SEECCM III

3rd South-East European Conference on Computational Mechanics-

an ECCOMAS and IACM Special Interest Conference
M. Papadrakakis, M. Kojic, I. Tuncer (eds.)
Kos Island, Greece, 12–14 June 2013

DESIGN OF RC SECTIONS IN THE ULTIMATE LIMIT STATE

UNDER BENDING AND AXIAL FORCE ACCORDING TO EC2
Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

National Technical University of Athens

Institute of Structural Analysis and Antiseismic Research
Zografou Campus, Athens 15780, Greece

e-mail: [email protected], [email protected], [email protected]

Keywords: Reinforced Concrete, RC, Section, Design, Eurocode, Reinforcement.

Abstract. In the imminent future the design of concrete structures in Europe will be governed
by the application of Eurocode 2 (EC2). In particular, EC2 – Part 1-1 [1] deals with the gen-
eral rules and rules for concrete buildings. An important aspect of the design is specifying the
necessary tensile (and compressive, if needed) steel reinforcement required for a Reinforced
Concrete (RC) section, in order to ensure that the RC member will be able to resist the design
loads.

According to EC2-Part 1-1 three different simplified diagrams for the stress-strain behavior
of concrete for RC sections design can be assumed: (a) the equivalent rectangular stress
block, (b) the parabolic-rectangular stress-strain relation and (c) the bi-linear stress-strain
relation as a simplification of the parabolic-rectangular case. In this study the three ap-
proaches suggested by EC2-Part 1-1 are investigated for the design of rectangular RC cross-
sections with tensile steel reinforcement to resist loading due to bending moment and axial
force. The tensile strength of concrete is neglected and concrete is supposed to work only in
compression. For each case analytical relations are extracted in detail with a step-by-step
detailed procedure, the relevant assumptions are highlighted and results for four beam design
examples are finally presented.

520
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

1 LITERATURE REVIEW
Rosca and Petru [2] study the design of a reinforced concrete section subjected to bending us-
ing the two stress–strain relationships mentioned in EC2, namely the parabola-rectangle stress
distribution and the rectangular distribution, and the differences are underlined. Two dimen-
sionless quantities are used to convert the parabola-rectangle stress distribution to an equiva-
lent concentrated force for the concrete in compression. Also analytical relations which
determine the limit between single reinforcement (only tensile) and double reinforcement
(tensile and compressive) are provided. The results drawn from the use of these two stress dis-
tributions, namely, parabola–rectangle and rectangle, showed that the differences between the
amounts of reinforcement are less than 1% for singly reinforced sections and less than 2% for
doubly reinforced sections.
Dulinskas and Zabulionis in [3] and [4] propose a method for the substitution of the non-
linear stress diagram with descending branch with an equivalent rectangular stress block when
the non-linear stress-strain relationship for concrete in compression is described according to
EC2. Analytical relationships in explicit form for area, the first moment of area, the coordi-
nate of centroid of the nonlinear stress diagram with descending branch, the ratio between the
depth of the rectangular stress block and that of the equivalent nonlinear stress diagram with
descending branch in respect to the concrete strength are given. Coefficients suitable for the
substitution of parabola stress diagram with descending branch given in EC2 with an equiva-
lent rectangular stress block are presented. These coefficients have to ensure that the substitu-
tion is equivalent, i.e. the carrying capacity of the compression zone calculated using either of
the two stress diagrams should be the same.
Židonis in [5] tries to replace the nonlinear stress-strain diagram of concrete adopted by
EC2 for structural analysis by another more general curvilinear diagram which relates stress
and strain of concrete. The new stress-strain diagram permits direct integration without the
need to discretize the stress-strain curve. Thus it makes the integration easier and can be ap-
plied to the concrete classes from C8/10 up to C90/105. Analytical stress-strain relations are
presented for concrete which can fit the stress-strain curves specified in EC2 within an error
of 1.5%. Finally, examples of application of the proposed stress-strain diagram are illustrated.
In [6] a method is presented and formulas are provided for application of non-linear con-
crete stress diagram for cross-section strength calculation in accordance with the limit state
(partial factors) method. Commonly reinforced concrete flexural members with rectangular
compression zone and the neutral axis within the cross-section are considered (beam-type
members); the effect of the descendant part of stress-strain diagram on strength of cross-
sections of beam type members is investigated and the limit between commonly and abun-
dantly reinforced concrete beams is determined. Finally the results of the new method are
compared with those of EC2, where rectangular compression zone stress diagram for concrete
is assumed. A table is extracted in which all necessary information needed to perform design
for bending of a reinforced concrete section for all concrete strength classes are shown.
Although the above studies deal with the application of the most suitable stress-strain dia-
gram for concrete for the “optimal” design of cross sections using different approaches, to the
authors’ knowledge, there is no study in which explicit formulas and/or graphs are provided
to achieve the design of RC concrete sections according to EC2-Part 1-1 for all the three
stress-strain relations provided. In the present study, the three suggested stress-strain diagrams
of EC2 are investigated and analytical formulas are given for the step-by-step design or RC
sections according to any one of the three design approaches.

521
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

2 DESIGN ASSUMPTIONS
The following design assumptions are made:

 Design is based on characteristic concrete cylinder strengths, not cube strengths.

 Plane sections remain plane.
 Strain in the bonded reinforcement, whether in tension or compression, is the same as
that in the surrounding concrete.
 Tensile strength of the concrete is ignored.
 Concrete stress distribution is considered according to the three cases of Eurocode 2,
as will be shown in detail in the next paragraphs.
 Stress in steel reinforcement is considered according to the stress-strain relation of Eu-
rocode 2 for steel, as will be shown in detail in the next paragraphs.

3 CONCRETE PROPERTIES
According to EC2-1-1 the compressive strength of concrete is denoted by concrete strength
classes which relate to the characteristic (5%) cylinder strength fck, or cube strength fck,cube, in
accordance with EN 206-1. Higher strengths of concrete are covered by Eurocode 2, up to
class C90/105. The strength classes for concrete are presented in the table below.

fck
12 16 20 25 30 35 40 45 50 55 60 70 80 90
(MPa)
fck,cube
15 20 25 30 37 45 50 55 60 67 75 85 95 105
(MPa)
Table 1. Strength classes for concrete.

where fck is the characteristic compressive cylinder strength of concrete at 28 days and fck,cube,
is the corresponding cube strength. The value of the design compressive strength is defined as
f ck
f cd  acc (1)
c
where:
 γc is the partial safety factor for concrete at the Ultimate Limit State, which is given in
Table 2.1N of EC2-1-1. For persistent and transient design situations, γc=1.5
 acc is the coefficient taking account of long term effects on the compressive strength
and of unfavourable effects resulting from the way the load is applied. The value of acc
for use in a country should lie between 0.8 and 1.0 and may be found in its National
Annex. The recommended value is 1.

It should be noted that higher concrete strength shows more brittle behavior, reflected by
shorter horizontal branch, as will be shown in the stress-strain relations, later.

4 CONCRETE STRESS-STRAIN RELATIONS FOR THE DESIGN OF CROSS-

SECTIONS
Eurocode 2 Part 1-1 suggests the use of three approaches for the stress-strain relations of con-
crete for the design of cross sections:

522
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

 Parabola-rectangle diagram (more detailed)

 Bi-linear stress-strain relation (less detailed)
 Rectangular stress distribution (simplest)

The three above approaches will be described in detail in the following sections.

4.1 Parabola-rectangle diagram for concrete under compression

According to EC2-1-1 (Paragraph 3.1.7), for the design of cross-sections, the following stress-
strain relationship may be used:
    n 
 f 1  1  c   for 0   c   c 2
 c   cd    c 2   (2)

 f cd for  c 2   c   cu 2
where:

 n is the exponent given by

 2.0 for f ck  50MPa

n  90  f ck 
4
(3)
1.4  23.4  100  for 50  f ck  90MPa
  
 εc2 is the strain at reaching the maximum strength given by

 2.0 for fck  50MPa
 c 2 (‰)   (4)
2.0  0.085  fck  50 
0.53
 for 50  fck  90MPa

 εcu2 is the ultimate strain given by

 3.5 for f ck  50MPa

 cu 2 (‰)    90  f ck 
4
(5)
2.6  35  100  for 50  f ck  90MPa
  

The above equation is depicted in the following figure, where compressive stresses (and
strains) are shown as positive.

523
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

Figure 1: Parabola-rectangle diagram for concrete under compression.

4.2 Bi-linear stress-strain relation

According to Paragraph 3.1.7(2) of EC2-1-1, other simplified stress-strain relationships may
be used if equivalent or more conservative than the one defined in Eq. (2), for instance bi-
linear according to the following figure, where again compressive stress and shortening strain
are shown as absolute values.

Figure 2: Bi-linear stress-strain relation.

where the values of εc3 and εcu3 are given by

 1.75 for fck  50MPa

 c3 (‰)   f  50 (6)
1.75  0.55  ck for 50  fck  90MPa
 40

524
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

 3.5 for f ck  50MPa


 cu 3 (‰)    90  f ck 
4
(7)
2.6  35  100  for 50  f ck  90MPa
  

4.3 Rectangular stress distribution

According to Paragraph 3.1.7(3) of EC2-1-1, a rectangular stress distribution may be assumed
for concrete, as shown in the following figure.

Figure 3: Rectangular stress distribution.

where:
 d is the effective depth of a cross-section
 x is the neutral axis depth
 As is the cross sectional area of the tensile steel reinforcement
 Fc is the concrete force (compressive, positive, as in the figure)
 Fs is the steel reinforcement force (tensile, positive, as in the figure)
 The factor λ defining the effective height of the compression zone and the factor η de-
fining the effective strength, are calculated from:
 0.8 for fck  50MPa

 fck  50 (8)
0.8  400 for 50  fck  90MPa

 1.0 for fck  50MPa


  fck  50 (9)
1.0  200 for 50  fck  90MPa

Note: If the width of the compression zone decreases in the direction of the extreme compres-
sion fibre, the value η∙fcd should be reduced by 10%.

5 STEEL PROPERTIES
The design strength for steel is given by
f yk
f yd  (10)
s
where:
 γs is the partial safety factor for steel at the Ultimate Limit State, which is given in Ta-
ble 2.1N of EC2-1-1. For persistent and transient design situations, γs=1.15

525
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

 fyk is the characteristic yield strength of steel reinforcement.

Table C.1 of Annex C of EC2-1-1 gives the properties of reinforcement suitable for use with
the Eurocode. The properties are valid for temperatures between -40ºC and 100ºC for the rein-
forcement in the finished structure. Any bending and welding of reinforcement carried out on
site should be further restricted to the temperature range as permitted by EN 13670.

Table 2. Properties of steel reinforcement (Table C.1 of Annex C of EC2-1-1).

The application rules for design and detailing in Eurocode 2 are valid for a specified yield
strength range, fyk from 400 to 600 MPa. The yield strength fyk is defined as the characteristic
value of the yield load divided by the nominal cross sectional area. The reinforcement shall
have adequate ductility as defined by the ratio of tensile strength to the yield stress, (ft/fy)k and
the elongation at maximum force, εuk.

5.1 Steel stress-strain relations for the design of cross-sections

For normal design, either of the following assumptions may be made for the stress-strain rela-
tion for steel (see figure below):

1. An inclined top branch with a strain limit of εud and a maximum stress of k∙fyk/γs at εuk,
where k=(ft/fy)k (see the above Table 2).
2. A horizontal top branch without the need to check the strain limit.

The value of εud for use in a country may be found in its National Annex. The recommended
value is 0.9εuk. The design value of the steel modulus of elasticity Es may be assumed to be
200 GPa.

526
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

Figure 4: Idealised and design stress-strain diagrams for reinforcing steel (for tension and compression)

In the present study we will use the second approach, assuming a horizontal top branch for the
steel stress-strain relation, but also limiting the maximum strain to εud, as shown in the follow-
ing figure.

Figure 5: Design stress-strain diagram for reinforcing steel (for tension and compression) used in the present
study.

527
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

6 INVESTIGATION OF THE SIMPLIFIED RECTANGULAR STRESS

DISTRIBUTION CASE
The figure below shows a typical rectangular cross section and the distribution of strains and
stresses (forces).
ηfcd
εc
Fc Fc
x λx
yN
Nd (tensile)
Μd
d
h z
ys

As εs Fs Fs
d1

b
Section Strains Forces Equilibrium
Figure 6: Cross section, strain and forces distribution and section equilibrium.

In the above figure:

 h is the height of the rectangular section

 d1 is the distance from the lower edge of the section to the center of the reinforcement
 εs is the tensile strain in the steel reinforcement
 εc is the compressive strain in the concrete upper edge
 λ is a factor defining the effective height of the compression zone
 η is a factor defining the effective strength of the compression zone
 Md is the applied external bending moment (puts the lower edge of the section in ten-
sion if positive)
 Nd is the applied external axial force (tensile for the section if positive), applied at a
position yN measured from the top of the section towards the lower edge of it. Note: If
we have central tension, then yn=h/2
 ys is the distance from the steel reinforcement to the position of the external applied
axial force
 z is the distance of the concrete force Fc from the steel reinforcement.

The goal of the design is to calculate the needed cross sectional area of steel reinforcement As.
In order to calculate As, we need first to calculate the unknown quantities x and z. We move
the external force Nd to the position of the steel reinforcement and we have the figure below.

528
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

Fc Fc
yN
Nd (tensile)
Μd
z
ys
Μsd=Md – Nd ys
Fs Fs Nd

Equilibrium Equilibrium

Figure 7: Equilibrium after moving the external force Nd to the position of the steel reinforcement.

From the equilibrium of the section in the x-direction, we have:

Fx  0  Fc  Nd  Fs  0  Fs  Fc  Nd (11)
We have also:
ys  yn  d  ys  d  yn (12)
The effective bending moment is:
M sd  M d  Nd  ys (13)

From the geometry of the section, we have:

x x
z d zd (14)
2 2
The concrete force is given by:
Fc   x  nfcd  b (15)

From the equilibrium of moments at the position of the steel reinforcement (Figure 7) we
have (clockwise moment taken as positive):
M steel  0  Fc  z  M sd  0  M sd  Fc  z (16)
By substituting Eq. (15) in Eq. (16), we have:
M sd   x  nfcd  b  z (17)

By substituting Eq. (14) in Eq.(17), we have:

 x   2 
M sd   x  nfcd  b   d    x     nf cd  b  d   x 2   nf cd  b    (18)
 2   2 

 nfcd  b   2  2
   x     nfcd  b  d   x  M sd  0 (19)
 2 

The above equation can be written as:

529
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

Ax 2  Bx  C  0 (20)
nfcd  b   2
where A , B      nfcd  b  d  , C  M sd (21)
2
The above quantities A, B and C are all known, so by solving the quadratic Eq. (20) we can
determine the quantity x. The discriminant of the quadratic equation is:
  B2  4 A  C     nfcd  b  d   2nf cd  b   2  M sd
2
(22)
The solution of the quadratic equation is:
 d 
 x1  
B   d    2A
x1,2     (23)
2A  2A  d 
 x2    2 A
Given the requirement that 0≤x≤d and the fact that λ=0.80 for fck≤50 MPa and λ<0.80 for
50<fck≤90 MPa, it is obvious that d/λ>d, and as a result x2>d which is not acceptable. So the
only acceptable solution is x=x1 and thus:
d 
x  (24)
 2A
After calculating x, it is easy to calculate also z with Eq. (14), Fc with Eq. (15) and Fs with Eq.
(11). We have also:
Fs
Fs  As   s  As  (25)
s
In the above equations, σs is the steel stress at the Ultimate Limit State (ULS) of the section.
The yield strain εys for steel is:
f yd
 ys  (26)
Es
where
f yk
f yd  (27)
s
if εs≥εys then the steel works in full stress and σs=fyd. Otherwise, if εs<εys then the steel does
not work in full stress and σs<fyd. In general, the steel stress σs is given by:

 s
 f yd    Es   s if 0 < ε s   ys
s   ys (28)
 f yd if ε s   ys


In order to determine the area of steel reinforcement As, we need to determine the steel
stress σs and thus we should determine the steel strain εs. In order to determine εs given the

530
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

value of x, we first need to know if the steel reinforcement or the concrete is critical (at the
ultimate strain) at the Ultimate Limit State of the section.
In the present study, the steel reinforcement is limited to εud. If the steel reinforcement is
critical, then εs=εud and εc<εcu3. Otherwise, if the concrete is critical, then εc=εcu3 and εs<εud. In
the special case where both materials are critical, then εs=εud and εc=εcu3. These three possible
states are presented in detail in the figure below.

εcu3 εc<εcu3 εc=εcu3

x<xult
xult xult xult
x>xult

εud εs=εud εs<εud

(a) (b) (c)

Figure 8: Three possible states of the strains of the cross section.

(a) Both materials are at the ultimate strain, (b) Steel at the ultimate strain, (c) Concrete at the ultimate strain.

In order to find out if the concrete or the steel is critical, we first calculate xult which is the
neutral axis depth for the special case of both materials being critical, as in the figure above –
Case (a). It should be noted that this is only a theoretical case and it does not correspond to
equilibrium of the cross section. Using similar triangles, we have:
 cu 3  cu 3   ud  cu 3
  xult  d (29)
xult d  cu 3   ud

Then we have 2 cases:

Case 1: x < xult, as shown in Figure 8 (b)

The steel reinforcement is at the critical strain, εs=εud, while εc<εcu3. In this case, it is ensured
that the steel works in full stress, and thus
 s  f yd (30)
And the steel area is given by:

Fs Fs
As   (31)
s f yd
The concrete strain εc in this case can be calculated by:
c  c   ud x
  c   c   ud  (32)
x d d

531
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

Case 2: x > xult, as in shown in Figure 8 (c)

The concrete zone is at the critical strain, εc=εcu3 while εs<εud. The steel strain εs can be calcu-
lated by:
s  cu 3   s  x
   s  1    cu 3   s  (33)
dx d  d
In this case, it is not sure whether the steel works in full stress or not. The steel strain εs has to
be checked if it is above or below the steel yield strain εys as follows:

 Case 2a: The steel works in full stress, εs≥εys

The steel reinforcement works in full stress, above the yield strain and as a result the steel ar-
ea can be calculated by Eq. (31).

 Case 2b: The steel does not work in full stress, εs<εys
The steel reinforcement does not work in full stress, as it works below the yield strain. The
steel stress σs can be calculated by Eq. (28) and the steel area can be calculated by Eq. (25).
In this case, although the reinforcement area can be calculated, the design with only tensile
reinforcement is not economic. Either compressive reinforcement should be considered, or an
increase in the effective depth of the cross-section d.

6.1 Comment on the above cases

Eurocode 2 allows the designer to not limit the ultimate strain for steel when a horizontal top
branch is assumed for the stress-strain diagram for steel. In this case, the concrete zone is as-
sumed to be at the ultimate strain at all times at the ULS of the section under any design loads
and the steel strain can take any value, without any limitation. If this is the case, then in the
above investigation we have to set xult=0 and we have always the case of Figure 8(c) where
the concrete zone is at the ultimate strain for all design cases.
This assumption can make things simpler and the calculations much easier. For reasons of
completeness, in the present study we will continue to assume that the ultimate strain for steel
is limited to εud. This is done in order for the proposed methodology to be able to be extended
also in the case where not a horizontal, but an inclined top branch is assumed for the steel
stress-strain relation, with a maximum stress of k∙fyk/γs at εuk, where k=(ft/fy)k. In this case Eu-
rocode 2-Part 1-1 enforces the use of the strain limit of εud for steel.
If the designer uses the proposed methodology and sets εud=∞, then we have the case of
forcing the concrete zone to the ultimate strain and letting the steel take any strain value, as
suggested by the Eurocode for the case of horizontal top branch stress-strain diagram. This
will be illustrated also in the numerical examples section.

532
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

7 INVESTIGATION OF THE BI-LINEAR STRESS-STRAIN RELATION CASE

εc fcd
Fc1 Fc1
x1
x Fc2 yN Fc2
εc3
x-x1
Μd
d Nd (tensile)
h z1
z2
ys

As εs Fs Fs
d1

b
Section Strains Forces Equilibrium

Figure 9: Cross section, strain and forces distribution and section equilibrium for the bi-linear stress-strain re-
lation case, assuming εc>εc3.

The goal of the design is again to calculate the needed cross sectional area of steel reinforce-
ment As. In order to calculate As, we need first to calculate the unknown quantities x, z1 and z2.
We move again the external force Nd to the position of the steel reinforcement and we have
the figure below.

Fc1 Fc1
Fc2 Fc2

Μd
Nd (tensile)
z1
z2
ys
Μsd=Md – Nd ys
Fs Fs Nd

Equilibrium Equilibrium

Figure 10: Equilibrium after moving the external force Nd to the position of the steel reinforcement.

We need to determine if at the ULS the concrete zone or the steel is at the critical strain.
Again, first we put both materials at the ultimate strain, so we have:
 c   cu3 (34)

 s   ud (35)

  cu 3
x  d
 cu 3  ud   cu 3  cu 3   c3   ud   cu 3
   (36)
x d x1  x   cu 3   c3  d
 1  ud   cu 3

Fc1  x1  fcd  b (37)

533
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

x1
z1  d  (38)
2
1
Fc 2   x  x1   fcd  b (39)
2
x  x1 x  2 x1
z2  d  x1  d (40)
3 3
Fc  Fc1  Fc 2 (41)
We will calculate the sum of moments at the steel reinforcement position. The sign of the sum
of moments will show us whether the concrete zone or the steel is at the ultimate strain at the
ULS. The sum of moments is (clockwise positive):
M steel  Fc1  z1  Fc 2  z2  M sd (42)
We have then two cases:

Case 1. ΣΜ≥0
The concrete force has to be decreased for the equilibrium of the cross section. The steel stays
at the ultimate strain (εs=εud), while for concrete εc≤εcu3, as shown in the figure below.
εcu3 εc εcu3

xult

εud εs=εud

Figure 11: (b) Case 1: ΣΜ≥0, Steel at the ultimate strain.

Case 2. ΣΜ<0
The concrete force has to be increased for the equilibrium of the cross section. Concrete stays
at the ultimate strain (εc=εcu3), while for steel εs<εud, as shown in the figure below.
εcu3 εc=εcu3

xult

εud εs<εud

Figure 12: Case 2: ΣΜ<0, concrete zone at the ultimate strain.

For both cases, we need to determine the value of x that satisfies the equilibrium of the
cross section. After having determined x, we can then proceed with the other calculations and

534
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

finally end up with the needed reinforcement area As. The value of x can be determined by
using trial and error iterations, or by using some kind of optimization in order to achieve sec-
tion equilibrium. Good tools for this are MS Excel (Goal Seek or Solver functions) and also
Matlab with its built-in root-finding and optimization tools. In the present study, we have used
three equivalent approaches, (a) Solver function of MS Excel, (b) Matlab and (c) a home-
made code which finds x by performing iterations, dividing the allowable height of the section
by two at each iteration until convergence (equilibrium). All three approaches provide the
same results at the end, as expected.
In the next sections, we will assume a value for x and we will end up with the equilibrium
equation, i.e. the sum of moments at the steel reinforcement position which has to be zero at
the equilibrium.

Case 1. ΣΜ≥0, Steel at the ultimate strain

We assume an initial value for x and we use the following equations:
 s   ud (43)
c  ud x
  c    ud (44)
x dx dx
 Case 1a: If εc>εc3
In this case we have the triangular diagram plus a rectangular diagram for the concrete zone
and the upmost fiber of the concrete section works at the ultimate stress fcd. From the similar
triangles we have:
x1 d  
  x1  c c3  d (45)
 c   c3  c   s c  s
Fc1  x1  fcd  b (46)
x1
z1  d  (47)
2
1
Fc 2   x  x1   fcd  b (48)
2
x  x1 2x  x
z2  d  x1  d 1 (49)
3 3
Fc  Fc1  Fc 2 (50)

M steel  Fc1  z1  Fc 2  z2  M sd (51)

After we reach the equilibrium (ΣΜsteel=0), and given that the steel reinforcement works in
full stress, above the yield strain, the steel area can be easily calculated by Eq. (31).

 Case 1b: If εc≤εc3

In this case we have only the triangular diagram for the concrete zone, there is no rectangular
part for the stresses and the upmost fiber of concrete works at stress σc≤fcd, as follows:
c
c   f  fcd (52)
 c3 cd

535
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

1
Fc 2  x  c  b (53)
2
x
z2  d  (54)
3
Fc  Fc 2 (55)

M steel  Fc 2  z2  M sd (56)
Again, after we reach the equilibrium (ΣΜsteel=0), and given that the steel reinforcement
works in full stress, above the yield strain, the steel area can be calculated by Eq. (31).

Case 2. ΣΜ<0, concrete zone at the ultimate strain

We assume an initial value for x and we use the following equations:
 c   cu3 (57)

 dx
 s    cu 3
 cu 3 s  cu 3   c3  x
   (58)
x dx x1  x1   cu 3   c3  d
  cu 3   s
Fc1  x1  fcd  b (59)
x1
z1  d  (60)
2
1
Fc 2   x  x1   fcd  b (61)
2
x  x1 2x  x
z2  d  x1  d 1 (62)
3 3
Fc  Fc1  Fc 2 (63)

M steel  Fc1  z1  Fc 2  z2  M sd (64)

 Case 2a: εs≥εys

The steel reinforcement works in full stress, above the yield strain and as a result the steel ar-
ea can be calculated by Eq. (31).

 Case 2b: εs<εys

The steel reinforcement does not work in full stress, as it works below the yield strain. The
steel stress σs can be calculated by Eq. (28) and the steel area can be calculated by Eq. (25).

8 INVESTIGATION OF THE PARABOLIC-RECTANGULAR STRESS-STRAIN

RELATION CASE
For the parabolic-rectangle stress-strain relation case, we use the same methodology as in the
case of the bilinear stress-strain relation. The only difference is the shape of the concrete
stress distribution where the triangular section becomes now parabolic, and also the ultimate

536
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

strain and the strain corresponding to the start of the rectangular section which become εcu2
(instead of εcu3) and εc2 (instead of εc3), respectively, as shown in the figure below.

εc fcd
Fc1 Fc1
x1
x Fc2 yN Fc2
εc2
x-x1
x2
Μd
d Nd (tensile)
h z1
z2
ys

As εs Fs Fs
d1

b
Section Strains Forces Equilibrium

Figure 13: Cross section, strain and forces distribution and section equilibrium for the parabolic-rectangular
stress-strain relation case, assuming εc>εc2.

In the above figure, x2 is the distance from the neutral axis to the centroid of the parabolic
section. The parabolic section is “full” in the figure, as εc>εc2. In the bi-linear case, the calcu-
lation of the area and centroid of the non-rectangular part was obvious, because of the triangu-
lar shape, but for the parabolic case, integration has to be used, as will be described in detail
later.
Again, we need to determine if at the ULS the concrete zone or the steel is at the critical
strain. First, we put both materials at the ultimate strain, so we have:
 c   cu 2 (65)

 s   ud (66)

  cu 2
 cu 2  ud   cu 2  cu 2   c 2  x    d
 cu 2
   (67)
ud

x d x1  x   cu 2   c 2  d
 1  ud   cu 2

Fc1  x1  fcd  b (68)

x1
z1  d  (69)
2
The above equations are almost the same as the ones used in the bi-linear case, but of course
in the parabolic-rectangle case we use εc2 and εcu2 instead of εc3 and εcu3. Yet, this time in or-
der to calculate Fc2 we need to integrate Eq. (2) to calculate the area of the parabolic part. For
the parabolic part of the stress, i.e. for strains εc in the region [0, εc2], we have the indefinite
integral:
n 1
 c 
 1 
 c 2 
     n c 2 f cd

  c d c   fcd 1  1   c 2   d c   c fcd 
c
n 1
(70)
 

537
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

Thus the area E1 of the full parabolic part [0, εc2] is given by the definite integral:
c 2
n
E1    d
0
c c 
n 1
 c 2 fcd (71)

The area E1 of the full parabolic part is shown in the figure below in black color.

Figure 14: Area E1 of the full parabolic part (for strains up to εc2) in black color.

If the integration is done on the cross section height, for the strain εc2 the corresponding height
of the section is (x-x1) and as a result the corresponding area of the full parabolic part A1 is
given by:
n
A1   x  x1  fcd (72)
n 1
The area A1 of the full parabolic part is shown in the figure below in black color.
εc=εcu2 fcd
Fc1
x1
x Fc2
εc2
x-x1
x2

z1
z2

εs=εud Fs

Figure 15: Strain and forces distribution.

The area A1 of the full parabolic part is shown in black color.

The concrete force Fc2 is given by:

n
Fc 2  A1  b   x  x1  fcd  b (73)
n 1
In order to calculate z2 we need to calculate the distance x2 defining the centroid of the A1 area.
In terms of strains, the centroid εcentroid of the E1 area is given by the definite integral:

538
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

c 2
1
 centroid1 
E1    d
0
c c c (74)

The indefinite integral in this case is given by:

   n   c 2 fcd fcd  c 2   c   c 2   c  n  1 
n 1

  c cd c    c fcd 1  1   c 2   d c  2 

c
 c 2n  n  2  n  1
(75)
 
Thus the centroid of the full parabolic part is given by:
c 2
1 n3
 centroid1 
E1    d
0
c c c 
2  n  2
 c2 (76)

If the integration is done on the section height, for the strain εc2 the corresponding height of
the cross section is (x-x1) and as a result the corresponding centroid of the full parabolic part
x2 is given by:
 centroid1 n3
x2    x  x1     x  x1  (77)
c2 2  n  2
Then we have
z2  d  x  x2 (78)

Fc  Fc1  Fc 2 (79)
Again, we will calculate the sum of moments at the steel reinforcement position. The sign of
the sum of moments will show us whether the concrete zone or the steel is at the ultimate
strain at the ULS. The sum of moments is (clockwise positive):
M steel  Fc1  z1  Fc 2  z2  M sd (80)
We then have again two cases:

Case 1. ΣΜ≥0
The concrete force has to be decreased for the equilibrium of the cross section. The steel stays
at the ultimate strain (εs=εud), while εc≤εcu2.
εcu2 εc εcu2

xult

εud εs=εud

Figure 16: (b) Case 1: ΣΜ≥0, Steel at the ultimate strain.

539
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

Case 2. ΣΜ<0
The concrete force has to be increased for the equilibrium of the cross section. The concrete
stays at the ultimate strain (εc=εcu2), while εs<εud.
εcu2 εc=εcu2

xult

εud εs<εud

Figure 17: Case 2: ΣΜ<0, concrete zone at the ultimate strain.

The methodology is exactly the same as the one of the bi-linear case. To start, we assume a
value for x and we should change it until we reach the final equilibrium. The equations below
end up with the calculation of the sum of moments which has to be zero at the equilibrium.

Case 1. ΣΜ≥0, Steel at the ultimate strain

We assume an initial value for x and we use the following equations:
 s   ud (81)
c  ud x
  c    ud (82)
x dx dx

 Case 1a: If εc>εc2

In this case we have the parabolic diagram plus a rectangular diagram and the upmost fiber of
concrete works at the ultimate stress fcd. From the similar triangles we have:
x1 d  
  x1  c c3  d (83)
 c   c3  c   s c  s
Fc1  x1  fcd  b (84)
x1
z1  d  (85)
2
In a similar way as previously (integrations), and since we have again a full parabolic part, we
have:
n
Fc 2   x  x1  fcd  b (86)
n 1
n3
x2    x  x1  (87)
2  n  2
z2  d  x  x2 (88)

Fc  Fc1  Fc 2 (89)

540
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

M  Fc1  z1  Fc 2  z2  M sd (90)
After we reach the equilibrium (ΣΜ=0), and given that the steel reinforcement works in full
stress, above the yield strain, the steel area can be easily calculated by Eq. (31).

 Case 1b: If εc≤εc2

In this case we have only part of the parabolic diagram, there is no rectangular diagram and
the upmost fiber of concrete works at stress σc≤fcd.
c
c   f  fcd (91)
 c 2 cd
Τhis time in order to calculate Fc2 we need to integrate Eq. (2) to calculate the area of the par-
abolic part, not for the full parabola (up to εc2), but for the region [0, εc] where εc≤εc2. Using
the indefinite integral of Eq. (70) we can calculate the corresponding area E2 of the parabolic
part for the region [0, εc] where εc≤εc2 as a definite integral as follows:
   n1      n1  
 c 2 fcd  1    1
c   c 2 1  1  c   
c    c2      c2   
E2    c d  c  f cd  c     f      (92)
n 1 n 1
cd c
0  
 
 
 
The area E2 of the parabolic part for the region [0, εc] is shown in the figure below in black
color.

Figure 18: Area E2 of the parabolic part for the region [0, εc] where εc<εc2, in black color.

If the integration is done on the section height, for a strain εc<εc2 the corresponding height of
the cross section is x while for the theoretical strain εc2 the corresponding height of the cross
section would be x∙εc2/εc and as a result the corresponding area of the parabolic part A2 is giv-
en by:

541
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

  c2    n1        n1  
 x 1  1    c2
1  1  c   
   c 2   
c

 c    c2   
A2  f cd  x      f  x 1  c    (93)
n 1 n 1
cd
   
   
   
   
The area A2 of the parabolic part in this case is shown in the figure below in black color.
εC<εc2 σc<fcd F
c2
x
x2

z2

εs=εud Fs

Strains Forces
Figure 19: Strains and forces distribution.
The area A2 of the parabolic part (for strains up to εc<εc2) is shown in black color.

The concrete force Fc2 is given by:

     n1  
 c2
1  1  c 
  c    c 2   
Fc 2  A2  b  f cd  x  b 1    (94)
 n  1 
 
 
 
The calculation of z2 for this case is the most difficult part. In order to calculate z2 we need to
calculate the distance x2 defining the centroid of the A2 area, as shown in Figure 19. In terms
of strains, the centroid εcentroid2 of the E2 area is given by the following formula:

1 c
E2 0
 centroid 2   c c d c (95)

c
The definite integral    d
0
c c c can be calculated from the indefinite integral of Eq. (75) as

follows:
c
 c 2 fcd fcd  c 2   c 
n 1
   c  n  1   fcd  c 2n2
   d
c2
  (96)
0
c c c
2  c 2n  n  2 n  1
If the integration is done on the section height, for the strain εc the corresponding height of the
cross section is x and as a result the corresponding centroid of the full parabolic part x2 is giv-
en by:

542
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

 centroid 2
x2  x (97)
c
Then we have
z2  d  x  x2 (98)

Fc  Fc 2 (99)
Again, we will calculate the sum of moments at the steel reinforcement position. The sign of
the sum of moments will show us whether the concrete zone or the steel is at the ultimate
strain at the ULS. The sum of moments is (clockwise positive):
M steel  Fc 2  z2  M sd (100)

Case 2. ΣΜ<0, concrete zone at the ultimate strain

We assume an initial value for x and we use the following equations:
 c   cu 2 (101)

 dx
     cu 2
 cu 2 s  cu 2   s  cu 2   c 2 s
x
    (102)
x dx d x1  x1   cu 2   c 2  d
  cu 2   s
Fc1  x1  fcd  b (103)
x1
z1  d  (104)
2
n
Fc 2   x  x1  fcd  b (105)
n 1
n3
x2    x  x1  (106)
2  n  2
z2  d  x  x2 (107)

Fc  Fc1  Fc 2 (108)

M  Fc1  z1  Fc 2  z2  M sd (109)

 Case 2a: εs≥εys

The steel reinforcement works in full stress, above the yield strain and as a result the steel ar-
ea can be calculated by Eq. (31).

 Case 2b: εs<εys

The steel reinforcement does not work in full stress, as it works below the yield strain. The
steel stress σs can be calculated by Eq. (28) and the steel area can be calculated by Eq. (25).

543
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

9 NUMERICAL RESULTS
Four concrete sections will be examined in total. All three approaches for the stress-strain re-
lations of concrete for the design of cross sections will be examined:

 Rectangular stress distribution

 Bi-linear stress-strain relation
 Parabola-rectangle diagram

Below are the common properties for all numerical examples:

 Steel class B500C (fyk=500 MPa)  εuk=75 ‰

 Es=200 GPa  εud=0.9∙εuk
 γc=1.50, γs=1.15  acc=1

9.1 Numerical example 1

The section of the first numerical example has the following properties:

 Concrete class C20/25  Md=60 kNm

 Height h=50 cm  Nd=0
 Width b=25 cm
 d1=5 cm

Below are the results of the design, for each of the three approaches for the stress-strain rela-
tions of concrete.

Rectangular Bilinear Parabola-rectangle

As (cm2) 3.22 3.22 3.22
εc (‰) 3.5 3.5 3.5
εs (‰) 26.5 24.6 26.8
x (m) 0.052 0.056 0.052
Critical material Concrete Concrete Concrete
Table 3. Design results for the 1st numerical example.

It is clear that all three approaches give the same final result for the needed steel reinforce-
ment area. Only minor differences can be found in the strains and the concrete zone height x.

9.2 Numerical example 2

The section of the second numerical example has the following properties:

 Concrete class C30/37  Md=100 kNm

 Height h=60 cm  Nd=50 kN
 Width b=30 cm  yN=h/2
 d1=5 cm

Below are the results of the design, for each of the three approaches for the stress-strain rela-
tions of concrete.

544
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

Rectangular Bilinear Parabola-rectangle

As (cm2) 4.90 4.91 4.90
εc (‰) 3.5 3.5 3.5
εs (‰) 53.1 49.6 53.8
x (m) 0.034 0.036 0.034
Critical material Concrete Concrete Concrete
Table 4. Design results for the 2nd numerical example.

The three approaches give again almost the same final result for the needed steel reinforce-
ment area. Again minor differences can be found in the strains and the concrete zone height x.

9.3 Numerical example 3

The section of the third numerical example has the following properties:

 Concrete class C70/85  Md=150 kNm

 Height h=70 cm  Nd=100 kN
 Width b=30 cm  yN=h/2
 d1=5 cm

Below are the results of the design, for each of the three approaches for the stress-strain rela-
tions of concrete.

Rectangular Bilinear Parabola-rectangle

2
As (cm ) 6.60 6.60 6.60
εc (‰) 2.1 2.4 2.3
εs (‰) 67.5 67.5 67.5
x (m) 0.020 0.023 0.023
Critical material Steel Steel Steel
Table 5. Design results for the 3rd numerical example.

Again, the results are the same for all three cases. This time the steel is the critical material
(at the ultimate strain) at the section equilibrium. The results above are calculated assuming a
horizontal top branch for the steel stress-strain relation, but also limiting the maximum strain
to εud. Eurocode 2 allows the designer to not limit the ultimate strain for steel when a horizon-
tal top branch is assumed for the stress-strain diagram for steel. In this case, the concrete zone
is assumed to be at the ultimate strain at all times at the ULS and the steel strain can take any
value, without any limitation. If we set εud=∞ (a very big number), then we have the results of
the following table.

Rectangular Bilinear Parabola-rectangle

2
As (cm ) 6.60 6.60 6.60
εc (‰) 2.7 2.7 2.7
εs (‰) 84.7 77.4 78.4
x (m) 0.020 0.022 0.021
Critical material Concrete Concrete Concrete
Table 6. Design results for the 3rd numerical example – No limitation for the steel strain.

545
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

We see that there is no difference in the required area of reinforcement As for the two cases,
the results are exactly the same and only the reported material strains change. Of course, this
time the critical material is the concrete zone, not the steel reinforcement.

9.4 Numerical example 4

The section of the fourth numerical example has the following properties:

 Concrete class C30/37  Md=378 kNm

 Height h=50 cm  Nd=0
 Width b=25 cm
 d1=5 cm

Below are the results of the design, for each of the three approaches for the stress-strain rela-
tions of concrete.

Rectangular Bilinear Parabola-rectangle

2
As (cm ) 26.14 33.78 26.67
εc (‰) 3.5 3.5 3.5
εs (‰) 2.14 1.69 2.13
x (m) 0.279 0.304 0.280
Critical material Concrete Concrete Concrete
εs/εys 0.98 0.78 0.98
Table 7. Design results for the 4rd numerical example.

This time the results are different and in the bilinear case the difference is big. Taking a
careful look at the results we can see that in this example, for all cases, εs<εys which means
that the steel reinforcement works below the yield strain and as a result the steel cannot work
in its full potential (fyd stress). These are cases where the design is poor and not economic and
the designer should either increase the height of the section or add compressive reinforcement,
also.
In such cases with εs<εys, the exact value of the steel strain εs is significant in the calcula-
tion of the final required reinforcement area as it affects directly the steel stress. In the bilinear
case, εs is calculated as 1.69 ‰, lower than in the other two cases, and that affects the required
reinforcement which is 33.78 cm2, much more than in the other two cases.
The difference is big, but these cases are theoretical since in practical cases we would nev-
er design a section in such a way that the steel reinforcement would work below the yield
strain.

10 CONCLUSIONS
 Eurocode 2-Part 1-1 gives us new tools in order to design concrete cross sections. Three
approaches may be used for the stress-strain relation of concrete and another two ap-
proaches for the stress-strain relation of steel reinforcement. The simplest approach for
concrete is the use of the Rectangular stress distribution. The other two approaches use
the Bi-linear stress-strain relation and the Parabola-rectangle diagram, respectively.
 This paper presents a detailed methodology for the design of rectangular cross sections
with tensile reinforcement, for all the three cases and for all concrete classes, covering all
concrete classes, from C12/15 to C90/105. The methodology is general and all other Eu-

546
 Vagelis Plevris, George Papazafeiropoulos and Manolis Papadrakakis

rocode parameters, such as γc, γs, acc, and others can be adjusted according to the prefer-
ences of the designer, without any limitation.
 The three approaches for concrete give almost the same results with each other for all
“normal” cases examined. The differences are very slight and not significant from an en-
gineering point of view. Big differences may occur in some “abnormal” cases where the
effective moment is big for the section and as a result the steel reinforcement works be-
low the yield strain εys. In any case, these cases have to do with bad section design and
they should be avoided. The best solution for these cases is to add height to the concrete
section, and/or add compressive reinforcement also.
 Eurocode 2 allows the designer to not limit the ultimate strain for steel when a horizontal
top branch is assumed for the stress-strain diagram for steel. In this case, the concrete
zone is assumed to be at the ultimate strain at all times at the ULS and the steel strain can
take any value, without any limitation. In the proposed methodology this can be achieved
by setting εud=∞ (a very big number) for the allowed steel strain. This was investigated in
a numerical example where the steel was the critical material and the result showed that it
made no difference in the final steel reinforcement area.
 A more detailed investigation has to be made regarding the three stress-strain approaches
for concrete in order to check if there are cases where the three approaches can lead to
different results. The next research step should be to use the proposed methodology in
order to generate dimensionless charts showing the required reinforcement for any load-
ing and any section. In this way, a general direct comparison of the three cases can be
performed.

REFERENCES
[1] EN 1992-1-1. Eurocode 2: Design of concrete structures. Part 1-1: General rules and
rules for buildings. Brussels, 2004.
[2] B. Rosca, M. Petru, Reinforced Concrete Section Design to Bending according to EN
1992-1-1/2004 – Eurocode 2. Buletinul Institutului Politehnic din Iasi, 53-66, 2009.
[3] E. Dulinskas, D. Zabulionis, Analysis of equivalent substitution of rectangular stress
block for nonlinear stress diagram. ISSN 1392 – 1207, Mechanika, 6 (68), 53-66, 2007.
[4] D. Zabulionis, E. Dulinskas, Analysis of compression zone parameters of cross-section
in flexural reinforced concrete members according to EC2 and STR 2.05.05. ISSN 1392
– 1207, Mechanika, 3 (71), 12-19, 2008.
[5] I. Židonis, Strength calculation method for cross-section of reinforced concrete flexural
member using curvilinear concrete stress diagram of EN-2. 11th International Confer-
ence on Modern Building Materials, Structures and Techniques (MBMST 2013), Vilni-
us, Lithuania, May 16 – 17, 2013.
[6] I. Židonis, A simple-to-integrate formula of stress as a function of strain in concrete and
its description procedure. ISSN 1392 – 1207, Mechanika, 4(66), 23-30, 2007.

547