accepts rather small values for the inclination angle :
EN 1992-1-1: 2004: 45,0 21,8
ANB:
met cp = 0: 45,0 26,6
met cp 0: 45,0 18,4
With the new prescriptions, the standard wants to take account of the observation that in
the case of shear failure of the structural member, the most important crack close to the
support and with the highest imposed shear load, is indeed characterized by a smaller
inclination angle; see figure 7.3.4-9. Adopting = 45 in this zone of the structural
member, leads to over-estimation of the necessary shear reinforcement.

Note:
Low inclination cracks only appear with lack of shear resistance. Figure
7.3.4-10 shows the crack pattern in a beam with sufficient shear
reinforcement; the beam has failed in bending and only nearly vertical
cracks are observed in the zone with shear loads.

Finally, it should be stressed that the design of shear reinforcement by means of a truss
model with variable strut inclination, is in full accordance with the principles of plastic
design which occupies a prominent place in the present version of the standard. Design
on the basis of the assumption of a strut inclination which does not fully coincide with
the real inclination, leads to a slightly different failure mechanism: the beam fails in a
way that is determined by the designer. With the adoption of a too small inclination
angle and thus the provision of less shear reinforcement, the designer asks to the
beam for a rearrangement of tasks with a heavier loading of the struts and of the main
reinforcement. Practically speaking, this rearrangement will be visible by a more
expressive development of cracks, because the design now asks for heavier loading of
the struts. The principles of plastic design are discussed in chapter 11 in these course
notes.

7-28

zone bending + shear

zone pure bending

d
= 45

zone with calculated shear

reinforcement

zone bending + shear

zone with technological

shear reinforcement

zone pure bending

zone with calculated shear

reinforcement

zone with technological

shear reinforcement

Figure 7.3.4-8
Comparison of truss models with the standard model ( = 45) on one hand and
the variable strut inclination method on the other hand

Figure 7.3.4-9
Shear failure of a beam without sufficient shear reinforcement: the most
important crack, associated with the maximum shear load, is characterized by an
inclination angle smaller than 45

7-29

Figure 7.3.4-10
Crack pattern in a beam subjected to a four point bending test; failure is in
bending and not in shear; shear reinforcement has been well designed

7.4 Design for shear

7.4.1

Introduction

This paragraph focuses on the ULS design calculation of shear reinforcement according
to EN 1992-1-1:2004; 6.2 and the complementary Belgian ANB prescriptions.
7.4.2

Definitions

The verification of shear resistance is based on three design values of resisting shear
forces:
design shear resistance of the member in a section without shear
VRd,c
reinforcement. VRd,c is calculated by means of expressions (7.2.5-3) and (7.2.5-4)
in these course notes;
design value of the shear force which can be sustained by the yielding
VRd,s
shear reinforcement. VRd,s is calculated by means of expressions (7.3.4-3) and
(7.3.4-4) in these course notes;
VRd,max design value of the maximum shear force which can be sustained by the
member, limited by crushing of the compression struts. VRd,max is calculated by
means of expressions (7.3.4-12) and (7.3.4-13) in these course notes.
VEd is the imposed design shear force in the section to be verified, resulting from
external loading on the structural member.

7-30

7.4.3 The principles of the shear verification procedure

In the regions of the member where VEd VRd,c no calculated shear reinforcement
is necessary. Yet, when on the basis of the design calculation, no shear
reinforcement is required, minimum (technological) shear reinforcement should
be provided. The minimum shear reinforcement may be omitted in certain special
cases such as:
- slabs where transverse redistribution of loads is possible;
- members of minor importance which do not contribute significantly to the
overall resistance and stability of the structure; example: lintels with span
2 m.

In regions where VEd > VRd,c sufficient shear reinforcement should be provided in
order that VEd VRd.. VRd is the resisting shear force and is equal to the smallest
of the two values VRd,s and VRd,max.

Important note:
It was already mentioned in figure 7.3.2-2, that there is experimental
evidence for the fact that the shear reinforcement only starts to work
as a member in a truss sytem, for a reduced value of the imposed shear
load. In the previous versions (1995, 1998) of EC2, it was accepted that
the reduction could be taken equal to the shear force VRd,c which is in
fact the shear load resisted without shear reinforcement, by the
following mechanisms:
- the shear resistance of the non-cracked compression concrete arch;
- the granulate interlocking effect along the shear crack;
- the dowel action of the main reinforcement.
It was thus accepted in the previous versions of EC2, in which the
standard method was used for shear verification (struts with constant
inclination angle of 45), to design shear reinforcement for the force
VEd - VRd,c.
This is not the case anymore in the present version (2004) of EC2, in
spite of the experimental evidence shown in figure 7.3.2-2; shear
reinforcement has now to be calculated for the full imposed shear force
VEd. The reason for this is that the present standard does not want to
accumulate too much favourable effects. Indeed, the present EC2
allows adopting small inclination angles in the regions where high shear
loads are applied; this leads already to smaller shear reinforcement
(while causing more severe loading of the concrete struts). EC2 does
not want to accumulate this positive effect on the shear reinforcement
with a second one generated by the reduction with VRd,c of the imposed
shear force VEd.

The longitudinal tension reinforcement should be able to resist the additional

tensile force caused by shear; in practice, the shift rule is used.

For members subjected to predominantly uniform distributed loading, the design

shear force need not to be checked at a distance less than d from the face of the
support; see figure 7.4.3-1. This rule takes into account that the loads applied
7-31

close to the support, are directly transmitted to the support without causing
bending and shear of the beam itself. The shear verification thus starts with the
first crack which is initiated at the tensile side of the beam and which develops
upwards with a certain inclination angle. Any shear reinforcement required in the
first verified section, should continue to the support. On top of that, it should
always be verified that the imposed shear force at the support is not larger than
VRd,max.

Figure 7.4.3-1
In the case of uniformly distributed loads, direct transmission to the support of the loads
applied close to the support may be assumed

Note 1:
For members with inclined chords (upper side and lower side), the value
of VRd should be increased with two additional resisting shear load
components (see figure 7.4.3-2):
VRd = minimum(VRd,s; VRd,max) + Vccd + Vtd
with:
- Vccd = the design value of the shear component of the force in the
compression area, in the case of an inclined compression chord (upper
side);
- Vtd = the design value of the shear component of the force in the
tensile reinforcement, in the case of an inclined tensile chord (lower
side).

7-32

D
Vccd
Vtd
T
Resisting
reaction

action

Figure 7.4.3-2
Additional shear resistance due to the presence of inclined chords in structural
members
(adaption of figure 6.2 in EN 1992-1-1:2004)

Note 2:
When loads are applied to the lower side of a structural member,
sufficient vertical reinforcement, in addition to the shear reinforcement,
is needed in order to transfer the loads to the upper side: see figure 7.4.33.

Figure 7.4.3-3
Additional reinforcement is needed to transfer loads applied to the lower side of
the beam to the upper side
7.4.4

Members not requiring design shear reinforcement: VEd VRd,c

Some prescriptions:
in the regions where VEd VRd,c no calculated shear reinforcement is necessary,
but minimum (technological) shear reinforcement should be provided: see further
in the paragraph on technological prescriptions;
for the design of the longitudinal reinforcement, the MEd-diagram should be
shifted over a distance al = d in the unfavourable direction;
for members with concentrated loads applied on the upper side and rather close to
the support, it may be assumed that a part of the load is transferred directly to the
support (without interaction of the beam itself), which gives a reduction of the
imposed shear force VEd. The prescriptions stipulate that when the concentrated
load is applied on the upper side within a distance 0,5d av 2d (see figure 7.4.41) from the edge of the support, the contribution of this load to the shear force VEd
7-33

may be reduced by multiplying the load by the factor = av/2d. This reduction is
only valid provided that the longitudinal reinforcement is fully anchored at the
support. For av 0,5d the value av = 0,5d should be used.
Important remark: the imposed shear force VEd, calculated without reduction by
the factor , should always satisfy the condition:
VEd 0,5.bw.d..fcd

(7.4.4-1)

where = the strength reduction factor for concrete cracked in shear:

f
(7.4.4-2)
v = 0,6.( 1 ck ) 0,5
250
in which fck is expressed in N/mm2.
This condition corresponds in fact to the verification of a fictive concrete
compression strut right above the support. Expression (7.4.4-1) is deduced from
expression (7.3.4-13) which defines the maximum shear force that can be resisted
from the point of view of strut failure:
V Rd ,max = v. f cd .b.z. sin . co s
By introduction of the lever arm z taken equal to d and = 45, expression (7.4.41) is obtained, which corresponds to a sort of upper limit for VEd. Indeed, taking
into account the following evolution of the term sin.cos:
sin.cos
0,5
0,49
0,47
0,43

45
40
35
30

one finds the condition VEd ...sin.cos more severe for VEd in order to take into
account that the strut is more heavily loaded in compression when it is less
inclined.

7-34

Figure 7.4.4-1
It may be assumed that a fraction of the loads applied near supports, is
transmitted directly to the support and does not give a contribution to the
imposed shear force VEd, on the condition that the main reinforcement is
sufficiently anchored (figure 6.4 in EN 1992-1-1:2004)

7.4.5

Structural members requiring design shear reinforcement: VEd > VRd,c

In regions where VEd > VRd,c, sufficient shear reinforcement should be provided in order
that VEd VRd. VRd is the resisting shear force and is equal to the minimum of the two
values VRd,s and VRd,max. The formulas for VRd,s and VRd,max are developed on the basis of
an analogous truss model which is once again represented in figure 7.4.5-1.

7-35

Figure 7.4.5-1
Truss model used for the calculation of the shear reinforcement in structural members
(figure 6.5 in EN 1992-1-1:2004)
7.4.5.1 When vertical shear reinforcement is used
VRd ,s =

Asw
.z. f ywd . cotg
s

VRd , max = bw .z. . f cd .

1
cotg + tg

(7.4.5-1)

(7.4.5-2)

with the additional condition that defines the maximum effective cross-sectional area of
shear reinforcement for = 45 ( cotg = 1 ):
1
Asw,max . f ywd . . f cd .bw .s
2

(7.4.5-3)

7.4.5.2 When inclined shear reinforcement is used

VRd ,s =

Asw
.z. f ywd .(cotg + cotg ). sin
s

VRd ,max = bw .z. . f cd .

cotg + cotg
1 + cotg 2

(7.4.5-4)

(7.4.5-5)

with the additional condition that defines the maximum effective cross-sectional area of
shear reinforcement for = 45 ( cotg = 1 ):
1
1
Asw,max . f ywd . . f cd .bw .s.
2
sin

(7.4.5-6)

7.4.5.3 Additional prescriptions

Just as in the case where shear reinforcement is not necessary, the rule for direct
transmission to the support of a fraction of the loads that are applied near supports, can
be applied here too. The prescriptions stipulate that when the load is applied on the
upper side within a distance 0,5d av 2d (see figure 7.4.4-1) from the edge of the
support, the contribution of the load to the imposed shear force VEd may be reduced by
multiplying the load by the factor = av/2d. This reduction may only be applied if the
main reinforcement is fully anchored above the support. For av 0,5d, the value
av = 0,5d may be adopted.

7-36

Important: the imposed shear force VEd, calculated in this way (thus with application of
the reduction factor ), should satisfy the condition:
VEd Asw . fywd . sin
where Asw . fywd = the resistance of the shear reinforcement crossing the inclined shear
crack between the loaded areas (see figure 7.4.5-2); only the shear reinforcement within
the central 0,75.av should be taken into account.

Note:
The reduction with the factor may only be applied
- for the calculation of shear reinforcement and not for the strut verification;
- provided that the longitudinal reinforcement is fully anchored at the
support.

Figure 7.4.5-2
Auxiliary figure for the calculation of shear reinforcement with direct strut action close
to the support (figure 6.6 in EN 1992-1-1:2004)

7.5 Overview of the shear reinforcement calculation for a beam with

uniformly distributed load and concentrated load
7.5.1

The first value of VEd to be considered: in principle

In principle, the largest value of VEd should be considered for the first shear design
calculation. To remember, design of shear reinforcement is performed in ULS; the
design values of the loads have thus to be considered (use of partial safety factors). See
example in figure 7.5.1-1.

7-37

Q.Q
A

g.g
B

Vgd

VQd

(Vg+Q)dA
(Vg+VQ)d

Figure 7.5.1-1
In principle, the largest value of VEd should be considered for the first shear design
calculation; that is the value at the support
7.5.2

Reduction of VEd to take account of direct load transfer to the supports

7.5.2.1 Uniformly distributed load

The shear calculation is performed by considering successive shear cracks, starting with
the first crack in the region with the largest value of the imposed shear force. The crack
is initiated on the tensile side (at the bottom of the beam in figure 7.5.1-1) at the edge of
the support; the crack is assumed to be vertical in the concrete cover, but starting from
the main reinforcement, develops upwards with an inclination angle. In reality, the
crack gets stuck in the concrete compression arch, but in the simplified truss model, the
crack is extended up to the upper side of the beam with a constant inclination angle.
When the first crack is considered with the inclination angle , the assumption is
adopted that the uniformly distributed load to the left of the upper end of the crack is
transferred directly to the support: see figure 7.5.2-1. The first value of VEd to be
calculated is thus the imposed shear force in the cross-section at the distance d.cotg (d
when = 45) from the edge of the support (cross-section C in figure 7.5.2-1). The
value of (VEd)C is equal to the maximum shear force in support A (due to g and Q)
reduced with the portion of shear load due to g over the distance a + d.cotg:

(VEd ) C = (V g +Q ) dA g .g.(a + d . cot g )

with
a: see determination of the design span for different support conditions in paragraph
2.5.2.2.

7-38

(VEd)C = (Vg+Q)dA - g.g.(a+d)

Q
C

45

Q.Q
g.g
A

B
(VEd)C' = (Vg+Q)dA - g.g.(a+d.cotg)
Q
C'

d.cotg

Figure 7.5.2-1
First shear verification in the cross-section with the largest imposed shear force VEd,
taking account of the reduction due to the direct transfer of g to the support

Note:
The adoption of smaller values of (than 45) leads to smaller values of VEd
(see figure 7.5.2-1). But with this assumption, it is very important to verify the
concrete compression strut above the support with the formula:
VEd (with reduction of g only) bw . z . . fcd . sin . cos
If this condition is OK, then further comparison of VEd with VRd,c can be
performed.

7.5.2.2 Concentrated loads close to supports

For concentrated loads which are applied at the distance av from the edge of the support,
with 0,5.d av 2.d, a reduction may be adopted of the imposed shear force. The
contribution of the concentrated force to the shear force VEd may be reduced by
multiplying the concentrated force with the reduction factor = av/2d ( 1). This rule is
independent from the choice of the inclination angle .
Figure 7.5.2-2 shows the example of the concentrated force Q applied within the
distance av = 2.d from the edge of the support. The imposed shear load (VQ)dA in cross7-39

section A may be reduced to .(VQ)dA; this means that the difference (VQ)dA - .(VQ)dA is
directly transferred to the support. The shear force (VEd)C to be considered in crosssection C, at the distance d.cotg from the edge of the support is thus:

(VSd ) C = (V g +Q ) dA g .g.(a + d . cot g ) (VQ ) dA (VQ ) dA

VSd

Q.Q
C

45

2d
av
Q
A

(VQ)dA
VSd

Q.Q
C'

Figure 7.5.2-2
Reduction of the imposed shear force due to direct transfer of a portion of the
concentrated force to the support

Important note:
The reduction by means of the factor is not allowed for the verification of the
concrete compression struts. The reduction is only considered for:
- the comparison between VEd and VRd,c;
- the calculation of the shear reinforcement.
It is thus recommended to calculate two distinct values of VEd for further use:
- VEdg in cross-section C (with only the reduction of the contribution of g);
- VEdg+Q in cross-section C (inclusive the reduction of g and Q).

7.5.3

The verifications to be performed for the first crack

The following verifications have to be performed.

7-40

7.5.3.1 VRd,c
Calculation of the resisting shear force VRd,c. The crack that has to be considered with
(VEd)C in cross-section C, starts from the tensile side at the edge of the support and
develops with an inclination angle up to the cross-section C. This crack is determining
for the area of reinforcement Asl that has to be considered for the calculation of VRd,c.
This value of VRd,c which is calculated in order to be compared with (VEd)C in crosssection C, has to take account of the main reinforcement that contributes to the shear
resistance by means of the dowel action in the crack. Again, only that reinforcement can
be considered that is sufficiently anchored with lbd beyond the section where the dowel
action takes place; this is thus Asl which continues over the length d+lbd such as pointed
out in figure 7.2.5-1(a) and in the definition of Asl in expression (7.2.5-1).
7.5.3.2 VEd VRd,c
Is VEd (with reduction of g and Q) VRd,c , then:
-

the technological shear reinforcement is sufficient (see further);

the strut above the support should be verified by means of expression (7.4.4-1):
VEd (with only reduction of g) 0,5.bw.d..fcd
This expression takes account of = 90, as the technological reinforcement is
composed of stirrups.

7.5.3.3 VEd > VRd,c

First, the strut-condition has to be verified. At this stage, if necessary, one can still make
a new choice of the inclination angle :
with stirrups
1
V Ed ( only reduction of g ) bw .z. . f cd .
(7.5.3-1)
cotg + tg
-

with bent-up bars

VEd ( only reduction of g ) bw .z. . f cd .

cotg + cotg
1 + cotg 2

(7.5.3-2)

Next, the necessary shear reinforcement per unit length can be determined by means of
the following formulas:
-

with stirrups
Asw VEd ( with reduction of g and Q )

s
z. f ywd . cotg

with bent-up bars

7-41

(7.5.3-3)

VEd ( with reduction of g and Q )

Asw

s
z. f ywd .(cotg + cotg ).sin

(7.5.3-3)

Once Asw/s calculated, one has to verify if this reinforcement is at least equal to the
minimum reinforcement ration (see further in detailing of reinforcement). The
practical translation of Asw/s into a suitable diameter and spacing, can be realized by
means of table 7.5.3-1.
Table 7.5.3-1
Values of Asw and

(mm)
2

Asw (mm )

6
56,5

8
100

1,131
0,942
0,808
0,707
0,628
0,565
0,471
0,404
0,377
0,353
0,314
0,283
0,226
0,188

2,011
1,676
1,436
1,257
1,117
1,005
0,838
0,718
0,670
0,628
0,559
0,503
0,402
0,335

s (mm)
50
60
70
80
90
100
120
140
150
160
180
200
250
300

Asw
for stirrups with two vertical legs
s
10
12
14
157
226
308
Asw
(mm2/mm)
s
3,141
4,524
6,158
2,618
3,770
5,131
2,244
3,231
4,398
1,963
2,827
3,848
1,745
2,513
3,421
1,571
2,262
3,079
1,309
1,885
2,566
1,122
1,616
2,199
1,047
1,508
2,053
0,982
1,414
1,924
0,873
1,257
1,710
0,785
1,131
1,539
0,628
0,905
1,232
0,524
0,754
1,026

16
402

8,042
6,702
5,745
5,027
4,468
4,021
3,351
2,872
2,681
2,513
2,234
2,011
1,608
1,340

Note:
The additional condition related to the use of the -factor for concentrated loads
should not be forgotten: the imposed shear force VEd, calculated with application
of the -factor, should always respect the following condition:
VEd Asw . fywd . sin

(7.5.3-5)

where Asw . fywd = the force that is generated by the shear reinforcement in this
area (at the support); only the reinforcement in the central part 0,75.av should be
taken into account (see figure 7.4.5-2).

7-42

7.5.4

Verification of further cracks

After the control of the first crack, the second crack has to be verified: as long as
VEd > VRd,c (and a calculated shear reinforcement is necessary), a next crack has to be
verified. In other words: if it is found in the considered cross-section that the calculated
shear reinforcement is larger than the minimum reinforcement, a next crack has to be
verified.
Figure 7.5.4-1 presents the example of an end support, with indication of the crosssections C1, C2, C3... that have to be verified successively. The cracks start in the tensile
region at the bottom side of the beam; this determines the areas of the main
reinforcement Asl to be considered in the calculation of VRd,c; see figure 7.2.5-1(a) and
the definition of Asl in expression (7.2.5-1): Asl1, Asl2, Asl3 are the areas of main
reinforcement that continue over the distance d+lbd to the left beyond the cross-sections
that are considered.
VEd1

VEd2

C1

VEd3

C2

C3

Asl1

Asl2

Asl3

Figure 7.5.4-1
The cross-sections to be considered successively in shear verification, in a beam close to
an end support

Figure 7.5.4-2 presents the example of an intermediate support in a continuous beam.

The first crack is easy to identify on the basis of the cone with direct load transfer to the
support. The first cross-section to be calculated is C1 with (VEd)C1. The crack that
corresponds to this (VEd)C1 starts in the tensile region, at the upper side in C2; this is the
place where the dowel action of the main reinforcement is activated. In accordance with
figure 7.2.5-1(b) and the definition of Asl in expression (7.2.5-1), the main
reinforcement Asl1 should continue over the distance d+lbd beyond C2, in order to assure
for 100% the dowel resistance in C2.

7-43

Asl3

Asl2

C3

Asl1

C2

VEd3

C1

VEd2

VEd1

Figure 7.5.4-2
The cross-sections to be considered successively in shear verification, in a beam close to
an intermediate support

7.6 Detailing of shear reinforcement

Reference: EN 1992-1-1:2004; 9.2.2
7.6.1 Shape and nature of shear reinforcement
The shear reinforcement should form an angle between 45 and 90 to the
longitudinal axis of the structural element.
The shear reinforcement may consist of a combination of:
links enclosing the longitudinal tension reinforcement and the concrete
compression zone: see figure 7.6.1-1;
bent-up bars;
cages, ladders, etc. which are cast in without enclosing the longitudinal
reinforcement but are properly anchored in the compression and tension
zones.

Figure 7.6.1-1
Examples of shear reinforcement (figure 9.5 in EN 1992-1-1:2004)
-

Links should be effectively anchored. Stirrups should form a closed rectangle; a

lap joint on the vertical leg is permitted provided that the link is not required to
resist torsion (see chapter on torsion).
At least 50 % of the necessary shear reinforcement should be in the form of links.

7-44

7.6.2

Minimum shear reinforcement

The geometric ratio of shear reinforcement is defined as:

w =

Asw
s.bw . sin

(7.6.2-1)

where:

: geometric ratio of shear reinforcement, with w w,min;

Asw : area of shear reinforcement within length s;

bw

: spacing of the shear reinforcement measured along the longitudinal axis of

the structural member;
: width (or breadth) of the web of the structural member;

: angle between the shear reinforcement and the longitudinal axis.

The value of w,min is identified by the following expression:

w,min =

0,08. f ck
f ywk

(7.6.2-2)

with fck and fywk in MPa.

w,min depends of the steel grade of the shear reinforcement and of the concrete class.
Table 7.6.2-1 presents the values for w,min for different combinations.

7-45

Table 7.6.2-1
Values of w,min for different concrete-steel combinations
Steel grade of shear reinforcement
S220
S400
S500
S600
Concrete class
(fywk = 220 MPa) (fywk = 400 MPa) (fywk = 500 MPa) (fywk = 600 MPa)
1,26E-03
6,93E-04
5,54E-04
4,62E-04
C12/15 (fck = 12 MPa)
1,45E-03
8,00E-04
6,40E-04
5,33E-04
C16/20 (fck = 16 MPa)
1,63E-03
8,94E-04
7,16E-04
5,96E-04
C20/25 (fck = 20 MPa)
1,82E-03
1,00E-03
8,00E-04
6,67E-04
C25/30 (fck = 25 MPa)
1,99E-03
1,10E-03
8,76E-04
7,30E-04
C30/37 (fck = 30 MPa)
2,15E-03
1,18E-03
9,47E-04
7,89E-04
C35/45 (fck = 35 MPa)
2,30E-03
1,26E-03
1,01E-03
8,43E-04
C40/50 (fck = 40 MPa)
2,44E-03
1,34E-03
1,07E-03
8,94E-04
C45/55 (fck = 45 MPa)
2,57E-03
1,41E-03
1,13E-03
9,43E-04
C50/60 (fck = 50 MPa)
2,70E-03
1,48E-03
1,19E-03
9,89E-04
C55/67 (fck = 55 MPa)
2,82E-03
1,55E-03
1,24E-03
1,03E-03
C60/75 (fck = 60 MPa)
3,04E-03
1,67E-03
1,34E-03
1,12E-03
C70/85 (fck = 70 MPa)
3,25E-03
1,79E-03
1,43E-03
1,19E-03
C80/95 (fck = 80 MPa)
3,45E-03
1,90E-03
1,52E-03
1,26E-03
C90/105 (fck = 90 MPa)

7.6.3

Spacing

7.6.3.1 In the longitudinal direction

The recommended value for the maximum longitudinal spacing sl,max between adjacent
stirrups, is determined by the formula:
sl,max = 0,75.d.(1 + cotg )

(7.6.3-1)

The recommended value for the maximum longitudinal spacing sb,max between adjacent
bent-up bars, is determined by the formula:
sb,max = 0,60.d.(1 + cotg )

(7.6.3-2)

Table 7.6.3-1 presents some values for bent-up bars with = 45 and vertical stirrups (
= 90).

Note:
The longitudinal spacing between stirrups may be quite large. The earlier
editions of EC2 were more severe at this point, with a maximum value of 300
mm. In practice, it is observed that the spacing criterion of 30 mm is still
adopted in workshops, because of practical considerations in the realization of
the reinforcement cages.

7-46

Table 7.6.3-1
Values of recommended maximum longitudinal spacing sl,max between adjacent bent-up
bars (with = 45) and vertical stirrups ( = 90)

d
(mm)
400
600
800
1000

Bent-up bars with = 45

( cotg = 1 )

Stirrups with
= 90
( cotg = 1 )

sb,max = 0,6.d.(1+cotg )
sb,max = 1,2.d
(mm)
480
720
960
1200

sl,max = 0,75.d.(1+cotg )
sl,max = 0,75.d
(mm)
300
450
600
750

7.6.3.2 In the transverse direction

The transverse spacing of the legs in a series of shear links should not exceed st,max,
determined by the formula:
st,max = 0,75.d 600 mm

(7.6.3-3)

7.7 Shear between web and flanges of T-sections

7.7.1

Analysis of the compression flange

7.7.1.1 Identification of the problem and the truss analogy

Figure 7.7.1-1 shows a T-beam subjected to simple bending. The NA is situated in
principle in the upper part of the cross-section; the assumption may be adopted that the
flange corresponds to the compression part of the cross-section. The force N' in the
compression zone (distributed over the whole width of the flange) changes over the
distance x with the value:
N ' =

M
z

The indication N'f ( f < flange) corresponds to the compression force in 1 part of the
flange left or right of the web (abstraction is made of an eventually compressed small
part of the web); N'f is thus a fraction of N' as expressed by the formula:

7-47

b f bw
N ' f = N '.

2
bf

(7.7.1-1)

The force N'f in 1 part of the flange left or right from the web, changes over the distance
x with the value:
b f bw
(N ' f ) x =

M
.
z

2
bf

N'

(7.7.1-2)

N'+N'
b w bf

N'

N'+N'

N'

N'+N'

bf
bf -bw
bf -bw
bw 2
2
I

hf
I

hf

RA

N'
N'
M/z

Figure 7.7.1-1
T-beam loaded in simple bending
One of the two hatched pieces of flange in figure 7.7.1-1 is now isolated: see detail in
figure 7.7.1-2. The difference in compression force (N'f)x has to be equilibrated by the
shear force distributed over the contact surface I-I with area: hf.x.

7-48

(N'f)x

(N'f)x

.hf.x =(N'f)x

Figure 7.7.1-2
Detail: equilibrium of a part of the flange, left or right of the web
The equilibrium equation leads to the expression for the mean shear stress in the
contact surface I-I, in the uncracked elastic phase:

And as V =

(N 'f ) x
h f .x

b f bw
=

1 M
.
.
z.h f x

2
bf

(7.7.1-3)

M
, one can thus also write:
x

V
.
z.h f

b f bw
2
bf

(7.7.1-4)

If this shear stress is too large, causing the principal tensile stress to reach the tensile
strength of the concrete, cracks will appear with an inclination angle with respect to the
longitudinal axis. Shear reinforcement is necessary in that case, in order to allow the
build-up of the longitudinal forces in the flange. The design of this shear reinforcement
can again be done by means of an analogous truss system in which shear is resisted by
the combined action of struts and tensile rods. Figure 7.7.1-3 shows the decomposition
of the shear force into two forces: a compression force which has to be resisted by the
concrete struts and a tensile force perpendicular to the longitudinal axis of the beam.

7-49

(N'f)x

(N'f)x

Fwapening
x

Fschoor

f
f

Figure 7.7.1-3
Decomposition of the shear force at the level of the contact surface flange-web, into two
forces: a compression force to be resisted by concrete struts and a tensile force
perpendicular to the longitudinal axis of the beam, which has to be resisted by steel
reinforcement
7.7.1.2 Transverse reinforcement
It is assumed that the orientation of the compression diagonals in the flange is
characterized by the inclination angle f. As:
tg f =

tensile force in transverse reinforcement

(N ' f ) x

(7.7.1-5)

one finds that the transverse reinforcement should be able to transfer a force (indicated
by Freinforcement in figure 7.7.1-3) equal to:

Freinforcement =

(N 'f ) x
cotg

It is assumed that the transverse reinforcement is composed of rods with sectional area
Asf and spacing sf; this means that the ratio Asf/sf represents the transverse reinforcement
per unit length. The necessary reinforcement is thus:
Asf
sf

.x. f yd =

(N ' f ) x
cotg

This leads to the transverse reinforcement in the flange per unit length of the beam
Asf
sf

(N ' f ) x
x. f yd . cotg

7-50

Asf
sf

(7.7.1-7)

or, when expression (7.7.1-2) is used:

Asf
sf

(M ) x
1
=
.
.
z.x f yd . cotg

b f bw
2
bf

(7.7.1-8)

or

Asf
sf

b f bw
V
1
.
.
z f yd . cotg

2
bf

(7.7.1-9)

7.7.1.3 The concrete struts

It is also necessary to verify if the concrete struts are able to withstand the imposed
compression force (indicated by Fstrut in figure 7.7.1-3). This means that the
compression stress in the strut should be limited to .fcd.
From figure 7.7.1-3, it appears that:
Fstrut =

(N ' f ) x

(7.7.1-10)

cos f

zodat
(N ' f ) x

strut =

cos f
Fstrut
=
strut section
h f .bstrut

with bstrut = x.sin f (see figure 7.7.1-4).

The condition thus becomes:

strut =

(N ' f ) x
cos f .h f .x.sin f

. f cd

(7.7.1-11)

As (N'f)x = .hf.x , expression (7.7.1-11) also allows to develop a condition for :

.h f .x
. f cd
h f .x. cos f . sin f
or

. f cd . cos f . sin f

7-51

(7.7.1-12)

bschoor = x.sinf

Figure 7.7.1-4
Auxiliary figure for the determination of the width of the compression strut
7.7.1.4 Alternative formulation of the shear problem between web and flanges by means
of the strut and tie method
The following reasoning is based on the philosophy of the strut and tie method, which is
discussed in a separate chapter in these course notes (see further); this method presents a
further generalization of the truss system analogy that was already introduced for the
shear verification of beams loaded in bending.
The analysis of the analogous truss system model, presented in figure 7.7.1.5, reveals
that the upwards inclined compression diagonal in the web makes equilibrium with the
tensile force in the vertical member and the compression force in the flange. In order to
install equilibrium in the flange, the force has to be spread out over the whole width of
the flange (effective width!). A new analogous truss model appears in the flange: the
force spreads out via two compression diagonals ab and ab', which on their turn have to
make equilibrium with the compression forces N'f in the flange. This equilibrium needs
the tensile member bb'. The further development of this model leads to the same
formulas as developed before.

(a)

(b)

N'f
b
a

b'

N'f

b'
a

7-52

Figure 7.7.1-5
Transfer of forces in the flange of a T-beam; (a) truss model for beam with
compression flange; (b) top vue of the beam (WALRAVEN, 1995)
7.7.2

The tension flange

The developments presented for the compression flange are also applicable to the
tension flange; see figure 7.7.2-1.

II
hf2
II
(Asl)II
Asl
Figure 7.7.2-1
The shear problem in the tension flange

In analogy with expression (7.7.1-4), the shear stress in the contact section II-II in figure
7.7.2-1, is defined by:

V ( Asl ) II
.
z.h f 2 Asl

(7.7.2-1)

with

flange;
( Asl ) II = the area of the longitudinal reinforcement in the isolated part of the
tension flange.

sl

= the total area of longitudinal reinforcement in the whole tension

In analogy with expression (7.7.1-9), the necessary transverse reinforcement per unit
length of the beam, is determined by:

7-53

Asf
sf

(A )
1
V
.
. sl II
z f s . cotg Asl

(7.7.2-2)

Figure 7.7.2-2 presents a model of force transfer in the tension flange and web
(philosophy of the strut and tie method).

side view

Projection of
bottom flange

Asl

Ast

Figure 7.7.2-2
Model of force transfer from web to tension flange, based on the philosophy of the strut
and tie method (WALRAVEN, 1995)
7.7.3

Prescriptions concerning the passage from the web to the flanges in T-beams

7.7.3.1 Principles
Reference: EN 1992-1-1:2004; 6.2.4
The shear strength of the flange may be calculated by considering the flange as a system
of compressive struts combined with ties in the form of tensile reinforcement. The
notations in figure 7.7.3-1 have to be applied:
Fd = the variation of the normal force in one part of the flange, over the length x
(notation (Nf)x was used in the text before);
hf = the heigth (or thickness) of the flange at the contact surface between web and
flange;
vEd = the longitudinal shear stress (notation was used in the text before) at the contact
surface between web and flange. Consequently: vEd = Fd / (hf. x);
Asf = the cross-sectional area of one transverse reinforcement bar;
the spacing between the transverse reinforcement bars.
sf =

Note:

7-54

The bar which is indicated with the letter B in figure (7.7.3-1) corresponds to
longitudinal reinforcement bars which are eventually present (for example to
take up occasional fixing moments 25% of the moment in the span).

Figure 7.7.3-1
Notations used in the standard for the analysis of the force transfer between web
and flange via compression struts and transverse reinforcement (figure 6.7 in EN
1992-1-1:2004)

For the length x to be used, the standard presents the following recommendations:
x the half of the distance between the cross-sections with bending
moments M = 0 and Mmax. Example: see figure 7.7.3-2;
where point loads are applied, the length x the distance between the point
loads.

M=0

M=0
M

Mmax

7-55

Figure 7.7.3-2
Indication of the length x to be used for the calculation of the transverse
reinforcement to assure the force transfer web-flange: in the particular case of a
simply supported beam with uniformly distributed load, there are at least four
regions to be considered
The value for the inclination angle f which determines the orientation of the struts,
should be in accordance with the following conditions:
for compression flanges
1 cotg f 2 or 45 f 26,5
for tension flanges
1 cotg f 1,25 or 45 f 38,6
7.7.3.2 The transverse reinforcement
The transverse reinforcement per unit length Asf/sf is determined by means of the
following expression (see also figure 7.7.3-3):
resisting tensile force tensile force due to the shear force
and thus:
Asf

v .h .x

.x . f yd Ed f
s

cotg f
f

or finally:
Asf
sf

vEd .h f

(7.7.3-1)

f yd . cotg f

7.7.3.3 The verification of the compression struts

The verification of the compression struts is performed in accordance with figure 7.7.33, which leads to the following formula:

strut =

Fd
cos f

Fstrut
=
section strut h f .bstrut

strut =

vE .h f .x
cos f .h f .(x. sin f )

7-56

 strut . f cd

As:
one finds:

vE . f cd . cos f . sin f

Fd+Fd= Fd +vE.hf.x

(7.3.3-2)

Tensile force to
be resisted by
transverse
reinforcement
= Fd.tg
bstrut=x.sinf

x
Fd=vE.hf.x

f
Fd

Compression force
to be resisted by
the strut
Fd
=
cos f

Figure 7.7.3-3
Auxiliary figure for the development of the formulas for the calculation of the
transverse reinforcement to assure the force transfer web-flange and for the
verification of the compression struts

7.7.3.4 Combined actions

In the case of combined
shear between the flange and the web, and
transverse bending (bending in the horizontal plane),
two possible solutions are available:
the area of steel should be greater than the one given by expression (7.7.3-1);
half of the area of steel given by expression (7.7.3-1) in combination with the
reinforcement needed to resist transverse bending.
7.7.3.5 Remark
Extra transverse reinforcement (in addition to the normal one for bending) is not
necessary when the following conditions are met:
vEd 0,4.fctd according to EN 1992-1-1:2004; 6.2.4(6)
vEd 0,5.fctd according to NBN EN 1992-1-1 ANB:2009; 6.2.4(6)

7-57

8 Chapter 8
Torsion
8.1 Introduction
Structural members in reinforced concrete may be subjected to torsional loading.
Distinction should be made between:
structural members whose torsional resistance is mandatory for the static
equilibrium of a structure; this may be called a primary torsion effect;
structural members in which torsion arises from compatibility considerations
only, because the member acts in the context of a statically indeterminate
structure; this may be called a secondary torsion effect.
The following text is composed on the basis of LAMBOTTE (1989) and WIGHT
(2009).
8.1.1

The primary torsion effect: torsional equilibrium

Figure 8.1.1-1 presents some examples of structural members which are loaded by
torsional moments and where the torsional resistance of the member is of vital
importance for the static equilibrium of the structure. The figure shows eccentrically
loaded beams and a tubular bridge girder or box girder; the last example is met in the
context of large motorway bridges which are composed of one central box girder,
supported by centrally positioned piles. Traffic loads may lead to important torsional
moments in the box girder which has to transfer the load to the columns and the
abutments without deformations that would cause perturbation of the traffic.

8-1

Figure 8.1.1-1
Examples of primary torsional effects: (a) and (b) eccentrically loaded beams; (c)
torsional loading of a beam with cantilever; (d) tubular bridge girder or box girder

8.1.2

Secondary torsional effect caused by compatibility considerations

Figure 8.1.2-1 presents typical examples of members which are subjected to

parasitical or secondary torsional loads caused by compatibility considerations; the
secondary effects are caused by the fact that the beam is connected to other structural
members and is subjected to imposed deformations.
Figure 8.1.2-1(a) represents the peripheral beam of a slab; due to the hogging moments
(fixed edge for the slab!), the slab imposes a rotation to the beam, which is hindered by
the bending stiffness of the supporting columns and walls that are rigidly connected to
the beam; this gives rise to a parasitical torsional moment in the peripheral beam.
Figure 8.1.2-1(b) represents a frame structure in which the main girder is connected to a
transverse beam. Again, the connection (compatibility considerations) leads to
secondary torsional loading of the main girder.
The main characteristic of the members loaded by a secondary torsional effect is that
neglecting the parasitical torsion moment does not have a major influence on the static
equilibrium; the safety of the member in ULS is not in danger. However, when the
stiffness of the connections and the secondary torsional moments are not taken into
account, problems may arise in SLS with the appearance of unacceptable crack widths.

Figure 8.1.2-1

8-2

Examples of secondary torsional effects due to compatibility considerations in statically

indeterminate structures; (a) the slab is fixed to the peripheral beam; (b) a main girder is
connected to a secondary beam in a frame structure

8.2 Reminder from the strength of materials course: basic formulas

for torsional loading
8.2.1

Unreinforced plain cross-section

In the strength of materials course, the basic formulas for members subjected to
torsional loads are typically developed for a prismatic beam with circular plain crosssection, loaded by the torsional moment T.
An important result of that theory is the expression of the shear stress in a cross-section:

T
.
Ip

(8.2.1-1)

where
T = the torsional moment (figure 8.2.1-1);
Ip

= the polar moment of inertia of the cross-section: I p =

= the radius.

.R 4
2

Consequently, the maximum stress is given by:

max =

T .R
T
=
4
.R
.R 3
2
2

8-3

(8.2.1-2)

Figure 8.2.1-1
Summary of results for a cylindrical bar with plain cross-section loaded by torsion

The relationship between stress and strain is given by:

= G.
where (see figure 8.2.1-1):

R.
1

and

= the torsion angle per unit length (or angle of twist), which is given by:

T
G.I p

(8.2.1-3)

The principal stresses 1 and 2 are oriented with an angle 45 with respect to the
longitudinal axis of the prismatic bar; the stresses are perpendicular to each other.
Moreover: 1 = -2 = -. Because of the principal tensile stress, helicoidally cracks
appear on the surface of a cylindrical bar composed of a brittle material (chalk,
unreinforced concrete).
The formula of DE SAINT- VENANT allows calculating the maximum shear stress for
various cross-sections:

max =

T
WT

(8.2.1-4)

where WT = the torsional modulus".

Table 8.2.1-1 presents values of max for several cross-sections.

Table 8.2.1-1
Torsion modulus WT and the maximum shear stress for torsional moments
(LAMBOTTE, 1987)

max =

Cross-section

T
WT

T
T
=
3
.d
0,196.d 3
16
8-4

T
.(d di 4 )
16.d
4

T
0,188.d 3

T
0,185.d 3

T
0,208.a 3
T
b .h 2
3.h + 1,8.b
2

T
b 2 .h
3

T
T
=
2. Ak .tmin 2.b1.h1.tmin

8.2.2

Tubular cross-section with variable shape and limited thickness of the walls of
the tube

Figure 8.2.2-1(a) presents a tubular cross-section with variable shape. The thickness t of
the wall of the tube may be variable but it is assumed that t is small with respect to the
mean radius of the cross-section. The torsional moment T generates the shear stresses 1
and 2 in the sections 1 and 2 respectively. It may be assumed that the shear stresses are
constant over the wall thickness (because of its small value). Figure 8.2.2-1(b)
represents an elementary part of the tube, isolated from the rest of the tube; the
elementary volume is situated between the planes 1 and 2 and between two crosssections with distance h between them.

8-5

Figure 8.2.2-1
Auxiliary figure for the development of the formula of BREDT for tubular crosssections

Exploitation of the reciprocity of the shear stresses and the expression of the vertical
equilibrium of the elementary volume, leads to:

1 .t1 = 2 .t 2 = .t = constant = q

(8.2.2-1)

The product .t is called the shear flow q; q is constant along the circumference of the
tube.
The expression of the rotation equilibrium (see figure 8.2.2-1(a)), leads to:

T = .t.a.dl = q. a.dl = q.(2. Ak )

(8.2.2-2)

where Ak is the area enclosed by the centre-line of the wall, including the inner hollow
area.
Expression (8.2.2-2) may be rewritten as follows:

q=

T
2. Ak

(8.2.2-3)

Note:
As .t = q is constant, the maximum shear stress corresponds to the minimum
thickness tmin ; this leads to the formula of BREDT:

8-6

T
2. Ak .tmin

(8.2.2-4)

WT = 2. Ak .tmin

(8.2.2-5)

max =
The torsional modulus WT is thus:

8.3 Members in reinforced concrete loaded by pure torsion:

fundamental aspects for design calculations
8.3.1

Some interesting experimental observations

8.3.1.1 Plain or tubular cross-section?

Experimental tests with torsional loading of prismatic beams in reinforced concrete

have been realized, in order to observe damage patterns and failure modes. The
reinforcement was composed of longitudinal bars and stirrups.
In general, it was observed that after the formation of inclined helicoidally cracks (with
inclination angle 45), the torsional resistance was assured only by a thin-walled tubular
shell at the surface of the member. This phenomenon is confirmed by the comparative
testing of reinforced concrete members with plain and with tubular cross-sections with
the same outer dimensions and the same reinforcement: after the cracking phase, both
members show the same rotational angle and the strains in the reinforcement are also
the same. In the case of the plain cross-section, the stirrups are put under stress only
when the cracks appear; the strain in the outer shell increases significantly which leads
to the dislocation of the shell from the core of the member: see figure 8.3.1-1. From that
point on, the contribution of the core to the torsional stiffness is negligible; torsional
stiffness is determined by the outer shell.
Conclusion: these experimental results show that the ULS design of structural members
loaded in torsion, should be based on the model of the tubular cross-section.

homogeneous
cross-section

Uncracked state

Concrete core
Tubular shell

Cracked state

Figure 8.3.1-1
After cracking (dislocation of the outer Shell, containing the reinforcement, from the
core), torsional resistance is determined by the outer shell

8-7

8.3.1.2 Truss models

The observed cracks are arranged according to the helicoidally pattern of the principal
tensile stress trajectories in the theoretical strength-of-materials-model. In analogy with
the shear verification philosophy, a truss model may be identified in the outer tubular
shell: the truss model is composed of
compressive struts with an inclination angle of 45 with respect to the
longitudinal axis, and
ties which correspond to the reinforcement.
The theoretical principal tensile stress trajectory leads to the observation that the ideal
reinforcement should follow the helicoidally pattern. Although helicoidally
reinforcement is possible in the form of spiral-like windings, the more practical solution
is composed of longitudinal bars and transversal links or stirrups. It should be stressed
that in many cases, the sign of the torsional moment is variable, which explains the
necessity to use orthogonal reinforcement meshes.

Note:
In the case of shear verification, it is possible to make beneficial use of the
arch-effect: one can take into account that the uncracked compression concrete
is able to withstand a certain shear load (VRd,c). This is not possible for torsion,
because the member is cracked all around its axis. This also means that the
design can be performed according to the pure truss system. This is confirmed
experimentally: measurements of the strains in the stirrups show that the stress
in the stirrups after cracking, indeed reaches the stress values calculated with the
theoretical truss-system.

8.3.1.3 Torsional stiffness in uncracked and cracked phase

Experimental results show that the torsional stiffness in the cracked phase is only 10%
of the stiffness in uncracked phase. As soon as the cracked phase is reached,
deformations increase drastically.
For structural members where torsional stiffness is needed (SLS conditions), it may be
necessary to increase the cracking moment to exceed the maximum imposed moment.
This may be realized by means of prestressing.
Deformations may also be limited by the choice of a cross-section with large torsional
rigidity: see figure 8.3.1-2.

8-8

Figure 8.3.1-2
Typical cross-sections with increasing torsional rigidity

8.3.2

1st truss model: rectangular tubular member with helicoidally arranged

reinforcement

The arrangement of reinforcement along a helical line is rather theoretical for girders
with rectangular cross-sections. Figure 8.3.2-1 presents the model on the basis of which
design formulas may be worked out. This theoretical solution is not further developed in
the present notes.

h1
b b1
sh

Figure 8.3.2-1
1st (theoretical) truss model for the design calculation of rectangular tubular members,
subjected to a torsional moment T

8-9

8.3.3

2nd truss model: rectangular tubular member with orthogonal reinforcement

mesh

8.3.3.1 The truss model

Figure 8.3.3-1 represents a tubular member with rectangular cross-section. The

reinforcement cage is composed of longitudinal bars and stirrups which are
perpendicular to the longitudinal bars. The spacing between the stirrups is s.

h1

Figure 8.3.3-1
Tubular member with rectangular cross-section and orthogonal reinforcement mesh

Figure 8.3.3-2 shows one face of the tubular member; the compression struts are
characterized by the inclination angle ; the reinforcement bars in the orthogonal mesh
are characterized by the angles and

with respect to the struts.

8-10

Figure 8.3.3-2
One face of the tubular member; the truss model is composed of compression struts
between inclined cracks and an orthogonal mesh of reinforcement bars

8.3.3.2 The stress in the compression struts

Figure 8.3.3-3(a) shows the unit length AB = 1 and the imposed shear flow q. The
equilibrium of forces is considered for the triangle ABC. The force acting on the side AB
is equal to:

.surface = .t. AB = q. AB = q
Taking account of the reciprocity of the shear stresses, one finds the forces acting on the
side BC:

.t.BC = q.BC = q.AB.cotg = q.cotg

The side AC is cracked (no loads applied). Consequently, the equilibrium of forces for
the triangle ABC leads to the expression of the force Fc in the compression strut (see
figure 8.3.3-3(b)):
Fc =

q
sin

or also:
Fc = q. sin + q. cotg . cos = q. cos .(tg + cotg )

8-11

(8.3.3-1)

T
t

1
q

Fc

Fc
(a)

q.cotg
(b)

Figure 8.3.3-3
Auxiliary figure for the determination of the stress in the compression struts;
(a) isolation of the triangle ABC; (b) equilibrium of forces

The compression stress in the strut is thus given by:

c =

Fc
F
= c
surface BD.t

where D is identified in figure 8.3.3-3(a);

With BD = AB.cos = cos and expression (8.3.3-1), one finds:

c =

With the shear flow q =

Fc
q
=
cos .t sin . cos .t

T
, the expression is finally written as:
2. Ak

c =

T
2. Ak .t. sin . cos

Note 1:
with = 45, expression (8.3.3-2) turns into:

c =

(8.3.3-2)

T
T
=
Ak .t b1.h1.t

Note 2:
8-12

(8.3.3-3)

It may be shown that the use of an orthogonal reinforcement mesh leads to the
doubling of the stress in the compression struts with respect to the case where
helicoidally reinforcement ( = 45) is used.

8.3.3.3 The reinforcement to withstand torsional loading

Starting from figures 8.3.3-2 and 8.3.3-3, the equilibrium of forces for the triangle ABC,
leads to the identifications of the forces that have to be taken up by the longitudinal
reinforcement (index l) and by the transverse reinforcement (index w - notations EC2).
1.

Transverse reinforcement (stirrups or links):

Fsw = q. AB = q
With s = the spacing between the stirrups and Asw = the cross-sectional area of one
stirrup (one leg only!), the number of stirrups that cross the corresponding part of a
crack with an inclination angle is equal to (see figure 8.3.3-4):
n=

BC cotg
=
s
s

1
A

cotg . AB = cotg

Figure 8.3.3-4
Auxiliary figure for the determination of the number of stirrups needed

8-13

The force Fsw has thus to be taken up by n stirrups distributed over the length
BC = cotg . With the assumption that the steel of the stirrups is loaded up to the
design yield strength fywd , one finds:

Fsw = q = n. Asw . f ywd

(8.3.3-4)

T
T
cotg
=
and n =
, expression (8.3.3-4) may be rewritten as
2. Ak 2.b1.h1
s
the formula that allows to determine the necessary transverse reinforcement per unit
length Asw/s :

With q =

Asw
T
T
=
=
s
2. Ak . cotg . f ywd 2.b1.h1. cotg . f ywd

Note:
With = 45, expression (8.3.3-5) leads to:
Asw
T
T
=
=
s
2. Ak . f ywd 2.b1.h1. f ywd

2.

(8.3.3-5)

(8.3.3-6)

Longitudinal reinforcement
The equilibrium of forces for the triangle ABC in figure 8.3.3-3, allows to
determine the tensile force Fsl that should be taken up by the longitudinal
reinforcement per unit length AB = 1 along the circumference of the cross-section:
Fsl = q. cotg
For the complete circumference uk along the centre-line of the thin-walled tubular
cross-section (uk is thus the circumference of the area Ak), one finds the total tensile
force:
Fsl ,tot = q. cotg .u k

(8.3.3-7)

where u k = 2.(b1 + h1 ) .
The total reinforcement area Asl of the longitudinal reinforcement which has to be
distributed along the circumference uk, is deduced from expression (8.3.3-7), taking
account of:
the assumption that the stress in the steel rises up to the design yield strength fyld:

8-14

Fsl ,tot = Asl . f yld

(8.3.3-8)

and
q=

T
2. Ak

Consequently:
Asl =

T . cotg .u k T . cotg .u k
=
2. Ak . f yld
2.b1.h1. f yld

(8.3.3-9)

Note:
With = 45, expression (8.3.3-9) leads to:
Asl =

T .u k
T .u k
=
2. Ak . f yld 2.b1.h1. f yld

(8.3.3-10)

8.4 Design procedure

Reference: EN 1992-1-1:2004; 6.3.
8.4.1

Principal recommendations

Where the static equilibrium of a structure depends on the torsional resistance of

elements of the structure, a full torsional design covering both ULS and SLS shall be
made.
Where in statically indeterminate structures, torsion arises from consideration of
compatibility only, and the structure is not dependent on the torsional resistance for its
stability, then it will normally be unnecessary to consider torsion at the ULS. In such
cases, a minimum reinforcement in the form of stirrups and longitudinal bars should be
provided in order to prevent excessive cracking.

8-15

8.4.2

Notations and definitions used in the EC2 prescriptions

The torsional resistance of a cross-section may be calculated on the basis of a thinwalled closed section. Solid cross-sections may be modelled by equivalent thin-walled
sections; see figure 8.4.2.-1.

TEd

tef,i

zi

wall i

tef,i/2

concrete cover

: longitudinal bars
: stirrups
: centre-line

Figure 8.4.2-1
The calculation model for solid members loaded in torsion is composed of a thin-walled
closed cross-section (figure 6.11 in EN 1992-1-1:2004)
The cross-section is composed of several walls i. The centre-lines of all the connecting
walls form together the centre-line of the tubular cross-section.
EN 1992-1-1:2004 introduces the following notations:
TEd
: design value of the imposed torsional moment;
Ak
: area enclosed by the centre-lines of the connecting walls, including inner
hollow areas; see figure 8.4.2-2;
A
: total area of the cross-section within the outer circumference, including
inner hollow areas; see figure 8.4.2-2;
u
: outer circumference of the cross-section; see figure 8.4.2-2;
uk
: circumference of the area Ak; see figure 8.4.2-2;
zi
: side length of wall i, measured along the centre-line;
tef,i
: effective thickness of wall i;
t,i
: torsional shear stress in wall i.

8-16

Ak

uk

Figure 8.4.2-2
Thin-walled closed cross-section: definitions of the areas A and Ak
and of the circumferences u and uk

In the case of solid cross-sections, EN 1992-1-1:2004 proposes to determine the

effective wall thickness for wall i as:
tef ,i =

A
u

(8.4.2-1)

but should not be taken as less than twice the distance between the outer edge and the
centre of the longitudinal reinforcement.

8.4.3

Design calculation of a single cross-section

The design procedure in EN 1992-1-1:2004 is based on a truss model. Figure 8.4.3-1

presents the truss model for a rectangular section. Two types of reinforcement bars may
be identified: stirrups and longitudinal bars. The concrete compression struts are
helicoidally arranged all around the section and show an inclination angle with respect
to the longitudinal axis.

8-17

struts
(upper side)

: longitudinal bars
: stirrups
: compression struts
: centre-line

struts
(side)

Figure 8.4.3-1
Truss model for the design calculation of a prismatic member with rectangular crosssection, subjected to torsional loading

As it was also the case for shear load verification (chapter 7 in these course notes), the
inclination angle of the struts is limited to the following:
1 cotg 2,5

(8.4.3-1)

45,0 21,8

(8.4.3-2)

which corresponds to:

The design is based on the shear force VEd,i, which is generated by the torsional moment
TEd in wall i:
VEd ,i = t ,i .tef ,i .zi

(8.4.3-3)

where t,i.tef,i is the shear flow qi in wall i:

TEd
2 Ak

(8.4.3-4)

TEd
. zi
2 Ak

(8.4.3-5)

t ,i .tef ,i =
Thus:
VEd ,i =

8-18

8.4.3.1 Design of the transverse reinforcement (stirrups)

The design of the necessary stirrups takes place in the same way as for the shear load
verification (see 7.3.4 in these course notes).
The shear force VEd,i in wall i has to be taken up by the following area of transverse
reinforcement per unit length (defined by expression (7.3.4-7)):
TEd
. zi
VEd ,i
Asw
2 Ak
=
=
s
z. f ywd .(cotg + cotg ). sin z. f ywd .(cotg + cotg ). sin

(8.4.3-6)

where is the inclination angle of the stirrups. EN 1992-1-1:2004 recommends using

the angle = 90; equation (8.4.3-6) then becomes:
Asw
TEd .zi
=
s
2. Ak .z. f ywd . cotg

(8.4.3-7)

For wall i, a first approximation for the depth z is zi; consequently:

Asw
TEd
=
s
2. Ak . f ywd . cotg

(8.4.3-8)

8.4.3.2 Design of the longitudinal reinforcement

The total longitudinal reinforcement area Asl is given by:
Asl =

TEd uk
cotg
2 Ak f yd

(8.4.3-9)

In the case of combined actions (torsion with bending):

- for the walls loaded in tension, the longitudinal reinforcement for torsion is
superimposed to the longitudinal reinforcement for bending;
- for the walls loaded in compression, the longitudinal reinforcement for torsion
may be reduced in proportion to the applied compression force.
8.4.3.3 Verification of the compression struts
The compression force in the concrete struts is in general the result of the combination
of torsion and shear. The maximum resistance of a member subjected to torsion and
shear is limited by the capacity of the concrete struts.

8-19

Where a section is loaded at the same time by the design torsional moment TEd and by
the design transverse force VEd, the following condition should be satisfied in order not
to exceed the strut resistance:
TEd
TRd , max

VEd
1
VRd , max

(8.4.3-10)

VRd,max is the maximum design shear resistance (the formula is developed in chapter 7 in
these course notes).
TRd,max is the design torsional resistance moment, given by:
TRd , max = 2. . cw . f cd . Ak .tef ,i .sin . cos

(8.4.3-11)

where
- is the reduction factor which takes account of the fact that the struts are not
loaded in uni-axial compression:

= 0,6.1
-

f ck
250

(with fck in MPa)

(8.4.3-12)

cw is a coefficient taking account of the state of the stress in the compression

strut:

cw = 1

for non-prestressed structures

cw 1

for prestressed structures

cw = 1 +

cp

for 0 < cp 0,25.fcd

f cd

cw = 1,25

cw = 2,5.1

for 0,25.fcd < cp 0,5.fcd

cp

f cd

for 0,5.fcd < cp < 1,0.fcd

Note:
For approximately rectangular solid sections, only minimum reinforcement is
required provided that the following condition is satisfied:
TEd
V
+ Ed 1
(8.4.3-13)
TRd , c VRd , c
where:
VRd,c : see chapter 7;
TRd,c : torsional cracking moment, which may be determined by setting
T,i = fctd.
8-20

8.4.4

Design calculation of complex cross-sections

Complex shapes, such as T-sections, may be divided into a series of sub-sections, each
of which is modelled as an equivalent thin-walled section. The torsional resistance is
taken as the sum of the capacities of the individual elements.
The distribution of the acting torsional moment TEd over the sub-sections, should be in
proportion to their uncracked torsional stiffnesses. This is in fact the expression of the
compatibility of deformations in torsion of the different sub-sections: all sub-sections
are subjected to the same rotation angle.
Each sub-section may be designed separately.
Figure 8.4.4-1 presents the application of the method to the example of a T-section. The
original cross-section (figure 8.4.4-1(a)) is subdivided into two simple sub-sections
(figure 8.4.4-1(b)): the flange with uncracked torsional stiffness G.IP1 and the web with
uncracked torsional stiffness G.IP2. The imposed torsional moment TEd is subdivided
among the two sub-sections according to the following:
T1
T
TEd
= 2 =
G.I P1 G.I P 2 G.I P1 + G.I P 2
Each sub-section is then replaced by an equivalent thin-walled closed tubular section
(figure 8.3.3-1(c)), as was explained in paragraph 8.4.2 in these course notes. The
design of each sub-section may be performed separately.

TEd

T1

T1

(a)

(b)

(c)

Figure 8.4.4-1
Calculation of the torsional resistance of cross-sections with complex shapes; example
of a T-section: (a) the full cross-section; (b) subdivision into two sub-sections;
(c) model with equivalent thin-walled closed tubular sections

8-21

8.4.5

Detailing of reinforcement

Reference: EN 1992-1-1:2004; 9.2.3.

- The torsion links (or stirrups)
- should be closed and be anchored by means of laps or hooked ends; see
examples in figure 8.4.5-1;
- should form an angle of 90 with the axis of the structural element.

Figure 8.4.5-1
Examples of shapes for torsion links
(figure 9.6 in EN 1992-1-1:2004)

- The minimum quantity of torsion links required, is determined by the same

provisions as for shear verification (see paragraph 7.6.2 in these course notes); a
minimum geometrical transverse reinforcement ratio is recommended in the
following way:

w,min =

0,08. f ck
f ywk

(8.4.5-1)

where fck and fywk are in MPa.

- The longitudinal spacing of the torsion links should not exceed:
u/8, where u is the outer circumference of the cross-section;
sl,max = 0,75.d.(1 + cotg ) (see paragraph 7.6.3.1, shear reinforcement);
the lesser dimension of the beam cross-section.
- The longitudinal bars should be arranged that there is at least one bar at each corner.
The other bars are distributed uniformly around the inner periphery of the links, with
spacing not greater than 350 mm.

8-22

9 Chapter 9
Calculation of crack widths and deflections
9.1 Crack width
9.1.1

The importance of the bond between steel and concrete

The fact that reinforced concrete may be used as construction material is essentially
determined by the composite action due to the bond between steel and concrete. This is
made clear by means of the example shown in figure 9.1.1-1. The figure shows three
geometrically identical beams, simply supported and loaded by the force Q in the
middle of the span:
beam 1 is characterized by the absence of adherence between the reinforcement bar
and the concrete; the bar has a smooth surface and is put into a cylindrical tube
filled with grease: figure 9.1.1-1(a);
beam 2 has the same arrangement as beam 1, but the ends of the steel bar are
anchored by means of anchor plates on both ends of the concrete beam: figure
9.1.1-1(b);
beam 3 contains a bar with ribbed surface and is characterized by good bonding
between steel and concrete all over the length of the bar: figure 9.1.1-1(c).

Beam 1

Beam 2

Beam 3

Figure 9.1.1-1
Three geometrically identical beams, loaded in bending;
(a) no bond between steel and concrete and no anchorage; (b) no bond between steel
and concrete but anchorage of the bar at its ends; (c) normal bond (WALRAVEN, 1995)

9-1

The behaviour of the three beams is different for the same type of loading.
1. Beam 1 fails when the tensile stress at the bottom side of the beam in the middle
cross-section, becomes equal to the flexural tensile strength of the concrete. At this
moment, for the load Q1, a crack appears. The steel bar starts to slide freely with
respect to the concrete; the bar is unable to take any tensile load and cannot prevent
that both concrete parts of the beam are separated as soon as the crack appears. The
sliding bar is thus absolutely useless because it does not increase the failure load of
the beam with respect to the one for the unreinforced beam.
2. The behaviour of the beam may be improved by the anchoring of the steel bar at the
ends of the concrete beam, such as for beam 2. The steel bar can now take up the
tensile force as soon as the crack in the concrete appears in the middle crosssection. Because of the lack of adherence all along the length of the bar, the
increasing load does not lead to additional cracks; there is only one crack whose
width corresponds practically to the increase in length of the whole bar. The failure
load for beam 2 is clearly larger than Q1 but the crack width is again unacceptable.
3. The behaviour of beam 3 is totally different from that of the beams 1 and 2. Again,
a crack appears in the middle cross-section when load Q1 is reached. A small slip of
the steel with respect to the concrete can be observed in the immediate
neighbourhood of the edges of the crack, but the parts of the beam are still
connected to each other thanks to the bond along the whole length of the steel bar.
When the load Q increases, new cracks appear in neighbouring cross-sections as
soon as the tensile stress in the concrete exceeds the concrete flexural tensile
strength. The crack widths of the first cracks increase with increasing load and
further cracks appear in the regions close to the supports. Finally, the beam fails for
load Q = Q3 according to one of the following mechanisms:
the tensile stress in the steel in the middle cross-section reaches the yield value,
which causes large steel strain and an unacceptable crack width in the middle
cross-section of the beam: Q3 = f (fyk), with Q3 significantly larger than Q1;
the compression stress in the concrete at the top side of the middle crosssection reaches the concrete compression strength and concrete crushing
appears: Q3 = f (fck), with Q3 significantly larger than Q1;
the bond between steel and concrete is disrupted and the steel bar slides with
respect to the concrete. This failure mechanism may be avoided by the
provision of sufficient anchorage.
The analysis of the behaviour of the three beams allows the identification of two
essential aspects for reinforced concrete:
the anchorage of the ends of the steel bar is necessary in order to avoid the 3rd
failure mode of beam 3 and to assure that failure in bending is only possible
either by the yielding of the steel in tension or by the crushing of the
compression concrete;
the bond between steel and concrete is necessary in order to obtain multiple
cracks well distributed over the length of the beam, allowing to keep the crack
widths reduced and to assure the applicability of the assumption of
BERNOULLI (no slip between steel and concrete).

9-2

9.1.2

Crack development: some experimental observations

In order to study the development of cracks in reinforced concrete, tensile tests have
been performed in the past (WALRAVEN, 1995) on prismatic concrete bars with
various reinforcement schemes: see figure 9.1.2-1. Parametric analysis included the
variation of the concrete strength, the reinforcement ratio, the bar diameter and the
number of bars. The results of the tension tests on axially loaded bars with centrally
located reinforcement are of fundamental importance for the development of models for
the verification of crack widths.

Figure 9.1.2-1
Schematic overview of crack patterns obtained by tensile tests on prismatic concrete
bars with centrally positioned reinforcement bars; the effects are illustrated for two
parameters: the increasing reinforcement ratio and the distribution of the crosssectional area of reinforcement in the concrete cross-section (WALRAVEN, 1995)
Figure 9.1.2-1 leads to the following observations:
a larger reinforcement ratio and a better distribution of the reinforcement within
the cross-section larger crack density with smaller crack widths;
the crack spacing seems to be constant.

9.1.3

Parametric analysis of the crack development in a tensile loaded prismatic bar

with centrally positioned reinforcement (WALRAVEN, 1995)

9.1.3.1 Uncracked situation

Figure 9.1.3-1 shows a prismatic concrete bar with centrally positioned reinforcement
bar, loaded by an axial load in a displacement controlled tensile test. When the bar is
uncracked, the strains in steel and concrete are the same. The portions of the load that
are taken up respectively by the steel and the concrete are:

9-3

N c c . Ac Ec . c . Ac

(9.1.3-1)

N s s . As Es . s . As

(9.1.3-2)

with c s

s=c=
c=Ec.

s=Es.

Figure 9.1.3-1
Concrete prism with central reinforcement bar loaded in tension; uncracked situation

The total force is thus:

N N c N s ( Ec . Ac E s . As )

(9.1.3-3)

or also:

N .Ec . Ac .(1
Introducing

Es As
. )
Ec Ac

Es
A
and s , leads to:
Ec
Ac
N .Ec . Ac .(1 . )

(9.1.3-4)

The uncracked situation is traditionally called state 1; strains and stresses are
indicated with the index 1: s1, c1, s1 en c1.

9.1.3.2 The first crack

With increasing imposed strain , at a certain moment, the stress in the concrete reaches
the concrete tensile strength fct. The concrete tensile strength is a stochastic property,
which explains why the first crack appears in the cross-section with the smallest
available tensile strength: figure 9.1.3-2(a). The stress in the concrete c = 0 in the
cracked cross-section; the steel bar has to transfer the full tensile force on its own.
Due to the bond between steel and concrete, the latter is gradually forced to take over a
part of the tensile force, on both sides of the crack. Outside the disturbed region, one
observes again the uncracked situation.
9-4

The length which is necessary to again build up the tensile force in the concrete depends
on the bond strength. Experimental results (WALRAVEN, 1995) show that the
simplified model with constant bond stress between steel and concrete, equal to two
times the mean concrete tensile strength ( = 2.fctm) may be assumed for predictions with
acceptable accuracy; see figure 9.1.3-2(e). With this assumption, one finds a linear
evolution of the stresses in steel and concrete over the disturbance length lt: see figure
9.1.3-2(c) and (d).
crack
disturbed region

(a)

(b)

s>s1
s1

(c)

c1
(d)

stress in
reinforcement

stress in
concrete

c=0

(e)

bond stress

Figure 9.1.3-2
The first crack appears in the tensile loaded prismatic specimen with centrally
positioned reinforcement bar; (a) prismatic specimen with the crack; (b) the disturbed
region; (c) the evolution of the steel stress in the disturbed region; (d) the evolution of
the concrete stress in the disturbed region; (e) evolution of the bond stress between steel
and concrete in the disturbed region (WALRAVEN, 1995)
When the prismatic specimen is subjected to a slowly increasing elongation and the first
crack appears, the tensile force which is necessary to produce the elongation drops
suddenly; indeed, because of the cracking of the concrete, the stiffness (E.A) is reduced
and the force that is necessary to realize the imposed is smaller; see figure 9.1.3-3.
9-5

N
Nr
N0

Figure 9.1.3-3
Drop of the tensile force with the first crack in the prismatic specimen (displacement
controlled tensile test!)

9.1.3.3 Further development of cracks

With further increasing strain, the tensile force increases again as well. Yet, when the
tensile force reaches the level of the first crack load, a second crack appears at the
second weakest spot in the prismatic specimen; as a matter of fact, the second crack
load is in principle somewhat higher than the first crack load. A new disturbed region
appears left and right of the second crack. This phase, during which new cracks are
progressively developed while the tensile force applied to the specimen does not
increase substantially, is called the crack development phase: see figure 9.1.3-4.

disturbed region

Figure 9.1.3-4
During the crack development phase, the tensile force N remains constant with
increasing imposed strain

9.1.3.4 Determination of the maximum tensile force during the crack development phase
The maximum force during the crack development phase Nr can be determined easily
on the basis of the stress distribution in steel and concrete in the disturbed region, right
at the moment that the stress in the concrete reaches the tensile strength fct again
somewhere: see figure 9.1.3-5. The steel stress in the crack reaches the maximum value

9-6

sr. During the cracking phase, the steel stress can never be larger than sr, because sr

corresponds to the fact that somewhere fct is reached and thus to the creation of a new
crack.

(a)

sr
s1

(b)

stress in
reinforcement

c1=fct

(c)

stress in
concrete

c=0

(d)

bond stress

Figure 9.1.3-5
Auxiliary figure for the determination of the maximum tensile force Nr in the crack
development phase; (a) disturbed region; (b) stress in the reinforcement bar in the
disturbed region; (c) stress in the concrete in the disturbed region; (d) evolution of bond
stress in the disturbed region: simplified model assuming constant bond stress

On the basis of the stress distributions shown in figure 9.1.3-5(b) and (c), one can write
two expressions for Nr:

in an undisturbed cross-section:

N N c N s c . Ac s . As
-

with c = the concrete tensile strength fct (the stress cannot be higher, because
with fct, a crack develops), and
s = s1 which is, because of BERNOULLIs assumption (s = c), equal to:

9-7

 s1

Es
. c . f ct
Ec

(9.1.3-5)

Consequently:
N r f ct . Ac . f ct . As
or (see (9.1.3-4)):
N r f ct . Ac (1 . )

(9.1.3-6)

in a cracked cross-section:
The steel bar has to take up the full tensile load Nr. Consequently:
N r sr . As

(9.1.3-7)

Conclusion:
N r f ct . Ac (1 . ) sr . As
which permits to find the expression for sr:

sr

f ct . Ac (1 . ) f ct .(1 . )

As

(9.1.3-8)

Note:
In reality, the maximum force Nr in the crack development phase, which is
reached every time just before a new crack is created, is not constant but
increases slightly during the crack development phase. Indeed, the first
crack develops in the weakest cross-section in the prismatic specimen and
the next cracks develop in sections where the tensile strength of the concrete
is a little bit higher. The tensile strength fct increases slightly during the
crack development phase and so does the tensile force Nr (no horizontal
plateau in figure 9.1.3-4), and so does the steel stress sr in the cracks.

9.1.3.5 The load transfer length lt

The load transfer length (NL: inleidingslengte; FR: longueur de transmission des forces)
is determined on the basis of figures 9.1.3-5(c) and (d). The concrete stress is 0 in the
crack and is equal to fct at the end of the disturbed region; hence, due to the bond
between steel and concrete, the following force is built up in the concrete along the load
transfer length lt:
N Ac . f ct
9-8

(9.1.3-9)

The force which is built up by the bond stress along the stress transfer length lt can
also be written as:
N .m. . .lt

(9.1.3-10)

with m = the number of reinforcement bars in the section Ac.

Introducing:
Ac

As

; As m.

. 2
4

; 2. f ct

expression (9.1.3-10) is rewritten as:

1
lt .
8

(9.1.3-11)

9.1.3.6 Maximum crack width during the crack development phase

The maximum crack width during the crack development phase is equal to the
difference between the elongation of the steel bar and the elongation of the concrete
along the disturbed region with length = 2.lt:
wmax 2.lt .( sm cm )

(9.1.3-12)

where
sm = the mean strain in the steel reinforcement along the stress transfer length lt;
cm = the mean strain in the concrete along the along the stress transfer length lt;
sm and cm are deduced from figures 9.1.3-5(b) and (c):

mean strain in reinforcement (figure 9.1.3-5(b)):

sm

1
.( sr s1 )
2.Es

with

sr

f ct

.(1 . )

s1 . f ct
Consequently

9-9

(see expression (9.1.3-8))

(see expression (9.1.3-5))

 sm

1
1
. f ct .( 2. )
2.Es

(9.1.3-13)

mean concrete strain (figure 9.1-7(c)):

cm

f ct

. f ct
2.Ec 2.Es

(9.1.3-14)

Introduction of expressions (9.1.3-13) and (9.1.3-14) into (9.1.3-12) leads to:

wmax 2.lt .

1
f
.( ct . f ct )
2.Es

Taking into account expression (9.1.3-11) leads to:

1 f
wmax . 2 . ct .(1 . )
8 Es

(9.1.3-15)

9.1.3.7 Smallest and largest stress transfer length (which is as a matter of fact the
distance between adjacent cracks)
New cracks are developed with the increase of the imposed deformation, until the whole
prismatic specimen is composed of disturbed regions. In the zones with many cracks,
the disturbed regions start to overlap each other.
The smallest possible crack spacing is observed when a new crack develops right at the
end of a disturbed region; see figure 9.1.3-6(a). The smallest crack spacing is thus equal
to the stress transfer length lt.
The largest possible crack spacing is observed when a new crack is developed right at
the distance 2.lt (in fact a little bit smaller than 2.lt) from the end of a disturbed region.
The distance between the two cracks is then just too small to allow the stress in the
concrete to reach again the tensile strength in between the two cracks; see figure 9.1.36(b).
Conclusion
The crack spacing always varies between lt and 2.lt.

9-10

(a)
fct

stress in
concrete

(b)
fct

stress in
concrete

Figure 9.1.3-6
Minimum (a) and maximum (b) crack spacing
9.1.3.8 Crack width after the crack development phase
The crack development phase is over when the prismatic specimen is fully composed of
disturbed regions. Although the stress in the reinforcement keeps growing with
increased imposed elongation, there are no new cracks developed anymore, at least in
theory. The only effect of increased imposed deformations, is that crack widths become
larger. The maximum crack width is observed for the cracks with the length of the
disturbed region equal to 2.lt on both sides of the crack: see figure 9.1.3-7.
Maximum crack
spacing = 2.lt

s
s1

sr

fct

s1
fct

stress in
reinforcement

stress in
concrete

disturbed region which

defines the crack width

Figure 9.1.3-7
Determination of the maximum crack width in the phase after the completion of the full
crack pattern

9-11

The stress in the reinforcement at the level of the cracks increases with the value s
with respect to the steel stress during the crack development phase (sr). The value of
the increase s is constant between the cracks; indeed, the concrete is cracked and does
not take up any additional load; additional load is taken up by the steel.
The value of s can be determined at the level of the concerned crack from:
s

N Nr
s sr
As

(9.1.3-16)

Using expression (9.1.3-8):

sr

f ct

.(1 . )

leads to
s s

f ct

. f ct

The increase of the crack width, which is only due to the elongation of the
reinforcement, is thus:
w 2.lt .

s 2.lt
f

.( s ct . f ct )
Es
Es

(9.1.3-17)

Finally, the maximum crack width is determined as the sum of:

the maximum crack width during the crack development phase (expression
(9.1.3-15), and
the contribution w from expression (9.1.3-17):
1 f
2.l

f
wmax . 2 . ct .(1 . ) t .( s ct . f ct )

8 Es
Es

1
which may be rewritten, taking into account expression (9.1.3-11): lt .
8
in the following form:
wmax

f

. s ct .(1 . )
4. .Es
2.

9-12

(9.1.3-18)

Conclusion:
Two phases have to be considered for the calculation of the crack width:
the crack development phase, during which the crack width is maximum at the
moment right before the development of the next crack. During this phase, the
tensile force N cannot exceed the cracking force Nr when imposed elongation
continues. The maximum crack width is determined with expression (9.1.3-15);
the phase after the completion of the crack pattern. The force N increases with
increased imposed elongation. There are no new cracks anymore, but the width of
the existing cracks increases. The maximum crack width depends on the steel stress
in the crack and is determined with expression (9.1.3-18).
9.1.3.9 The load-deformation curve for tensile loaded prismatic specimen
Figure 9.1.3-8 presents in a simplified way the theoretical relationship between the
imposed deformation of the prismatic specimen and the corresponding tensile force that
is necessary to realize the deformation.
cracking
phase

crack pattern is
completed

(e)
(c)
(b)

Nr

0,375.

(d)

f ct
Es .

(a)

f ct .

0,625 .
Es .

Figure 9.1.3-8
Schematic representation of the tensile behaviour of the prismatic specimen with
centrally positioned reinforcement bar; (a) uncracked phase; (b) crack development
phase; (c) phase after completion of the crack development; (d) the reinforcement bar
without concrete cover; (e) plastic deformation of the steel bar

The first linear part (a) of the diagram corresponds to the tensile behaviour of the
uncracked specimen. The crack development phase (b) starts when the cracking force Nr
is reached; during this phase, the tensile load N does not exceed the cracking force Nr
with increasing deformation (the curve is simplified without presentation of the force
jumps each time a crack appears, as was the case in figure 9.1.3-4). After completion of
the crack pattern, the tensile force increases again with further deformation.
9-13

The dashed line in figure 9.1.3-8 represents the relationship N l/l for the same
reinforcement bar but without concrete cover. The load-deformation diagram for the
reinforced prismatic specimen in cracked situation, is parallel with the diagram for the
isolated bar. The horizontal distance between both parallel lines has a particular
meaning: this distance represents the so-called tension stiffening effect. Indeed, the
tensile loaded part of concrete between two adjacent cracks, assures a small contribution
in rigidity which explains that for a load N > Nr, the strain is smaller than for the
uncovered isolated reinforcement bar.

Note:
The term that links N with = l/l is the stiffness E.A. This explains why the
slope for the uncracked specimen ((a) in figure 9.1.3-8) is much steeper than for
the uncovered isolated reinforcement bar ((d) in figure 9.1.3-8).

The horizontal distance between the diagrams for the prismatic specimen and the
isolated steel bar may be calculated on the basis of a simplified model assuming a mean
spacing of cracks equal to 1,5.lt: see figure 9.1.3-9.
1,5.lt = mean crack spacing

s.As

s.As

s
sm

Figure 9.1.3-9
Prismatic specimen with centrally positioned reinforcement bar loaded in tension with
N > Nr; simplified model assuming crack spacing = 1,5.lt; evolution of the stress in the
steel bar

The stress in the steel bar at the level of the cracks is s = N/As, which is the same stress
as in the uncovered isolated bar. The mean steel stress, at the distance 0,375.lt from the
crack, is determined taking into account the force transfer mechanism via the bond
stress :

9-14

 sm s

0,375.lt . . .
. 2
4

1
With = 2.fct and lt . , the formula is rewritten as:
8

sm s 0,375.

f ct

Consequently, the mean strain is:

sm

sm
Es

s
Es

0,375. f ct
Es .

(9.1.3-19)

Conclusion
For the applied tensile force N, the difference between the mean strain of the cracked
concrete prismatic specimen sm and the strain of the uncovered isolated bar s (= s/Es),
0,375. f ct
; see indication in figure 9.1.3-8.
is thus equal to
Es .
Another interesting information is the strain for which the crack development phase can
be considered as completed. For N = Nr, one finds the following strain in the uncovered
steel bar:

s
Es

Nr
Es . As

Taking into account expression (9.1.3-6), the strain is:

f ct . Ac .(1 . ) f ct .(1 . )

Es . As
Es .

This strain has to be reduced with the tension stiffening effect in order to determine the
strain corresponding to the completion of the crack development phase:

f ct .(1 . )
f
0,375. ct
Es .
Es .

f ct .

0,625 .
Es .

9-15

(9.1.3-20)

See also the indication in figure 9.1.3-8.

In practice, it is generally observed that the strain which is the result of an imposed
deformation (for example due to a temperature difference) is smaller that the one
predicted by expression (9.1.3-20). The maximum crack width is then determined by
expression (9.1.3-15). If the mean strain is larger than the one predicted by expression
(9.1.3-20), the crack width should be calculated by means of expression (9.1.3-18). If
the tensile force N is larger then the cracking force Nr, expression (9.1.3-18) may be
used.
9.1.3.10 Influence of shrinkage of concrete on the crack width
The effect of shrinkage on the crack width is different in the two important phases: the
crack development phase and the phase after completion of the crack pattern.
During the crack development phase, and assuming the same total
deformation, shrinkage of concrete leads to the increase of the external
tensile force and thus to the development of new cracks. However, the
maximum width of these cracks does not increase. During the crack
development phase, shrinkage does not influence the maximum crack width.
In the phase after completion of the crack pattern, shrinkage does have an
influence on crack width. Indeed, as there are no new cracks developed in
this phase, the shrinkage of the concrete inevitably leads to the widening of
the existing cracks. The concrete shrinkage r leads to an increase of the steel
stress with r.Es; expression (9.1.3-18) thus has to be extended to:
wmax

f

. s ct .(1 . ) r .Es
4. .Es
2.

(9.1.3-21)

9.1.3.11 The influence of the position of the reinforcement on the crack width
The model that has been used in the developments presented up to now, starts from the
assumption that the steel reinforcement is uniformly distributed over the cross-section
and that the forces which are transferred to the concrete by the bond stresses do not have
to be spread out over a large cross-sectional area. If this were the case, the problem of
crack widths is not only determined by the bond properties but also by the geometry of
the structural element and the position of the reinforcement.
The expressions above are developed with the assumption that the load transfer length lt
is large with respect to the breadth of the concrete cross-section. With that assumption,
the concrete stresses are, in the section where the tensile strength is reached, almost
uniformly distributed over the cross-section (figure 9.1.3-10); the crack develops
instantly over the whole cross-section.
This is not the case anymore when the width of the concrete cross-section is large,
because the distribution of the tensile stresses is irregular: see figures 9.1.3-10(b) and
(c); the crack does not develop over the whole surface. Figure 9.1.3-10 (b) shows that
around the reinforcement bar, the mechanisms are identical to the ones discussed in the
previous paragraphs concerning the prismatic specimen. A lot of small cracks develop
along the bar but away from the bar, several cracks join together; only a few cracks

9-16

appear at the outer surface and generally show unacceptable widths. The reinforcement
scheme illustrated in figure 9.1.3-10(b) is thus not recommended for thick structural
elements when crack widths are an issue.
Figure 9.1.3-10(c) presents a better alternative with steel bars close to the outer side of
the structural member. It is observed that a lot of small cracks are distributed along the
surface; the wider cracks are situated on the inner side but they are invisible and are not
really an issue for aspects such as durability and permeability.
It may be concluded that the influence of a reinforcement bar on the control of the crack
width is limited to an effective zone immediately around the bar. This is taken into
account by the standard.
The phenomena described above also show up in structural elements loaded in bending.
It is thus clear that in structural members with large total depth h, the main
reinforcement (which is situated close to the edges) is only able to control crack width
in a limited region of the cross-section. Additional measures are necessary to prevent
the development of wide cracks outside the effective zones around the bars; this
explains the need for skin reinforcement.

(a)

convergence of several cracks

(b)

(c)

Figure 9.1.3-10
The effect of reinforcement bars on the control of crack width, is limited to an effective
region around the bars; (a) distribution of tensile stress in the concrete from a crack (b)
concentration of small inner cracks in the outer skin appears when the width of the
structural element is large with respect to the load transfer length; (c) favourable
position of the reinforcement.

9-17

9.1.4

The principle of the calculation of the crack width

9.1.4.1 Basic formula

The development is based on the behaviour of the prismatic specimen with centrally
positioned reinforcement bar, loaded in tension. The phase under consideration is the
one after the completion of the crack development: figure 9.1.4-1. The model is
simplified by assuming a constant spacing of the cracks: the mean value for the spacing
is indicated by the symbol srm (= a distance between the load transfer lengths lt and 2.lt
!).
The compatibility between steel and concrete deformations permits to identify the
relationship between the mean crack spacing srm and the mean crack width wm:
wm

s rm
0

s .dx

s rm
0

c .dx

or:
wm srm .( sm cm )

(9.1.4-1)

Figure 9.1.4-1
Prismatic specimen with centrally positioned reinforcement bar, loaded in tension; the
tensile force N exceeds the cracking force Nr; schematic representation of the simplified
model with constant crack spacing equal to the mean value srm
9.1.4.2 The mean value of the steel strain sm
Figure 9.1.4-2 shows a schematic representation of the relationship between the steel
stress N/As (= s2) and the steel strain s; the diagram is deduced from the one in figure
9.1.3-8 (with s instead of N).

9-18

Figure 9.1.4-2
Schematic representation of the tensile behaviour of the prismatic specimen with
centrally positioned reinforcement bar
Several points in this diagram are discussed in the following:

the first linear part in the diagram corresponds to the behaviour in uncracked
situation (the so-called state 1); the uncracked behaviour is represented by the
straight line with indication s1. The stiffness which determines the slope of the
diagram, is determined from expression (9.1.3-3):
N .Ec .( . As Ac )
or:
N
A
Ec .( c ).
As
As
with:

c s
The relationship between
(

N
(=s2) and s is determined by the stiffness
As

Ac
).Ec ;
As

the tensile behaviour of an uncovered isolated reinforcement bar (the so-called

state 2), is represented by the straight line with indication s2. The stiffness
which determines the slope of the straight line is deduced from:

9-19

N
Es . s
As
The relationship between s and s is determined by the stiffness Es (which is
A
smaller than the stiffness ( c ).Ec in state 1!);
As

the cracking stress sr is the stress in the tension reinforcement calculated on the
basis of a cracked section under the loading conditions causing first cracking;

the crack development phase, which is not represented here by a simplified

horizontal plateau but by a slightly inclined curve which is prolonged smoothly into
the curve representing the tensile behaviour after completion of the crack
development and which is marked in figure 9.1.4-2 by the indication sm (mean
steel strain).
The value s corresponds to the contribution of the concrete in tension between
the cracks (tension stiffening effect). Experimental analysis has shown that s may
be characterized by a hyperbolic evolution and that the following relationship offers
a good approximation (WALRAVEN, 1995):
s s ,max .

sr
s2

(9.1.4-2)

Starting from Nr, it is observed that s decreases with increasing N and evolves
asymptotically with respect to the straight line indicated by s2.
Figure 9.1.4-2 shows that for N > Nr (thus after crack development), sm is given by:

sm

l
s 2 s
l

Introduction of expression (9.1.4-2) into expression (9.1.4-3) gives:

sm s 2 s ,max .

sr
s2

For N = Nr , one finds s ,max s 2 r s1r

consequently:

sm s 2 s 2 r .

sr

s1r . sr
s2
s2

or:

9-20

(9.1.4-3)

sm s 2 .1

s 2 r sr
sr
.
s1r .
s2 s2
s2

On the basis of the similar triangles in figure 9.1.4-2, one may write (see figure 9.1.4-3):

s2

N
As

sm

s1

s2

s2
s,max

sr

s2

sr

s2
sr
s

s1r

s1

s2

s2r

Figure 9.1.4-3
Auxiliary figure for the development of the expression for the mean steel strain sm:
exploitation of similar triangles

s 2 r sr

s2 s2

and

s1r sr

s1 s 2

and thus:

sm

2
2
sr
sr
s1.

s 2 .1
s 2
s2

(9.1.4-4)

This expression can be rewritten as:

sm (1 ). s1 . s 2
The distribution coefficient has the following values:

when s 2 sr

9-21

(9.1.4-5)


1 sr
s2

(9.1.4-6)

when s 2 sr : 0

A correction coefficient may be introduced in the formula for in order to take

account of the effects of the duration of the load and of repeated loading; sustained
loading and repeated loading is likely to reduce the quality of the bond between steel
and concrete. Therefore, EN 1992-1-1:2004; 7.4.3 specifies the formula of as:

1 sr
s2

(9.1.4-7)

with = 1,0 for a single short-term loading and = 0,5 for sustained loads or many
cycles of repeated loading.

Note:
The elongation of an axially tensile loaded element is thus calculated as follows:

when Nk < Nr :
N k .l
expression (9.1.3-4) leads to: l l. c1 l. s1
( Ac . As ).Ec

with = Es/Ec equal to 6 or 7 for short-term loads and 15 to 20 for longterm loads in SLS;
when Nk Nr:
l l. sm with sm from (9.1.4-5) and from (9.1.4-7).

9.1.4.3 The mean steel strain with respect to the adjacent concrete
Next to expression (9.1.4-3), expression (9.1.4-5) should be seen as an alternative
formulation for the tension stiffening effect. According to expression (9.1.4-5), the
mean steel strain is composed of:
a fraction of the steel strain s2 which would appear if the bar were not covered by
concrete; this fraction becomes more important with increased applied stress (above
the cracking stress);
a fraction of the steel strain s1 which would appear if there werent any cracking of
the concrete and thus if there werent any slip between steel and concrete (s1 = c1);
this fraction becomes smaller with increased applied stress (above the cracking
stress).
The mean strain in the steel reinforcement with respect to the adjacent concrete is thus
only composed of the term s2, because the term (1-)s1 corresponds to the
deformation of the concrete.
Conclusion: the mean strain in the reinforcement with respect to the adjacent concrete
(notation sm,r) is equal to:

9-22

 sm,r sm cm . s 2
and with expression (9.1.4-7):

sm,r

sr s 2
.
1 1. 2 .

s 2 Es

(9.1.4-8)

9.1.4.4 The mean value of the crack spacing srm

The mean value of the tensile stress in the concrete at the distance x from the crack (see
figure 9.1.4-1) is:

ct , x

. . x ,m .x
Ac ,eff

with Ac,eff = the area of concrete surrounding the reinforcement and which effectively
cooperates. Indeed, stress analysis shows that the tensile stress in the concrete in a
cross-section, is not uniformly distributed but decreases towards the outer edges (z = c)
of the prismatic specimen. The stress at the edge approaches the value of the stress in
the neighbourhood of the bar for small values of c/x so that ct,x(z=c) can be expressed as:
c
x

ct , x ( z c ) (1 K . ). ct , x
K is a factor which takes account of the stress distribution in the cross-section and with
the bond conditions: good quality bond good distribution of the load stress
distribution is more uniform.
The mean value of the crack spacing is obtained by expressing that the tensile stress in
the concrete at the edge (which is built up by the bond shear stress ), reaches the tensile
strength at the spot where the crack appears. For the distance x equal to the mean crack
spacing (srm), the following condition is fulfilled when a new crack appears:

ct , x ( z c ) f ct
or:
(1 K .

c . . m .srm
).
f ct
srm
Ac ,eff

Consequently:

srm

f ct . Ac ,eff

. . m

9-23

K .c

With r defined as:

As
Ac ,eff

the expression for the mean crack spacing (with As = 2) becomes:

srm K .c

. 2 . f ct
4. r . . . m

or finally:

srm K .c

. f ct
4. r . m

(9.1.4-9)

9.1.4.5 The mean crack width

The mean crack width is found by introducing expressions (9.1.4-8) and (9.1.4-9) in the
expression (9.1.4-1), which leads to:
2

sr s 2
. f ct
.
.1 1. 2 .
wm K .c
4. r . m
s 2 Es

(9.1.4-10)

9.1.4.6 Characteristic value of the crack width

Experimental analysis of crack widths reveals that in first approximation, the
assumption of a normal distribution may be applied to the crack width. This allows
defining the characteristic value of the crack width (95% of the crack widths are smaller
than wk):

wk .wm

(9.1.4-11)

The value of wk has to be limited to certain limit values wlim depending on the specific
nature and function of the structure, on the environmental conditions and on the
considered load combination.

9.1.5

Practical calculation of the crack width according to NBN B15-002; 4.4.2.4.

(EC2:1998)

The prescriptions for the crack width calculation have been subjected to an evolution
during the last decade. This is the reason why this paragraph presents a short overview
of prescriptions from the previous (Belgian) version of EC2.

9-24

9.1.5.1 Basic formula

The characteristic value of the crack width is determined by means of the following
expression:

wk .srm . sm

(9.1.5-1)

where:
wk = the characteristic value of the crack width;
srm = the mean crack spacing;
sm = the mean value of the strain which takes into account the contribution
(stiffness, shrinkage, etc) of the concrete, determined for the quasi-permanent load
combination. Important: the meaning of sm in expression (9.1.5-1) is indeed the
mean strain in the reinforcement with respect to the surrounding concrete, given by
sm,r in expression (9.1.4-8)!
= coefficient which makes the link between the mean value and the characteristic
value of the crack width; the following values are recommended:
= 1,7 for cracking due to loading and due to obstruction of deformations in crosssections with minimum dimension (depth or width) larger than 800 mm; the
probability of having a uniformly distributed tension stress is limited with large
dimensions;
= 1,3 for cracking due to obstruction of deformations in cross-sections with
minimum dimension (depth of width) smaller than 300 mm.
The values for for dimensions in-between are obtained by interpolation.
The influence of shrinkage on the crack width may be taken into account by:

wk .( sm ,r cs ).srm

(9.1.5-2)

9.1.5.2 The mean value of the strain in the reinforcement with respect to the
surrounding concrete
Notation: sm in NBN B15-002 (EC2:1998)!
sm is determined with the following equation:

sm

.1 1. 2 . sr
Es
s2

0,4. s
Es

(9.1.5-3)

where:
s = stress in the reinforcement, calculated on the basis of the cracked crosssection;
sr = stress in the reinforcement, calculated on the basis of the cracked crosssection for the load that causes the first crack;

9-25

1 : coefficient to take into account the bond characteristics:

1 = 1,0 for bars with good adherence;
1 = 0,5 for bars without ribs.
2 : coefficient to take into account the duration of the load or repeated loading:
2 = 1,0 for a single short-term loading;
2 = 0,5 for sustained loads or many cycles of repeated loading.

For construction elements which are only subjected to imposed internal deformations,
s may be taken equal to sr.
9.1.5.3 The mean crack spacing srm
The mean value of the crack spacing in structural elements subjected mainly to tension
and bending, may be determined by means of the following equation:
srm 50 0,25.k1.k 2 .

(9.1.5-4)

where:

= bar diameter in mm;

k1 = coefficient which takes account of the bond properties:
k1 = 0,8 for high bond bars;
k1 = 1,6 for bars with plain surface.
k2 = coefficient which takes account of the distribution of strain:
k2 = 0,5 for bending;
k2 = 1,0 for pure tension.
r = effective reinforcement ration As/Ac,eff where As is the area of reinforcement in
the effective tensile concrete section Ac,eff ; figure 9.1.5-1 presents some acceptable
choices for Ac,eff.

9-26

(a) beam
h

2,5.(h-d)
centre of gravity of the reinforcement
Ac,eff
x

(b) slab

h
c
t

smallest value of 2,5.(c ) and

2

Ac,eff

hx
3

(c) member in tension

c
Ac,eff
t

smallest value of 2,5.(c ) and

2
2

Figure 9.1.5-1
Effective tension area (typical cases)
(Figure 4.33 in NBN B15-002 (EC2:1998))

Note:
The formula for the mean crack spacing srm does not take into account the
spacing of the bars. The calculated crack width at a larger distance away from
the longitudinal bars, is under-estimated. The method thus allows to find the
characteristic values of the crack width in the direct neighbourhood (effective
tensile zone) of the longitudinal bars.

9.1.6

Practical calculation of the crack width according to EN 1992-1-1:2004; 7.3.4

9.1.6.1 Basic formula

The characteristic value of the crack width is obtained by means of the following
equation:
wk sr ,max .( sm cm )
where:
sr,max

= the maximum crack spacing

9-27

(9.1.6-1)

 sm

= the mean strain in the reinforcement under the relevant combination of

loads, including the effect of imposed deformations and taking into account
the effects of tension stiffening.

cm

= the mean strain in the concrete between cracks.

sm cm may be calculated from the expression:

s kt
sm cm

f ct ,eff

p ,eff

(1 e p ,eff )
0,6

Es

s
Es

(9.1.6-2)

where:
s

= the stress in the tension reinforcement assuming a cracked section.

= the ratio Es/Ecm

p,eff

As
Ac ,eff

with: Ac,eff = the effective area of concrete in tension surrounding the

reinforcement (where hc,ef is the lesser of 2,5.(h-d) , (h-x)/3 or
h/2; see figure 9.1.6-1).
kt

= a factor dependent on the duration of the load

kt = 0,6 for short-term loading
kt = 0,4 for long-term loading

sr,maxmay be calculated from the following expression:

s r ,max 3,4.c 0,425.k1 .k 2 .

,eff

(9.1.6-3)

where:

= the bar diameter in mm. Where a mixture of bar diameters is used in a section,
an equivalent diameter should be used;
c = the cover of the longitudinal reinforcement;
k1 = coefficient which takes account of the bond properties of the bonded
reinforcement:
k1 = 0,8 for high bond bars;
k1 = 1,6 for bars with an effectively plain surface (typically prestressing tendons).
k2 = coefficient which takes account of the distribution of strain:
9-28

k2 = 0,5 for bending;

k2 = 1,0 for pure tension.
For cases of eccentric tension or for local areas, intermediate values of k2 should be
used, which may be calculated from the relation:

k2 1 2
2. 1
where 1 is the greater and 2 is the lesser tensile strain at the boundaries of the
section considered, assessed on the basis of a cracked section;
p,eff

As
Ac ,eff

with: Ac,eff = the effective area of concrete in tension surrounding the

reinforcement, and
hc,ef = is the lesser of 2,5.(h-d), (h-x)/3 or h/2; see figure 9.1.6-1.

Figure 9.1.6-1
Effective tension area Ac,eff, for typical cases
(Figure 7.1 in EN 1992-1-1:2004)

9-29

9.2 Deflection control

9.2.1

Principle of the calculation of deflections

9.2.1.1 Introduction
The deflection of a beam made of homogeneous, continuous, elastic and isotropic
materials, loaded in bending, may be determined on the basis of the strength of
materials relationship:
d 2v 1 M

dx 2 r EI

(9.2.1-1)

The application of this formula, developed for ideal material assumptions is

questionable for reinforced concrete for which the flexural stiffness E.I is dependent of
loading and time. Indeed, the elastic modulus E is determined by the duration of the
loading, which compels immediately to distinguish between deflections for short-term
loads and for long-term loading. In addition to that, the moment of inertia I is highly
determined by the uncracked or cracked status of the concrete in the structural element;
I is large when the section is not cracked (M < Mr) and smaller in the cracked regions
(M Mr): see figure 9.2.1-1. E.I is thus determined by the load and by the value of the
cracking moment Mr. The relationship between the bending moment M and the
curvature 1/r is represented in a schematic way in figure 9.2.1-1(c).
The bending moment diagram in figure 9.2.1-1(b) and the relationship M 1/r in
figure 9.2.1-1(c) allow to determine the diagram of the local curvatures 1/r along the
beam: see figure 9.2.1-1(d).

9-30

(a)

(b)

(c)

(d)

Figure 9.2.1-1
The evolution of the curvature along a cracked beam in reinforced concrete;
(a) crack pattern; (b) bending moment diagram, with indication of the cracking moment
Mr and the SLS bending moment Mser; (c) diagram Moment M - curvature 1/r; (d)
evolution of the curvature 1/r along the beam (LAMBOTTE, 1987)
Conclusion:
The calculation of the deflections is based on the following factors:
the cracking moment Mr;
the bending stiffness (E.I)1, in uncracked state;
the bending stiffness (E.I)2, in cracked state;
the influence of the shrinkage, which will be taken into account (see further) by the
introduction of the shrinkage coefficient for deflection predictions for long-term
loading.
This leads to the conclusion that the geometry and the material properties of the
structural element have to be known and that the deflection control is essentially a SLS
verification after design. In this context, it is worthwhile to note that, due the many
uncertainties regarding the influencing factors and due to the assumptions on the basis
of the calculation models, the calculated deflection will generally present an estimation
of the real deflection; differences up to 20% are sometimes encountered in practice.

Note: simplified model for the relationship between the curvature 1/r and the
strain .
Figure 9.2.1-2 presents two cross-sections separated by the elementary distance
dx, in a structural element loaded in bending. The figure allows deducing:

9-31

r.d dx

or:
1 d

r dx
with:
d tg d

c .dx s .dx
d

Consequently:
1 c s

r
d
or with s > 0 in tension and c < 0 in compression:
1 c s

r
d

(9.2.1-2)

Figure 9.2.1-2
Auxiliary figure for the deduction of the relationship between the local curvature 1/r of
a beam in bending and the strain

9.2.1.2 Determination of the deflection in all cross-sections (determination of the

deformed axis of the beam)
The deflection v(x) along a structural element loaded in bending may be determined by
the double integration of the curvatures along the structural element, obtained by means
of the basic formula (9.2.1-1).
Because of the tension stiffening effect, the real mean curvature in the cracked regions
is situated in between the two extreme values:
the curvature in non-cracked state (1/r1);
the curvature in fully cracked state (1/r2).
By means of expression (9.2.1-2), one thus finds the expression for the mean curvature:

9-32

 cm
M
1

sm
rm EI m
d

(9.2.1-3)

The calculation of the deflection of the beam is thus performed on the basis of:
the curvatures 1/r1 (in non-cracked regions), and
the mean curvatures 1/rm (in cracked regions).

Note: deflection in one section

If one knows the evolution of the curvatures along the structural element, the
deflection vx=xo in an arbitrary cross-section can also be determined by the
application of the principle of virtual energy (the theorem of CASTIGLIANO):
l 1 M
vx x0 .
.dx
(9.2.1-4)
0 r F
i
where Fi = the generalized force associated with the deflection vx=xo: see figure
9.2.1-3; or:
vx x0

l
0

1 M *
.
.dx F * 0
i
r Fi *

in case a fictive generalized force is necessary.

The calculations are much shorter with the application of this principle method,
in the case where one has to determine the deflection in one particular crosssection only.

Figure 9.2.1-3
Application of CASTIGLIANOs theorem for the calculation of the deflection in one
particular cross-section

9.2.1.3 The cracking moment Mr

1.

The formula

9-33

The cracking moment is determined with the formula:

Mr

f r .I1
f r .W1
a1

(9.2.1-5)

where:
I1 = the moment of inertia of the uncracked fictive section (thus of the entire
concrete section Ac and .As);
a1 = the distance of the extreme fibres loaded in tension towards the centre of
gravity of the uncracked fictive section;
W1 = the bending modulus of the uncracked fictive section;
fr = the concrete tensile strength. In a first approximation, one can use the mean
tensile strength fctm.

Note:
Reminder (EN 1992-1-1:2004; table 3.1):
f ctm 0,30. f ck

2/3

for concrete classes C50/60

f ctm 2,12. ln1 cm

10

2.

for concrete classes > C50/60

(9.2.1-6)
(9.2.1-7)

Flexural stiffness for the uncracked rectangular section

In the case of the uncracked, doubly reinforced rectangular cross-section, one finds
(see figure 9.2.1-4):
b.x 3 b.(h x)3
I1

. As1.(d x) 2 . As 2 .( x d 2 ) 2
3
3

W1

I1
hx

(9.2.1-8)

(9.2.1-9)

The value of x which characterizes the position of the centre of gravity of the
uncracked fictive section, is determined from the translation equilibrium equation:

b.x 2
b.(h x) 2
. As 2 .( x d 2 )
. As1.(d x)
2
2

(9.2.1-10)

For relatively low reinforcement ratios s1 and s2 (< 0,8 %), sufficient accuracy
may be obtained with the simplified assumption I1 = bh3/12.

9-34

Figure 9.2.1-4
The uncracked doubly reinforced rectangular cross-section

9.2.1.4 The curvature 1/r1 in uncracked regions

For bending moments M smaller than the cracking moment Mr, the curvature in
uncracked state (state 1) is defined by:

M
1

r1 Ec I1

(9.2.1-11)

where:
I1 = the moment of inertia of the uncracked fictive section (thus of the entire
concrete section Ac and .As);
Ec = the modulus of elasticity of concrete.

9.2.1.5 The curvature 1/r2 in full cracked state

1.

The formula
In the assumption of a fully cracked situation, which means that the entire concrete
section is cracked (state 2), the curvature is defined by:
1
M

r2 ( E.I ) 2

(9.2.1-12)

where (E.I)2 = the flexural rigidity of the cracked section, composed of the area of
compression concrete and .As.
2.

The flexural rigidity of the cracked doubly reinforced rectangular cross-section

See figures 9.2.1-5 and 5.3.2-3 (chapter 5 SLS analysis)

9-35

Figure 9.2.1-5
The cracked doubly reinforced rectangular cross-section

In this case:
( E.I ) 2 Ec .I 2
where:
Ec = the modulus of elasticity of concrete;
I2 = the moment of inertia of the cracked fictive section with respect to the
NA (see paragraph 5.3.2.2 (4) in these course notes):

b.x 3
I2
. As1.(d x) 2 . As 2 .( x d 2 ) 2
3

(9.2.1-13)

The value of x identifying the position of the centre of gravity of the cracked
section, is determined by equation (5.3.2-7).
3.

Note: alternative reasoning for the determination of (E.I)2

See figure 9.2.1-6.

NAs2+Nc

x
h

NAs1

Figure 9.2.1-6

9-36

Auxiliary figure for the determination of the flexural stiffness (E.I)2 of the cracked
section
The behaviour of cracked sections may also be analyzed by means of the simplified
model of an equivalent homogeneous steel section, with the flexural rigidity determined
by the requirement to reach the steel stress s for the imposed bending moment M; this
is an alternative method for determining the bending stiffness of the cracked section.
The starting point is the relationship M for a homogeneous steel cross-section (for
example: the typical I-shaped cross-section of an I-beam):

M
M

.(d x)
W2 h I 2 h

(9.2.1-14)

where W2h and I2h are the bending modulus and the moment of inertia of the equivalent
homogeneous steel section (index h < homogeneous is used) which, for the imposed
bending moment M, lead to the stress s (at the distance (d - x) of the NA).
The rotation equilibrium written around the compression reinforcement of the cracked
concrete cross-section in figure 9.2.1-6, leads to:

M As1. s .z

(9.2.1-15)

with z defined by (5.3.2-8).

M in expression (9.2.1-14) is replaced by expression (9.2.1-15), which leads to:

As1. s .z
.(d x)
I 2h

or:

I 2 h As1.z.(d x)

(9.2.1-16)

This moment of inertia of the equivalent homogeneous steel cross-section has to be

multiplied by Es in order to find the bending stiffness of the equivalent steel section and
thus also the bending stiffness of the cracked concrete section.

Conclusion:
( E.I ) 2 Es . As1.z.(d x)

9-37

(9.2.1-17)

9.2.1.6 The mean curvature in the cracked regions

The moment-curvature relationships for the uncracked state and fully cracked state are
represented by the straight lines in accordance with expressions (9.2.1-11) and (9.2.112) respectively. For bending moments which are larger than the cracking moment Mr,
the real relationship is situated between the two lines because of the tension stiffening
effect: the uncracked concrete in between the cracks gives a contribution to the bending
stiffness: figure 9.2.1-7.

Figure 9.2.1-7
Moment-curvature diagram for the calculation of deflections in bending

The notion of mean curvature 1/rm is introduced for the calculation of the deflections in
the cracked regions. The mean curvature is defined by expression (9.2.1-3) starting from
the mean concrete strain cm and the mean steel strain sm over a sufficiently large length
of the structural element. In paragraph 9.1.4, the mean steel strain was already presented
under the form of expression (9.1.4-5), which is rewritten here:

sm (1 ). s1 . s 2
where the distribution factor is given by expression (9.1.4-7):

1 . sr
s2

In general, the simplifying assumption is accepted to adopt an analogous expression for

the mean concrete strain:

cm (1 ). c1 . c 2

9-38

The mean curvature is then given by:

1
1
1
(1 ). . for M > Mr
rm
r1
r2

(9.2.1-18)

Note:
As sr and s2 are calculated on the basis of the linear theory of elasticity,
expression (9.1.4-7) may also be written as:
2

Mr

1 1 . 2 .

M
0

for M M r

(9.2.1-19)

for M M r

9.2.1.7 Influence of an axial load

For structural elements which are subjected to bending combined with an axial load, the
expressions (9.2.1-11) and (9.2.1-12) for the curvatures in uncracked and fully cracked
state remain valid. The cracking moment is now also dependent on the axial load and is
defined as:

N
M r W1. f r
A1

(9.2.1-20)

where A1 represents the area of the uncracked fictive section.

9.2.2

Calculation of deflections according to Eurocode 2

Reference: EN 1992-1-1:2004 ; 7.4.3.

9.2.2.1 General
Where a calculation is necessary, the deformations should be calculated for the load
conditions which are appropriate to the purpose of the check. This implies that the
calculation method should represent the true behaviour of the structure under relevant
actions to an accuracy appropriate to the objectives of the calculation. By saying this,
the authors of the standard want to stress that deflections are influenced by many
parameters. Therefore, it is really necessary to adapt the calculation method to the level
of required detailing (if the influencing factors are not well known, detailed calculation
is not necessary). Main influencing factors are:
- shrinkage and creep;
- the tension stiffening effect caused by the concrete between cracks;
- eventual cracks caused by earlier loading;
- indirect actions such as temperature;
- the static or dynamic nature of the loading;
- the real value of the modulus of elasticity of the concrete, taking into account the
nature of granulates and the degree of hardening when the load is applied.

9-39

In order to avoid too elaborate (but useless) methods, EN 1992-1-1:2004 proposes a

simplified method which is suitable for structural elements such as beams, frames and
slabs loaded in bending. For buildings, it is acceptable to calculate deflections for the
quasi-permanent load combination where the loads are considered as long-term loads.

9.2.2.2 Simplified calculation method

The following two states determine the extreme values for the deflections of structural
elements:
the uncracked state, which is characterized by steel and concrete presenting elastic
behaviour in tension as well as in compression;
the fully cracked state, in which the tensile concrete is neglected.
Members which are not expected to be loaded above the level which would cause the
tensile strength of the concrete to be exceeded anywhere within the member, should be
considered uncracked.
Members which are expected to crack, but may not be fully cracked, will behave in a
manner intermediate between the uncracked and fully cracked conditions; for members
subjected mainly to flexure, an adequate prediction of behaviour is given by:

.11 (1 ).1

(9.2.2-1)

where:
= the deformation parameter considered which may be the deflection of the
member, a strain, a curvature or a rotation angle;
1, 11 = the values of the parameter considered, calculated for the uncracked and
fully cracked conditions respectively;
= distribution coefficient given by the expression:

1 . sr
s

(9.2.2-2)

with:
= 0 for uncracked sections;
= a coefficient taking account of the influence of the duration of the loading or of
repeated loading on the average strain:
= 1 for a single short-term loading;
= 0,5 for sustained loads or many cycles of repeated loading.
s = the stress in the tension reinforcement calculated on the basis of a cracked
section;
sr = the stress in the tension reinforcement calculated on the basis of the cracked
section under the loading conditions causing first cracking.

Note:
s/sr may be replaced by M/Mr for flexure or N/Nr for pure tension, where Mr
and Nr are the cracking moment and the cracking force respectively.
9-40

Deformations due to loading are calculated using the tensile strength and the effective
modulus of elasticity of the concrete; the knowledge of these properties is of
fundamental importance for deflection calculations:
EN 1992-1-1:2004 proposes to use fctm (see table 3.4.3-1 in these course
notes) for the determination of the best estimate of the deflection behaviour;
when there are no axial tensile stresses (for example caused by shrinkage or
thermal effects) the flexural tensile strength fctm,fl may be used. Reminder:
EN 1992-1-1:2004; 3.1.8 defines the flexural tensile strength by the
following expression (where h = total depth of the cross-section):

h
f ctm, fl max 1,6
f ctm ; f ctm
1000

EN 1992-1-1:2004 proposes to use Ecm (see table 3.4.3-1 in these course

notes);
for loads with a duration causing creep, the total deformation including creep
may be calculated by using an effective modulus of elasticity for concrete
according to the following expression:
E c ,ef

E cm
1 (, t 0 )

(9.2.2-3)

where (,t0) = the creep coefficient for stabilized deformations (at time =
) with the load is applied at time t0. EN 1992-1-1:2004; 3.1.4 proposes a
procedure for the determination of the effective creep coefficient by the
exploitation of a series of curves; however, an order of magnitude of the
creep coefficient may be determined from table 9.2.2-1.

Table 9.2.2-1
Creep coefficient (,t0) for normal concrete classes
(NBN B15-002 EC2:1998)
Fictive dimension of the cross-section 2Ac/u (in mm)
Age when load is applied,
50
150
600
50
150
600
after t0 (days)
Dry atmospheric
Humid atmospheric
conditions (inside)
conditions (outside)
(RH = 50 %)
(RH = 80 %)
1
5.5
4.6
3.7
3.6
3.2
2.9
7
3.9
3.1
2.6
2.6
2.3
2.0
28
3.0
2.5
2.0
1.9
1.7
1.5
90
2.4
2.0
1.6
1.5
1.4
1.2
365
1.8
1.5
1.2
1.1
1.0
1.0

Note:

9-41

EN 1992-1-1:2004; 7.4.3(7) specifies that the most rigorous method of assessing

deflections by means of expression (9.2.2-2) is to compute the curvatures at
frequent sections along the member and then calculate the deflection by
numerical integration. However, in most cases it is acceptable to compute the
deflection twice, assuming the whole member to be uncracked and fully cracked
in turn, and then interpolate using expression (9.2.2-2). This simplified method
is not directly applicable for cracked members subjected to high axial forces
(prestressed members).

9.2.3

Example

See annex A9.2.3.

9-42

12 Chapter 10
Non-linear and plastic analysis of indeterminate
members and frames: a brief overview
10.1 Introduction
Non-linear and plastic analysis of beams is the subject of an optional course in the
BRUFACE program, presented in the 2nd year of the Master program.
Yet, when commercial software is used for the calculation of indeterminate beams in
reinforced concrete, one is inevitably confronted with questions about the possibility of
moment redistribution.
The objectives of this chapter are limited to the following:
to present the possibility and background of bending moment redistribution
in continuous beams;
to draw attention to limitations in the use of this approach.
The chapter is composed of the following paragraphs:
the relationship between bending moment M and local curvature K in a
cross-section of a reinforced concrete member; this paragraph focuses on the
existance of plastic rotation capacity, which is at the basis of non-linear and
plastic design methods;
the elastic-plastic analysis: the paragraph is limited to the formulation of the
principle of the method;
the plastic design method: the paragraph first explains the basic principles
which lead to the distinction between the static and kinematic method; the
principle behind the two methods is explained by means of the example of
analysis of an indeterminate member. This paragraph is of use in view of the
further application (in chapter 12) of the strip method for slab analysis;
the bending moment redistribution method: the paragraph explains the
significance and advantage of the moment redistribution method by means of
a worked out example;
the identification of the ductility conditions which have to be fulfilled in
order to be sure that enough rotation capacity is available if one wants tu
apply moment redistribution or plastic methods.
The course notes in this chapter are essentially based on the books of O'BRIEN (1995),
MOY (1996), FAVRE (1990) and MOSLEY (2007).

10.2 Non-linear and plastic structural analysis: why is that possible?

The linear elastic structural analysis is based on the following assumptions:
- the assumption of BERNOULLI: flat sections remain flat;
- the assumption of linear relationship between stress and strain: = E..
For a beam with homogeneous cross-section, which is loaded in bending (figure 10.21), these assumptions lead to the linear relationship between bending moment M and

12-1

curvature K:
M = E .I .K

(10.2-1)

where:

- E = the modulus of elasticity;

- I = the moment of inertia of the cross-section;
- K = the local curvature = 1/R.
Figure 10.2-1 (c) shows the linear relationship between moment en curvature.

Figure 10.2-1
Linear elastic analysis; (a) beam loaded in bending; (b) segment of the beam loaded in
bending; (c) relationship between bending moment M and curvature K for beam with
linear elastic homogeneous material (O'BRIEN, 1995)
Reinforced concrete is not homogeneous because it is composed of steel and concrete
which have different values for the elastic modulus; however, it is possible to identify
an equivalent homogeneous concrete section with an equivalent moment of inertia.
When the bending moment M is small, the concrete is uncracked and the equivalent
moment of inertia is designated by Iu (undamaged). With the elastic modulus for
concrete Ec, expression (10.2-1) becomes:
M = Ec .I u .K

(10.2-2)

Increasing values of M lead to the cracking of the section; the moment of inertia drops
to a lower value; expression (10.2-2) is written as:

12-2

M = Ec .I c .K

(10.2-3)

where Ic = the equivalent moment of inertia of the cracked section. The relationship
between M and K is represented in figure 10.2-2. The expression (10.2-3) is valid untill
the steel reinforcement starts to yield.

Figure 10.2-2
Relationship between the bending moment M and the curvature K for a section in
reinforced concrete (O'BRIEN, 1995)
The curve M-K may be calculated for every given cross-section in reinforced concrete;
this is typically done by the calculation of some salient points:
M and K just before the appearance of the flexural crack in the cross-section;
M and K just after the appearance of the flexural crack;
M and K when concrete and/or steel start to yield;
M and K when failure is reached (normally due to the crushing of the compression
concrete);
Figure 10.2-3 shows the calculated and experimentally registered curves M-K for two
different cross-sections which are defined in figure 10.2-4, with steel fyk = 560 MPa and
concrete C60/75. The experimental curves are measured during four-point bending tests
on two beams with constant cross-section.

12-3

Moment
in function
of curvature
for
Moment
in functie
van kromming
voor beide
two different
beams
balken

225

beam 2

Mom ent (kNm )

200
175
150
125
100

beam 1

75
50
25
0
0

10

15

20

25

-6

Curvature
(10 /mm)
Kromming (10^(-6)/mm)
Beam11 Analytisch
: calculated
Balk

Beam
: calculated
Balk 22 Analytisch

Beam11Experimenteel
: experimental
Balk

Beam 22 Experimenteel
: experimental
Balk

Figure 10.2-3
Calulated and experimentally measured curves M-K for two different cross-sections (see
figure 10.2-4) (VUJOVIC, 2009)

stirrups
12 each 200 mm

(a)

stirrups
12 each 200 mm

(b)

Figure 10.2-4
Cross-sections corresponding to the curves shown in figure 10.2-3;
(a) = beam 1; (b) = beam 2; width b = 200 mm; total depth h = 500 mm;
steel fyk = 560 MPa; concrete C60/75 (VUJOVIC, 2009)
For a properly designed reinforced concrete section, the steel yields before the concrete
crushes. The yielding of the steel occurs at an applied moment My (yield); see figure
10.2-2. As steel is a ductile material, the reinforced concrete cross-section too is ductile:
beyond the yield point, the curvature increases greatly for a relatively small increase in
the applied moment. Complete failure of the section occurs when the concrete at the
extreme fibres in compression crushes. The curvature at that stage is Kult (< ultimate).
The moment-curvature relationship M-K shown in figure 10.2-2 may be idealized by the

12-4

simplified relationship shown in figure 10.2-5. Linear elastic analysis is based on the
assumption that K < Ky in all sections; non-linear analysis are based on the assumption
that there are no restrictions on K.

Figure 10.2-5
Idealized relationship between the bending moment M and the curvature K for a crosssection in reinforced concrete (O'BRIEN, 1995)
Conclusion
The exploitation of the plastic rotation capacity of reinforced concrete sections allows
for a more accurate prediction of the real failure load of statically indeterminate
structures.

10.3 Elastic-plastic analysis

The elastic-plastic method may be used for the prediction of failure of structural
members and of indeterminate members in particular. The method is based on the
assumption that curvatures in certain cross-sections are allowed to exceed the yield
curvature Ky when the member is close to collapse; this means that the formation of so
called plastic hinges is accepted before the collapse of the beam occurs. The method
comprises two steps :
- the determination of the ultimate failure load of the member, taking into account
that plastic hinges may be realized in several cross-sections before collapse of
the member (which corresponds to the formation of a mechanism);
- the verification of the available rotation capacity in the cross-sections where
plastic hinges are created. Indeed, the method is based on the assumption that
the curvature in the plastic hinges may increase beyond Ky but without
exceeding Kult. The question is if the cross-section has enough ductility or plastic
rotation capacity to allow the development of the plastic hinge to allow the beam
transforming itself into a mechanism. This is an important verification for crosssections in reinforced concrete, because (especially for deep beams) the
difference Kult - Ky is in general not very large.
The elastic-plastic analysis method is developed in the optional course Inelastic design
of concrete structures - 2nd year Master BRUFACE.

12-5

10.4 Plastic analysis

10.4.1 Introduction
This paragraph is limited to a brief introduction to explain the difference between the
different types of plastic analysis methods.
Plastic analysis is based on three conditions:
static equilibrium should be fulfilled: all external forces (reaction forces
included) that are applied to the structural member, have to be in equilibrium. In
each cross-section, the internal forces have to be in equilibrium with the external
forces. This condition is mandatory;
maximum resistance (maximum load) should not be exceeded: in each crosssection, the internal forces have to be smaller than the values corresponding to
failure. For example, the bending moment has to be smaller to Mult ; however, as
the simplified model shown in figure 10.2-5 is applied, one may also say that in
all cross-sections, the bending moment should be limited to My (M My);
failure should take place on the basis of a mechanism: failure should only occur
after the formation of a series of plastic hinges, transforming the structural
member into a mechanism.
Plastic structural analysis only considers the phase of plastic deformations and does not
take account of the history (the elastic phase) in structural behaviour. In that case, it is
impossible to satisfy to the three conditions mentioned above all at the same time. Each
plastic method thus starts from two fulfilled conditions and tries (by successive
approximations) to satisfy in the best way to the third condition. This gives rise to the
distinction between two families of plastic methods:
the static methods: these methods start from the first condition (equilibrium) and
the second condition (maximum resistance); when bending moments are
considered along structural members, the static approach starts from a so-called
statically acceptible moment distribution for which the two first conditions
are reached;
the kinematic methods: these methods start from the first condition (equilibrium)
and the third condition (mechanism). The kinematic approach starts from a socalled kinematically acceptible mechanism .
Both approaches work on iterative basis: one has to show that the assumed moment
distribution on the one hand and the adopted failure mechanism on the other hand, leads
to the ultimate load which approximates as much as possible the exact failure load.

10.4.2 Static methods

The static methods are based on the so-called "static theorem": all loads Qi that
correspond to a static acceptible moment distribution Mi, are smaller than or equal to the
exact failure load Qu (u < ultimate); see figure 10.4.2-1.

12-6

(distr)1

Distribution of moments

(distr)u

Figure 10.4.2-1
Schematic representation of the static theorem
As already mentioned above, the static methods start from the equilibrium and
resistance conditions without fulfilment of the third condition requiring that a
mechanism is realized when failure appears. Design of structural members is performed
in a first stage on the basis of the bending moment; the static approach starts with the
adoption of statically acceptible moment distribution; this is a moment distribution for
which in each cross-section:
- equilibrium is reached with the external loads on the member ;
- M My.
Figure 10.4.2-2 presents the example of a continuous beam for which several statically
acceptible moment distributions are presented. It should be noted that the condition of
equilibrium
d 2M
=q
dx 2

Is always fulfilled, whatever the diagram considered.

Figure 10.4.2-2
Continuous beam with several statically acceptible bending moment diagrams

One may observe that there are in fact an infinite number of statically acceptible
bending moment diagrams. As long as the member does not transform into a

12-7

mechanism, the load Qi (qi) determined by means of the static method is smaller than
the exact failure load Qu (qu). For this reason, the static theorem is called the lower
bound theorem , leading to a lower bound solution (NL: theorema van de
ondergrens ; FR: solution limite infrieure ; DU: unterer Grenzwert). The static method
is thus a conservative (or safe-solution) method which leads to over-estimated
predictions of the failure load. The accuracy of the prediction depends on the
appropriate choice of the statically acceptible moment distribution.

10.4.3 Kinematic methods

These methods are based on the so-called "kinematic theorem": all loads Qi that
correspond to a kinematically acceptible failure mechanism, are larger than or equal to
the exact failure load Qu (u < ultimate); see figure 10.4.3-1.

Mechanism

Figure 10.4.3-1
Kinematic theorem
The kinematic method starts from the choice of a mechanism for which the equilibrium
condition is fulfilled. In principle, an infinite number of kinematically acceptible
mechanisms may be identified, for which the value of My is not reached in all sections
except for the plastic hinges (where M = My off course); it is however important to
choose a plausible one.
As long as the condition on the maximum resistance is not fulfilled, the load Qi (qi)
obtained for the chosen mechanism, is larger than the exact failure load Qu (qu). For this
reason, the kinematic theorem is called the upper bound theorem , leading to an
upper bound solution (NL: theorema van de bovengrens ; FR: solution limite
suprieure ; DU: oberer Grenzwert). The kinematic method is thus an unsafe method
which leads to under-estimated predictions of the failure load. The accuracy of the
prediction depends on the appropriate choice of the kinematically acceptible
mechanism.

10.4.4 uniqueness of the solution

Starting from the two theorems mentioned before:

Qu , static Qu ,exact

12-8

(10.4.4-1)

Qu ,cinematic Qu ,exact

(10.4.4-2)

One may formulate the theorem of the uniqueness of the solution: if it is possible to
identify a kinematically acceptible mechanism which matches with a statically
acceptible moment distribution, the corresponding load Qi is equal to the exact failure
load Qu; see figure 10.4.4-1.

solution for the kinematic method

solution for the static method

(distrib)u

Distribution of moments
Mechanisms

Figure 10.4.4-1
Theorem of the uniqueness of the solution

The validity of the lower-bound and upper-bound theorems is demonstrated in the

course on Theory of plasticity in materials by the demonstrations of PRAGER
(KACHANOV, 2004).

10.4.5 Example: analysis of a statically indeterminate beam (FAVRE, 1990)

10.4.5.1

Definition of the problem

Given
The statically indeterminate beam AB, fixed in the end A and simply supported in the
end B, loaded by two concentrated loads Q: see figure 10.4.5-1. It is assumed that the
member is reinforced in such a way that the plastic moment M p+ = M p = constant over
the whole length of the beam (notation M for bending moments in beams).
Question
The determination of the failure load Qu by means of the static and kinematic method.

12-9

Figure 10.4.5-1
Schematic representation of the statically indeterminate beam for which the failure load
Qult has to be determined

10.4.5.2

Solution obtained by means of the static method

A statically acceptible moment distribution is chosen (conditions of equilibrium and

maximum resistance are fulfilled): see figure 10.4.5-2 (c). The chosen M-distribution
obviously differs from the elastic solution; the chosen distribution could be inspired by
by the idea that hinges could be expected in clamped sections. Figure 10.4.5-2 (d) leads
to the following conclusions:

support reaction forces:

R A = RB = Q

the bending moment to the right of section D:

l Q.l
M p = RB . =
3
3
This leads to the estimation of the failure load (Qu)i :
(Qu ) i =

3.M p
l

This is a solution obtained by the static approach; thus:

Qu

3.M p
l

12-10

(10.4.5-1)

Figure 10.4.5-2
Static method;
(a) the beam for which the ultimate load has to be determined; (b) the elastic Mdiagram; (c) first choice of a statically acceptible M-distribution;
(d) development of the bending moment to the right of section D
10.4.5.3

Solution obtained by means of the kinematic method

The procedure starts with the choice of a kinematically acceptible mechanism; two
plastic hinges are introduced in the sections where the point-loads are applied: figure
10.4.5-3. The value of the failure load Qu may be determined by means of the virtual
energy method: the work done Ae by the external forces = the work done (= energy
absorbed) Ai by the internal forces. From figure 10.4.5-3 (c), one easily identifies:
l
Ae = Q. .
3

12-11

Ai = M p . + M p .2.

By expressing:
Ae = Ai
l
Q. . = M p .3.
3

One finds an estimation for the failure load (Qu)i:

(Qu ) i =

9.M p
l

This is a solution obtained by means of the kinematic appraoch; thus:

Qu

9.M p
l

(10.4.5-2)

Figure 10.4.5-3
Kinematic method; (a) the beam for which the ultimate load has to be determined; (b)
first choice of a kinematic acceptible mechanism; (c) development of the virtual energy
12-12

method
10.4.5.4

Improvement of the static solution

The exact value of the failure load Qu should be situated in between the two by now
obtained solutions:
3.M p
l

Qu

9.M p
l

(10.4.5-3)

One may identify a better M-distribution, which is statically acceptible and which leads
to a better estimation of the failure load Qu ; see the M-distribution in figure 10.4.5-4
(b). This diagram is far more closer to the elastic M-diagram in figure 10.4.5-2 (b); in
the section where M is maximum, it is assumed that M = Mp. Figure 10.4.5-4 (c) shows
an alternative representation of the statically acceptible M-diagram. From figure 10.4.54 (c), one may find:

translation-equilibrium:
R A + RB = 2.Q

(attention: R A RB Q !)

moment-equilibrium around B :
1
2
Q. .l + Q. .l + M p R A .l = 0
3
3
or

Q.l + M p = R A .l

the support reaction forces are:

RA = Q +
RB = Q

Mp
l
Mp
l

the bending moment in section D is :

M = M p = RB .

l
3

And thus:
Mp l

.
M p = Q
l 3

12-13

Finally, the failure load (Qu)i is:

(Qu ) i =

4.M p

l
This is a solution obtained by means of the static approach; thus:
Qu

4.M p
l

(10.4.5-4)

Figuur 10.4.5-4
Static method; (a) the beam for which the failure load has to be determined; (b) refined
choice of the staically acceptible M-diagram; (c) development for the determination of
the failure load
10.4.5.5

Improvement of the kinematic solution

A better (more logical) choice of failure mechanism is presented in figure 10.4.5-5 (b).

12-14

The elastic analysis shows in what sections the largest bending moments occur and thus
where plastic hinges may appear. Figure 10.4.5-5 (c) permits to find:
2l
l
Ae = Q. . + Q. . = Q. .l
3
3
Ai = M p . + M p .3.
The expression:
Ae = Ai

Q.. .l = M p .4.
leads to the estimation of the failure load (Qu)i :
(Qu )i =

4.M p
l

This is a solution obtained by means of the kinematic approach; thus:

Qu

4.M p
l

12-15

Figure 10.4.5-5
Kinematic method; (a) the member for which the failure load has to be determined; (b) a
better choice of the kinematically acceptible mechanism; (c) development of the virtual
energy method
10.4.5.6

Uniqueness of the solution

The exact failure load is:

Qu =

4.M p
l

10.4.6 Important note

Plastic analysis of members is rather easy to perform. However, the results do not give
any information about the amount of plastic rotation in the plastic hinges prior to
failure. This explains why plastic analysis is not really indicated as method for analysis
of beams and frames in reinforced concrete, because of the limited availability of ductile
behaviour. This is completely different for slabs for which plastic methods are indeed
very useful.

10.5 Linear analysis with moment redistribution

10.5.1 Introduction

When a member has sufficient ductility (plastic rotation capacity) to achieve failure by
means of a mechanism, it is characterized by a failure load which is larger than the
failure load determined by means of the linear elastic method. Yet, elastis-plastic
analysis is not so practical for everyday design because it requires the knowledge of the
moment capacity of all members in advance of the structural analysis. This difficulty is
circumvented by the simplified method of the linear analysis with plastic moment
redistribution. This is an approximate method by which the elastic bending moment
diagram is adjusted to account for the ductile behaviour of reinforced concrete members
in bending. EN 1992-1-1:2004 allows the original elastic moment in continuous
members to be reduced by an amount of up to 30% and to redistribute it to other
sections in the member.
The amount of redistribution allowed is dependent on:
- the grade of the concrete;
- the ductility charcteristics of the reinforcement;
- the position of the neutral axis (see further).
It should be noted that bending moment redistribution
- leads to the reduction of the bending moment in one cross-section and thus to a
reduction of the reinforcement area in that cross-section ;
- also leads to the increase of the bending moment in other cross-sections (because
there should always be equilibrium between internal and external forces);

12-16

also leads to an adaption of the shear loads diagram.

10.5.2 Example showing the usefullness of bending moment redistribution

10.5.2.1

Given

A continuous beam with rectangular cross-section is represented in figure 10.5.2-1 (a).

It is assumed that two arrangements of reinforcement are available for design:
- two bars 20 mm, giving a resisting moment capacity of 254 kNm;
- twee bars 25 mm, giving a resisting moment capacity of 306 kNm.
Either pair of bars can be placed near the top of the cross-sectionto resist hog or near the
bottom to resist sag.

Figure 10.5.2-1
Continuous beam; (a) geometry and loading;
(b) elastic bending moment diagram for P = 150 kN and L = 10 m;
(c) reinforcement necessary to resist the elastic bending moments (O'BRIEN,1995)
10.5.2.2

Question

Using the plastic moment redistribution, determine the most suitable arrangement of
reinforcement, for the load P = 150 kN and L = 10 m.
It may be assumed that up to 30% redistribution is allowed.
10.5.2.3

Solution

12-17

The linear elastic analysis of the beam leads to the bending moment diagram shown in
figure 10.5.2-1(b). With P = 150 kN and L = 10 m, one finds a hogging moment at the
central support equal to 281 kNm and a sagging moment in the spans equal to 235 kNm.
To resist these elastic moments, two bars 20 are required in the interior of each span
(giving a moment capacity of 254 kNm) and two bars 25 are required over the central
support (giving a moment capacity of 306 kNm): see figure 10.5.2-1(c).
Based on the course on stability of structures, it may be said that the bending moment
diagram of the continuous member (figure 10.5.2-1(b) is constructed by superimposing
the bending moment diagrams for simply supported spans on the bending moment
diagram associated with the support moments. That is to say that the elastic bending
moment diagram is the sum of the bending moment diagrams illustrated in figure
10.5.2-2(a) and (b).

Figure 10.5.2-2
Components of bending moment diagram;
(a) elastic bending moment diagram if spans are simply supported; (b) elastic bending
moment diagram associated with the support moment; (c) bending moment diagram
associated with a reduced support moment
When a plastic hinge forms at the central support, it prevents any further increase in
moment at that point. It is now assumed in this example that the moment capacity in the
central support is limited to 254 kNm (2 bars 20); this means that a hinge is formed
already for an applied load P which is smaller than 150 kN, while for further increase
in load, the spans AB and BC behave as if they were simply supported. The new bending
moment diagram is obtained by superimposing the diagram in figure 10.5.2-2(a) on the
diagram in figure 10.5.2-2(c). The result of this 10% redistribution in support B is
shown in figure 10.5.2-3: the in-span moments have increased up to 248 kNm, but it is
observed that this new sagging moment can still be resisted with 2 20 (254 kNm).
Hence, for a redistribution of 10%, a saving in reinforcement has been made. In this
example, it is obvious that further redistribution is not useful as it would result in larger
reinforcement needed in the spans (sagging moments > 254 kNm).

Note:

12-18

Figure 10.5.2-3 shows that, on the basis of the bending moment diagram after
redistribution, the upper reinforcement is only needed in the region XX. But it
should not be forgotten that for normal loads, the elastic bending moment
diagram has to be applied, with moment equal to 0 in section Y; the upper
reinforcement is thus also necessary between the sections X and Y, when one
wants to avoid cracking. One thus has to use the envelope of the bending
moment diagrams.

Figure 10.5.2-3
Bending moment diagram before and after plastic moment redistribution

10.5.3 Usefulness of bending moment redistribution

Two strategies may be followed in order to use bending moment redistribution in a

positive way:
- for members with a rectangular cross-section, moment redistribution may be useful
in order to arrange for similar (in absolute value) sagging and hogging moments,
which results in more simple arrangements of reinforcement;
- for members with T- and L-sections, moment redistribution may be useful to keep the
hogging moments as small as possible, as the available moment capacity is much
smaller at the intermediate supports than in the span (smaller area of concrete in
compression).

10.6 Requirements for ductility: plastic rotation capacity is needed

The question arises what are the conditions that have to be verified in order to be sure
that sufficient plastic rotation capacity is available. As a reminder: the elastic-plastic
method comprises the verification by calculation. The question here is what simple
conditions have to be fulfilled in order to be sure that plastic moment redistribution may
be applied.

12-19

10.6.1 The singly reinforced cross-section

Figure 10.6.1-1 presents calculated relationship between the bending moment M and the
curvature K for a singly reinforced cross-section, in function of the reinforcement ratio
(MOY, 1996). Each curve corresponds to a particular value of the reinforcement ratio.
The curves show interesting results:
at low reinforcement percentages, the curves have a similar shape to those for
steel beams, with:
- an initial almost straight line showing rapid increase of moment with only a
small increase in curvature (this is the elastic region);
- a horizontal plateau showing a large curvature with only a small increase in
moment (this corresponds to the plastic rotation);
at higher reinforcement percentages, the maximum moment becomes
proportionally larger, but there is also a decrease in maximum curvature.
The rapid curvature increases with low reinforcement are caused by the steel yielding
before the concrete reaches its maximum strain and crushing. Such behaviour is called
under-reinforced. The large curvatures give rise to large deflections and cracks so that
there is warning of impeding failure, which eventually occurs by the concrete crushing.
At the higher percentages, failure is sudden and often explosive, with no large
deflection increase prior to failure. The concrete crushes before the steel yields. This is
called over-reinforced behaviour.
Under-reinforced sections are effectively controlled by the reinforcement and behave
like steel beams. Ultimately, the concrete fails when its maximum strain reaches a
limiting value (0,35% for normal concrete grades) which is considerably less than the
maximum possible steel strain, so that the capacity for plastic rotation is limited
compared with steel beams.
Over-reinforced sections are controlled by the brittle concrete and have no plastic
rotation capacity.
Finally, for very low percentages of reinforcement, a pronounced increase in moment as
well as curvature just before failure may be observed; this correponds to the onset of the
strain hardening in the reinforcement: see figure 10.6.1-3.
Figure 10.6.1-3(b) also shows clearly that the plastic rotation capacity decreases when
higher steel grades are used.

12-20

Figure 10.6.1-1
Calculated relationship between the bending moment M and the curvature K for a singly
reinforced cross-section, in function of the reinforcement ratio (MOY, 1996);
(a) concrete grade C25/30, steel fyk = 250 N/mm2, cross-section is under-reinforced for
ratios smaller than 5,49%; (b) concrete grade C25/30, steel fyk = 460 N/mm2, crosssection is under-reinforced for ratios smaller than 2,85%. The dashed line takes account
of the strain hardening effect in the tensile curve of the steel reinforcement

10.6.2 The doubly reinforced cross-section

In fact, all reinforced concrete beams have some compression reinforcement:

- in order to boost the load capacity of the compression concrete, or
- for technological reasons, in order to provide a framework on which to place the
shear links.
The moment-curvature relationship in a cross-section may be calculated in a similar
way as for singly reinforced sections. Figure 10.6.2-1 presents calculated relationships

12-21

between the bending moment M and the curvature K for singly and doubly reinforced
cross-sections (MOY, 1996). It is observed that adding compression reinforcement to
the cross-section leads to an increase of the plastic rotation capacity; for example: the
section with 2% tensile reinforcement and 1 % compression reinforcement shows a
rotation capacity which is very much similar to the one for a section with only 1 %
tensile reinforcement.
In an approximative way, the rotation capacity of a section with compression
reinforcement can be found by taking the difference between the percentages of tension
and compression reinforcement (equiv = ') and taking the rotation capacity of the
beam with just that percentage of tension reinforcement.

Figure 10.6.2-1
Calculated relationship between the bending moment M and the curvature K for a
doubly reinforced cross-section, in function of the reinforcement ratio (MOY, 1996);
the curve with indication 4 % corresponds to a percentage of 4 % of tensile
reinforcement only; the curve with indication 2 %, 1 % corresponds to 2 % tensile
reinforcement and 1 % compression reinforcement. The dashed line takes account of the
strain hardening effect in the tensile curve of the steel reinforcement

10.6.3 What happens if there is not enough rotation capacity?

This question is answered by means of the simple example of an indeterminate beam.

Figure 10.6.3-1(a) presents a continuous beam with two spans with length L = 5m
which are both loaded by a concentrated load P = 1600 kN. The corresponding linear
elastic bending moment diagram is presented in figure 10.6.3-1(b). It is assumed that the
maximum bending moment is equal to the failure moment: Mult = My = 1500 kNm (see
12-22

figure 10.2-5).
Figure 10.6.3-1(c) presents what happens if there is sufficient rotation capacity:
redistribution of bending moments is possible when the load on each span is increased.
The loads may be increased until three plastic hinges appear.
Figure 10.6.3-1(d) presents what happens if there is insufficient plastic rotation
capacity: brittle failure shows up at the central support and the beam is subdivided into
two simply supported statically determinate beams. For the load P = 1600 kN, one
finds the maximum moment in each span to be equal to:

P L 1600 kN .5 m
4

= 2000 kNm

This bending moment is larger than the resisting bending moment (1500 kNm) and
consequently, both spans collapse.
The two possibilities discussed above are off course extreme situations. In general, there
is allways a certain amount of plastic rotation capacity which allows moment
redistribution. But as soon as this rotation capacity is fully exploited, fatal collapse
appears.
As plastic rotation capacity is not evident for reinforced concrete beams, plastic design
with beams and frames is not a common practice.

12-23

Figure 10.6.3-1
Example of continuous beam in reinforced concrete; (a) the beam;
(b) the elastic bending moment diagram for the given loads and dimensions;
(c) if plastic rotation capacity is available, increase of the load leads to the formation of
a plastic hinge in the central support cross-section. Moment redistribution allows to
increase the loads up to 1800 kN on each span before a mechanism is formed (and
collapse appears; (d) this figure represents the case that plastic rotation capacity is
insufficient: brittle failure appears in section B; the continuous beam is transformed into
two separate beams

10.6.4 Criteria for the verification of plastic rotation capacity: practical rules

10.6.4.1

Observations

Figure 10.6.4-1 shows the evolution of the position of the NA in ULS in function of the
percentage of reinforcement for a given singly reinforced cross-section. This figure
shows that the NA gets a lower position with higher reinforcement ratios. The transition
point corresponds to the passage from under-reinforced sections to over-reinfrced
sections. EN1992-1-1:2004 formulates the condition to assure that enough ductility is
present (= the condition for having under-reinforced cross-sections) as a condition
regarding the x-value.
The transition point in figure 10.6.4-1 (passage from under-reinforced to overreinforced cross-sections) also depends of the concrete strength: it is shown by
calculation and by experimental results that a higher concrete grade leads to a small
increase of the reinforcement percentage (and to a small decrease of x/d).

12-24

Figure 10.6.4-1
Singly reinforced concrete cross-section; evolution of the position of the NA in ULS in
function of the reinfporcement percentage
10.6.4.2

Standard provisions for linear elastic analysis with limited moment

redistribution

Reference: EN 1992-1-1 :2004 ; 5.5

-

Linear analysis with limited redistribution may be applied to the analysis of

structural members for the verification in ULS, on the condition that sufficient
rotation capacity is available; in practice, this means that the member should be
under-reinforced.
The moments at ULS calculated using a linear elastic analysis may be
redistributed, provided that the resulting distribution of moments remain in
equilibrium with the applied loads. Whatever method for analysis is used, the
equilibrium condition should alwayse be satisfied.
Where moment redistribution is applied, one has to take account of its
consequences for all aspects of the design: for example, moment redistribution
leads to the adaption of the shear load diagram and thus to a revised shear load
verification.
It should be noted that a moment redistribution which differs largely from the
linear elastic bending moment diagram, increases the risk for unacceptible cracks
and disaccordance with the SLS conditions.
In continuous beams or slabs which are predominantly subjected to flexure and
which have the ratio of the lengths of adjacent spans in the range of 0,5 to 2,
redistribution of bending moments may be carried out without explicit check on
the rotation capacity, provided that the following conditions are satisfied:

12-25

0,0014 xu
.
for f ck 50 MPa
0,44 + 1,25. 0,6 +
cu 2 d

0,0014 xu
.
for f ck > 50 MPa
0,54 + 1,25. 0,6 +

cu 2

d
0,7
for steel classes B and C (high ductility)

for steel class A (normal ductility)

0,8
where:
= the ratio of the redistributed moment to the elastic bending moment before
redistribution;
xu = depth of the NA at the ULS after redistribution.
Table 10.6.4-1 presents the -conditions for all concrete grades. Figure 10.6.4-2 shows a
graphical representation of the -conditions for all concrete grades ans steel classes.

Table 10.6.4-1
Conditions concerning the moment reduction coefficient to be fulfilled when
redistribution of moments is applied
betonklasse
cu2

x
t.e.m. C50/60
0,35 %
0,44 + 1,25. u
d
xu
C55/67
0,313 %
0,54 + 1,31.
d
x
C60/75
0,288 %
0,54 + 1,36. u
d
x
C70/85
0,266 %
0,54 + 1,41. u
d
x
C80/95
0,260 %
0,54 + 1,42. u
d
x
C90/105
0,260 %
0,54 + 1,42. u
d

12-26

C50/60

C55/67

C60/75

C70/85

C80/95 & C90/105

Series6

Series7

Series8

Series9

Series10

1
normal
gewone ductility
ductiliteit

0,8

hogeductility
ductiliteit
high

0,7
0,54
0,44

x u /d

0,32/0,33/0,34/0,35

0,45

Figure 10.6.4-2
Graphical representation of the conditions concerning the moment reduction coefficient
to be fulfilled when redistribution of moments is applied
10.6.4.3

Standard provisions for plastic analysis

Reference: EN 1992-1-1 :2004 ; 5.6

-

Plastic analysis should be based either on the lower bound (static) method or on
the upper bound (kinematic) method.
Plastic analysis shall only be used for the check at ULS. The ductility of the
critical sections shall be sufficient for the expected mechanism to be formed
(plastic hinges should be possible in the critical sections in order to reach the
formation of the mechanism).
The required ductility may be deemed to be satisfied without explicit verification
if all the foillowing conditions are fulfilled:
the area of tensile reinforcement is limited such that, at any section:
- xu/d 0,25 for concrete strength classes C50/60;
- xu/d 0,15 for concrete classes C55/67;
the reinforcing steel is either class B or C;
the ratio of the moments at intermediate supports to the moments in the span
should be between 0,5 and 2.
EN 1992-1-1:2004; 5.6.3 presents a simplified procedure for the verification of
the plastic rotation capacity. The procedure prescribes the calculation of the

12-27

rotation angle of the beam (over a certain distance) and the comparison of this
with an acceptible value pl,d which may be determined by means of a nomogram
(pl,d is presented in function of xu/d, the concrete grade and the steel class).

12-28

11 Chapter 11
Slabs: linear elastic analysis
11.1 Introduction and classification of slabs
11.1.1 Importance of slabs in the context of buildings
Slabs are important load bearing structural elements in house-building and in industrial
and high-rise buildings. A very high percentage of the total volume of concrete used for
the construction of buildings is used for the realisation of slabs; see table 11.1.1-1.
Consequently, as slabs have an important impact on the total cost of the construction,
optimal design is necessary.
Table 11.1.1-1
Percentage of the total concrete volume used for the realisation of the main load bearing
structural elements in buildings (WIGHT, 2009)
Percentage of total
Structural element or load bearing member
volume of concrete
22%
Foundations and ground supported slabs
Bearing walls

4%

Columns

5%

Slabs (elevated or framed slabs)

59%

Others

10%

11.1.2 Typical terminology and classification of slabs

In general, distinction can be made between:
ground-supported slabs, also called slabs-on-ground or slabs-on-grade,
which bear directly on a compacted ground (grade) with organic top soil
removed. These slabs are typically used for foundation purposes;
elevated slabs, which rest on and are part of the structural frame of the
building; they are also called framed slabs. Elevated slabs are used as floor
slabs in buildings.
Elevated concrete floor systems can be classified as beam-supported floors and
beamless floors.
Beam-supported floor slabs are further divided into several types:
one-way solid slab; see figure 11.1.2-1(b). The load on the slab is transferred
along one direction (= the short direction). The main reinforcement is placed
along the short direction and is referred to as the primary reinforcement;

11-1

figure 11.1.2-1(a) shows that the primary reinforcement is very analogous to

the reinforcement in continuous beams;
one-way joist floor; see figure 11.1.2-1(c). The concrete floor consists of
closely spaced, narrow ribs in one direction supported on beams in the other
direction. Because the ribs are narrow and closely spaced, the floor
resembles a wood joist floor. It is therefore called a joist floor or a ribbed
floor;
two-way solid slab; see figure 11.1.2-2(a). Both directions participate in
carrying the load; this is particularly the case when the ratio of the two spans
is close to 1. Reinforcement is therefore provided in both directions as
primary reinforcement. Figures 11.1.2-2(b) and 11.1.2-2(c) show two-way
slabs without supporting beams:
o flat plate (NL: vlakke plaat; FR: plancher dalle); see figure 11.1.2-2(b).
A flat plate system allows for low floor-to-floor height; its formwork is
economical. At first sight, because the beams are concealed within the
slab thickness (see further), columns need not be arranged on a regular
grid. However, a flat plate is a two-way system; hence, the column
spacing in both directions should be approximately the same. The flat
plate is of course vulnerable for the punching shear phenomenon;
o flat slab (NL: paddestoelvloer; FR: plancher champignon); see figure
11.1.2-2(c). A flat slab is similar to a flat plate, but it has column heads,
referred to as drop panels. The primary purpose of the drop panels is to
provide greater resistance to punching shear.
two-way joist floor; see figure 11.1.2-2(d). A two-way joist floor, also called
waffle slab, consists of joists in both directions.

11-2

Figure 11.1.2-1
One-way slabs: (a) the typical scheme of the primary reinforcement in one-way slabs;
(b) one-way solid slab supported by beams; (c) one-way joist slab or ribbed slab

11-3

Figure 11.1.2-2
Two-way slabs: (a) two-way solid slab supported by beams;
(b) flat plate; (c) flat slab; (d) waffle slab

11-4

The figures 11.1.2-3 to -8 represent some practical floor systems (Belgian Building
Research Institute BBRI, 2002):
-

one-way slab composed of precast, prestressed hollow-core slabs: figure 11.1.2-3;

one-way joist slab or ribbed slab, site-cast in figure 11.1.2-4 and precast in figure
11.1.2-5;

precast floor plates (NL: breedplaatvloeren; FR: prdalles): figure 11.1.2-6;

slabs composed of ribs and hollow blocks: figure 11.1.2-7. The structural part of
the slab is composed of the ribs, which support also a series of hollow blocks. The
hollow blocks are made of clay tile or with concrete containing lightweight
aggregate. The principal advantage of these floors is the reduction in weight
achieved by removing part of the concrete below the NA;

mixed steel deck concrete floor: figure 11.1.2-8.

Figure 11.1.2-3
Slab composed of precast hollow-core elements; 1 = tie connection; 2 = plastic lid to
prevent the cast on-site concrete filling the hollow core; 3 = transversal tie; 4 = sleeves
opened from the top; 5 = cast on-site concrete; 6 = bearing strip in neoprene (BBRI,
2002)

11-5

Figure 11.1.2-4
Site-cast one-way joist slab (or ribbed slab) (BBRI, 2002)

Figure 11.1.2-5
Precast elements in (reinforced or) prestressed concrete; (1) T-beam; (2) double-T
element; (3) inverted U-beam (BBRI, 2002)

Figure 11.1.2-6
Schematic representation of a slab realized with precast floor plates; 1 = precast floor
plate; 2 = site-cast concrete or second phase concrete; 3 = lattice girder; 4 = connecting
tie (BBRI, 2002)

11-6

Figure 11.1.2-7
Slab composed of hollow filler blocks; 1 = floor covering; 2 = floor screed; 3 =
structural concrete topping; 4 = precast inverted T-beam in reinforced concrete; 5 =
hollow filler bloc in concrete or terra cotta (BBRI, 2002)

Figure 11.1.2-8
Mixed steel deck concrete floor; left: steel deck is integrated in the floor structure and is
used as formwork for the site-cast concrete; right: concrete floor screed is separated
from the steel deck by an insulation plate (BBRI, 2002)

11.1.3 Principle: slabs are calculated for distributed loads

The design of slabs is performed in principle for uniformly distributed loads only. If
concentrated loads are present, alternative systems are generaly added or build-in to
transfer the concentrated loads towards the supports, for example by means of ribs or
heavier reinforced strips integrated in the slab (see strip method). If the slab itself is
made responsible for the transfer of the concentrated loads, design ends up with rather
thick slabs and thus a less economical solution.

11.2 One-way slabs

11.2.1 Terminology
NL: in n richting dragende plaat; FR: dalle portant dans un seul sens; DE: einachsig
gespannte Platten.

11.2.2 When?
Reference: EN 1992-1-1:2004; 5.3.1(5)
11-7

A slab subjected to dominantly uniformly distributed loads may be considered to be

one-way spanning if either:

it possesses two free (= unsupported) and sensibly parallel edges, or

it is the central part of a sensibly rectangular slab supported on four edges with a
ratio of the longer to the shorter span greater than 2.
Distinction is made between isolated one-way slabs and continuous slabs.

11.2.3 Principle of the calculation

The calculation follows the analogy between one-way slabs and beams.

11.2.4 Principal (or primary) reinforcement

The principal reinforcement in one-way slabs is calculated such as for beams. Figure
11.2.4-1 represents a slab which is supported on two opposite edges (simply supported
or clamped); the principal reinforcement is calculated by considering a beam that is
parallel to the free edges, with simple supports or with fixed edges, and with unit width
(typically 1m).

11-8

Figure 11.2.4-1
The principal reinforcement in the one-way slab is calculated such as for a beam, taking
into account the maximum and minimum bending moments

11-9

The design problem may thus be formulated in the following way (see figure 11.2.4-1):

given:
width b = 1 (m);
total depth h or effective depth d: see the paragraph on the first
estimation of the depth, presented in chapter 5 of these course notes;
the design bending moment md, which is determined by the static
analysis of the beam, taking into account the span in the direction in
which the load is transferred (span a in the x-direction in figure
11.2.4-1). The notation for the design moment is not with the capital
M (such as in beam theory) but with the small letter m, because it
represents a bending moment per unit length (kNm/m) as the beam is
considered with unit width.

question:
to determine the principal reinforcement As (bottom reinforcement in the
span and eventually top reinforcement at a fixed support) per unit length in
the y-direction.

solution:
the principal reinforcement As (cm2/m) is calculated by application of the
design rules that are presented in chapter 4 (ULS) in these course notes.
Attention: in the case of the slab with four simply supported edges and with one span
much larger than the span in the other direction (a/b 2), the principal reinforcement in
the span is parallel to the short edge! This is in accordance with the principle that loads
are always transferred towards the nearest supports; the short span corresponds to the
load bearing direction. An alternative reasoning is shown in figure 11.2.4-2: both virtual
beams AA and BB present the same deflection in the middle because they are part of one
slab; the short beam BB presents the largest curvature and is thus subjected to the largest
bending moments; the reinforcement is thus to be put in the short direction.

11-10

A
B

B
A
b

B
f

A
f

Figure 11.2.4-2
Auxiliary figure to show that the bending moments are larger in the short
direction
The design calculation of the principal reinforcement is performed in ULS. Additional
verifications in SLS are performed according to the same standard prescriptions as for
beams.

Note:
ULS design of the reinforcement for slabs leads in general to lower
reinforcement ratios than for beams; this is logical because of the much larger
width. Consequently, SLS conditions are in general more restrictive than ULS
conditions and may thus be determining for the final design.

11.2.5 Secondary reinforcement

11.2.5.1 Necessity
The reasoning above, which considers a slab as a series of parallel beams without
transverse interaction, assumes in fact that the POISSON coefficient is 0. As 0, the
deformation of the cross-section of a beam (loaded in bending) is characterized by the
lateral expansion of the longitudinally compressed zone above the NA and the lateral
contraction of the longitudinally tensioned zone under the NA: see figure 11.2.5-1.
Because of the continuity with the neighbouring beams, bending moments around the xaxis appear; according to the notations for bending moments in slab theory, this new
bending moment is indicated by my; it is a moment in the y-direction! This bending
moment is rather small; moreover, the concrete below the NA is in general cracked,
which tempers even more the phenomenon of this transverse moment. This justifies that
the calculation of the primary reinforcement is performed with the assumption that

11-11

 = 0. Yet, in order to take account of the POISSON effect, a secondary reinforcement

is arranged in the transverse direction (parallel to the y-axis).
Terminology: secondary reinforcement; NL verdeelwapening; FR: armature de
rpartition.
Two additional arguments may be put forward to stress the usefulness of the secondary
reinforcement:
resistance against shrinkage in the y-direction;
the design of slabs is performed by means of the model with uniformly distributed
loads all over the slab. This is of course an idealisation of the reality. In practice,
it might be that the imposed load is only applied on a part of the slab surface, over
a limited part of the span b. Although the loads are smaller, in that case, a
particular curvature may appear in the y-direction and thus also a bending moment
in that direction.

Figure 11.2.5-1
With one-way slabs in bending, one observes also a bending moment in the transverse
direction, due to the POISSON-effect
11.2.5.2 Practical determination of the secondary reinforcement
EN 1992-1-1:2004; 9.3.1.1(2) stipulates that secondary transverse reinforcement of
minimum 20% of the principal reinforcement should be provided in one-way slabs.

Note:
- this rule is valid for one steel grade; if different steel grades are applied for
the principal and secondary reinforcement, the reinforcement areas have to
be adapted in proportion to the yield value;

11-12

the secondary reinforcement might be increased when highly concentrated

mobile loads are encountered (for example the wheels of a fork lift truck).

11.2.5.3 Justification of the 20% rule

The 20% rule wants to take account of the POISSON effect. Figure 11.2.5-1 presents a
slab with two opposite simply supported edges and load transfer in the x-direction. The
maximum bending moment mx in a strip with unit width is equal to:
mx

q.a 2
8

where a = the effective span.

The maximum compression stress in the x-direction occurs at the top side and equals:

mx
Wx

with Wx = the flexural modulus.

The corresponding strain is:

x
E

mx
E.Wx

If the strip could deform freely in the transverse direction (along the y-axis), the
corresponding strain would be equal to:

y . x
This deformation cannot take place because the strip is part of a slab; the edges of the
slab remain vertical (the slab remains flat in the y-direction), which leads to the
occurrence of a bending moment in the y-direction that has to reduce the strain y back
to 0; the bending moment is thus characterized by:
my
E.W y

y . x .

mx
E.Wx

For uncracked solid slabs: one has Wx = Wy, and thus:

m y .mx
With = 0,2 for concrete in general, one finds:
11-13

m y 0,20.mx
In reality, the concrete slab presents cracks, which changes the ratio between E.Wx and
E.Wy. The reasoning presented above is conservative in nature and the result (20%) is
acceptable for practical calculations.

11.2.6 Prescriptions and detailing of principal and secondary reinforcement

Reference: EN 1992-1-1:2004; 9.3.

11.2.6.1 General
The detailing of the main reinforcement in one-way slabs has to be in accordance with
the prescriptions for the longitudinal reinforcement in beams: minimum and maximum
percentages, lengths, anchorage, lap length, etc.
For the particular case of thick slabs where shear reinforcement is necessary,
verification of the anchorage is prescribed as well as the application of the shift rule
with a simplified shift length al = d instead of al = z.(cotg cotg )/2.
11.2.6.2 Spacing of bars
The maximum spacing of bars, smax,slabs, is determined by the following prescriptions:
for the principal reinforcement: min[3.h; 400 mm] according to EN and
min[2,5.h; 400 mm] according to ANB, where h = the total depth of the
slab;
for the secondary reinforcement: min[3,5.h; 450 mm] according to EN and
min[3.h; 450 mm] according to ANB.
In areas with concentrated loads or areas of maximum moment, the provisions regarding
smax,slabs are more severe:
for the principal reinforcement: min[2.h; 250 mm] according to EN and
min[1,5.h; 250 mm] according to ANB, where h = the total depth of the
slab;
for the secondary reinforcement: min[3.h; 400 mm] according to EN and
min[2,5.h; 400 mm] according to ANB.
All provisions regarding smax,slabs are summarized in table 11.2.6-1.
Table 11.2.6-1
Maximum spacing smax,slabs for reinforcement in slabs
Principal reinforcement
Secondary reinforcement
EN
ANB
EN
ANB
General prescriptions
3.h
2,5.h
3,5.h
3.h
400 mm
400 mm
450 mm
450 mm
11-14

Prescriptions for regions with 2.h

concentrated
loads
and 250 mm
regions
with
maximum
moment

1,5.h
250 mm

3.h
400 mm

2,5.h
400 mm

11.2.6.3 The 50% - rule for simply supported slabs

Half of the bottom reinforcement in the span should be prolonged to the edges and has
to be anchored at the support; see figure 11.2.6-1. In this way, one gets a reserve in load
bearing capacity. However, in the areas where the 50% rule is applied, attention should
be paid to the application of the provisions for maximum spacing of bars.

Figure 11.2.6-1
Principal reinforcement in a one-way slab; half of the bottom reinforcement in the span
should be prolonged to the edges
11.2.6.4 Accidental hogging moments along simply supported edges
Accidental hogging moments may arise along a simply supported (= simplified model)
edge of a slab; a wall that is build on top of the edge (see figure 11.2.6-2) limits the free
rotation capacity of the edge. However, the degree of clamping that is caused by such a
wall is very difficult to calculate. A simplified solution is thus necessary: if accidental
hogging moments exist which are not taken into account in the calculation, top
reinforcement should be arranged along the support which has to withstand minimum
25% of the largest moment in the adjacent span. The hogging reinforcement should be
provided over a distance which is minimum equal to 0,2 times the length of the adjacent
span, and the application length should start from the edge of the support.

11-15

Note:
For the extreme edge of a continuous slab with several spans, the accidental
hogging reinforcement may be limited to 15% (instead of 25%) of the moment
capacity in the adjacent span.

Accidental hogging moments also appear due to the monolithic connection of the slab
with supporting beams; they may be induced by the torsion stiffness of the peripheral
beams. From a practical point of view, the following philosophy may be followed for
the evaluation of bending moments in slabs:
for the calculation of the sagging reinforcement (the bottom reinforcement) in the
span, it is assumed that the peripheral supporting beams are without torsion
stiffness and consequently that edges are simply supported. This assumption leads
to an overestimation of the bending moments in the span;
as the supporting beams do have (a limited) torsion stiffness in reality,
(accidental) hogging moments are inevitable at the slab edges. The earlier
described simplified approach is then applied. If the slab would be totally without
hogging reinforcement, cracks would appear on the upper side of the slab in the
support regions;
the connection between the slab and the peripheral beams leads to torsion
loading of the beams by compatibility. The beams are not designed for this type
of loading (see chapter on torsion);
the existence of the accidental hogging moments has also to be taken into account
for the dimensioning of columns. Because of these moments, the eccentricity of
the normal load increases, which leads to heavier column reinforcement.

Figure 11.2.6-2
The masonry wall on top of the slab causes accidental hogging moments along the
edge which has been considered as simply supported for the span reinforcement
calculation
11.2.6.5 Detailing along a free edge
Figure 11.2.6-3 shows the position of the longitudinal and transverse reinforcement that
has to be arranged along a free (= unsupported) edge. The purpose of the additional bar
in the top corner and the links is to prevent spalling of the edge corner due to occasional
impact loads. The whole free edge of the slab is indeed characterized by the presence of
two vulnerable edge corners.

11-16

U-shaped stirrup
or link
Secondary reinforcement
Main reinforcement

Figure 11.2.6-3
Detailing of reinforcement along a free unsupported edge of a slab
(figure 9.8 in EN 1992-1-1:2004)

11.2.7 Shear reinforcement in slabs

In general, for normal dimensions and loads, slabs are designed in a way to avoid shear
reinforcement; indeed, the arrangement of stirrups in slabs is not easy in view of their
limited thickness.
With thick slabs, space is available and shear reinforcement might also be necessary!
The following prescriptions have to be applied (see also shear in beams in chapter 7 in
these course notes):
a slab in which shear reinforcement has to be provided, should have a total depth
of at least 200 mm;
minimum values of shear reinforcement have to be respected;
if VEd 1/3.VRd,max, the shear reinforcement may consist entirely of bent-up bars;
maximum longitudinal spacing of successive links (stirrups) is:
smax = 0,75.d.(1 + cotg );
maximum longitudinal spacing of bent-up bars is: smax = d;
maximum transverse spacing of shear reinforcement is: smax = 1,5.d.

Note:
In the case of flat plates and flat slabs, particularly concentrated shear
reinforcement might be necessary to prevent the punching shear phenomenon of
the slabs (see further).

11-17

11.3 Two-way slabs particular case of an isolated rectangular slab

11.3.1 Terminology - classification

NL: in twee richtingen dragende plaat; FR: dalle portant dans deux sens; dalle
armature croise, DU: zweiachsig gespannte Platten.
This paragraph 11.3 is limited to the analysis of an isolated rectangular slab with simply
supported edges or fixed (clamped) edges. This gives rise to 6 different combinations of
support conditions, which are presented in figure 11.3.1-1.
The load on the slab is uniformly distributed.

Figure 11.3.1-1
The 6 support combinations for an isolated rectangular slab; single hatched = simply
supported edge; double hatched = fixed edge

11.3.2 General principle for the calculation of slab reinforcement

Because of the more or less square shape of the isolated slab that is supported on all
edges, the load is transferred in two directions. The design of the reinforcement is
performed such as for beams with unit width, on the basis of sagging and hogging
bending moments. The main problem in this chapter is thus the determination of the
bending moments for slabs, which is in fact more a problem of structural analysis.
Different methods for the determination of the bending moments in slabs are presented
in the following paragraphs.

11-18

11.3.3 Methods for structural analysis of slabs in reinforced concrete

11.3.3.1 Experimental observations

The analysis of slabs in reinforced concrete may start from two interesting experimental
observations:
on the condition that the slab is oversized from the point of view of the concrete,
the slab fails after the appearance of a set of yield lines which lead to a
mechanism. The slab fails when all reinforcement bars that cross the yield lines,
are yielding; see figure 11.3.3-1;
a slab with a certain distribution of reinforcement bars and loaded by a uniformly
distributed load, is considered; the slab is characterized by a certain failure load. If
the reinforcement area is increased in certain positions along the yield lines, then
it is observed that the failure load also increases, even if the increase of
reinforcement is situated in these regions where the bending moments are not
maximum according to elastic analysis; this observation leads to the conclusion
that a redistribution of moments has taken place.

Figure 11.3.3-1
11-19

Failure of a slab with four fixed edges; failure is characterized by the appearance of a
typical yield line pattern; the load is applied to the bottom side of the slab; (a) top side
of the slab; (b) bottom side; (c) schematic representation of the yield line pattern (MOY,
1996)

11.3.3.2 Methods for analysis

Slabs are characterized by a large capacity for bending moment redistribution. This
explains why the standard accepts several methods for structural analysis. EN 1992-11:2004; 5 makes distinction between the following methods:
linear analysis without or with redistribution;
plastic analysis based on the kinematic method (upper bound method) or the static
method (lower bound method);
numerical methods which allow to take into account the non-linear behaviour of
materials.
Linear analysis methods are applicable for calculation in ULS as well as in SLS.
Because of their high degree of simplification, plastic methods are only used for ULS
calculations. Well known plastic methods for slab analysis are the yield line method
(which is a kinematic method) and the strip method (which is a static method); see
further in these course notes.
The present chapter is limited to the elastic calculation of slabs, with:
a short reminder of the thin plate theory (based on the equation of LAGRANGE),
which may be applied straightforward to circular and elliptical slabs and which
may be applied to rectangular slabs via the approximate solution of NAVIER;
simplified methods which are applicable to rectangular slabs (isolated slabs as
well as continuous slabs):
- method of equal deflections;
- method of MARCUS;
- tables of CZERNY.
the explanation of the principle of the elastic numerical analysis by means of the
finite element method.
The non-linear and plastic analyses are considered in a separate chapter in these course
notes.

11.3.4 Theory of elasticity of thin isotropic slabs with small deflections (theory of
KIRCHHOFF): scheme of the calculations

Preliminary remark:
Annex A11.3.4 presents an overview of the assumptions which are at the basis
of the theory of KIRCHHOFF, the definitions of the internal forces in slabs and
the development which leads to the basic slab equation (equation of
LAGRANGE). The annex also contains the explanation why the shear force
does not equal the support reaction along the slab edges.

11-20

11.3.4.1 The equation of LAGRANGE

The assumptions at the basis of the thin plate theory allow to develop a linear partial
differential equation, expressed in w(x,y) which is the deflection of the middle surface
of the slab; this is the equation of LAGRANGE (see annex A11.3.4):
4w
4w
4w q

2
x 4
x 2 y 2 y 4 D

(11.3.4-1)

with
q = the uniformly distributed load on the slab;
D = the bending stiffness of the slab defined by:

h3 E
.
12 1 2

(11.3.4-2)

where
h = the total depth of the slab;
E and are the elastic modulus and the POISSON coefficient of the material.
The solution of this differential equation, taking account of the edge conditions, leads to
the knowledge of w(x,y), on the basis of which the following effects may be determined:
all displacement components (u and v);
the strains (x, y, xy);
the stresses (x, y, xy);
and finally the internal forces mx, my, mxy, vx and vy.
The internal forces are most wanted for further calculation of reinforcement areas in
slabs in reinforced concrete.
11.3.4.2 Solutions of the equation of LAGRANGE
The exact" solution of the equation of LAGRANGE is only possible for elliptical (and
thus also circular) slabs and only for particular load distributions. A general solution of
the equation of LAGRANGE (and consequently the expressions for the internal forces)
has been presented by TIMOSHENKO (1959) for circular slabs, taking account of the
boundary conditions; the formulas are expressed in polar coordinates (, ). BROUCKE
(1969) developed a general formulation for the case of circular slabs with symmetrical
and anti-symmetrical load distribution by a judicious grouping of terms in the formulas
for stresses, strains and internal forces. On the basis of this general formulation,
BROUCKE composed tables which allow identifying closed form solutions for circular
slabs. Figure 11.3.4-1 presents in a schematic way a water tower as an example of
structure where quite a lot of circular slabs with axi-symmetric or anti-symmetric
loading conditions are met.

11-21

Figure 11.3.4-1
Schematic representation of a water tower. The structure may be decomposed into
several simple loading conditions for circular slabs
An exact solution of the differential equation of LAGRANGE is not possible for
rectangular slabs; approximate solutions may be found by means of series. The solution
of NAVIER (1823) and BARRE DE SAINT VENANT is well known for rectangular
slabs with four simply supported edges, for which double trigonometric series are used:

mx ny
w( x, y ) amn . sin
. sin

a b
m n

(11.4.3-3)

This series converges rather quickly. Generally, the solution is limited to the first term
for w and to the first three terms for the moments (3% approximation). The boundary
conditions for a simply supported edge are: w = 0 and m = 0 (m with the moment vector
along the supporting edge).
An alternative solution is developed by LEVY who replaced the double trigonometric
series of NAVIER by single series (TIMOSHENKO, 1959). The result is presented in
figure 11.3.4-2: one distinguishes the evolutions of the sagging moments in the span mx
and my, the torsional moment (or twisting moment) mxy, the shear forces vx and vy, and
the reactive forces at the supporting edges qx and qy. The internal forces are defined in
figure 11.3.4-3.

11-22

Figure 11.3.4-2
Distribution of bending moments mx and my, torsional moment (or twisting moment)
mxy, shear force vy and support reaction force qy for a uniformly loaded slab, simply
supported at all four edges; assumption: POISSON-coefficient = 0,3 (LEVY,
TIMOSHENKO, 1959)

11-23

Figure 11.3.4-3
Definition of the internal forces in slabs (including the sign convention)

11.3.4.3 Boundary conditions

It appears from figure 11.3.4-2 that shear forces and support reaction forces are not
equal; the explanation for this is found in the boundary conditions (see also annex
A11.3.4). Indeed, the torsional moment along the edge may be replaced by a series of
statically equivalent couples of vertical forces which have to be considered together
with the shear forces: see figure 11.3.4-4; the resultant forces for each edge segment dy
are thus equal to:
mxy
y

dy

Consequently, the support reaction forces (per unit length) qx and qy are equal to:

mxy
y

mxy
qy vy
x
q x vx

11-24

(11.3.4-4)

Figure 11.3.4-4
Auxiliary figure in which the torsional moment along the edge is replaced by statically
equivalent couples of forces

11.3.4.4 Conclusion
The theory of elasticity is almost not used for practical calculations of slabs. In the
rather frequently occurring case of a rectangular slab, a simple solution can be identified
on the basis of approximations of physical nature, which is shown in the following
paragraphs.

11.3.5 The method of equal deflections

11.3.5.1 The method of equal deflections general model

In the general model, the slab is replaced by two perpendicular systems of identical
strips A and B: see figure 11.3.5-1; the connections between the strips are composed of
hinges (spheres) and only centric forces Z can be transmitted in each crossing point of
the strips A with the strips B. A series of equations may be elaborated, expressing that in
all crossing points, the deflection of the strip A is equal to the deflection of the strip B.
The calculation of the slab is thus replaced by the calculation of the perpendicular grid
of strips (or beams) by means of a huge system of equations in which the contact forces
Z are unknown. Finally, each strip A (or B) is then calculated as a beam, supported at
11-25

both ends and loaded by a number of concentrated loads Z. In the case of a large number
of loads, they may be replaced by a uniformly distributed load qa along a and another
distributed load qb along b.
In spite of its overal nature, the general model is not used for parctical calculations
because of the large number of equations.

Figure 11.3.5-1
General model for the method of equal deflections: the slabs is replaced by two
perpendicular layers of beams with hinges in the connection points

11.3.5.2 The method of equal deflections simplified model

Terminology: NL = methode van de gelijke doorbuigingen; FR = mthode de la double
flexion.
1.
Slab with four simply supported edges
A simplified model consists in the replacement of the slab by two perpendicular beams
A and B (see figure 11.3.5-2):
centrally positioned in the slab;
with unit width;
loaded by the uniformly distributed linear loads qa and qb.

11-26

Figure 11.3.5-2
Simplified method of equal deflections: replacement of the slab by a simple model
composed of two perpendicular, centrally located beams
Equivalence of forces:

q .dx.dy q .dx.dy q.a.b

a

leads to:
qa qb q

(11.3.5-1)

Moreover, the ratio between qa and qb is not arbitrary, because of the complementary
condition that the deflections of both the strips A and B should be equal in the centre of
the slab (point 0 in figure 11.3.5-2).
Taking into account that:
w A, 0

5 qa .a 4
.
384 E.I

wB , 0

5 qb .b 4
.
384 E.I

the condition wA, 0 wB ,0 leads to:

qa .a 4 qb .b 4

11-27

(11.3.5-2)

Introduction of expression (11.3.5-1) into expression (11.3.5-2), gives:

b4
.q
4
4
a b

a4
qb 4
.q
a b 4
qa

Figure 11.3.5-3 represents the ratio

(11.3.5-3)

qa
a
in function of the ratio of spans
(indicated by
q
b

the label support condition 1):

qa

1
a
1
b

f ( a / b)

(11.3.5-4)

Values of a/b

Figure 11.3.5-3
Abacus for the identification of the load transfer in rectangular slabs according to the
method of equal deflections; the curves correspond to 6 different support combinations;
a hatched edge in the figure corresponds to a fixed edge of the slab
Some observations can be made for the curve associated with the support combination 1
in figure 11.3.5-3:
11-28

as soon as the slab evolves towards a rectangular shape (with a/b < 1), qa
corresponds to a larger part of q and qb is reduced in value;
for a/b = 0,5 one finds qa = 0,94.q and qb = 0,06.q. The minimum secondary
reinforcement (20% of the primary reinforcement) which is prescribed by EN
1992-1-1:2004, thus leads to a conservative solution;
for the particular case of the square slab, one finds:
a = b ; qa =qb = q/2
mx

qa .a 2
8

; my

qb .b 2
8

; mx m y

q.a 2
16

2. Slab with other support combinations

The simplified method of equal deflections that was applied for the simply supported
slab, may also be applied for slabs with one of the five other support combinations.
Therefore, it is again expressed that the deflections wA,0 and wB,0 in the centre of the slab
(position 0 in figure 11.3.5-4) are the same for both beams A and B:
wA,0 = wB,0.
Consequently:
w A, 0

a qa a 4
.

384 E I a

wB , 0

b qb b 4

. .
384 E I b

Introduction of Ia = Ib and qb = q - qa , leads to:

qa
b .b 4

q a .a 4 b .b 4

a4
1 a . 4
b b

(11.3.5-5)

where the coefficients a and b are characteristic for each support combination. This
leads to the identification of a curve that presents qa/q in function of the ration a/b for
each support combination; see figure 11.3.5-3. These curves are helpful in identifying
the distribution of q among qa and qb.

11-29

Figure 11.3.5-4
The strength of materials formula for the calculation of the deflection in the central
point 0 of the strip A
3.
Conclusion: determination of the bending moments
With qa and qb determined, the sagging moments per unit length (strips are considered
with unit width!) mx and my, as well as the hogging moments per unit length mX and/or
mY, may be determined by means of the formulas presented in figure 11.2.4-1 and
taking into account the conventional notations in figure 11.3.5-5.

11-30

Figure 11.3.5-5
Conventional notations used for the indication of sagging moments in the span and
hogging moments along fixed edges of a rectangular slab

4.
-

Critical remarks concerning the simplified method of equal deflections

The simplified method of equal deflections does not lead to an accurate
determination of the bending moments, among other reasons because the equality
of the deflections is only considered in the centre of the slab, without taking
account of the effect of the connection with the adjacent strips. Each strip in the xdirection "helps" each strip in the y-direction, not only by offering a support
reaction force, but also by the provision of torsional stiffness: see figure 11.3.5-6.
This additional stiffness is not considered in the simplified method of equal
deflections which leads to the over-estimation of the bending moments. For
example, in the case of a square slab, the simplified method leads to the sagging
moment per unit length equal to:
mx m y

q.a 2
16

but when the torsional stiffness is taken into account (solution of the equation of
LAGRANGE by the series method of LEVY), one finds:
mx m y

11-31

q.a 2
21

The elastic deformation of a simply supported rectangular slab shows

maximum curvatures and bending moments in the middle of the slab.
However, for practical reasons, the reinforcement that is calculated for the
maximum bending moments mx,max and my,max, is provided in a certain area
which is not limited to the pure middle of the slab. That is also a reason
why the slab will not directly collapse when the design load would be
exceeded by accident. Indeed, even if the yield strength in the steel
reinforcement would be reached in a narrow strip in the middle of the slab,
the deflection of this strip cannot increase unlimitedly because it is
connected to the adjacent strips which have an equal area of reinforcement
but with less loading. This additional reserve in strength is also not taken
into account in the simplified method of equal deflections.

bending of a strip

torsion of a transverse strip

Figure 11.3.5-6
The twisting effect in a deformed slab (FAVRE, 1990)

11.3.6 The abaci of MARCUS

11.3.6.1 Principle
In order to take account of the criticisms formulated in the preceding paragraph,
MARCUS determined reduction coefficients x and y which are to be applied to the
11-32

bending moments mx and my that are obtained with the simplified method of equal
deflections, in order to obtain a better estimation of the moments in the slab. MARCUS
considered rectangular slabs with the six support combinations; he calculated the
bending moments
with the simplified method of equal deflections on the one hand, and
by solution of the equation of LAGRANGE by means of series, on the
other hand.
By comparison of both results, MARCUS identified the following formulas for his
reduction coefficients:
2
5 a mx
x 1 . .

6 b Mx

2
5 b my
y 1 . .
6 a M y

(11.3.6-1)

where:

mx and my are the bending moments obtained with the simplified method of equal
deflections;

Mx

q.a 2
8

My

q.b 2
8

11.3.6.2 Slab with four simply supported edges

The simplified method of equal deflections leads to the following result:
mx

qa

qa .a 2
8

b4
a
.q f ( ).q
4
4
a b
b

By introduction of this into expression (11.3.6-1), x can also be written as:

a
5 a 2 .b 2
5
f1 ( )
1
4
4
2
2
b
6 a b
a
b
6. 2 2
a
b

x 1 .

Consequently:

11-33

m x ,max x .(m x ,max equal deflections )

q a .a 2
8
a2
a
a
f 1 ( ). f ( ).q.
8
b
b
x.

MARCUS has written this in a more exploitable way:

m x ,max K x .q.a 2

(11.3.6-2)

m y ,max K y .q.b 2

(11.3.6-3)

and:

a
.
b
Reference books present the Kx and Ky values in table format (which asks for
interpolation efforts) or in graphical format (which is largely sufficient for the necessary
accuracy level in slab design). Figure 11.3.6-1 presents the Kx and Ky values for the 1st
support combination.
where Kx and Ky are functions of the ratio of the slab dimensions

Note:
It should be stressed that the abaci of MARCUS also lead to approximate values
of the bending moments in slabs. For example, by application of the abaci of
MARCUS, one finds for a square slab the following results:
mx

q.a 2
16

Mx

q.a 2
8

5
6

x 1 .1.

8
0,583
16

mx ,max 0,583.(mx ,max gelijke doorbuiging )

0,583.

q.a 2
16

q.a 2
27,4

11-34

This has to be compared with the elastic solution

q.a 2
!
21

11.3.6.3 Slabs with other support combinations

MARCUS has developed tables and abaci for all six support combinations. For each
support combination, the abacus presents three or four curves in case clamping
moments (hogging moments) mX and/or mY have to be considered next to the sagging
moments mx and my. The abaci of MARCUS are presented in the series of figures
11.3.6-1 to 11.3.6-6, using the conventional notations presented in figure 11.3.5-5.

Figure 11.3.6-1 Support combination 1

Abacus for the determination of the bending moments in rectangular slabs according to
the approximation of MARCUS
Support combination 1: rectangular slab with four simply supported edges

11-35

Figure 11.3.6-2 Support combination 2

Abacus for the determination of the bending moments in rectangular slabs according to
the approximation of MARCUS
Support combination 2: rectangular slab with one fixed edge

Figure 11.3.6-3 Support combination 3

Abacus for the determination of the bending moments in rectangular slabs according to
the approximation of MARCUS
Support combination 3: rectangular slab with two opposite fixed edges

11-36

Figure 11.3.6-4 Support combination 4

Abacus for the determination of the bending moments in rectangular slabs according to
the approximation of MARCUS
Support combination 4: rectangular slab with two adjacent fixed edges

Figure 11.3.6-5 Support combination 5

11-37

Abacus for the determination of the bending moments in rectangular slabs according to
the approximation of MARCUS
Support combination 5: rectangular slab with three fixed edges

Figure 11.3.6-6 Support combination 6

Abacus for the determination of the bending moments in rectangular slabs according to
the approximation of MARCUS
Support combination 6: rectangular slab with four fixed edges

11.3.7 Other references for determination of bending moments in rectangular slabs

Some alternative methods, which lead to very similar results, are available:

the tables and graphs of CZERNY; some examples are presented in annex
A11.3.7;

the tables of BARES (1973);

the formulas presented by ROARK (1954).

11.3.8 The reinforcement in the isolated slab

The area of reinforcement is calculated according to the ULS design method for all
hogging and sagging moment on the basis of the cross-section with unit width (1m).
Furthermore, the SLS provisions as well as the rules for detailing of reinforcement have
to be applied concerning minimum reinforcement, bar diameters and spacing, in the
same way as for beams and one-way slabs.
Curtailment of reinforcement bars may be considered in function of the evolution of
bending moments, but still taking into account:

11-38

the prescriptions regarding anchorage lengths (the same rules as for the
longitudinal reinforcement of beams);
the prescription saying that half of the maximum sagging moment reinforcement
should be prolonged up to the edges of the slab.

Finally, one may also take advantage of the positive effect of the supporting edges
parallel to the considered direction of the reinforcement. In general, the reinforcement in
the x-direction is reduced with 50% over a width of (0,2.a) along both edges //x; the
width of the strip in which the reduction is applied is taken equal to 0,2.(the smallest
span of the slab). The same procedure may be applied to the reinforcement in the ydirection, which is also reduced with 50% in the two peripheral strips (two edges //y!)
with width equal to (0,2.a). This so called 50% rule is illustrated in figure 11.3.8-1.
It should be noted that practical considerations in the workshop and on site often lead to
the adoption of simplified reinforcement schemes in slabs.

Figure 11.3.8-1
50% rule for the sagging moment reinforcement along the edges of the isolated slab

11.3.9 Torsional reinforcement in a corner defined by two simply supported edges

11.3.9.1 Necessity of additional reinforcement
The slab in figure 11.3.9-1, with four simply supported edges, is subdivided into a grid
of strips. The reasoning focuses on the prism that is part of the strips A0 and B0. The
actions (twisting couples) applied by the adjacent strips A1, A2, B1 and B2 onto the
prism, cause torsional deformation of the prism:
11-39

the strip B2 drags the strip B0 downwards while the strip B1 pushes the strip B0
upwards;
the strip A2 drags the strip A0 downwards while the strip A1 pushes the strip A0
upwards.

The diagonals 2-4 and 5-7 of the considered volume element are in compression while
the diagonals 1-3 and 6-8 are in tension; extra tension effects require extra
reinforcement which is called torsional reinforcement.

Figure 11.3.9-1
Auxiliary figure to explain the necessity of additional reinforcement in the corner
defined by two simply supported edges

11.3.9.2 Explanation on the basis of the theory of slabs

The need for additional reinforcement in corners defined by simply supported edges is
explained in the preceding paragraph starting from the existence of the connection
moments between the different strips in a slab. These twisting moments are maximum
in the region where there is a large variation of the deflection, thus close to the supports.
The phenomenon may be formulated in another way, by consideration of the support
reaction forces of the slab. Figure 11.3.9-2 shows a corner which is defined by two
simply supported edges. Torsional moments which find their origin in the shear stresses
xy and yx (theory of plates) are acting along the edges. The torsional moments may be
replaced by statically equivalent couples of vertical forces (see also figure 11.3.4-4).
The torsional moment along the elementary length dy may be replaced by a couple of
forces with values equal to:
m xy .dy
dy

mxy

11-40

For the adjacent surface elements, the couple is equal to:

mxy

mxy
y

dy

The forces mxy compensate each other at the edge between the adjacent surface
mxy
dy . The value of the
elements; the only subsisting forces are the differential forces
y
mxy
.
force per unit length is equal to
y
In this way, a continuous distribution of additional shear forces is observed along the
edges, which have to be added to the already existing shear forces vx and vy. However, it
is also observed that there is no compensation of mxy in the corner; moreover, an
additional force mxy is identified in the corner for the other side parallel to the x-axis. A
concentrated load with intensity 2 mxy is thus observed in the corner; this is illustrated in
figure 11.3.9-3 for the case of the square simply supported slab (FAVRE, 1990). If the
slab would simply be put on the edges without further connections, the corners of the
slab would deform upwards due to the application of the distributed load on the slab:
see figure 11.3.9-4. Pushing down the corner will thus give rise to cracks in the upper
side of the slab, such as is illustrated in figure 11.3.9-5.

Figure 11.3.9-2
Corner of a simply supported slab: replacement of the torsional moment by couples of
vertical forces

11-41

Figure 11.3.9-3
Shear forces and support reaction forces along the edges of a square simply supported
slab; special attention is given to the four reaction forces in the corners (FAVRE, 1990)

Figure 11.3.9-4
Application of a distributed load on the slab leads to the upward curling of the corners
of the simply supported slab (BBRI, 2002)

11-42

Figure 11.3.9-5
Pushing down the corners of a simply supported slab may lead to a typical crack pattern
on the upper side of the slab

11.3.9.3 Torsional reinforcement is not necessary in corners with fixed edges and free
edges
In the case of a slab with fixed edges, the boundary conditions for the edge parallel to
the y-axis are:
w=0

and

w
0
x

w
2w
has also to be 0 because the tangent
is horizontal in each point of
xy
x
this edge. This is also the case for an edge parallel to the x-axis. Knowing that one of
the relationships developed in the theory of plates (KIRCHHOFF) is the following:
The term

mxy D.(1 ).

2w
xy

(11.3.9-1)

one may conclude that mxy = 0 along fixed edges. There is no isolated concentrated
reaction force in the corner. The support reaction forces are equal to the shear forces;
this is illustrated in figure 11.3.9-6 for a square slab (FAVRE, 1990).
In the case of corners defined by free (non-supported) edges, the boundary conditions
are: m = 0 and support reaction forces = 0. There is no isolated reaction force in the
corner and mxy = 0; consequently: no torsional reinforcement necessary!

11-43

Figure 11.3.9-6
Support reaction forces for a square slab with fixed edges: there are no concentrated
forces in the corners (FAVRE, 1990)

11.3.9.4 Technological aspects related to the torsional reinforcement in the corners of

simply supported slabs
From a theoretical point of view, the torsional reinforcement should be arranged in the
corners defined by simply supported edges, on both the top and bottom sides, in square
regions with side equal to 0,3.(the smallest span of the slab). Figure 11.3.9-7 shows the
theoretical reinforcement scheme. The density of the torsional reinforcement should be
the same as for the largest reinforcement in the span. However, although the positioning
of the reinforcement bars in the direction of the tensile stresses corresponds to the most
efficient use of the materials, it is undoubtedly not the most economical solution from
the point of view of practice on the construction site. The theoretical scheme is therefore
practically only applied in the case of thick slabs (total depth more than 1m in order to
have enough place for the positioning of all the different layers of reinforcement!) such
as for example foundation slabs.
Reinforcement meshes are more commonly used in slabs with normal dimensions. An
orthogonal grid is put in every corner, with in each direction the density of the largest
span reinforcement: see figure 11.3.9-8. This solution is normally adopted at the upper
side of the slab; at the bottom side however, in order to reduce the number of layers, it
is often preferred to simply extend the full span reinforcement until the edges (thus no
application of the 50% rule).

Notes:
- for simplifying reasons, it is sometimes preferred to abandon also the
arrangement of additional reinforcement grids at the upper side of the slab; in
that case, it is recommended to arrange for upper reinforcement anyway to
take up accidental hogging moments and to neglect the use of the reduction
coefficients of MARCUS to the span moments;
11-44

- in principle, there is no additional reinforcement needed in corners defined by

two fixed edges. However, in such case, one has to be sure about the real
nature of the clamping!

upper side

bottom side

Figure 11.3.9-7
Theoretical scheme for the torsional reinforcement in corners defined by simply
supported edges; (a) upper side; (b) bottom side

0,3.a

0,3.a

x
ax
ax

ax
ay

11-45

Figure 11.3.9-8
Torsional reinforcement by means of orthogonal meshes: the density of the
reinforcement in both directions is the one of the largest span reinforcement (ax in this
example)

11.4 Shear in slabs

EN 1992-1-1:2004 does not make a distinction between beams and slabs regarding the
verification method for shear. The specific rules that are applicable for one-way slabs
are also valid for two-way slabs (recapitulation of paragraph 11.2.7 in these course
notes):
a slab in which shear reinforcement has to be provided, should have a total depth
of at least 200 mm;
minimum values of shear reinforcement have to be respected;
if VEd 1/3.VRd,max, the shear reinforcement may consist entirely of bent-up bars;
maximum longitudinal spacing of successive links (stirrups) is:
smax = 0,75.d.(1 + cotg );
maximum longitudinal spacing of bent-up bars is: smax = d;
maximum transverse spacing of shear reinforcement is: smax = 1,5.d.
The following prescription (ANB) has to be added to the list: according to NBN EN
1992-1-1 ANB:2009; 6.2.2(1) the value of VRdc may be multiplied by 1,25 for slabs
where the edges are supported in a continuous way.

11.5 Overview of the design process for slabs

Given:

structural concept: one- or two-way slab, support conditions;

functional dimensions: span;

environmental conditions: exposure class, environment class;

imposed loads;

concrete class and steel grade.

Question:

slab design: thickness, reinforcement.

Solution scheme:
1.
Effective span

EN 1992-1-1:2004; 5.3.2.2; figure 2.5.2-1 in these course notes

2.

Thickness of the slab

useful preliminary prescriptions

The previous version of EC2 recommended a minimum thickness of 50 mm,
but this recommendation is not mentioned anymore in the actual version.
11-46

3.

Another useful indication is that a slab in which shear reinforcement has to

be provided, should have a total depth of at least 200 mm.
prescriptions about deflections (SLS)
EN 1992-1-1:2004; 7.4; formulas (5.5.2-1) and (5.5.2-2) and table 5.5.2-2 in
these course notes.
concrete cover

transmission of bond forces: EN 1992-1-1:2004; 4.4.1.2; paragraph

5.6.3.4 in these course notes; this leads to the CHOICE of the bar
diameter;

environmental conditions: NBN EN 1992-1-1 ANB:2009; tables

5.6.3-1 and 5.6.3-2 in these course notes;

allowance for deviation;

fire safety (if applicable)

Conclusion: the verifications mentioned above lead to the following
properties:
effective depth d
nominal cover (which has to be mentioned on drawings)
bar diameter
total depth h

Loads (combinations of loads): chapter 2 in these course notes

characteristic values

self-weight (which is determined by the total depth)

imposed loads

SLS (quasi-permanent combination)

design values (ULS)

4.

Internal forces in ULS: chapter 2 in these course notes

5.

Main reinforcement Asx

(a) theoretical necessary area for ULS
Note: verification of

max x/d : paragraph 4.2.5 in these course notes;

limit value x/d : need for compression reinforcement?

(b) verification of Asx for SLS

limitation of stresses:

provisions for minimum reinforcement: EN 1992-1-1:2004;

9.2.1.1; paragraph 6.1.1 in these course notes;

verification of the stresses in concrete and steel; paragraph 5.3 in

these course notes;

11-47

(c)

verification of SLS crack opening: verification of steel stress, bar

diameter and/or spacing; paragraph 5.4 in these course notes;

verification SLS deflections: paragraph 5.5 in these course notes.

Conclusion for main reinforcement Asx:

choice of number and

verification of spacing: EN 1992-1-1:2004; 9.3.; paragraph 11.2.6.2 in

these course notes.

6.

2nd main reinforcement or secondary reinforcement

7.

Torsional reinforcement

8.

Verification for shear

9.

Detailing of reinforcement (chapter 6 in these course notes), including:

choice type of anchorage

anchorage length

lap length

11.6 Continuous slabs

Classical resident and office buildings often contain continuous floor slabs, which are
continuous over the whole floor surface and are supported by a grid of beams.
In principle, the abaci of MARCUS, which are developed for isolated plates, cannot be
used for the determination of the design moments (sagging and hogging) in continuous
slabs. However, the objective of the following paragraphs is to show that the abaci of
MARCUS indeed can be used in the context of continuous slabs, on the condition that
the dimensions of each slab area between the supporting beams are not too different
(ratio a/b larger than 0,8).

11.6.1 Sagging moments

A trick is used in order to be able to apply the abaci of MARCUS. The total distributed
load q is subdivided into two parts:
q' comprises the permanent load g and half of the imposed loads p/2 (this is in fact
the quasi-permanent part of the imposed loads; example: the desks and furniture
in a classroom);
q'' comprises the second half of the imposed loads p/2 (this is in fact the variable
part of the imposed loads; example: the self-weight of the students in a
classroom).
Consequently:

11-48

1 1

q q ' q" g p p
2 2

(11.6.1-1)

This paragraph presents the school example of a continuous slab with four simply
supported peripheral edges and two intermediate supports: see figure 11.6.1-1. The
objective of the analysis is to determine the sagging moment in the central part of the
slab, along the longest span.
The slab may be replaced by a continuous beam (figure (a)). The study of the influence
line of the sagging moment in section K shows that the maximum sagging moment mK
is obtained with maximum load on the central part and minimum load on the peripheral
parts of the slab (figure (b)). This loading case may be considered as the combination of
the two loading cases in figure (c). The two continuous slabs in (c) may be replaced by
two series of isolated slabs, because the combination of the boundary conditions with
the nature of the loading leads to the same elastic deformations for (c) and (d) (and thus
also the same curvatures and the same bending moments).
The bending moment mK for the continuous slab may thus be calculated as the sum of
two moments in K which may be determined by means of the abaci of MARCUS:
the abacus for the slab with 2 opposite fixed edges (figure 11.3.6-3
support combination 3), and
the abacus for the slab with four simply supported edges (figure 11.3.6-1
support combination 1):
2

max mKx K x 3 .q'.l x K x1.q".l x

max mKy K y 3 .q '.l y K y1.q".l y

In the same way, one finds:
min mKx K x 3 .q '.l x K x1.q".l x

min mKy K y 3 .q'.l y K y1.q".l y

A similar reasoning may be applied to determine the sagging moment in the left part of
the slab in figure 11.6.1-1:
2

max mx K x 2 .q '.l x K x1.q".l x

max m y K y 2 .q '.l y K y1.q".l y

11-49

Conclusion: the sagging moment in one part of the continuous slab is obtained by taking
the sum of:
-

K .q '.l 2 ; this is the sagging moment due to q' in the part with at least 1 fixed edge.
The factor K is determined by means of the abaci of MARCUS for the support
combinations 2 to 6;

K1.q".l 2 ; this is the sagging moment due to q'' in the considered part of the slab,
but with all edges simply supported; the factor K1 is determined by means of the
abacus of MARCUS for support combination 1.

11-50

simply supported edge

Infl mK

Figure 11.6.1-1
Determination of sagging moments in continuous slabs by means of the abaci of
MARCUS; (a) schematic representation of the slab as a continuous member; (b) loading

11-51

for max m K ; (c) subdivision into two loading cases applied to the continuous slab; (d)
equivalence with isolated slabs

11.6.2 Hogging moments

The continuous slab in figure 11.6.2-1 is considered again. The question is now to
determine the maximum hogging moment at an intermediate support.
The study of the influence line of the hogging moment in the intermediate support
shows that the adjacent spans have to be loaded with the maximum load (figure (b));
this leads to the subdivision of the load into the two cases in figure (c). The analogy
between the continuous slab and the series of three isolated slabs is straightforward for
the load q'. For load q'' however, it is not possible to identify a subdivision in isolated
slabs that leads to equivalent bending moments. An approximate solution is found by
choosing the application of the full load q onto the series of isolated slabs with fixed
edges at the intermediate supports. The approximate value of the hogging moment is
thus determined as:
max mx K X .q.l x

max m y K Y .q.l y

where the factors KX and KY can be determined with the abaci of MARCUS for the
support combinations 2 to 6. The following cases are possible (assuming all parts of the
slab with identical dimensions and edge conditions - figure 11.6.2-2):
1.
two adjacent spans:
both parts correspond to the support combinations 2 (or 4 or 5), and thus:
m X K X 2 .q.l x

or ( K X 4 or K X 5 )

2.

three adjacent spans:

the left part corresponds to the support combination 2 (or 4 or 5); the right part
corresponds to the support combination 3 (or 5 or 6).
It is recommended to:
- reduce ( 0,8) the hogging moment obtained for the part with 1 fixed edge;
- increase ( 1,2) the hogging moment obtained for the part with 2 opposite fixed
edges (along the considered direction).
Finally, the largest among these two hogging moments is selected.

3.

four adjacent spans:

- for m X 1 : see the 2nd case discussed above;

11-52

- for m X 2 : the two adjacent parts are identical (support combination 3 (or 5 or 6)).
For adjacent spans with different dimensions, one always selects the largest value
of the hogging moment.
simply supported edge

Infl mK

Figure 11.6.2-1

11-53

Determination of hogging moments in continuous slabs by means of the abaci of

MARCUS; (a) schematic representation of the slab as a continuous beam; (b) loading
for max mK ; (c) subdivision into two loading cases applied to the continuous slab; (d)
equivalence with isolated slabs

Figure 11.6.2-2
Determination of the hogging moments in continuous slabs by means of the abaci of
MARCUS; analysis of different cases with different numbers of adjacent spans

11.6.3 Particular cases

When the continuous slab is fully surrounded by stiff peripheral beams in

reinforced concrete, the different spans of the slab may be considered with the
support combination 6.

When the parts of the continuous slab have very different dimensions, each part
should be calculated as an isolated slab:

in order to determine the hogging moments, the isolated slab is considered

with fixed edges on the side of the adjacent spans (support combinations 2
to 6). For each intermediate support, the largest value of the two hogging
moments is finally selected;

11-54

in order to determine the sagging moments, each isolated slab is considered

with simple supports (support combination 1). This corresponds to the idea
of considering the continuous slab cracked along all intermediate supports.
This leads to an overestimation of the sagging moments.

11.7 Flat plates and flat slabs

11.7.1 Terminology

Distinction is made between (see figure 11.7.1-1):

a two-way slab supported by a grid of beams; NL: in 2 richtingen dragende plaat;

FR: dale portant dans deux directions ou dalle armatures croises; DU:
zweiachsig gespannte Vollplatte;

a flat plate ; NL: vlakke plaatvloer; FR: plancher-dalle, sans chapiteaux; DU:
Flachdecke;

a flat slab or mushroom floor; NL: paddestoelvloer; FR: plancher-champignon,

avec chapiteaux; DU : Pilzdecke.

Figure 11.7.1-1
Various column-slab systems used in reinforced concrete structures:
(a) beam-supported two-way solid slab, (b) flat slab, (c) flat plate

With a flat plate, the load is transferred by the slab to the columns in a direct way,
without the intervention of beams. A critical point with this slab system is the
application of concentrated support reaction forces, which might give rise to a punching
shear problem. This is the reason why, in general, flat plates have a somewhat higher
total depth than beam-supported two-way slabs. With large loads applied to the slab,

11-55

column heads (also called drop panels) might be necessary as additional measure to
avoid punching shear.
The connections between the plate and the columns are stiff connections. Flat plates and
flat slabs are of interest when the load is uniformly distributed, but arrangements are
possible to transfer concentrated loads. The repetitive nature of this type of floor as well
as the flat aspect of the lower side of the floor, offers the advantage of formwork
economy. Flat plates and flat slabs are easy to recognize in plans and drawings because
of their repetitive nature: see figure 11.7.1-2.

Figure 11.7.1-2
Ground plan of an office building, showing the regular scheme of columns and
rectangular column heads

11.7.2 Simplified design rules (rules of thumb)

Several reference books present general rules of thumb which allow quick pre-design of
flat slabs. One example of such rules of thumb is presented in figure 11.7.2-1 which
focuses on a general idea of dimensions.

11-56

Figure 11.7.2-1
Simplified pre-design rules for flat slabs (VAN WAMBEKE, 1984)
A manual edited by the Institution of Structural Engineers (ISE) presents formulas
which allow preliminary estimation of the total depth of a flat slab in order to avoid
punching shear:
1250.q.area supported by column
0,6 N/mm 2
column perimeter 9.h .d

(11.7.2-1)

1250.q.area supported by column

0,9. f ck and 5 N/mm 2
column perimeter.d

(11.7.2-2)

and

where:
- q = load per unit surface (in kN/m2);
- d = effective depth of the slab at the columns (in mm);
- h = total depth of the slab at the columns (in mm);
- area supported by the column = in m2;
- column perimeter = in mm.

11.7.3 The mechanisms of load transfer in flat slabs

11.7.3.1 Distribution of bending moments in a flat slab

The following results are obtained by analysis of elastic behaviour of flat slabs:

11-57

figure 11.7.3-1 presents the scheme of tensile stress trajectories at the lower side
of the flat plate; one observes that reinforcement is necessary from a theoretical
point of view in four directions;
figure 11.7.3-2 presents the evolution of the bending moment mx in the x-direction
of the flat plate (VAN GEMERT, 1982): it is observed that the hogging and
sagging moments are maximum in the side strips parallel to the x-direction, along
the columns.

The scheme of stress trajectories on the one hand and the bending moment distribution
for mx (as well as the analogous distribution for my) on the other hand, lead to a typical
reinforcement scheme such as the one presented in figure 11.7.3-3. In case it is
necessary to reduce the number of layers of reinforcement (because of the limited
available depth), the four layers may be replaced by one (adapted) orthogonal mesh.

Figure 11.7.3-1
Scheme of tensile stress trajectories at the lower side of a flat plate

11-58

Figure 11.7.3-2
Evolution of the bending moment mx

Figure 11.7.3-3
Typical reinforcement scheme for flat slabs; left = bottom side;
right = upper side

11-59

11.7.3.2 Calculation of the main reinforcement of a flat plate

There are several approximate methods for the determination of the bending moments
that are necessary for the reinforcement calculation. The present course notes only
consider the so called method of the equivalent frame, which is one of the simplest
methods and which is also most often used in design offices and control organisms.).
This method transforms the calculation of a flat plate into the calculation of an
equivalent frame; VERMEESCH (1999) has shown that this simplified method leads to
results which are very similar to the ones obtained by finite element models or with
advanced plastic methods.
The flat plate is subdivided in strips in the x-direction, in-between the vertical planes
where each plane comprises a row of columns in the x-direction: see figure 11.7.3-4.
Each strip in the x-direction comprises n spans with span length equal to lx; the width of
the strips is equal to ly. The selected strip is now considered as a multiple degree
indeterminate frame. The floor slab with cross-section [(depth of the slab).ly]
corresponds to the girder of the frame; two half columns give one column in the
equivalent frame. The frame is composed of
- the selected floor level, together with
- the columns of the upper and lower floor levels; the columns may be
considered as perfectly clamped in both the upper and lower adjacent floor
level.
The frame analysis is typically performed
- with the concrete of the floor slab in uncracked situation;
- with the total load applied to the frame (and with consideration of the
unfavourable load combinations on the adjacent spans);
For quick calculations, the frame is sometimes replaced by a continuous beam.
The analysis of the frame (or of the continuous beam) leads to the values of the
maximum sagging moment in each span (in absolute value) mx and of the maximum
hogging moments at the columns mX. All these bending moments refer to the whole
width ly of the considered strip; in order to take account of the moment distribution
shown in figure 11.7.3-2, the moments mx and mX are divided over the width ly with an
appropriate distribution: see figure 11.7.3-4.
The calculation of the strip // x-direction should be repeated for the strip // to the ydirection, again for the total load on the slab.
The bending moments are then used for determination of the reinforcement areas in the
different zones of the flat plate or flat slab.

11-60

Figure 11.7.3-4
Basic figure showing the principle aspects of the simplified method of the equivalent
frame, which may be used for the calculation of flat plates

11.7.3.3

Detailing of reinforcement in the slab

Reference: EN 1992-1-1:2004; 9.4.1

11-61

At internal columns, top reinforcement of area 0,5.At should be placed in a width equal
to the sum of 0,125 times the panel width on either side of the column; see figure
11.7.3-5. At represents the area of reinforcement required to resist the full hogging
moment and which is distributed over the sum of the two half panels each side of the
column.
At internal columns, bottom reinforcement (at least two bars) should be provided in
each orthogonal direction; this reinforcement should pass through the column.
Specific rules for the detailing of reinforcement in the slab at edge and corner columns
are presented in EN 1992-1-1:2004; 9.4.2.

top reinforcement
= At

ly

0,125.ly
0,125.ly
top reinforcement
= 0,5.At

lx

ly

lx

Figure 11.7.3-5
Basic figure to illustrate the standard provisions for detailing of reinforcement in
flat plates; specific rules for reinforcement in the slab at internal columns

11.7.4 Punching shear

11.7.4.1 Phenomenologie
Flat plates are in general somewhat thicker than beam-supported slabs, mainly because
of the heavy concentrated load applied to the plate at the columns (upwards support
reaction force). Punching shear (NL: pons; FR: poinonnement; DU: Durchstanzen) can
result from a concentrated load or reaction force acting on a relatively small area of the
floor slab or foundation slab. Punching of plates by columns is in principle a shear
problem which leads to the observation of a failure mechanism around the loaded area:
see figure 11.7.4-1. One observes a typical punching shear cone.

11-62

The shape of the support (the loaded area) determines the shape of the crack surface: a
circular column leads to a conical wedge and a square or rectangular column leads to a
pyramidal wedge. Shear verification is not performed along the inclined crack surface
but in a vertically positioned critical cross-section around the column, at a certain
distance from the column, thus along the so-called critical perimeter". The shear load
per unit length is verified in this critical section. Outside the area determined by the
critical perimeter, the plate should be designed according to the provisions for shear
defined in EN 1992-1-1:2004; 6.2 (see chapter 7 in these course notes).
If the plate cannot withstand the support reaction force without a risk for punching
shear, additional reinforcement is necessary in the plate at the columns. In a first stage,
one may think of a distribution of inclined bars across the shear crack; see figure 11.7.42; such a crown of bars is a quite theoretical solution and is labour intensive. Solutions
that are used today on the construction site, are:
- a mesh composed of two orthogonal bundles of plied bars which cover the
column head;
- stirrups, as illustrated in figure 11.7.4-3;
- shear studs, as illustrated in figure 11.7.4-4.
Finally, instead of adding specific reinforcement, one may also choose to increase the
total depth of the plate in the areas at the columns, by means of column heads (or drop
panels) which transforms the flat plate into a flat slab.

Figure 11.7.4-1
Punching = a shear problem

11-63

stirrups

radial
reinforcement

Figure 11.7.4-2
Punching shear reinforcement composed of a crown of inclined bars, radially positioned
around the column

11-64

Figure 11.7.4-3
Punching shear reinforcement composed of stirrups; stirrups are tied to orthogonal bars
or are welded to shear hoops

Figure 11.7.4-4
Punching shear reinforcement composed of shear studs welded to a base plate

11.7.5 ULS design rules for punching shear verification in column-supported flat
plates and slabs

Reference: EN 1992-1-1:2004; 6.4

11-65

11.7.5.1 The basic verification model

The standard presents a simplified verification model for checking punching shear
failure at ULS in column-supported flat plates: see figure 11.7.5-1. In its most simple
formulation, the model only considers a uniformly distributed load on the plate and a
vertical column reaction force (as if the column was part of a braced frame). The
refinement of the model necessary to take account of bending moments at the support, is
presented later in these course notes.
Figure 11.7.5-1 allows formulating some definitions:
loaded area Aload = the contact area between column and plate = the cross-section
of the column;
basic control section = the area over which the support reaction force the shear
load) is smeared out; this leads to a shear stress which has to be limited;
basic control perimeter u1 = the perimeter around the column along which the
shear stress has to be verified;
basic control area Acont = the area comprised within the basic control perimeter;
d = the effective depth of the plate above the column; when two orthogonal
reinforcement layers are present with effective depth dx and dy respectively, the
verification is performed with the mean value for the effective depth:
d dy
d x
2
the inclination angle defines the extent of the punching shear cone. The previous
version (1998) of EC2 recommended to adopt = 33,7. In the actual version
(2004), an even less inclined shear crack is accepted: = 26,6. The reason for the
choice of a rather flat punching shear crack is the presence of already large
reinforcement areas in the plate above the column support, because of the rather
large values of the hogging moments at column supports. The presence of a dense
reinforcement mesh forces the crack to find its way through the plate at a larger
distance from the loaded area. This explains the simplified model assumption to
identify the basic control perimeter at the distance 2.d from the column edge
(from the edge of the loaded area).

11-66

Figure 11.7.5-1
Simplified model for the ULS analysis of the punching shear problem in columnsupported flat plates (figure 6.12 in EN 1992-1-1:2004)

11.7.5.2 The basic control perimeter u1

The standard assumes to identify the basic control perimeter at the distance 2.d from the
edge of the loaded area. For columns with irregular shape, the basic control perimeter u1
has to be taken as small as possible (worst vase for the verification!): see figure 11.7.52. Indeed, a larger perimeter leads to a larger basic control section and thus to a smaller
shear stress.
Some particular cases are presented in the following figures:
figure 11.7.5-3: model for the reduction of the basic control perimeter when the
column is close to an opening in the slab; in this way, it is taken into account that
the part of the control perimeter close to the opening is not able to resist
efficiently to the punching shear stresses;
figure 11.7.5-4: models for basic control perimeters for edge and corner columns;
the principle is: verification is performed on the basis of the smallest perimeter u1!

Figure 11.7.5-2
Typical basic control perimeters around loaded areas (figure 6.13 in EN 1992-1-1:2004)

11-67

Figure 11.7.5-3
Models for the reduction of the basic control perimeter when the column is close to an
opening in the slab (figure 6.14 in EN 1992-1-1:2004)

Figure 11.7.5-4
Models for basic control perimeters for edge and corner columns (figure 6.15 in EN
1992-1-1:2004)
Sometimes, it is preferred to arrange for column heads between the plate and the
column, in order to solve the punching shear problem; this leads to the concept of flat
slab. The dimensions of the column heads determine the basic control perimeter, as
illustrated in the following figures for the case of a circular column:
figure 11.7.5-5: small column head (lH < 2.hH; the symbol H < Head); it is not
necessary to consider the basic control perimeter in the column head because it is
too small. The basic control section is situated in the slab; its effective depth is d;
figure 11.7.5-6: intermediate case (lH = 2.hH); same conclusion as in the former
case;
figure 11.7.5-7: rather large column head (lH > 2.hH but lH < 2.(d+hH)). Two
control perimeters may be identified. However, both perimeters are characterized
by the same effective depth d; hence, only the smallest perimeter has to be
verified;
figure 11.7.5-8: large column head (lH > 2.hH and lH > 2.(d+hH)). Two basic
control perimeters have to be considered, one in the column head and one in the
slab. The effective depth of the inner control section is dH = d + hH ; the effective
depth of the outer control section is d.

11-68

The philosophy adopted for circular columns may also be applied for columns with
rectangular cross-section. EN 1992-1-1:2004; 6.4.2(8) discusses the case of rectangular
columns with dimensions c1 and c2 with a column head characterized by the dimensions
l1 and l2 (l1 = c1 + 2.lH1, l2 = c2 + 2.lH2, l1 l2): it is recommended to consider a circular
basic control perimeter with a radius which may be taken as the lesser of
2.d + 0,56.(l1.l2) and 2.d + 0,69.l1.

2.d

d
hH
lH

lH < 2.hH

2.hH

Figure 11.7.5-5
Small column head (lH < 2.hH; H < Head). The basic control section is situated in the
slab at the distance 2d from the edge of the column head

2.( d + hH )

d + hH
hH

lH=2.hH

lH = 2.hH

Figure 11.7.5-6
Intermediate case (lH = 2.hH). The basic control section is situated in the slab at the
distance 2d from the edge of the column head

11-69

2.( d + hH )
2.d

d + hH

hH

lH

lH > 2.hH
but
lH < 2.(d+hH)

Figure 11.7.5-7
Rather large column head (lH > 2.hH but lH < 2.(d+hH)). The shortest perimeter is the
basic control perimeter

lH+2.d
2.( d + hH )

2.d

d + hH

hH

lH

lH > 2.(d+hH)

Figure 11.7.5-8
Large column head (lH > 2.hH and lH > 2.(d+hH)). Two basic control perimeters have to
be considered, one in the column head and one in the slab

As explained above, the shear stress is verified in the basic control section. When the
shear stress is too large, measures have to be taken: the depth of the slab may be
increased (globally or locally by means of column heads) or punching shear
reinforcement may be installed. Next, shear stress has to be verified in the second
control section which is situated at a distance 2.d from the basic control section and
11-70

which has a similar perimeter (in shape) as the first one. Verifications of the shear stress
have to take place in successive control sections until shear stress are under control.

11.7.5.3 Principle aspects of punching shear verification in ULS

The punching shear verification is based on the comparison of four different design
values of shear stresses, which are indicated by means of the small letter v (as the shear
force V is smeared out all over the control section):
the design value of the imposed shear stress in the control section
- vEd =
considered, which is due to the shear force VEd or the support reaction force
of the column;
the design value of the punching shear resistance of a slab without punching
- vRd,c =
shear reinforcement, along the control section considered;
- vRd,cs = the design value of the punching shear resistance of a slab with punching
shear reinforcement, along the control section considered;
- vRd,max = the design value of the maximum punching shear resistance along the
control section considered.

Note:
Additional comment regarding the last shear stress vRd,max.
The punching shear calculation is performed on the basis of the model of the
punching shear cone defined by an inclined shear crack. In analogy with the
shear model for beams, two different calculations have to be made:
- the determination of the necessary reinforcement to bridge the crack (when
reinforcement is necessary);
- the verification of the inclined concrete compression strut defined by two
adjacent cracks.
In the context of the beam analysis, the attention was already drawn to the
consequences of considering less inclined cracks: more efficient use of the shear
reinforcement (thus less reinforcement) and an increased loading of the concrete
compression struts. In the case of punching shear, which is also characterized by
the small inclination angle of the crack, verification of the compression struts is
absolutely necessary and should even be arranged at the start of the whole
verification procedure. This verification should start in a particular way: the
imposed load vEd has to be compared with the maximum resistance which may
be generated from the point of view of the compression strut at the edge of the
column (see further).

In general, the punching shear verification comprises the following steps (which are
worked out in the next paragraphs):
identification of the imposed shear load (support reaction force of the
column) VEd;
calculation of the absolute maximum resistance which can be generated
against the punching shear (vRd,max) and comparison with the imposed shear
stress vEd at the edge of the column (the edge of the loaded area Aload), thus
by smearing out the imposed load VEd all over the section at the edge of the
column:

11-71

(v Ed ) edgecolumn ? v Rd ,max

This means that the verification of the concrete compression strut is realised
in very conservative conditions by selecting a small perimeter (and thus vEd
large), especially because this verification is so critical considering the small
inclination angle of the crack!;
identification of the first control perimeter (the basic control perimeter)
around the column; determination of the depth of the basic control section
and the length of the basic control perimeter u1;
calculation of the imposed shear stress vEd due to the shear force (support
reaction force), smeared out over the basic control section; comparison with
vRd,c:
(v Ed ) basic control perimeter ? v Rd ,c

when vEd vRd,c, there is no need for punching shear reinforcement and the
verification principally stops here;
when vEd > vRd,c, punching shear reinforcement is necessary; the verification
should be continued by considering the next control perimeter.

11.7.5.4 Verification of vRd,max at the edge of the column

In principle VEd is the support reaction force of the column. This force is smeared out
over the section at the edge of the column and the resulting stress is compared to vRd,max.
The formula is:
v Ed

.VEd
u 0 .d

v Rd ,max

where
VEd =
u0 =

(11.7.5-1)

the support reaction force of the column (in N)

the perimeter (mm) of the edge of the column (= the perimeter of the loaded
area Aload);
u0 =
the perimeter (mm) of the column itself for an internal column;
u0 =
c2 + 3.d c2 + 2.c1 (mm) for an edge column; see figure 11.7.5-9;
u0 =
3.d c1 + c2 (mm) for a corner column; see figure 11.7.5-9;
d=
the effective depth of the slab (mm);
=
increase coefficient which takes account of the effect of eventual bending
moments at the slab-column connection (see further);
vRd,max = the design value of the maximum punching shear resistance of the slab in the
section at the edge of the column, given by the formula:

11-72

v Rd ,max 0,5. . f cd
and
=

(11.7.5-2)

the reduction factor for the concrete strength, to be used for compression struts
in the context of shear verification:

0,6.1

f ck
(fck in MPa)
250

(11.7.5-3)

When vEd is found to be larger than vRd,max, the design has to be adapted. If this is not the
case, the punching shear verification may be continued with the verification of the
successive basic control perimeters.

c2
c2

c1

c1

Figure 11.7.5-9
Auxiliary figure: dimensions to be considered for edge and corner columns

11.7.5.5 Calculation of vEd for the first control perimeter (the basic control perimeter)
VEd is the support reaction force of the column. In analogy with the procedure for shear
control in beams where direct transfer of part of the loads towards the supports was
considered, one might have the idea to adopt similar rules for punching shear
calculations, assuming that the load distributed within the basic control perimeter is
directly transferred to the column. However, EC2 recommends abandoning this idea for
the case of punching shear, based on the following arguments:
punching shear is really a critical phenomenon;
one cannot be sure in the design phase of a project, that the distributed load will
always be present within the basic control perimeter. This is illustrated by the
following example: it might be possible that in the later renovation process of a
building, a reinforced concrete wall on top of the columns is replaced by a lightweight partition wall which reduces drastically the magnitude of the service loads
within the basic control perimeter.
Conclusion: see figure 11.7.5-10.

11-73

vEd

VEd
u1.d

VEd

Figure 11.7.5-10
The value of the shear stress vEd along the basic control perimeter u1, for punching shear
verification of column-supported slabs; direct load transfer is not considered

Notes:
- Even when concentrated loads are applied close to the column head, the EC2
still recommends not using direct load transfer to the supports.
- An exception is made for the punching shear calculation of a foundation slab,
where direct load transfer is accepted. Indeed, the distributed load (the soil
bearing pressure) within the control perimeter is always present.

11.7.5.6 Calculation of vEd along a control perimeter in the case of an eccentrically

positioned support reaction (case of frame action)
Reference: EN 1992-1-1:2004; 6.4.3-(3), (4) en (5)
In the case of framework structures, bending moments and axial forces are present at the
ends of the columns. The presence of a bending moment corresponds to an eccentricity
of the support reaction force and thus to non-uniform distribution of the shear stress
around the column: see figure 11.7.5-11. When the support reaction is eccentric with
regard to the considered control perimeter ui, the maximum shear stress along this
control perimeter ui should be taken as:

vEd

VEd
ui .d

(11.7.5-4)

where:
d : the mean effective depth of the slab; d = (dx + dy)/2 with dx and dy the effective
depths in the x- and y-directions respectively;
ui : the length of the control perimeter being considered;

11-74

increase coefficient to take account of the unfavourable effects of eccentricity,

given by the formula:

1 k

M Ed u1
.
VEd W1

(11.7.5-5)

where:
length of the basic control perimeter;
u1 :
k:
coefficient dependent on the ratio between the column dimensions c1 and c2;
see figure 11.7.5-11 and table 11.7.5-1;
W1 : factor which corresponds to a distribution of shear as illustrated in figure
11.7.5-11. W1 is a function of the control perimeter considered:
W1 e .dl
u1

(11.7.5-6)

where:
dl : length increment of the perimeter;
e : distance of dl from the axis about which the moment MEd acts.

Figure 11.7.5-11
Shear distribution due to a moment at the slab-column connection (figure 6.19 in EN
1992-1-1:2004)

Table 11.7.5-1
Values of k for rectangular loaded areas (Table 6.1 in EN 1992-1-1:2004)
c1/c2
0,5
1,0
2,0
3,0
0,45
0,60
0,70
0,80
k

Some particular cases are now discussed:

11-75

For a rectangular column, the factor W1 is given as:

2

W1

c1
c1.c2 4.c2 .d 16.d 2 2. .d .c1
2

(11.7.5-7)

where:
c1 : is the column dimension parallel to the eccentricity of the load (the
support reaction force);
c2 : is the column dimension perpendicular to the eccentricity of the load.
For internal circular columns, follows from:

e
D 4.d

1 0,6.

(11.7.5-8)

where D is the diameter of the circular column.

For an internal rectangular column where the loading is eccentric to both axes, the
following approximate expression for may be used:
e
1 1,8 y
bx

ex


by

(11.7.5-9)

where:
ex and ey = the eccentricities MEd/Ved along the x- and y-axes respectively (ey
results from a moment about the x-axis and ex from a moment about the yaxis);
bx and by = dimensions of the control perimeter.

For edge column connections, where the eccentricity perpendicular to the slab edge
(resulting from a moment about an axis parallel to the slab edge) is toward
the interior and there is no eccentricity parallel to the edge, the punching
force may be considered to be uniformly distributed along the control
perimeter u1*: see figure 11.7.5-12.

11-76

Figure 11.7.5-12
Reduction of the length of the basic control perimeter u1 to u1* for edge and
corner columns in the case of eccentrically positioned support reaction
forces (figure 6.20 in EN 1992-1-1:2004)

Where there are eccentricities in both orthogonal directions, may be

determined using the following expression:

u1
u
k 1 e par
W1
u1*

(11.7.5-10)

where:
the basic control perimeter (considered when the load is uniformly
u1 :
distributed); see figure 11.7.5-4;
u1* : the reduced basic control perimeter (see figure 11.7.5-12(a) for the
case of the edge column);
epar : the eccentricity parallel to the slab edge resulting from a moment
about an axis perpendicular to the slab edge;
k:
may be determined from table 11.7.5-1 with the ratio c1/c2 replaced
by c1/2c2 ;
W1 : is calculated for the basic control perimeter (see figure 11.7.5-2).
For a rectangular column (as shown in figure 11.7.5-12(a)), the factor W1 is
given by the formula:
2

W1

c2
c1.c2 4.c1.d 8.d 2 .d .c2
4

11-77

(11.7.5-11)

If the eccentricity perpendicular to the slab edge is not toward the interior,
expression (11.7.5-5) may be applied. When calculating W1, the eccentricity
e should be measured from the centroid of the control perimeter.
For corner column connections where the eccentricity is toward the interior
of the slab, it is assumed that the punching force is uniformly distributed
along the reduced control perimeter u1*, as defined in figure 11.7.12(b). The
value of is taken as u1/u1*. If the eccentricity is toward the exterior,
expression (11.7.5-5) should be applied.
11.7.5.7 Calculation of vEd for the first control perimeter for the case of an eccentrically
positioned punching force, but not in the context of frame action
For structures
where the lateral stability does not depend on frame action between the slab and
the columns;
where the adjacent spans do not differ in length by more than 25%,
vEd is determined by means of the following expression:
vEd

VEd
ui .d

(11.7.5-12)

where approximate values for may be used: see figure 11.7.5-13.

Figure 11.7.5-13
Recommended values for the increase coefficient which allows taking account of nonuniformly distributed shear stress around columns
(figure 6.21N in EN 1992-1-1:2004)

11-78

11.7.5.8 Calculation of vRd,c

The calculation of the punching shear resistance of slabs which do not contain punching
shear reinforcement, may be performed by means of an expression which is very similar
to the one used for determining VRd,c in the context of shear verification in beams. This
is evident because the mechanisms that offer shear resistance are the same (see figure
7.2.4-1 in chapter 7).
v Rd ,c C Rd ,c .k .(100. l . f ck )1 / 3 k1 . cp v min k1 . cp

(11.7.5-13)

where:
is in MPa;
fck
k=
1 + (200/d) 2,0 with d in mm;
l =
(lx.ly) 0,02 where the geometric reinforcement ratios lx and ly are related
to the tension steel in the x- and y- direction respectively. The values of lx and
ly should be calculated as mean values taking into account a slab width equal
to the column width plus 3.d each side
(cx + cy)/2 where cx and cy are the normal concrete stresses in the critical
cp =
section in the x- and y-direction respectively (MPa, positive if compression),
which are generated by applied loads such as prestressing loads:
N Ed , y
N
and c , y
c , x Ed , x
Acx
Acy
CRd,c = 0,12
k1 =
0,1
vmin = 0,0035.k3/2.fck1/2

With all coefficients filled in, the expression transforms into:

v Rd ,c 0,12.k .(100. l . f ck )1 / 3 0,1. cp v min 0,1. cp

(11.7.5-14)

11.7.5.9 Calculation of the punching shear reinforcement calculation of vRd,cs

The value of vRd,cs which is the design value of the punching shear resistance of a slab
containing punching shear reinforcement, should be determined with the following
expression:
v Rd ,cs 0,75.v Rd ,c 1,5.(d / s r ). Asw . f ywd ,ef .(1 /(u1 .d )). sin

(11.7.5-15)

where:
Asw = the area (in mm2) of one perimeter of shear reinforcement around the column;
see figure 11.7.5-14;
sr =
the radial spacing of successive perimeters of shear reinforcement (in mm);

11-79

fywd,ef = effective design strength of the punching shear reinforcement; EC2 proposes
fywd,ef = 250 + 0,25.d fywd (in MPa);
d=
the mean of the effective depths in the slab, taking into account the two dvalues in the orthogonal directions (in mm);
=
the angle between the shear reinforcement and the plane of the slab (vertical
stirrups: = 90).

Discussion of this expression:

the determination of the area of reinforcement that is necessary to resist to vEd,
starts from the assumption that a part of the resistance is provided by the punching
shear resistance of the slab without shear reinforcement; EC2 even accepts that
75% of vRd,c may be taken into account;
the contribution from the reinforcement is determined on the basis of the
assumption that the reinforcement is arranged in several perimeters with spacing
sr within the radius 1,5.d. The ratio (1,5.d)/sr determines the number of perimeters.
With only 1 active perimeter, the ratio is (1,5.d)/sr = 1;
the force that is generated by the active reinforcement is smeared out over the
control section with surface equal to u1.d.

11-80

2.d

2.d

d
sr

sr

Figure 11.7.5-14
Schematic representation of the punching shear reinforcement in a flat plate: the figure
presents three reinforcement perimeters with spacing sr
11.7.5.10

The outermost control perimeter uout

The outermost control perimeter at which shear reinforcement is not required anymore
is indicated in figure 11.7.5-15 by the notation uout for a complete perimeter and by
uout,ef for a perimeter composed of effective parts. uout corresponds to the condition
vEd vRd,c. Taking into account expression (11.7.5-12), one finds the value of uout as:
V
u out Ed
(11.7.5-16)
v Rd ,c .d

11-81

The outermost perimeter of shear reinforcement should be placed at a distance not

greater than 1,5.d from the outermost control perimeter uout: see figure 11.7.5-15. This
provision is already part of the detailing of the punching shear reinforcement;
complementary provisions are presented in the next paragraph.

Figure 11.7.5-15
Schematic representation of the punching shear reinforcement in a flat plate above an
internal column: the distance between the outermost reinforcement perimeter and the
outermost control perimeter uout or uout,ef should be less than (k.d) = 1,5.d

11.7.6 Detailing of punching shear reinforcement

Reference: EN 1992-1-1:2004; 9.4.3

Some provisions about detailing of punching shear reinforcement are summarized in
figure 11.7.6-1. Attention is drawn to the following points:
(as was already pointed out before) the distance between the outer perimeter of
punching reinforcement and the outermost control perimeter (for which punching
shear reinforcement is not necessary anymore and thus for which the condition
vEd vRd,c is fulfilled) is maximum (k.d) = 1,5.d;
when punching shear reinforcement is composed of stirrups, at least two
reinforcement perimeters have to be arranged; one perimeter is sufficient with
inclined bars;
radial spacing between the punching shear reinforcement perimeters should not
exceed 0,75.d;
tangential spacing between the reinforcement legs along one perimeter should not
exceed 1,5.d within the 1st control perimeter and 2.d for further control
perimeters.

Note:
The provisions mentioned above and the figures shown in the standard present
rather ideal arrangements of punching shear reinforcement. In practice, it is not
unusual to notice on real construction sites that punching shear reinforcement is
composed of groups of stirrups which are put in a rather arbitrary way within the
outer control perimeter uout.
11-82

Figure 11.7.6-1
Detailing of punching shear reinforcement

11.7.7 Punching shear verification of a foundation slab (also called column base)

A foundation slab is typically loaded on one side by a concentrated force transferred by

a column and on the other side by the soil bearing pressure. The punching shear
verification for a foundation slab is thus very similar to the one for a flat slab: see figure
11.7.7-1. The only (big) difference between the two cases is that for foundation slabs, it
is allowed to take account of direct load transfer: the shear force (the action of the
column) may be reduced by the force generated by the soil bearing pressure within the
control perimeter. The justification for this: the soil bearing pressure is always present
and thus offers a favourable effect on the punching shear problem for foundation slabs.
The formula for the calculation of vEd is:
v Ed
where
g =

area =

VEd area. g
u i .d

VEd ,red
u i .d

(11.7.7-1)

the upwards directed soil pressure which is only due to the service loads
applied to the foundation slab (without consideration of the self-weight
of the foundation slab);
the surface within the control perimeter.

For an eccentrically applied load, one may use the expression:

vEd

VEd ,red
u.d

M Ed u
.
1 k
VEd ,red W

11-83

(11.7.7-2)

Significance of the used symbols: see expression (11.7.5-5).

Note:
The punching shear resistance of column bases without punching shear
reinforcement should be verified at control perimeters within 2.d from the
periphery of the column. The standard recommends an adapted expression for
the calculation of vRd,c at the distance a from the edge of the column (thus
smaller than the usual distance 2.d !):
vRd ,c 0,12.k .(100. l . f ck )1/ 3 x 2.d / a vmin x 2.d / a

(11.7.7-3)

with a = the distance of the control perimeter to the edge of the column.
With a = 2.d, one finds back again the original expression.
VEd

(a)

g
2.d

VEd

2.d

(b)

(c)

VEd

Figure 11.7.7-1
The punching shear verification for a foundation slab (or a column base)
(a) : idealisation (constant soil pressure g)
(b) and (c) : it should be noted that the distribution of the real soil pressure is not
constant, but is determined by the stiffness of the foundation slab and the soil nature

11-84

13 Chapter 12
Hillerborgs strip method for ULS design of slabs
12.1 Introduction
The chapter starts with the justification for using plastic design methods for slabs: slabs
have a large capacity for plastic moment redistribution.
The chapter then focuses on the strip method (HILLERBORG) which is presented as an
interesting tool for slab design in ULS. The chapter is limited to the ULS problem of
bending; shear and punching shear as well as SLS are considered in other chapters.

12.1.1 Behaviour of slabs up to failure: large plastic deformation capacity is observed

The linear elastic analysis of a structure is important in the context of SLS and for the
understanding of the behaviour of the structure under imposed (service) loads.
Reinforced concrete slabs are structures which show large plastic deformation capability
when they are loaded up to failure; slab behaviour is characterized by the following
phases (see figure 12.1.1-1):

the elastic phase: which corresponds to the behaviour of elastic, homogeneous,

continuous and isotropic material;

the cracking phase: cracks appear in the tensile region; they lead to the progressive
reduction of the inertial moment of the cross-sections. A first redistribution of
bending moments is observed: the moments in the uncracked sections increase
faster (than in the elastic phase) with the same increase of loads. As long as the
reinforcement deforms elastically, crack opening remains limited;

the phase of plastic deformation of steel: in the sections where steel starts to yield,
deformations continue to increase while the bending moment remains constant; this
is the second (and important) redistribution of moments. The plastic deformation
appears along lines which correspond to the big cracks in the tensile region. These
lines are considered as plastic hinges (which thus still transfer the yield moment).
The pattern formed by these so-called yield lines (NL: vloeilijnen; FR: lignes de
rupture) depends on the shape of the slab, the edge conditions, the distribution of
the reinforcement and the load pattern;

the failure: an instable state of equilibrium is reached when the progressive

formation of yield lines ends up into a mechanism. It is observed that the
deformations in the slab are concentrated along the yield lines: the straight parts of
the slab, which are delimitated by the yield lines, rotate around the supported edges
and the yield lines. This process leads to the crushing of the compressed concrete
along the yield lines: the yield lines become visible on the compression side of the
slab.

13-1

load
rupture
plastic deformation
of steel

cracking
service

elastic
behaviour
deflection

Figure 12.1.1-1
Schematic representation of the load-deformation diagram of a reinforced concrete slab

Figure 12.1.1-2 shows in a schematic way the process of the formation of a yield line
pattern. Figure 12.1.1-3 shows a real yield line pattern observed during the loading of a
rectangular simply supported slab by a concentrated load in the middle; the yield lines
are identified where large crack openings are observed as well as the concentration of
the plastic deformation of the steel reinforcement.
The ability to form a yield line pattern before failure leads to the conclusion that slabs
have a large capacity for moment redistribution. Plastic methods are thus suited for slab
analysis and design.

13-2

Figure 12.1.1-2
Schematic representation of the formation of a yield line pattern

13-3

Figure 12.1.1-3
Example of a real yield line pattern

13-4

12.1.2 The plastic moment along the yield line and ductility conditions
12.1.2.1

The plastic moment mp

The ultimate moment capacity of a cross-section in reinforced concrete is indicated by

the symbol mult; mult corresponds to failure.
The plastic moment mp is the moment which is reached when the plastic deformation of
the steel reinforcement starts.
On the condition that:
- the cross-section is under-reinforced, and that
- the ultimate resistance is determined by the yielding of the steel,
one finds the following design value for the plastic moment per unit length of the yield
line mpd (see figure 12.1.2-1):

m pd = As . f yd .z

(12.1.2-1)

where z is the lever arm between the resultant compression force Nc and the resultant
tensile force Ns.

b=1

fcd
Nc

x
h

Ns

As = .d.1

fyd

Figure 12.1.2-1
Rectangular section with unit width, singly reinforced, loaded in simple bending
It is shown in chapter 4 that ULS analysis often leads to the adoption of the simplifying
assumption: z 0,9.d; consequently, the plastic moment may be calculated by means of
the following expression:

m pd = As . f yd .0,9.d

13-5

(12.1.2-2)

12.1.2.2

Ductility conditions to be fulfulled for use of plastic methods

Equation (12.1.2-1) could give rise to the idea that mp may be increased by increasing
the area of reinforcement As. It has already been discussed in chapter 10 (plastic
methods for beam analysis) that sections have to be under-reinforced if plastic methods
are considered for use. As a reminder, the reinforcement ratio has to be limited
between an upper and lower limit value:
has to be large enough in order to avoid brittle rupture when concrete in
tension cracks;
has to be small enough in order to assure plastic rotation capacity; failure by
concrete crushing may only appear after plastic deformation of the steel.
For rectangular sections in reinforced concrete, one may use the practical limit values
for the reinforcement ratio formulated by FAVRE (1989):

0,15% 1,5%
EN 1992-1-1:2004; 5.6 adds to that the following: the required ductility may be deemed
to be satisfied without explicit verification if all the following conditions are fulfilled:
- the area of tensile reinforcement is limited such that at any section:
xu/d 0,25 for concrete strength classes C50/60 ;
xu/d 0,15 for concrete strength classes C55/67 ;
- the reinforcing steel is either class B or C.

Note:
The value xu/d = 0,25 corresponds to the limit between the domains 1b and 2a
(see chapter 4): full exploitation of the concrete strength (ultimate strain in
compression is reached: 0,35 %) and full exploitation of the steel strength
(ultimate strain in tensile steel is reached: 1 %).

12.1.3 Plastic methods for slab analysis

The basic principles of plastic methods for structural analysis are introduced in chapter
10, in the context of beam and frame analysis. Two families of plastic methods are
identified:
- the lower bound methods, also called static methods;
- the upper bound methods, also called kinematic methods.
These methods are also applicable for slabs (which is in fact much more justified than
for beam analysis):
the static method starts from the choice of a statically acceptable moment
distribution and leads to a lower bound value of the real failure load. This leads
to a conservative (safe) solution because the slab is designed as if it were less
resistant than it is in reality. The static method is well known for slab analysis
thanks to the practical formulation presented by HILLERBORG in 1956, called
the strip method.

13-6

the kinematic method starts from the choice of a kinematically acceptable

mechanism for which energy balance calculations (energy input from the loads =
energy absorption by the plastic hinges) lead to an upper bound value of the real
failure load. This is in principle an unsafe approach and conclusions have to
be well interpreted in practice. A well known kinematic method is the yield line
method which was developed by JOHANSEN in the years 1950.

12.2 Static methods for slab analysis: general introduction

12.2.1 General principle of the static method

According to the lower bound theorem, a structure with sufficient ductility resists in a
safe way to the applied loads, on the condition that a distribution of internal moments
can be identified which satisfies to the equilibrium conditions and for which mp is not
exceeded anywhere (reminder; see chapter 10). The determination of the exact failure
load needs the identification of the kinematically acceptable mechanism which
corresponds to the moment distribution. In principle, many moment distributions can be
found which satisfy the two conditions (equilibrium and mp); it may thus be necessary
(for complex boundary conditions and slab shapes) to use numerical tools for support of
the iteration process in order to find THE distribution that corresponds to THE
mechanism.
Anyway, one can always start the plastic analysis with a plausible choice of elastic
bending moment distribution, knowing that this will lead to an under-estimation of the
real failure load; one thus thinks that the slab is weaker than it really is.

12.2.2 An extreme example to illustrate the principle of the static method

The principle of the static method is now illustrated by means of an example. Figure
12.2.2-1 shows a square slab in reinforced concrete; the slab is:
- simply supported along all four edges;
- loaded by a uniformly distributed load q = 12 kN/m2;
- reinforced in two directions by means of an orthogonal reinforcement mesh.
With an equally large moment resistance capacity in both directions, such that no plastic
hinges appear, the moment distribution is the one obtained by linear elastic analysis. A
finite element calculation leads to a sagging moment in the span equal to 18,4 kNm/m in
both directions; if the moment resistance capacity of the slab were larger than
18,4 kNm/m in both directions, the slab would not show any yielding for the imposed
load.
It is now assumed that the slab is reinforced only in one direction, for example in the xdirection. The bending moment for a beam (with unit width) in the x-direction, loaded
by 12 kNm/m, equals 54 kNm. By putting a series of beams in the x-direction, one next
to the other, each beam with a bearing capacity of 54 kNm, one obtains a safe structure
to transfer the imposed load, even without having reinforcement in the y-direction. The

13-7

slab could thus be designed on the basis of a moment distribution which is typical for
beam structures; by doing this, one forces the slab to act as a series of beams.
Off course, a slab is a structural member which generally transfers the load in two
directions; the slab has thus to transform itself into a series of beams, which is
accompanied by severe cracking in the direction parallel to the reinforcement until the
load is fully transferred by the beams. If one is willing to reinforce the slab as if it were
a series of beams, he has to be sure that enough ductility is present in order to permit the
transition form slab to beam system (ductility is needed to allow the formation of the
necessary hinges). This discussion also shows that a bad choice of load transfer system
leads to problems with SLS requirements: unacceptable cracks are developed associated
with the transfer from one bearing system to another (an imposed one).
It is generally accepted that when the chosen moment distribution is close to the linear
elastic distribution, the slab has sufficient ductility to be designed by the static method;
moreover, this is practically a condition sine qua non to satisfy to the SLS requirements.

13-8

Figuur 12.2.2-1
Extreme example of application of the static method to a simply supported rectangular
slab, loaded by a uniformly distributed load; (a) plan; (b) result of the linear elastic
analysis; (c) distribution of the bending moments for a beam like structure

13-9

12.3 Static methods for slab analysis: the strip method

12.3.1 Background of the method

Before introducing the specific assumptions and characteristics of the strip method, it is
necessary to remind some results from the elastic analysis of beams and slabs.
Figure 12.3.1-1 expresses the equilibrium of an elementary part of the beam or slab.

Equilibrium of a beam-element
The vertical translation equilibrium is assured by the shear forces at the ends:
figure 12.3.1-1 (a).
q.dx + V +

dV
.dx V = 0
dx

dV
q=
dx

(12.3.1-1)

The moment equilibrium considers all the moments at the ends of the considered
element: see figure 12.3.1-1 (b). Moments and shear forces are represented
positively with respect to right hand system of axis (x, y, z) in accordance with
the sign convention in theory of elasticity. The moment equilibrium around G
leads to:

M M

dM
dx
dx dQ
dx
.dx + Q. + Q. +
.dx. = 0
dx
2
2 dx
2

After elimination of the higher order terms:

Q=

dM
dx

(12.3.1-2)

Introduction of expression (12.3.1-2) into expression (12.3.1-1) leads to the

equilibrium equation for beams:

d 2M
= q
dx 2

(12.3.1-3)

Equilibrium of a slab-element
The vertical translation equilibrium is now assured by the shear loads at the four
edges (shear forces per unit length!): see figure 12.3.1-1 (c)

13-10

q.dx.dy + vx .dy +

v
vx
.dx.dy vx .dy + v y .dx + y .dy.dx v y .dx = 0
x
y

or
v v
q = x + y
x y

(12.3.1-4)

The moment equilibrium concerns all moments generated at the four edges of
the slab-element: figure 12.3.1-1 (d). The difference with the beam is that
torsional moments mxy are introduced, next to the bending moments mx and my;
the torsional moments are related to the 2-dimensional character of the slab. The
moment equilibrium around the x-axis leads to:
m y
m
v

dy
dy
m y +
.dy .dx m y .dx + mxy + xy .dx .dy mxy .dy v y .dx. v y + y .dy .dx. = 0
y
x
y
2
2

After elimination of the higher order terms:

vy =

m y
y

mxy
x

(12.3.1-5)

In an analogous way, the moment equilibrium around the y-axis leads to:
vx =

mx m yx
+
x
y

(12.3.1-6)

Introduction of expressions (12.3.1-5) and (l2.3.1-6) in (12.3.1-4) leads to the

equilibrium equation of slabs:
2
2 m yx
2 mx m y
+
+2
= q
x 2
y 2
xy

13-11

(12.3.1-7)

Figure 12.3.1-1
Auxiliary figure for the elaboration of the equilibrium equations for beams and slabs;
(a) and (c) shear forces; (b) and (d) moments

12.3.2 Principle aspects of the strip method

The equilibrium equation (12.3.1-7) is the starting point for the introduction of the strip
method. HILLERBORG assumes that torsional resistance may be neglected; expression
(12.3.1-7) is transformed into:
2
2 mx m y
+
= q
x 2
y 2

(12.3.2-1)

Neglecting the torsional resistance inevitably leads to the under-estimation of the global
bearing capacity of the slab. The practical consequence of this assumption is that the
calculation of the slab is transformed into a simple calculation of beams; indeed,
according to expression (12.3.2-1) the load q is transferred to the edges by means of
bending moments mx and my, thus by means of beams or strips in the x- and y-direction.
The remaining question however is: with what distribution ratio?
HILLERBORG accepts a second important principle in order to solve this problem: the
load is transferred to the closest supporting edge by a bending mechanism in the plane
perpendicular to the supporting edge.

13-12

For a slab with edges parallel to the x- and y-direction, HILLERBORG proposes that a
part .q of the load q is transferred in the x-direction (by bending), and that the part (1).q is transferred into the y-direction (also by bending); consequently:
2 mx
= .q
x 2
2my
y 2
where =

= (1 ).q

(12.3.2-2)

(12.3.2-3)

the distribution ratio coefficient for the load applied in every point of the
slab; 0 1 and may in principle change from one location in the slab
to another. For practical reasons (economic aspect and simplicity of
reinforcement meshes) the coefficient is taken constant in regions of
the slab, most of the time with = 0, 1/2 or 1 (see examples further in
this chapter).

The expression (12.3.2-2) and (12.3.2-3) are identical to the equilibrium equation
(12.3.1-3) for beams.

12.3.3 A basic example of the application of the strip method

Figure 12.3.3-1(a) represents a rectangular slab, simply supported along all four edges,
loaded by a uniformly distributed load q. In accordance with the basic assumptions in
the method of HILLERBORG, the slab may be subdivided into several regions in
function of the preferential direction of load transfer: see figure 12.3.3-1(b).
The load q applied in zone 1 and 2, is transferred in the y-direction to the nearest
support which are parallel to the x-direction; the value of in (12.3.2-2) is equal to 0 in
zone 1 and 2. The load q applied in zone 3, is transferred in the x-direction to the nearest
support which is parallel to the y-direction; the value of in (12.3.2-2) is equal to 1.
The next step in the method is illustrated in figure 12.3.3-1(c): the analysis of the slab is
transformed into the analysis of beam elements with particular loading dispositions.
It may be observed that the selection of the direction of load transfer is highly related to
the identification of the so called tributary areas in the slab. A tributary area is related to
a supporting edge: it determines for what slab surface the load is transferred to that
supporting edge. As the edges are simply supported in the example, the separation lines
between the tributary areas correspond to the bisector lines of the rectangular corners
between the adjacent edges.

13-13

Beam 1

Beam 2

Beam 3

Figure 12.3.3-1
Basic example of the application of the strip method: (a) rectangular simply supported
slab; (b) identification of the directions of load transfer for the different regions in the
slab; (c) analysis of the strips

13-14

Figure 12.3.3-1(c) shows that a continuous evolution of bending moments has to be

considered for the strips 2 and 3; indeed, the distances b and c vary continuously from
one strip to another. This would lead to a continuous varying area of reinforcement in
the slab, which is not practical.
In complete accordance with the principle of the static method, one may choose to work
with a slightly adapted load transfer model, in view of the simplification of the practical
arrangement of reinforcement in the slab. An alternative load transfer model is shown in
figure 12.3.3-2. More simple models are shown in figures 12.3.3-3 and 12.3.3-4, in
which a bi-directional solution is adopted for the corners; the value of in expression
(12.3.2-2) is equal to in the corner regions. This load transfer model leads to the
analysis of only four strips with unit width; it is observed that the strips are loaded in the
corner zones with the load q/2. The result of the strip analysis (beam analysis) is a set of
four maximum sagging moments, which are kept constant in the regions of the slab
which are related to the strips considered (see figure 12.3.3-5):
(m yf )11 , (m yf ) 22 , (m yf ) 33 and (m yf ) 44 (notation span moments with f < field)
These span moments then allow to determine the reinforcement areas to be arranged in
the different zones of the slab.

Figure 12.3.3-2
Strip method; first alternative load transfer model: one looks for a rational subdivision
of the slab in zones or regions with practical dimensions (where the reinforcement
meshes will be put later on)

Figuur 12.3.3-3

13-15

Strip method; second alternative load transfer model: two-directional load transfer is
considered in the corners of the slab

Figuur 12.3.3-4
Strip method; further exploitation of the second alternative load transfer model;
4 types of strips are calculated; this leads to 4 maximum span moments (notation f <
field), 2 in the x-direction and 2 in the y-direction
13-16

Figure 12.3.3-5
Strip method; the resulting distribution of the design bending moments

Note:
The strip method does not consider the large torsional moments that are anyway
present in the corners of a simply supported slab (see chapter 11). It may be
necessary to add reinforcement in the corners in order to avoid unacceptable
cracking.

12.3.4 Application of the strip method in particular cases

Preliminary note: up to now in these course notes, attention has been paid to rather
simple slabs with rectangular shape, with simple edge conditions and loaded by a
uniformly distributed load. The strip method offers an interesting approach to tackle
also slabs with:
irregular shapes, such as L-shaped slabs, slabs with openings, etc,
variabel support conditions,
concentrated loads.
It should again be stressed that this approach:
only supports the design calculation at ULS;
delivers a lower bound solution (conservative approach!);
leads to a solution which should be further completed in order to take
account of effects which are not covered by the strip method, such as
torsional moments and accidental (parasitic) fixing moments.

13-17

12.3.4.1

Rectangular slabs with fixed edges

Fixed edges attract much more load than simply supported edges. The tributary areas
may be determined by means of the simple 60-30 and 45-45 model: see figure
12.3.4-1.

Figuur 12.3.4-1
Study of the transfer of the uniformly distributed load on the slab towards the nearest
edges (= subdivision of the slab into tributary areas): fixed edges attract more load than
simply supported edges; the tributary areas may be determined by means of the simple
60-30 and 45-45 model
12.3.4.2

Rectangular slabs with one free (unsupported) edge

As one edge is free (unsupported), the strips that are perpendicular to that edge (and
which have to transfer load!) have no support at the free edge. This may be solved by
adding an internal supporting strip along the free edge. The design procedure is then
completely similar as for a slab with four supported edges: figure 12.4.4-2 shows that
the 60-30 and 45-45 model is used for the study of the load transfer, also at the free
edge, because this one is now replaced by a supporting strip.
The supporting strip is first considered with unit width. The load applied to it, is
composed of:
- the support reaction forces of the strips perpendicular to the free edge and which are
supported by the supporting strip along this free edge;
- eventual applied loads which have to be transferred to the real supports by the
supporting strip itself; this is the case in figure 12.3.4-2(a) where the slab has a
short free edge.

13-18

A more appropriate width of the supporting strip may be determined in order to reduce
the bending moments in the supporting strip. It is recommended anyway to foresee a
concentration of reinforcement in the supporting strip.

45

45

60

45

30

60

45

30

60

45
30

45
(b)

(a)

Figure 12.3.4-2
Strip method applied to a uniformly loaded slab with 1 free edge;
study of the load transfer = subdivision of the slab into tributary areas; (a) short free
edge; (b) long free edge
12.3.4.3

Rectangular slabs with two adjacent free (unsupported) edges

(1) All other edges are simply supported

Slabs with two free edges are characterized by complex static behaviour, especially
when the other edges are simply supported: figure 12.3.4-3. As the strips which are
parallel to the edges, do not have supports at both ends, the major part of the load is
transferred by torsional moments. One may argue that this type of structure (which
appears sometimes in small balconies) is not really a safe way of design.
HILLERBORG (1996) presents a solution based on the introduction of an internal
supporting strip between A and C (see figure 12.3.4-3). He also presents an
alternative (expensive) solution with orthogonal reinforcement parallel to the edges.
This case is not further worked out in the present course notes.

13-19

Figure 12.3.4-3
Rectangular slab with two adjacent free edges and two adjacent simply supported edges

(2) One fixed edge

If a support is fixed, a strip at right angles to the support can act as a cantilever,
carrying load by means of pure bending moments: see figure 12.3.4-4.

Figure 12.3.4-4
Uniformly loaded rectangular slab with 2 adjacent free edges and 1 fixed edges; strip
method: strips act as cantilever beams carrying load by means of bending moments
In principle, if the load is carried by cantilever action, there is only top
reinforcement needed in the slab. This is unacceptable as this may result in wide
bottom cracks (SLS requirements are not fulfilled). A simple way of taking care of
this problem is to assume that a part of the load is carried without cantilever action
but as if the slab were simply supported. The amount of load carried in this way is
chosen to ensure that a suitable minimum amount of bottom reinforcement is
provided, in order to limit crack openings and to take care of the torsional moments
in the SLS. For a slab with regular dimensions, one could start with a choice of
20 % of the total load; this corresponds to the subdivision in tributary areas as is
shown in figure 12.3.4-5. The strips in the x-direction are supported to the right by a
supporting strip which acts itself as cantilever beam.

13-20

Figuur 12.3.4-5
Uniformly loaded rectangular slab with 2 adjacent free edges and 1 fixed edges;
strip method: subdivision in tributary areas load transfer by means of strips in two
directions
12.3.4.4

Concentrated loads on a slab with four supported edges

A concentrated load has a too high value per unit area to be taken directly by a one-way
strip or by two crossing one-way strips, without giving rise to too excessively high local
moments; it may be a point load, a line load or a high load on a limited area.
Concentrated loads have to be considered in combination with the distributed load (selfweight and service loads). It is generally accepted that a separate calculation of the slab
for the concentrated load is only needed when its intensity is larger than 25% of all the
other loads.
The general way of taking care of a concentrated load is by distributing it over a suitable
width by means of specially designed distribution reinforcement.
(1) Solution 1: load transfer in 1 direction
Figure 12.3.4-6 presents the load transfer of a concentrated load to the nearest
supports by only 1 strip. This is a simple and economical way of working, but the
solution might give problems with respect to the SLS requirements.

13-21

Figure 12.3.4-6
Strip method: transfer of a concentrated load to the supports by means of 1 strip; the
load is distributed over a suitable width by means of specially designed distribution
reinforcement

The width a1 of the strip in the y-direction is taken sufficiently large in order to
reduce the bending moment in the strip. In order to be sure that the whole width
cooperates effectively, distribution reinforcement is necessary in the x-direction,
over a limited distance b1 in the y-direction but large enough to limit the moment in
the distribution band. The static analysis of the distribution band (which is
illustrated in figure 12.3.4-7), leads to the maximum total moment in the
distribution band:
m=

F a1 F .a1
. =
2 4
8

(12.3.4-1)

This moment has to be distributed over the width b1; this leads to the bending
moment per unit length in the distribution band, equal to:
m=

F .a1 1
.
8 b1

(12.3.4-2)

The dimensions a1 and b1 are chosen taking into account that reinforcement areas
have to be limited in order to assure sufficient ductility.

13-22

Figure 12.3.4-7
Auxiliary figure for the static analysis of the distribution band that has to distribute the
concentrated load over an effective width a1

(2) Solution 2: load transfer in two directions

It may be necessary (in view of the SLS requirements) to transfer the concentrated
load by means of two strips. The procedure comprises the following steps:
the determination of the distribution ratio between the two parts of the
load that are transferred in the two directions (see chapter 11; method of
equal deflections);
the design of the two orthogonal strips such as in the preceding
paragraph.
12.3.4.5

Concentrated loads on a slab with a free edge

(1) Concentrated load is close to the free edge

The principle is: a concentrated load near a free edge should be transferred by
means of a supporting strip along the free edge: see figure 12.3.4-8. The width of
the band is chosen in order to reduce the moments and thus the reinforcement area.

13-23

Figure 12.3.4-8
A concentrated load close to a free edge is transferred by means of a supporting strip
along the free edge
The application of the concentrated load very closely to the edge, results in a
hogging moment and thus the necessity to arrange top reinforcement at right angles
to the edge: see figure 12.3.4-9. This distribution reinforcement is in essence
necessary to
distribute the load over the supporting strip and to assure the cooperating
effective width of the band;
to limit crack width at the upper side of the slab.
In order to limit crack widths at the top side, HILLERBORG (1996) proposes to
adopt a band width equal to minimum 1/10 to 1/5 of the length of the free edge. The
hogging moment which has to resist by the distribution reinforcement, can be
determined from figure 12.3.4-9, and is equal to:
m = F .b2

width b1 of the
support band

Figure 12.3.4-9

13-24

(12.3.4-3)

Auxiliary figure for the static analysis of the distribution reinforcement in the
supporting strip which has to transfer a concentrated load close to the free edge
(2) Concentrated load at a large distance from the free edge
Figure 12.3.4-10 shows that the transfer of the concentrated load needs a supporting
strip along the free edge.
Support band

Support band

Figure 12.3.4-10
Transfer of a concentrated load in 1 direction (a) or in 2 directions (b)
12.3.4.6

Slab with an opening

The important question is how much the static behaviour of the slab is influenced by the
presence of the opening; determining factors are the shape of the opening, its size and
its position.
Where the static behaviour of the slab is only slightly changed by the opening, the
design may be based on the analysis of the slab without an opening. The reinforcement
which would be cut by the opening must be arranged along its edges and properly
anchored. It is generally accepted that this approximate approach may be used if the
opening can be inscribed in a square with side equal to 0,2 times the smallest span of the
slab; see figure 12.3.4-11. Yet, HILLERBORG (1996) notes that one has to be careful
with this simplification in case of plate regions with large torsional moments and free
edges.

13-25

Figure 12.3.4-11
Slab with less important opening; analysis is performed such as for slab without
opening

Figure 12.3.4-12 represents a slab with a large opening. The following points may be
highlighted:
a possible solution of load transfer in proposed in the figure. The tributary areas
are indicated; it may be observed that the dimensions of the tributary areas have
been adjusted in order to reduce the number of strips to be considered later on;
the proposed solution is based on the use of supporting strips along the edges of
the opening. One may choose between two possibilities:
a supporting strip may be supported at the edges of the slab; this is the case
of supporting strips 1-1 and 2-2 in figure 12.3.4-12, or
a supporting strip may be supported at the crossings with the supporting
strips that are perpendicular to the first ones; this is the case for the
supporting strips AB and CD which are supported by the supporting strips 33 and 4-4 in figure 12.3.4-12.
The last solution seems to be the most logical one for the slab in figure 12.3.4-12.

Figure 12.3.4-12
Slab with large opening ABCD; a possible solution for load transfer is proposed, on the
basis of supporting strips along the edges of the opening

13-26

The analysis then continues with the determination of the bending moments in the
supporting strips, for which unit width is chosen. In practice, the width of the supporting
strips is chosen in such a way that the moments in the supporting strip are distributed
away from the edge of the opening; the width may be chosen up to 1/3 of the distance
towards the nearest parallel edge of the slab.
Figure 12.3.4-13 represents a slab with fixed edges. When openings are present, the
load transfer may also take place by means of cantilever beams; this way of working
allows avoiding supporting strips along the opening. Another consequence is that, in
principle, reinforcement is needed on the upper side; this is the case for the upper part of
the slab in figure 12.3.4-13(b), where sagging moment reinforcement is not necessary in
the x-direction; it is observed that such sagging reinforcement is present in the slab
without opening! (see figure 12.3.4-13(a)). It is also observed that sagging
reinforcement in the x-direction is necessary in the lower part of figure 12.3.4-13(b),
where the supporting strip supports the reaction forces of the strips in the y-direction.
Figures 12.3.4-13(b) and (c) represent two alternative models for load transfer.
Finally, the Annex A12.3.4 shows the results of a finite element calculation (linear
elastic) of a slab with an opening. The necessity of the supporting strips around the
opening is clearly shown; moreover, the figures also give a good indication of the
dimensions and position of the supporting strips.

Figure 12.3.4-13
Slab with fixed edge: (a) force transfer in the slab without opening;
(b) and (c) alternative solutions for the load transfer in the slab with an opening; there is
only one supporting strip!

12.3.5 Case study

13-27

Annex A12.3.5 presents a case study where the strip method is applied to an L-shaped
slab with variable edge conditions.

13-28

11 Chapter 13
Columns
13.1 Introduction - Classification
13.1.1 General
When members are loaded in compression, the question rises if second-order effects
may appear which may lead to instability or to failure of the member. Buckling of
columns with nearly centrally applied compression loads, is the subject of EN 1992-11:2004; 5.8: "analysis of second-order effects with axial load".
The sensitivity of a column for buckling is evaluated on the basis of the slenderness of
the structural member. Second-order effects are more likely to appear with members
who allow larger deformations; therefore, one should be careful in using simplifying
assumptions for the analysis of columns with large values of the slenderness. Inversely,
buckling may be neglected with very stiff columns.

13.1.2 Classification of structures and structural members

Load bearing structures and structural members are classified as:
braced or not, in function of the presence or not of stiffening elements (such as
cores);
sway or non-sway, in function of the sensitivity for second-order effects as a
result of lateral deflections.
The following systems may be identified:
- braced frame (NL: geschoord raamwerk; FR: une ossature contrevente; )
Load bearing structures which contain stiffening elements, which are characterized
by high bending and shear stiffness and which are fixed in rigid foundations, are
considered as braced. The stiffening elements have to transmit the horizontal loads to
the foundation and assure the stability of the braced members which are part of the
structure.
- non-sway frame (NL: niet schrankend raamwerk; FR: ossature noeuds dplacables)
Load bearing structures or structural members with or without stiffening elements,
for which the influence of the displacement of the nodes on the design values of
moments and forces is negligible, are considered as non-sway structures and
elements.
Frames are considered as non-sway structures when the first-order displacements do
not lead to an increase of more than 10% of the internal forces. When the
displacements of the nodes of the frame lead to a substantial increase of moments
and forces, the structure is termed a sway frame.
EN 1992-1-1:2004 only considers the distinction between braced and unbraced
members in its development of simplified criteria for second order effects.

11-1

13.1.3 Types of columns

Whatever the bearing structure may be, the determination of the slenderness of a
column as well as the buckling analysis is performed on the basis of simplified criteria
for an isolated column.

13.2 Determination of the sensitivity of columns for second-order

effects
Reference: EN 1992-1-1:2004; 5.8
13.2.1 The slenderness ratio of a column
For a column with constant cross-section, the slenderness ratio is defined by:

l0
i

(13.2.1-1)

where:
- l0 : the effective length determined from the linear theory of elasticity; in a practical
way, one may say that the effective length l0 corresponds to the distance between
the two cross-sections with curvature = zero;
- i : the radius of gyration (smallest value) of the uncracked concrete cross-section:
i

Ic
Ac

13.2.2 The effective length l0 of a column

13.2.2.1

Isolated column

Figure 13.2.2-1 represents the effective length of isolated columns with different end
conditions.

11-2

Figure 13.2.2-1
Examples of different buckling modes and corresponding effective lengths for isolated
members (figure 5.7 in EN 1992-1-1:2004)

13.2.2.2

Column in the context of a framework

According to EN 1992-1-1:2004; 5.8.3.2 (3), for compression members in regular

frames, the effective length l0 may be determined by means of the following formulas:
braced members:

k1
k2
.1

l0 0,5.l. 1
0,45 k1 0,45 k 2

(13.2.2-1)

unbraced members:

k .k
k
k
l0 l. max 1 10. 1 2 ; 1 1 .1 2
k1 k 2 1 k1 1 k 2

(13.2.2-2)

where:
- l:
the clear height of the compression member between end restraints;
- k1 ,k2: the rotation-flexibilities at ends 1 and 2 respectively.
The rotation-flexibility k is defined by:
k

EI
M

where:
- : the rotation angle (see figure 13.2.2-1);
11-3

(13.2.2-3)

- EI:

bending stiffness of the compression member.

The extreme values of k are:

- k = 0: the theoretical limit for rigid rotational restraint (= perfectly clamped);
- k = : the limit for no restraint at all (= no limitation of the rotation).
EN 1992-1-1:2004 recommends to adopt the minimum values for k1 and k2: 0,1 instead
of 0, in order to take into account the difficulties encountered when perfect clamping
has to be realized in practice.
If an adjacent compression member in a node is likely to contribute to the rotation at
buckling (= the rotation flexibility of the node is determined by two columns), the
rotation-flexibility is defined by:
k

EI

EI
.
M l a l b

(13.2.2-4)

where a and b represent the compression member above and below the node.
It may be noted that the recommendations in EN 1992-1-1:2004; 5.8.3.2 are very
concise, especially about the definition of the rotation flexibility k of a node at the end
of a column in the context of a framework. The practical calculation of the rotation
flexibility k needs back ground information (AENEAS, 2010). The factor k is called
rotation flexibility of the node in the standard; however, k is in fact defined as the
ratio of the stiffness of the columns that meet in the node considered to the stiffness of
the beams that meet in the node considered. With the scheme and notations presented in
figure 13.2.2-2, the factor k for node 1 is defined as:
E.I
E.I

l ca l cb
k1
M

beams

11-4

(13.2.2-5)

column a (ca)
beam 1 left (b 1l)

beam 1 right (b 1r)

node 1
column b (cb)

beam 2 left (b 2l)

beam 2 right (b 2r)

node 2
column c (cc)

Figure 13.2.2-2
Schematic representation of a column in the context of a frame
The factor k gives an indication of the relative importance of the rotational restraint
caused by the adjacent beams, with respect to the stiffness of the columns. A larger
value of k means that the node has more rotational flexibility, which results in a larger
value of the effective length (more buckling danger). Expression (13.2.3-5) may also be
written as:
k

stiffnes columns
stiffness beams

(13.2.2-6)

Large values of k mean that there is less influence of the beams; small values of k
indicate that there is a large influence of the beams.
Figure 13.2.2-3(a) presents the example of a column in the context of a braced frame. It
is assumed that the analysis of the structure leads to the adoption of a clamped edge to
the left of beam 1l, and to the adoption of a simple support at the right of beam 1r. The
relationship between M and the rotation angle of node 1 (see figure 13.2.2-3(c)) may be
determined by means of the energetic theorem of CASTIGLIANO:
f

l
0

M M
.
.dx
E.I F

where
F = the generalized force; in this example, F is the imposed moment M;

11-5

(13.2.2-7)

f = the deformation associated with the generalized force; in this example, f is the
rotation angle caused by the imposed moment M.
light glass
curtain
faade

ca

central
core
b1l

b1 r
1

cb

(a)
Model:

(b)

b1 r

b1l
1
M

(c)

Mr

Ml
(d)

Figure 13.2.2-3
Schematic representation of a column in the context of a braced frame; (a) analysis of
the beams that are concurrent in node 1; (b) the model to be examined; (c) rotational
restraint of node 1 is the result of the stiffness of the two beams; (d) simplified model
for the calculation of the rotational restraint of node 1
The application of CASTIGLIANOs theorem may be avoided by the simplified
approach which considers the sum of the stiffness of the two beams left and right of the
node: see figure 13.2.2-3(d). With that simplification, expression (13.2.2-5) becomes:

11-6

 E.I
E.I
E.I
E .I

l ca l cb
l ca l cb
k1

M
E .I
E .I
bl .

br .

beams
l bl
l br

(13.2.2-8)

where bl and br are coefficients which take into account the nature of the supports at
the other edges of the beams.
For a braced frame, in which the position of the nodes is not changed, the two cases
presented in figure 13.2.2-4 may be used; this figure presents the stiffness as the
moment which is needed for the realization of a unit rotation angle in the node, taking
into account the nature of the other edge of the beam (fixed end or simply supported
end).

Moment required to realize a

unit rotation angle

4.EI
l

3.EI
l

Figure 13.2.2-4
Two frequently occurring cases for the determination of the rotation restraint in a node

Application of the two cases presented in figure 13.2.2-4, leads to the following factor
k1 for the example presented in figure 13.2.2-3:
E.I
E.I
E .I
E .I

l ca l cb
l ca l cb
k1

M
E .I
E .I
4.

3.

beams
l bl
l br
In the case that the right beam would be a cantilever beam without any support at the
right side (see figure 13.2.2-5), the contribution of the right beam to the rotational
restraint of node 1 is zero. The factor k1 is then:

11-7

 E.I
E.I
E.I
E .I

l ca l cb
l ca l cb
k1

M
E .I
E .I
4.

0.

beams
l bl
l br
ca
b1l

b1r
1

(a)

cb

Model:
b1l

b1r
1

Mb1l

Mb1r

(b)

(c)

EI
M 4. .
l b1l

stiffness = 0

Figure 13.2.2-5
Study of a column edge in the context of a braced frame; (a) one of the beams that are
concurrent in node 1 is a cantilever beam; (b) the model to be examined; (c) rotational
restraint of node 1 is the result of the stiffness of the left beam only
For columns in unbraced frames, it is necessary to consider other models, such as the
one presented in figure 13.2.2-6, in order to evaluate the influence of the beams on the
rotational restraint of a node.

11-8

Figure 13.2.2-6
Study of a column in the context of an unbraced frame; the buckling scheme of the
column leads to more elaborated models for the evaluation of the influence of the beams
on the rotational restraint of the nodes of the column (AENEAS, 2010)

13.2.3 The slenderness criterion for isolated members

Reference: EN 1992-1-1:2004; 5.8.3.1.

Second-order effects may be ignored if the slenderness is below a certain limit value
lim, defined by the following expression:

lim

20. A.B.C
n

(13.2.3-1)

where:
-

N Ed
Ac . f cd

1
1 0,2. ef

ef = effective creep coefficient (see EN 1992-1-1:2004; 5.8.4 and

3.1.4); if ef is not known, A = 0,7 may be used.
-

B 1 2.

11-9

As . f yd
Ac . f cd

If is not known, B = 1,1 may be used.

-

C 1,7 rm
where rm

M 01
; M01 and M02 are the first-order moments at the ends
M 02

of the column and M 02 M 01 . If the end moments M01 and M02 have
the same sign, rm is positive. If M02 = M01, then is rm = 1 and C = 0,7;
if M02 = -M01, then is rm = -1 and C = 2,7.
If rm is not known, C = 0,7 may be used.
In the following cases, rm should be taken as 1 and C = 0,7:
- for braced members in which the first-order moments arise only
from imperfections or transverse loading;
- for unbraced members in general.
In its most simple formulation (with A = 0,7; B = 1,1 and C = 0,7), the expression
(13.2.3-1) is formulated as follows:

lim

10,78
N Ed
Ac . f cd

(13.2.3-2)

13.2.4 A worked out example of the determination of the sensitivity of a column for
second order effects

See annex A13.2.4.

13.2.5 An alternative method for the evaluation of the sensitivity of columns for
second order effects

The EC 2 is a living document. Recommended simplified methods are often the result
of PhD research and are derived from elaborated analytical, experimental and numerical
analysis. This explains why models are updated throughout the successive versions of
the standard.
The method for the determination of the effective length of a column in the context of a
framework, which is presented above, is included in the version 2004 of the EC2. This
model differs from the simplified method presented in the previous version of EC2 (EN
1992-1-1:1998), where the effective length of a column was determined by means of an

11-10

easy to use nomogram. The results of both methods (versions 2004 and 1998) are quite
similar, but the approach in the 2004 version is more fundamental.
The simplified method for the determination of the effective length of a column in the
context of a framework, according to the previous version of EC2 (EN 1992-1-1:1998)
is presented in annex A13.2.5. This annex also includes the slenderness criterion which
was recommended by the 1998 version of the standard.

13.3 Design of a non-slender column

See Chapter 4: bending combined with axial force. Indeed, it was already highlighted in
chapter 4 that there is always uncertainty about the real position of the centrally
applied axial force on a column; therefore, the standard proposes to use a an additional
eccentricity and thus to perform the design calculation in ULS in the situation of
combined bending.

13.4 Buckling analysis of slender column in reinforced concrete

Buckling is not considered in this course, but is a topic discussed in the optional course
on Inelastic design of concrete structures (2nd Master BRUFACE).

13.5 Detailing
Reference: EN 1992-1-1:2004; 9.5
13.5.1 Longitudinal reinforcement
- Minimum bar diameter: EN: 8 mm; ANB: 12 mm
- Minimum amount of longitudinal reinforcement As,min is given by the following
expression:
As ,min

0,10.N Ed
0,002. Ac
f yd

(13.5.1-1)

where:
fyd : the design yield strength of the reinforcement;
NEd : the design axial compression force;
Ac : area of concrete section.
- Maximum amount of longitudinal reinforcement As,max

As ,max 0,04. Ac (0,08. Ac at lap locations)

11-11

(13.5.1-2)

- The longitudinal bars should be distributed along the circumference of the column
cross-section. For columns having a polygonal cross-section, at least one bar should
be placed at each corner. The number of longitudinal bars in a circular column
should not be less than four.
13.5.2 Transverse reinforcement

In order to prevent buckling of the longitudinal reinforcement in columns, every

longitudinal bar (or bundle of bars) placed in a corner should be held by transverse
reinforcement. In practice, stirrups are most commonly used as ties; sometimes, a
continuous helical spiral is used as an envelope of the longitudinal bars all along the
column length. No bar within a compression zone should be further than 150 mm from a
restrained bar. Examples of cross-sections with ties are shown in figure 13.5.2-1.

stirrup

tie

Figure 13.5.2-1
Tie arrangements in a typical rectangular or a square reinforced concrete column

- Minimum diameter for links (= stirrups), loops or helical spiral reinforcement: the
largest of the following values should be taken:
6 mm;
1/4 . maximum diameter of the longitudinal bars.
For wires of welded mesh fabric used for transverse reinforcement: minimum 5 mm.
- Maximum spacing: not more than the smallest of the following three distances:
20 times (EN) or 15 times (ANB) the minimum diameter of the longitudinal
bars;
the lesser dimension of the column;
400 mm (EN) or 300 mm (ANB).
The maximum spacing should be reduced by a factor 0,6:
in sections within a distance equal to the larger dimension of the column
cross-section, above or below a beam or slab;

11-12

 near lapped joints, if the maximum diameter of the longitudinal bars is

greater than 14 mm. A minimum of 3 bars evenly placed in the lap length is
required.

Note 1:
Where the direction of the longitudinal bars changes (for example at changes in
column size), the spacing of transverse reinforcement should be calculated
taking account of the lateral forces involved.

Note 2:
The prescriptions related to the positioning of stirrups in laps of longitudinal
bars have to be applied.

11-13

14 Chapter 14
Design with strut-and-tie models
14.1 Introduction
14.1.1 Plastic analysis
The analysis of shear in reinforced concrete members is based on the so called truss
analogy (see chapter 7 in these course notes). The truss model helps to understand how
forces are transferred throughout a structural member. The equivalent truss model is
identified on the basis of the crack pattern that appears when the member is loaded; the
crack pattern helps to describe load transfer in the member by means of forces that are
concentrated in compressive struts and tensile loaded bars.
The truss analogy proposed by RITTER-MRSCH (1899) for the calculation of shear
reinforcement in reinforced concrete members is the oldest and well known example of
reasoning by means of strut-and-tie models (NL: staafwerkmodellen; FR: modles
composs de bielles et tirants). However, the method is also very well suited for the
design calculation of compact structural members such as foundation slabs (or blocs)
supported by piles, corbels, deep beams, walls and anchorage regions in prestressed
members. The strut-and-tie method is essentially founded on the publications (years
1980-90) of SCHLAICH (Univ. Stuttgart).
Designing members by means of the strut-and-tie method belongs to the domain of the
application of the plastic design philosophy, and more in particular to the domain of
(lower bound) static methods. Indeed, the design is based on a truss model in which
acceptable stresses are not exceeded in the members of the truss, and is thus based on a
statically acceptable system, which may transfer the loads with sufficient safety. If on
top of that, the directions of the compression struts coincide with the orientations of the
stresses in uncracked situation, there is no need for large redistributions in order to
activate the truss system; consequently, deformations and crack widths remain limited
which is a visual proof of the well functioning of the truss system in service conditions.
The philosophy described above is the one of the lower bound methods for plastic
design in ULS. According to the lower bound theorem, the structural member does not
fail if for the imposed load combination, a stress distribution can be identified for
which:
equilibrium is guaranteed (the system of forces is in equilibrium with a given set
of loads), and
the threshold for plastic behaviour of materials is not exceeded.
The present course notes are essentially based on the following references:
LAMBOTTE (1989), WALRAVEN (1995), PROVOST (1998) and WIGHT (2009).

14.1.2 Application domain

Every structural member can be subdivided in so-called B-regions and D-regions.

14-1

B stands for BERNOULLI and corresponds to those parts of the member where the
assumption of BERNOULLI can be accepted (planar sections remain plane after load
has applied). For example: the assumption is accepted for the analysis of long beams
loaded in bending where the load transfer is continuous over long distances: cracks and
compression struts between the cracks are parallel to each other over the whole region.
D stands for Discontinuity and corresponds to the regions where the assumption of
BERNOULLI is no longer valid; this is typically the region around:
a geometrical discontinuity: changes in cross-section, openings, nodes in
frames, connections between girders and beams, etc.;
a static discontinuity: isolated loads, supports, temperature changes, anchorage
of prestressing tendons, etc.
D-regions are characterized by the disturbance of the load transfer system, due to
sudden changes in dimensions, by changes in orientation of load transfer due to the
presence of openings, the connection with other structural members, the introduction of
concentrated loads, etc.
Figure 14.1.2-1 shows the truss model in a beam; the figure shows the regular pattern of
cracks and struts between the cracks in the B-region and the fan-like distribution of the
cracks at the position of the concentrated loads (support reaction force included).
Figure 14.1.2-2 shows a frame structure with subdivision into B- and D-regions. In this
figure, the principle of BARRE DE St VENANT is used, saying that the localized effect
of a disturbance dies out by about one member-depth from the point of disturbance. On
this basis, D-regions are assumed to extend one member-depth each way from the
discontinuity.
Although the strut-and-tie method may also be applied in B-regions (see chapter 7 for
shear design by means of the truss model of MRSCH), it is above all THE method to
be used for the design of D-regions. Strut-and-tie models are helpful tools to design the
necessary reinforcement which helps to arrange the transfer of loads through the Dregions towards the B-regions.

B-region

D-region

Figure 14.1.2-1
Truss model for a beam loaded in bending with B- and D-regions

14-2

B-region

D-region

Figure 14.1.2-2
Frame structure with indication of B- and D-regions. The conceptual principle of DE St
VENANT is used as a quantitative guide in selecting the dimensions of D-regions: Dregions are assumed to extend one member-depth each way from the discontinuity

14.2 Principle aspects of the strut-and-tie method, explained by means

of an example
14.2.1 Identification of the strut-and-tie model
The principle aspects of the strut-and-tie method are explained in this paragraph by
means of the example of a deep beam which has to transfer the concentrated load P of a
column towards two foundation piles.
In accordance with the definitions presented in paragraph 2.5.2.1 in these course notes,
the beam is characterized by a length l < 3.h, and is thus identified as a deep beam
(NL: dragende wand; FR: poutre-cloison). The principal stress trajectories in the deep
beam, in uncracked situation, are presented in figure 14.2.1-1(a); this permits to deduce
a plausible truss model (figure 14.2.1-1(b)), composed of:
two concrete compression struts with inclination angle with respect to the
horizontal axis;
a horizontal tie in the lower part of the deep beam.

14-3

Figure 14.2.1-1
Deep beam; (a) stress trajectories in uncracked situation; (b) truss model

Each compression strut has to transfer the load:

Pc =

P
2. sin

(14.2.1-1)

P
2. tg

(14.2.1-2)

The tie has to transfer the tensile force:

Ps =

It is important to limit the concrete compression stress in the nodes and the struts as well
as to assure the anchorage of the ties in the nodes. The truss analysis has thus to
consider the performance of the struts, the ties and also the nodes.
14-4

14.2.2 The ties

The tensile force Ps may be resisted by a bundle of parallel bars. It is of essential

importance that the bars are efficiently anchored in the nodes; in such situation, the
design yield strength of the steel fyd may be used to design (in ULS) the reinforcement
area As by the expression:
As =

Ps
f yd

(14.2.2-1)

The anchorage of reinforcement bars is discussed in chapter 6 in these course notes,

with topics such as anchorage plates (limited usefulness), anchorage length, reduction of
anchorage length by means of hooks, loops and welded transverse bars; loops are very
efficient because they create a favourable stress distribution in the concrete enclosed
within the loop (multi-axial loading of the concrete in the node).

Note:
The truss model in itself does not take care of crack development considerations;
two strategies may be applied to limit crack development:
the stresses in the steel reinforcement should be limited to values lower than
fyd;
the tie-bars should not be concentrated around the axis of the tie in the truss
model, but should be distributed in the zone where, according to the elasticity
theory, the largest tensile stresses appear in the concrete. It is preferred to
arrange for more bars with smaller diameter, but lack of space may hamper
this.

14.2.3 The compression struts

The identification of the axis of the concrete compression struts in a truss model is in
general not a difficult problem. It is less evident to determine reasonable values for the
dimensions of the cross-section of a strut, which in reality has a bottle-like shape (see
ffigure 14.2.1-1(a)). A simplified solution is to consider an idealized prismatic strut. The
rectangular cross-section of the strut is determined by the smallest value of b1 and b2 on
the one hand (see figure 14.2.1-1(b)) and by the dimensions of the nodes on the other
hand.
With the assumption (for reasons of simplicity) that the width of the struts equals b, the
cross-section of the compression strut is (see figure 14.2.3-1):
b.a3 = b.

a1 / 2
sin

14-5

(14.2.3-1)

The force Pc in the strut defined by expression (14.2.1-1) is distributed over the crosssection, which leads to the compression stress in the strut:

c ,schoor

P
Pc
P
=
= 2. sin =
b.a1
b.a3 b. a1
2. sin

(14.2.3-2)

In this example, the stress in the compression strut is equal to the stress in the contact
area between the deep beam and the column (the area used for the introduction of the
load into the deep beam). Consequently, the node is subjected to a bi-axial stress
situation where stresses are the same in all directions; this corresponds to a (planar) state
of hydrostatic pressure (represented by 1 point in MOHRs circle).

Figure 14.2.3-1
Detailed representation of the node where the load of the column is introduced into the
deep beam. A hydrostatic stress state is observed

The compression stress in the strut is compared to a threshold value, which is taken as a
fraction of fcd (analogous reasoning as in chapter 7 dealing with struts in the analogous
truss system for the calculation of shear). EN 1992-1-1:2004; 6.5.2(2) recommends to
limit the compression stress in the struts to:

c ,schoor 0,6. . f cd
where:

14-6

(14.2.3-3)

f cd =

f ck
(ANB: coefficient 0,85 is not used for strut-and-tie design!)
1,5

=1

f ck
250

(14.2.3-4)

and fck in MPa.

The reasons for the reduction of fcd are mainly:
the fact that the strut is subjected to a two-dimensional stress state with
compression along the axis of the strut and tension in the transversal direction
(which will be taken up by reinforcement). The tensile stress may lead to splicing
cracks parallel to the struts axis;
the uncertainty about the exact dimensions of the compression strut. Indeed, the
idealized prismatic strut is surrounded by concrete that cooperates in the
transmission of a part of the load (see the bottle-like shape of the compression struts
in figure 14.2.1-1(a)); this is even becoming clearer when cracks appear
perpendicular to the tie: see figure 14.2.3-2.

Figure 14.2.3-2
Deep beam: principal stress trajectories in cracked situation

Note 1:
The design strength for a concrete strut in a region with transverse compressive
stress or no transverse stress may be fcd, without reduction coefficient. It may
even be appropriate to assume higher design strength in regions where multiaxial compression exists.

Note 2:
More elaborated models for struts are available in EC2, where the bottle-like
shape is translated into a sub-truss system with additional ties: see figure 14.2.33.

14-7

Figure 14.2.3-3
More elaborated models for compression struts (figure 6.25 in EN 1992-1-1:2004;
6.5.3)

14.2.4 The nodes

The dimensioning and detailing of concentrated nodes are critical in determining their
load-bearing resistance. Concentrated nodes may develop where point loads are applied,
at supports, in anchorage zones with concentration of reinforcement or prestressed
tendons, at bends in reinforcing bars and at connections and corners of members.
It was already observed in the preceding paragraphs that distinction has essentially to be
made between:
nodes where only compression struts are concurrent; a multi-axial
compression stress state is observed in the node, which is a far more
comfortable position for the concrete than in the case of the onedimensionally loaded concrete strut;
nodes where ties are anchored. The main issue is off course the anchorage of
the tensile reinforcement, but attention has also to be paid to the verification
of the compression stress in the concrete in the node, which appears to be
less resistant due to the presence of the tie.
EN 1992-1-1:2004; 6.5.4 recommends limited design values for the compression
stresses within nodes:
in compression nodes where no ties are anchored at the node (see figure
14.2.4-1): Rd,max = .fcd where fcd = fck/1,5
in compression-tension nodes with anchored ties provided in one direction
(see figure 14.2.4-2): Rd,max = 0,85..fcd where fcd = fck/1,5
14-8

in compression-tension nodes with anchored ties provided in more than one

direction (see figure 14.2.4-3): Rd,max = 0,75..fcd where fcd = fck/1,5

Figure 14.2.4-1
Concrete compression stress verification in a compression node without ties (figure 6.26
in EN 1992-1-1:2004; 6.5.4)

Figure 14.2.4-2
Concrete compression stress verification in a compression-tension node with anchored
ties provided in one direction (figure 6.27 in EN 1992-1-1:2004; 6.5.4)

14-9

Figure 14.2.4-3
Concrete compression stress verification in a compression-tension node with anchored
ties provided in more than one direction (figure 6.28 in EN 1992-1-1:2004; 6.5.4)

Note 1:
EN 1992-1-1:2004; 6.5.4 (5) allows to increase the design compressive stress
values defined above, by up to 10% where at least one of the following applies:
triaxial compression is assured;
all angles between struts and ties are 55;
the stresses applied at supports or at point loads are uniform, and the node is
confined by stirrups;
the reinforcement is arranged in multiple layers;
the node is reliably confined by means of bearing arrangement or friction.
Note 2:
EN 1992-1-1:2004; 6.5.4 (6) allows to increase the design compressive stress
value in a triaxially compressed node, by up to Rd,max = 3..fcd if for all three
directions of the struts the distribution of load is known.

14.2.5 Final principle remarks

In many cases, it appears that the limitation of the concrete stress in the struts is not
a critical issue.
The objective of the design by means of the strut-and-tie method is to get first
yielding of the steel reinforcement, before the compression stress limit is reached in
the struts.

14-10

14.3 Practical development of a strut-and-tie model: a procedure in 6steps

In the following paragraphs, the elaboration and exploitation of a strut-and-tie model is
explained on the basis of a procedure composed of six steps.

14.3.1 Step 1: the identification of the B- and D-regions

The structural member is subdivided into:

regions with regular transmission of forces according to BERNOULLIs
assumption (B-regions), and
regions around discontinuities related to changes in dimensions, introduction
of concentrated forces, etc (D-regions). The D-region is identified by means
of St.-VENANTs principle: the disturbance of the regular stress distribution
extends one member-width from the discontinuity. Yet, figure 14.3.1-1
shows that the D-region not only depends on the depth of the member (a
foundation bloc), but also on the width over which the loads have to be
smeared out. Figures 14.3.1-2 to 14.3.1-4 present several examples of Dregions caused by geometrical discontinuities (changes of dimensions,
presence of corners and openings) or by static discontinuities (application of
concentrated loads) (LAMBOTTE, 1989).

Figure 14.3.1-1
Example of a foundation block with two different loading conditions: the size of the Dregion is also determined by the load distribution into the member

14-11

Figure 14.3.1-2
Some examples of D-regions caused by geometrical discontinuities

Figure 14.3.1-3
Some examples of D-regions caused by static discontinuities (application of
concentrated loads)

14-12

static and
geometrical

geometrical

Figure 14.3.1-4
Some examples of D-regions for members with combination of static and geometrical
discontinuities

14.3.2 Step 2: isolation of the D-region and identification of the loads at the edges of
the D-region

The loads that are applied onto the edges of the D-region have to be identified:
on one edge, the applied forces;
on the other edges, the forces which are the result of the stress distributions
along these edges, determined according to the elasticity theory in the
adjacent B-region.
Figures 14.3.1-2 to 14.3.1-4 present several examples of forces and resulting stress
distributions along the edges of the D-regions.
For further calculation purposes, distributed loads along edges are replaced by their
resultant forces, in order to obtain later on, the transmission scheme of forces
throughout the D-region. The D-region is calculated for the forces acting on it; it is
evident to say that all applied forces (on all edges) have to verify the conditions of static
equilibrium.

14.3.3 Step 3: analysis of the load transfer throughout the D-region in view of the
identification of a suitable truss model

A suitable truss model has to be identified for the design calculation. However, in order
to find a solution which meets the SLS requirements (avoiding excessive crack widths),
the chosen truss system should be as much as possible in agreement with the real stress
state situation according to linear elasticity theory (lower bound theorem for plastic
14-13

methods). The ideal situation is that the scheme of principal stress trajectories would be
known (as a result of a preliminary FEM calculation for example). The stress
trajectories show how the imposed loads are transferred throughout the structural
member to the supports. Figure 14.3.3-1 illustrates that point for two corbels.
compression trajectories
tension trajectories

zone without stress

Figure 14.3.3-1
Principal stress trajectories for two types of corbels

It should be noted that in many cases, it is possible to draw a workable approximate

scheme of load transfer without the knowledge of the exact linear elastic stress
trajectories. This is illustrated in figure 14.3.3-2 where simple (almost intuitively
determined) load paths are shown. It is important to notice that when curvatures in the
load path appear (where the action line of the forces changes orientation), reaction
forces have to be resisted.

Note:
The passage of the forces between B- and D-regions should be compatible over
the whole structural element: detail design should be performed taking account
of the adjacent regions.

14-14

tension
compression

compression

compression

tension

tension

Figure 14.3.3-2
Quickly sketched load paths throughout D-regions

14.3.4 Step 4: identification of the strut-and-tie model

The analysis performed in step 3 allows to identify the strut-and-tie model which has to
assure the internal equilibrium of the member for the imposed loading conditions. In
most cases, the aim is to identify a statically determinate system, which allows the
determination of the tensile and compression member forces by means of equations
expressing the equilibrium of the nodes in the truss. The following points should help
with the identification of a suitable strut-and-tie system:
each time the load path orientation changes, a node should be added to the strutand-tie system: see figure 14.3.3-2;
the ties should have orientations that lead to easy arrangement of the reinforcement
bars in the execution phase. The reinforcement corresponding to one tie may be
realized in practice by means of a bundle of bars so that the resultant force
corresponds with the tie force in the idealized truss system. However, it may also be
argued that trying to respect scrupulously the tie orientation in the model, is not
necessary because the scheme of stress trajectories changes when cracks appear;
the strut-and-tie model should be as simple as possible, with a minimum value for
the total length of the tensile reinforcement. Indeed, the model has to assure load
bearing capacity in ULS (thus after crack development) with the least possible
forces and deformations. In order to make a choice between two or more truss
models which seem to be plausible solutions, the criterion of minimum energy
(minimum work done) may be used in its most simple formulation:

F .l
si

= minimum

where: Fsi = tensile force in tie N i of the truss model (s< steel);
li = elongation of the tie N i.

14-15

(14.3.4-1)

This criterion is used without consideration of the compression struts, because

deformations are rather small with respect to the ones of the ties. Figure 14.3.4-1
presents the example of a deep beam for which, on the basis of the stress trajectories,
two plausible strut-and-tie models may be identified. It may be shown that the second
model (characterized by a longer total tie-length) leads to larger deformations.

Figure 14.3.4-1
Deduction of plausible strut-and-tie models from the scheme of the stress trajectories:
(a) compression arch model; (b) suspension model
Some examples of frequently used strut-and-tie models are presented in figure 14.3.4-2.

Note:
Sometimes, statically indeterminate systems have to be considered. Figure
14.3.4-3 shows a beam which transmits a support reaction force to a wall (beam
is perpendicular to the wall). It is recommended to include the part of the wall on
top of the beam in the load transfer system; otherwise, cracks may appear in this
part.

14-16

Figure 14.3.4-2
Examples of frequently used strut-and-tie models

14-17

tie bar

possible cracking

Figure 14.3.4-3
Example of a statically indeterminate strut-and-tie system, which is needed for the
modelling of the transfer of the support reaction force of the beam to the wall

14.3.5 Step 5: computation phase

On the basis of the chosen strut-and-tie model, the strength of struts, ties and nodes is
verified according to the principles presented in paragraph 14.2. If the method is used
for design purposes (determination of dimensions, reinforcement areas), overall
dimensions are generally determined on the basis of the conditions for stress limitation
in the concrete struts. This leads to the dimensions of the nodes and to limit values of
the inclination angles and to the overall dimensions b and h of the structural member.
In other cases, overall dimensions are given, except for the reinforcement:
reinforcement areas and position of the reinforcement have to be determined. Although
the stresses in the compression struts are in general not determining in such cases,
sufficient attention should always be paid to the verification of the struts.

14.3.6 Step 6: identification of the reinforcement

The tensile forces in the ties have to be taken up by reinforcement bars. The
reinforcement may be composed of several bars which are oriented according to the axis
of the tie in the strut-and-tie model. In order to prevent surface cracking, additional
reinforcement may be necessary, such as links and technological reinforcement.
.

14-18

14.4 Applications and examples

14.4.1 Deep beams

This paragraph focuses back again on the example of deep beam that was already
presented in figure 14.2.1-1(b). It is assumed that the span (2a) and the support
dimensions (a2) are known as well as the imposed load P.
Question: to determine h (and d) and As.
The solution starts with the choice of the strut inclination angle . With a choice = 45o

, one finds:

d a. tg
The next step is the determination of the cross-section of the prismatic compression
struts and the verification of the compression stress in the struts:

c =

P
c , strut ,max
2. sin . Astrut

where:

c , strut ,max = 0,6. . f cd and = 1

f ck (MPa )
250

With c much smaller than the maximum value, the depth d may be reduced (which
means that less inclined struts are possible); when c > 0,6..fcd, the depth d has to be
increased.
Finally, the reinforcement area As is determined by:
As =

P
2. tg . f yd

EN 1992-1-1:2004; 9.7 specifies that deep beams should be provided with an

orthogonal reinforcement mesh near each face, with a minimum As,dbmin (db < deep
beam):
As ,db ,min 0,1%. Ac
150 mm 2 /m
in each face and each direction (vertically and horizontally).
where:
As,db,min = minimum reinforcement in each face and each direction of the deep beam
bruto concrete cross-section for the vertical or horizontal direction.
Ac =
14-19

The distance between two adjacent bars (between horizontal bars and between vertical
bars):
2 the deep beam thickness;
300 mm.

Note:
Figure 14.4.1-1 shows the transition from deep beams (short span beams)
towards long span beams (characterized by the presence of the regularly spaced
stirrups).

Figure 14.4.1-1
Transition from deep beams (short span beams) towards long span beams (characterized
by the presence of the regularly spaced stirrups): (a) strut-and-tie model; (b) main
reinforcement

14.4.2 Walls

Walls are large D-regions for which typical D-models may be applied.

14-20

14.4.2.1

Example 1

Figure 14.4.2-1 shows the example of a wall supported by two columns, loaded by a
uniformly distributed load on the top side of the wall. A typical strut-and-tie model is
presented in figure 14.3.4-2 (case e.). Figure 14.4.2-1 also presents the evolution of the
horizontal tensile stress in the mid-section of the wall.

Figure 14.4.2-1
Typical strut-and-tie model for a wall supported by 2 columns, loaded by a uniformly
distributed load on the top side of the wall

The main tensile reinforcement is concentrated at the bottom side of the wall. For a wall
with a ratio d/l approximately equal to 1 and a/l 0,1, the tensile force reaches the value
of about 0,2.q.l. The main horizontal tensile reinforcement should be arranged over the
whole length of the wall and should be anchored preferably with hooks and loops, in
order to avoid the tearing off of the corners above the supports. For walls with large
dimensions, the main tensile reinforcement should also be distributed over a depth
0,15.d in order to assure a better crack distribution: see figure 14.4.2-2.
Finally, the region where the load is applied should be reinforced by means of vertical
links (see further in these course notes).

Figure 14.4.2-2
Reinforcement of the wall presented in figure 14.4.2-1

14-21

Prescriptions EN 1992-1-1:2004; 9.6:

vertical reinforcement
- should be provided at both sides of the wall;
- the area of the vertical reinforcement (for both sides together) should lie
between 0,002.Ac and 0,04. Ac;
- the distance between two adjacent vertical bars (s1 in figure 14.4.2-2) shall not
exceed (3 the wall thickness t) or 400 mm, whichever is the lesser (the
smallest value should be used).

horizontal reinforcement
- horizontal reinforcement running parallel to the faces of the wall (and to the
free edges) should be provided at each surface;
- the area of the horizontal reinforcement (for both sides together) should be:
25% of the total vertical reinforcement for both sides
0,001.Ac
- the spacing between two adjacent horizontal bars (s2 in figure 14.4.2-2) should
not be greater than 400 mm.

transverse reinforcement
- where the total area of the vertical reinforcement in the two faces exceeds
0,02.Ac, transverse reinforcement in the form of links should be provided in
accordance with the requirements for columns: the links are necessary to avoid
buckling of the vertical bars. The technological provisions for walls are
identical to the ones for columns (EN 1992-1-1:2004; 9.5.3); figure 14.4.2-3
presents a summary of the provisions for walls.
6 mm
0,25.max,vertical bar

20.max,vertical bar
( 15.max,vertical bar in ANB)
wall thickness
400 mm ( 300 mm in ANB)

slab 2

4 thickness
l

slab 1

wall
4 thickness

in these regions:
spacing 0,6
(danger of splitting with
introduction of forces)

minimum 4 links / m2 wall area

Figure 14.4.2-3
Detailing of the transverse reinforcement necessary to link the reinforcement meshes at
both sides of the wall

14-22

14.4.2.2

Example 2

Figure 14.4.2-4 presents the case of a wall supported by two columns and loaded by a
uniformly distributed load applied at the bottom side of the wall.
A typical strut-and-tie model is represented in figure 14.3.4-2 (case f). The model shows
the development of a compression arch from which the distributed load and the selfweight is suspended. Consequently, vertical reinforcement is necessary up to the upper
side of the wall: figure 14.4.2-5. The stresses in the horizontal direction are practically
the same as in example 1 (load applied on top of the wall) which leads to an identical
horizontal reinforcement scheme.

Figure 14.4.2-4
Typical strut-and-tie model for a wall supported by two columns and loaded by a
uniformly distributed load at the bottom side of the wall

Figure 14.4.2-5
Reinforcement of the wall presented in figure 14.4.2-4
14-23

14.4.2.3

Example 3

Figure 14.4.2-6(a) presents a wall with large height. The wall is supported by two
columns. A concentrated load (a column) is applied in the middle of the upper side of
the wall. The depth of both D-regions at the upper and lower side of the wall, is equal to
the width b (applying St VENANTs principle). At the interface between the B-region
P
and the two D-regions, one observes the compression stress
where t = wall
b.t
thickness; see figure 14.4.2-6(b).

Analysis of the upper D-region

The distributed stresses at the B-D interface are replaced by two resultant forces
1
P . This allows drawing the strut-and-tie model: see figure 14.4.2-6(c). This leads
2
to the identification of the horizontal tensile force N, which is translated into a
certain number of horizontal ties. The ties embrace the peripheral reinforcement
and are distributed over the region where N is active: see figure 14.4.2-6(d).

Analysis of the lower D-region

An analogous reasoning leads to the identification of tensile reinforcement at the
bottom of the wall.

stirrup

tie

Figure 14.4.2-6

14-24

Analysis of a high wall by means of strut-and-tie models; (a) subdivision in B- and Dregions; (b) identification of the load transfer; (c) strut-and-tie models; (d) scheme of
the reinforcement

14.4.3 Reinforced concrete blocs with concentrated loads

14.4.3.1

Centrically applied load

This case has already been discussed in the paragraph above (the upper D-region in
figure 14.4.2-6). This case is frequently encountered in practice, for example in the
context of anchorage regions in prestressed girders, column heads and column footings
where concentrated compression forces are applied, etc.
The following notes present the well known results of the analysis by SCHLAICH
(1981) of a reinforced concrete bloc with limited thickness, loaded by a centrically
applied load.
Figure 14.4.3-1 represents the stress trajectories, the tensile stress in the plane of
symmetry and the evolution of the stress in a plane, parallel to the plane of symmetry, at
the distance 0,3.h, for different values of a1/a: see figure 14.4.3-1(b). Figure 14.4.3-1(c)
represents a plausible strut-and-tie model, figure 14.4.3-1(d) represents the values of
F1/P for h = a and different values of a1/a; figure 14.4.3-1(e) represents (for a1/a = 0,1)
the influence of the ratio a/h on the values of F1/P and . SCHLAICH developed the
following empirical expression for the maximum stresses x shown in figure 14.4.31(b):

x ,max = 0,5.

P a1
.1 (for h = a)
b.a
a

(14.4.3-1)

The curve that represents the evolution of F1/P in figure 14.4.3-1(d) may be
approximated (in a safe way) by the straight line passing through the end points; the
equation of this straight line is:

a
F1 = 0,3.P.1 1
a

(14.4.3-2)

Another possible assumption is to put the action line of the tensile force F1 at a distance
0,4.a from the bottom edge of the bloc. With this assumption, almost the same
approximate relationship is found (with 0,31 instead of 0,3).
F1
is arranged by means of closed stirrups, put in horizontal
f yd
position. The stirrups are distributed in a zone with an approximate height of a/2,
symmetrically with respect to the action line of F1.

The reinforcement As =

14-25

Stresses

Stress trajectories

and for

Strut-and-tie model

and for

Figure 14.4.3-1
Analysis of a reinforced concrete bloc with limited thickness (SCHLAICH, 1981)

14.4.3.2

Eccentrically applied loads

Figure 14.4.3-2 presents the stress trajectories as well as a plausible strut-and-tie model.

14-26

Figure 14.4.3-2
Application of the strut-and-tie model method for the analysis of a thin bloc in
reinforced concrete, loaded by an eccentrically applied concentrated load:
(a) stress trajectories; (b) strut-and-tie model

This model may be applied to the wall presented in figure 14.4.3-3; the length of the Dregion is once again b. The stress distribution at the B-D-interface may be determined
from theory of elasticity; this leads to the determination of the resultant forces D1
(compression) and N (tensile). There are no discontinuities at the B-D-interface, which
allows the compression force D1 to enter in an undisturbed way into the D-region
(dashed line passing by point c). Regarding the position of the tensile force N, as
reinforcement is always needed along the edges of walls and blocs, the reinforcement
for N is arranged as close as possible to the edges, but taking into account the cover
requirements. On the basis of these edge conditions and in combination with the scheme
of the stress trajectories, a suitable strut-and-tie model may be identified: figure 14.4.33(b).
Special attention has to be paid to point d. The compression diagonal cd is situated in
the mid-plane of the wall, while the reinforcement is put along the side edges. This
means that a transversal tensile component N0 is necessary to assure equilibrium; extra
links are thus needed to take up N0.

14-27

links

Figure 14.4.3-3
Analysis of a D-region in an eccentrically loaded wall; (a) stress trajectories;
(b) strut-and-tie model; (c) reinforcement

14.4.3.3

Three-dimensional case

With large three-dimensional reinforced concrete blocs, load paths for the concentrated
load may be considered in two directions; this is the approach adopted in figure 14.4.34, where the strut-and-tie method is applied in two perpendicular planes. The
reinforcement is composed of meshes put in planes that are perpendicular to the action
line of the concentrated load.

14-28

Figure 14.4.3-4
Foundation bloc loaded by a centrically applied concentrated load; the strut-and-tie
method is applied in two different planes

Figure 14.4.3-5 presents the case of foundation blocs supported by 3 or 4 piles. The load
is transferred in multiple directions. The strut-and-tie model should be based on the
principle that loads are always transferred to the piles along the shortest load paths. The
ties in the model are thus situated along the shortest distance between the pile heads.
The better solution is thus to concentrate the reinforcement above the pile heads instead
of having reinforcement distributed over the whole foundation bloc.
For widely spaced piles (l > 4.p in figure 14.4.3-5 (d)), the regions in between the piles
are also reinforced. It is necessary to arrange for supporting stirrups along the edges
(supporting stirrups are also necessary with indirect supports; see further).

14-29

sectional
view AA
suspension
stirrups

Figure 14.4.3-5
Foundation blocs supported by three or four piles; (a) top view; (b) strut-and-tie models;
(c) scheme of the reinforcement; (d) case of blocs with large distances between the pile
heads (l > 4.p)

14-30

Note: suspension stirrups for indirect supports

The load transfer from a transverse beam to a main girder (figure 14.4.3-6 (a)),
may be realized by means of compression struts (figure 14.4.3-6 (b)) or by
means of an inclined tie (figure 14.4.3-6 (c)). With compression struts, the load
enters into the main girder at the bottom layer, but has to be suspended to the
upper layer by means of stirrups (called suspension stirrups). The design of the
main girder is then performed in the same way as for a beam with a concentrated
load applied on top of the beam.
Reminder: it was already observed in figure 14.3.4-1 that an inclined suspension
tie is less efficient than an inclined strut (which provides a stiffer load transfer
mechanism).
Figure 14.4.3-6 (d) shows the process of the load transfer. The load is fixed in
the upper layer by the stirrups (bb') in the transverse beam. This load is
transferred by the inclined compression strut to the main girder. Suspension
stirrups (aa') are used for the transfer of the load to the upper part of the main
girder. It is important to note that the suspension stirrups are situated in the same
plane as the transverse beam. Figure 14.4.3-6 (e) represents the reinforcement
scheme.

14-31

Figure 14.4.3-6
Indirect support; (a) two main girders and one tranverse beam;
(b) load transfer by means of a compression strut; (c) load transfer by means of an
inclined tie; (d) detailed view of the load transfer mechanism from the transverse beam
to the main girder; (e) reinforcement scheme with side and top view

14-32

14.4.4 Corbels

A corbel (or console) is a small cantilever fixed on a column or wall, used to support
other horizontal structural members such as transverse beams. In historical masonry or
wood buildings, corbels are pieces of stone or timber jutting out of a wall to carry any
superincumbent weight. The word "corbel" comes from the latin corbellus, a diminutive
of corvus (a raven) which refers to the beak-like appearance.
The load applied to the corbel is mainly the vertical load induced by the self-weight.
Yet, corbels may also be subjected to important horizontal actions, for example where
temperature variations lead to shortening of the transverse beam and to transmission of
horizontal actions to the corbel by friction resistance in the contact area between the two
members.
Design recommendations are given in Annex J to EN 1992-1-1:2004. The essential
recommendations are summarized in the following.
The basic model for corbel design is presented in figure 14.4.4-1. The proposed model
is a combination of two strut-and-tie systems:
a direct strut, charcterized by the inclination angle ;
a secondary system, including a horizontal tie, which takes account of the
bottle shape of a real strut.
EC2 specifies that the inclination of the direct strut should be sufficiently high:
1 tg 2,5
or:
45 68,2
If the inclination angle is too small, the model with one direct strut is not appropriate
anymore; the corbel should then be treated as a cantilever beam. This condition is
reinforced by a second one (see figure 14.4.4-1 for the significance of the symbols):
ac < z0

Note: the horizontal force HEd is indicated in figure 14.4.4-1, but specific models
and rules concerning this load are not discussed in EC2 (version 2004).

14-33

Figure 14.4.4-1
Basic model for the design of a corbel by means of the strut-and-tie method (figure J.5
in EN 1992-1-1:2004; annex J)
Two cases are considered in function of the length of the corbel:
if ac 0,5.hc, closed horizontal or inclined links should be provided in
addition to the main tension reinforcement (see figure 14.4.4-2(a)), with:
As,lnk 0,25.As,main

if ac > 0,5.hc and FEd > VRd,c closed vertical links should be provided in
addition to the main tension reinforcement (see figure 14.4.4-2(b)), with:
As,lnk 0,5.FEd/fyd

Note: for corbels with large depth, strut-and-tie models such as illustrated in figure
14.3.4-2 should be used.

14-34

Figure 14.4.4-2
Corbel detailing; (a) reinforcement for ac 0,5.hc ; (b) reinforcement for ac > 0,5.hc
(figure J.6 in EN 1992-1-1:2004; annex J)

14.4.5 Half joints in beams

A corbel is a very visible support system. Sometimes, for aesthetic reasons, it is

preferred to work with so called half joints, where the corbel is integrated into the
beam: see figure 14.4.5-1.

Note:
In case a horizontal force is applied to the beams end at the support, it is
necessary to increase the vertical stirrup reinforcement: see figure 14.4.5-1 (d).

14-35

Figure 14.4.5-1
Beams with half joints (NL: tandoplegging van balken; FRXXX); (a) D and B-regions;
(b) strut-and-tie model; (c) reinforcement scheme; (d) a horizontal tensile force at the
support leads to increased vertical reinforcement at the beams end

14.4.6 Corners in frames

14.4.6.1

Corners subjected to positieve moments

Figure 14.4.6-1 shows the tie-and-strut model as well as the reinforcement for a corner
where column and beam have approximately the same dimensions.
More elaborated strut-and-tie models are necessary where the dimensions of the beams
and columns are very different: see figure 14.4.6-2; indeed, it is recommended to choose
the inclination angle of the compression struts not too different from 45.
14-36

Figure 14.4.6-1
Corner in a frame structure, subjected to a positive moment; (a) strut-and-tie model;
(b) reinforcement scheme

Figure 14.4.6-2
Corner in a frame structure, subjected to a positive moment; (a) the simplified strutand-tie model is characterized by a non-realistic inclination angle of the strut; (b)
improved strut-and-tie model; (c) reinforcement scheme

14.4.6.2

Corners subgjected to negatieve moments

Figure 14.4.6-3 presents a strut-and-tie model, with two plausible solutions for the
reinforcement scheme. More elaborated strut-and-tie models are available in literature
(SCHLAICH): figure 14.4.6-4.

14-37

Figure 14.4.6-3
Corner in a frame structure, subjected to a negative moment; (a) strut-and-tie model;
(b) first solution: a loop is used for the main reinforcement; (c) second solution
composed of straight bars and inclined stirrups

Figure 14.4.6-4
Corner in a frame structure, subjected to a negative moment; more elaborated strutand-tie model and reinforcement schemes

14.4.7 Nodes in frame systems

Different load combinations are possible in the nodes of frame structures; the
reinforcement scheme should thus be adapted for all these load combinations. This leads
to "universal solutions" such as shown in figure 14.4.7-1.

14-38

cold joint

cold joint

Figure 14.4.7-1
Nodes in framework systems; (a) load combinations; (b) strut-and-tie models;
(c) universal reinforcement scheme

14-39