TECHNICAL SCIENCES

Abbrev.: Techn. Sc., No 12, Y 2009

DOI 10.2478/v10022-009-0007-6

EFFECTIVE LENGTHS OF REINFORCED CONCRETE

COLUMNS IN SINGLE-STOREY FRAME STRUCTURES
IN THE LIGHT OF THE EUROCODE

Michał Knauff1, Krzysztof Klempka2

1
Institute of Building Structures
Warsaw University of Technology
2
Department of Civil Engineering and Building Structures
University of Warmia and Mazury in Olsztyn

K e y w o r d s: effective length of column, method based on nominal stiffness, second order effects,
reinforced concrete buildings, reinforced concrete structures.

Abstract

In the following article we present the application of the Eurocode rules for calculating the
effective length of columns in single-storey buildings. Assuming the same effective lengths for
cantilevers (i.e. l0 = 2lcol) and the construction columns, as in the example, will not always be correct.
The problem of calculating the effective length mostly presents itself when using the simplified
method. In the exact method it is not required to determine the effective lengths. This is why it is
advisable while designing to use the exact method based on second order analysis and taking into
account the nominal stiffness.

DŁUGOŚCI OBLICZENIOWE SŁUPÓW ŻELBETOWYCH W PARTEROWYCH

UKŁADACH RAMOWYCH W ŚWIETLE PRZEPISÓW EUROKODU

Michał Knauff1, Krzysztof Klempka2

1
Instytut Konstrukcji Budowlanych
Politechnika Warszawska
2
Katedra Budownictwa i Konstrukcji Budowlanych
Uniwersytet Warmińsko-Mazurski w Olsztynie

S ł o w a k l u c z o w e: długość obliczeniowa słupa, metoda nominalnej sztywności, efekty

drugiego rzędu, hale żelbetowe, konstrukcje żelbetowe.

Abstrakt

Rozpatrzono zastosowanie zasad Eurokodu do wyznaczania długości obliczeniowych słupów

w jednokondygnacyjnych budynkach halowych. Przyjmowanie długości obliczeniowych l0 jak dla
72 Michał Knauff, Krzysztof Klempka

wsporników (tzn. l0 = 2lcol) dla słupów w takich halach, jak w przykładzie, nie zawsze będzie
prawidłowe. Problem wyznaczania długości obliczeniowej dotyczy przede wszystkim metody upro-
szczonej. Zastosowanie metody ścisłej nie wymaga określania długości obliczeniowych, dlatego do
projektowania zaleca się stosowanie metody ścisłej polegającej na analizie II rzędu z uwzględnieniem
sztywności nominalnych.

Introduction

In 2010, Eurocode EN 1992-1-1 (abbr: EN) will replace the Polish norm
PN-B-03264 (abbr: PN) for designing reinforced concrete and pre-stressed
constructions, which will in turn lead to many other changes in the design
process. The EN calculating methods differ in many ways from the PN
methods, and one of the differences are the principles for calculating the
effective lengths of columns. The EN contains extensive rules referring to the
effective length of isolated elements, but it does not contain any rules
corresponding to the Polish rules for calculating single-storey structures
according to the Appendix C of the PN. According to the PN, it can be assumed
that l0 = 1,6lcol (when the roof construction is rigid), while according to the
older versions of the PN even l0 = 1,2lcol “when there are four or more
columns”. After implementing the EN the recommendations regarding the
effective lengths enclosed in the Appendix C of the PN will no longer act as
rules of the norm. While calculating columns for single-storey buildings, as in
the example, according to the EN it is required to assume that the effective
length is the same as for the cantilevers (i.e. l0 = 2lcol) – the EN has no rules
which would allow engineers to assume any other effective length. We are thus
faced with a question whether the assumptions of the EN referring to the
effective lengths are correct. To answer this question we compared the values
of bending moments for columns in a single-storey building obtained according
to the method based on nominal stiffness (a simplified method in the EN) with
the values obtained according to the exact method based on the second order
analysis and taking into account the nominal stiffness. Later on in the article
we present a short description of the method based on nominal stiffness and
a derivation of the formula for a coefficient increasing the moment. We finish
with examples of calculating a single-storey building.

Method based on nominal stiffness

This method is based on the fact that in the second order analysis are
applied constant (i.e. load independent) values of stiffness, also called nominal
stiffness, obtained from simple approximations of the flexural stiffness, smaller
than the initial stiffness, calculated taking into account the influence of
 Effective Lengths of Reinforced... 73

cracking, non-linear qualities of materials and creep. In order to apply this

method to the whole framed structure it is necessary to carry out static
calculations according to the first order theory taking into account the nominal
stiffness of columns and then to increase the derived bending moments M0Ed
according to the following formula:

MEd = η M0Ed η= 1+
( NB
NEd
β
–1) (1)

where

π2
β= ,
c0

c0 is a coefficient which depends on the distribution of first order moment (for

instance, c0 = 8 for a constant first order moment, c0 = 9,6 for a parabolic and
12 for a symmetric triangular distribution etc.).

Formula (1) takes into account the second order effects and was derived for
a column like in Figure 1 loaded with the longitudinal force NEd and any
transverse load. w0 is a deflection calculated according to the first order theory
and w is a total deflection of column. The deflection half way through the
length of the element wmax is a sum of the deflection according to the first order
theory w0max and the increment ∆w caused by the moment of force NEd towards
the deformed axis of the element. It was assumed that the deflection can be
approximately calculated from the formula

πx
w = wmax sin
lo

After applying the Maxwell-Mohr formula we obtained:

NEd wmaxl2o wmax

∆w = = NEd
π2 EI NB
74 Michał Knauff, Krzysztof Klempka

w0
NEd
w

lo
x

NEd

Fig. 1. A diagram of a column

In the above formula EI is the flexural rigidity of a compressed element

(nominal stiffness), and NB is a buckling load obtained from the Euler’s
formula

π2 EI
NB = (2)
l2o

Total deflection

NEd
wmax = w0max + ∆ w = w0max + wmax
NB

thus

1
wmax = w0max (3)
NEd
1–
NB

Half way through the length of the element the bending moment according
to the second order theory

MEd = M0Ed + NEd wmax (4)

Substituting in (3) wmax derived from (4) we obtain

 Effective Lengths of Reinforced... 75

MEd – M0Ed NB
= w0max ,
NEd NB – NEd

( )
w0max NB
NB M0Ed
MEd = M0Ed + w0max = M0Ed 1+ (5)
NB NB
–1 –1
NEd NEd

The deflection w0max may be expressed by the following formula

M0max l2o
w0max = (6)
c0 B

Placing (6) in (5) we obtain formula (1) recommended by the norm. The
derivation of formula (1) was also presented in other articles, e.g. KLEMPKA,
KNAUFF (2005).
The effective length of a column according to the EN may be derived by
transfroming formula (2) to the following form

√
EI
l0 = π (7)
NB

Then substituting NB derived from (3) in formula (7) we obtain:

√ ( )
EI w0max
l0 = π 1–
NEd wmax

Using the above formula we can derive the buckling coefficient

√
π
µ=
l0
EI
NEd
1– (
w0max
wmax ) (8)

The examples of calculations

In the following examples the values of bending moments were calculated

in columns of two-nave buildings using the method based on nominal stiffness
and the exact method. We assumed that the horizontal force caused by wind
76 Michał Knauff, Krzysztof Klempka

pressure and suction equals H = 30 kN. We also assumed that the rigid
construction of the roof forces equal horizontal shifts of the top ends of
columns. The columns have identical cross-sections b = 40 cm, h = 45 cm,
concrete C40/50, and steel A-III. The edge columns’ reinforcement is 4 φ 16 (As
= 8,04 cm2), and the internal columns’ reinforcement is 7 φ 16 (As = 14,07 cm2)
on each side of the cross-section, a = 3,5 cm.
The calculations were carried out for two different longitudinal loads:
Case 1
P1 = 200 kN in edge columns,
P2 = 900 kN in internal column.
Case 2
P1 = 450 kN in edge columns,
P2 = 790 kN in internal column.
The calculations for case 1 are presented below.
Imperfections according to point 5.2 of the EN

2 2 2
αh = = = 0,756, ≤ 0,756 ≤ 1,0, α m = √0,5(1+1/m) =
√l √7,0 3

= √0,5(1+1/3) = 0,816,

1
θ h = θ0 α hα m = 0,756 · 0,816 = 0,00308.
200

Horizontal forces caused by imerfections:

– In edge columns H1 = θ i P1 = 0,00308 · 200 = 0,616 kN,
– In internal column H2 = θ i P2 = 0,00308 · 900 = 2,772 kN.

Calculations according to the EN – the method based on nominal

stiffness
Moments according to the first order theory caused by forces Hi, assuming
that the stiffness of columns is identical, are presented in the diagram marked
as ENIconst (Fig. 2).
The nominal stiffness (point 5.8.7.2 of the EN) depends on the amount of
reinforcement. While designing columns the reinforcement should always be
taken into account because it allows us to calculate the increased bending
moments, and based on these moments we can calculate the necessary
reinforcement. The result is obtained through iteration after reaching a rea-
sonable conformity between the assumed and calculated reinforcement.
 Effective Lengths of Reinforced... 77

P1 P2 P1
H

7.0

2×16.0

ENIconst

79.34 79.34 79.34

ENIvar

59.8 118.5 59.8

(82.1) (352.6) (82.1)

ENII

114.4 210.1 114.4

Fig.2. Static diagram and bending moments in columns [kNm]- case 1, the values in brackets were
determined using formula (1), the abbreviations ENIconst, ENIvar, ENII are explained in the text

The design value of modulus of elasticity of concrete Ecd = 29170 MPa, the
moment of inertia Ic = 3,038 × 10-3 m4. Coefficients k1 and k2 according to point
5.8.7.2 of the EN:
k1 = √fck/20 = √40/20 = 1,414. For l0 = 2lcol = 2 · 7 = 14,0 m, the radius
h 0,45
of inertia i = = = 0,1299 m. Slenderness λ = l0 / i = 14,0 / 0,1299
2√3 2√3
= 107,8. We assumed the effective creep ratio ϕef = 1,945. The coefficient k2 for
the edge columns
78 Michał Knauff, Krzysztof Klempka

n = NEd / (Acfcd) = 200/(0,40 × 0,45 × 26,7 × 103) = 0,0416,

λ 107.77
k2 = n · = 0,0416 = 0,0264 ≤ 0,20.
170 170

The moment of inertia of the reinforcement

Is = 2As(h/2 – a1)2 = 2 × 8,04 × 10-4(0,45/2 – 0,035)2 = 5,8 × 10-5 m4.

Coefficient Kc = k1k2 / (1 + φ ef) = 1,414 × 0,0264 / (1 + 1,945) = 0,0127 i Ks = 1

Nominal stiffness of the edge columns:

EI = KcEcd Ic + KsEsIs = 0,0127 × 29170 × 103 × 3,0375 × 10-3 + 1,0 ×

× 200 × 106 × 5,8 × 10-5 = 12,73 MNm2.

In the internal column

n = NEd / (Ac fcd) = 900/(0,40 × 0,45 × 26,7 × 103) = 0,1873,

λ 107.77
k2 = n · = 0,1873 = 0,118 ≤ 0,20,
170 170

Kc = k1k2 / (1 + ϕef) = 1,414 × 0,118 / (1 + 1,945) = 0,0566 i Ks = 1.

Moment of inertia of the reinforcement

Is = 2As(h/2 – a1)2 = 2 × 14,07 × 10-4(0,45/2 – 0,035)2 = 10,15 × 10-5 m4.

Nominal stifffness of the internal column:

EI = KcEcd Ic + KsEsIs = 0,0566 × 29170 × 103 × 3,0375 × 10-3 + 1,0 ×

× 200 × 106 × 10,15 × 105 = 25,31 MNm2.

The result of calculations carried out according to the first order theory
with nominal stiffness of columns is presented in Figure 2 – the diagram
marked as ENIvar.
Incresed bending moments:
a) In the edge column:
 Effective Lengths of Reinforced... 79

π2 π2
Buckling load NB = 2
EI = 12,73 = 0,64077 MN,
l0 142

π2
MEd = M0Ed 1 +
( 12
NB
NEd
–1 ) (
= 59,8 1 +
0,8225
640,77
200
–1)= 82,1 kNm.

b) In the internal column:

π2 π2
Buckling load NB = 2
EI = 25,31 = 1,27471 MN,
l0 142

π2

MEd = M0Ed
(
1+
NB
12

NEd
–1 ) = 118,515 1 +
( 0,8225
1274,71
900
–1 )
= 352,6 kNm.

Calculations according to the EN – second order theory,

nominal stiffness
Using the stiffness derived in the previous point we carry out static
calculations according to the second order theory – the result diagram marked
as ENII is presented in Figure 2.
Additionally, the buckling coefficients were calculated using formula (8) for
the edge columns

√ √
π π
µ=
l0
EI
NEd (
1–
w0max
wmax
=
7 ) 12725,14
200 (
1–
0,0765
0,1546 )
= 2,5

and the internal column

√ √
π π
µ=
l0
EI
NEd (
1–
w0max
wmax
=
7 ) 25314,41
900 (
1–
0,0765
0,1546 )
= 1,69
80 Michał Knauff, Krzysztof Klempka

P1 P2 P1
H

7.0

3×16.0

ENIconst

82.14 82.14 82.14

ENIvar

65.68 115.07 65.68

(158.72) (279.30) (158.72)

ENII

157.92 276.58 157.92

Fig. 3. A static diagram and bending moments in columns [kNm]- case 2, the values in brackets were
determined using formula (1), the abbreviations ENIconst, ENIvar, ENII are explained in the text

Values of w0max and wmax were obtained using the exact method.
The result of calculations for case 2 are presented in Figure 3.
The buckling coefficients for the edge columns:

√ √
π π
µ=
l0
EI
NEd (
1–
w0max
wmax
=
7 ) 14124,92
450
1– (
0,0759
0,2047 )
= 2,0
 Effective Lengths of Reinforced... 81

and the internal column:

√ √
π π
µ=
l0
EI
NEd(1–
w0max
wmax
= )
7
24729,69
790
1–(0,0759
0,2047 )
= 2,0

In the second case the reliable moments are similar to the reliable moments
calculated using the method of increasing the moment, while in the first case
these values are different. This diffeerence results from a false assumption
that in case 1 each of the columns in the frame behaves in the same way as an
isolated cantilever, which means that identical effective lengths were assumed
for all the columns and the cantilevers (µ = 2). Such an assumption can only be
made when the ratios of the columns’ stiffness EI to the longitudinal forces
acting in them NEd are identical, which results from formula (8). This
conclusion refers to cases in which the roof construction forces equal horizon-
w0max
tal shifts of the top ends of columns, i.e. cases in which the ratio
wmax
is identical for each column.

Conclusions

Considering the presented analyses, we can conclude that:

1. In a single-storey building with columns joined monolithically with the
foundations and via hinges with the roof construction the buckling coefficient
for each column is 2,0 only when the ratios of the columns’ stiffness EI to the
longitudinal forces acting in them NEd are identical. It may not always be
possible to meet such a condition, so assuming that µ = 2,0 for each column will
not always be correct.
2. The problem of calculating the effective length mostly refers to the
simplified method. While applying the exact method it is not required to define
the effective lengths – the shape of deflection and the resulting increase of the
moments are calculated directly. This is why while designing it is advisable to
use the exact method based on the second order analysis and taking into
account the nominal stiffness.

Accepted for print 12.10.2009

82 Michał Knauff, Krzysztof Klempka

References

Eurocode 2: Design of Concrete Structures. Part 1-1. General Rules and Rules for Buildings.
EN 1992-1-1: 2008.
JASTRZĘBSKI P., MUTERMILCH J., ORŁOWSKI W. 1986. Wytrzymałość materiałów. Arkady, Warszawa.
KLEMPKA K., KNAUFF M. 2005. Design of slender RC columns according to Eurocode and polish code
compared with the improved numerical model. Archives of Civil Engineering. LI, 4.
Konstrukcje betonowe, żelbetowe i sprężone. Obliczenia statyczne i projektowanie. PN-B-03264: 2002.
KORZENIOWSKI P. 1997. Effectiveness of increasing load bearing capacity of rc columns by raising the
strength of concrete and amount og reinforcement. Archives of Civil Engineering, XLIII, 2.
KUKULSKI W., SULIMOWSKI W. 2006. Stan graniczny nośności z udziałem efektów odkształceń kon-
strukcji.W: Podstawy projektowania konstrukcji żelbetowych i sprężonych według Eurokodu 2.
Dolnośląskie Wydawnictwo Edukacyjne, Wrocław.