Engineering Structures 31 (2009) 28652872

Contents lists available at ScienceDirect

Engineering Structures
journal homepage: www.elsevier.com/locate/engstruct

Seismic analysis of reinforced concrete frame-wall systems considering ductility

effects in accordance to Eurocode
Juraj Krlik , Juraj Krlik Jr
Faculty of Civil Engineering, Slovak University of Technology in Bratislava, Radlinskho 11, Bratislava 813 68, Slovakia

article

info

Article history:
Received 5 October 2008
Received in revised form
30 January 2009
Accepted 2 July 2009
Available online 26 August 2009
Keywords:
Seismic
Reinforced concrete
Frame-wall
Ductility
Eurocode

abstract
This paper presents the results of experimental and numerical analysis of the seismic resistance of a
reinforced concrete coupling system considering the plastic capacity in accordance with the standard
requirements STN ENV 1998 (2005), ENV 1998 (2003), norm B4015 (2002) and FEMA 368 (2001). The
plastic capacity of the structure can be established by parameter q in the case of the spectral analysis
to determine the seismic response. The experience from dynamic analysis of a hospital structure in
accordance with standard requirements is presented in this paper. Dynamic parameters of the building
structure are checked by experiment and the calculation model is modified on the basis of the experiment.
The nonlinear analysis of the coupling system was realized in the program CRACK under system ANSYS
for Kupfers failure condition and ervenkas model of the concrete strength reduction.
2009 Elsevier Ltd. All rights reserved.

1. Introduction

2. Inelastic design spectrum

The new seismic-resistant construction standards [1] and STN

P ENV 1998 enable one to consider the seismic load effect in structures with regard to their partial damage, and eventually collapse.
With respect to computational and economical complexity
there was a method established, which permits transforming a
nonlinear dynamic calculation to a linear domain. The process of
the failure is described by so-called ductility factor. The evaluation
of the seismic response based on spectral analysis performs the
ductility factor to design spectrum, as main feature of the seismic
load.
The ductility factor is well described in standards, although
nonessential ignorance of physical background could tend to incorrect results and performance errors [2,3]. Incorrect and unsuitable
interpretation can occur especially when hybrid structural systems, irregular geometry shapes and masses are considered [47].
Nonlinear behavior of a hybrid system became an aim of
this analysis. The framework of the hospital facility consists of a
combination of frames, shear walls and a core wall system (Fig. 1).
Nonlinear analyses of walls coupled with frames were realized
by software called CRACK, using Kupfers biaxial stress failure
criterion and also the reduction of concrete strength described by
ervenka [8]. Program CRACK runs as a subroutine of the software
package ANSYS [9].

If a structure or its elements occurs in the plastic domain,

the design spectrum can be reduced by a ductility factor [2].
The standardized ultimate strength characterizes elasto-plastic
behavior (Fig. 2).

Corresponding author. Tel.: +421 252494332.

E-mail addresses: juraj.kralik@stuba.sk, kralik@svf.stuba.sk (J. Krlik).

0141-0296/$ see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.engstruct.2009.07.029

fy =

fy
fo

uy
uo

1
Ry

(1)

where fy and uy are the yield values of the force and displacement of
the structure, fo and uo are peak values of a seismic response force
and deformation at the linear behavior system corresponding to an
elasto-plastic system. The value of the parameter fy is evaluated as
a function fo through the decrease factor Ry of the ultimate strength
value.
The ductility factor is defined [2] as a ratio which normalizes
the displacement in relation to the maximum displacement value
in the elastic domain. (For instance there is a case when the first
plastic hinge occurs in the frame.):

um
uy

(2)

where the displacement value um defines either a peak or an

absolute value of a seismic displacement induced in the elastoplastic system.
The ductility factor is equal to 1( = 1) if fy = fo in the case of
elasto-plastic behavior of the structure. Moreover the following is
valid
um

= fy = .
(3)
uo
Ry

2866

J. Krlik, J. Krlik Jr / Engineering Structures 31 (2009) 28652872

Fig. 1. The calculation model of the reinforced concrete structure of the hospital facility.
Table 1
The calculation model of reinforced concrete structure of the hospital building.
Period

Elastic response spectrum

ENV 1998-1 (2003)

Design response spectrum

ENV 1998-1 (2003)

Horizontal spectrum

T
TB

. (.2, 5 1)

Se (T ) = ag .S ..2, 5

TC
T

Se (T ) = ag .S ..2, 5

T C .T D
T2

0 T TB

Se (T ) = ag .S . 1 +

TB T TC

Se (T ) = ag .S ..2, 5

TC T TD
TD T

Sd (T ) = ag .S .

2
3

T
TB


i
. . 2q,5 23

Sd (T ) = ag .S . 2q,5

.
i

Sd (T ) = ag .S . 2q,5

TC
T

2,5
q

TC TD
T2

2
3

Sd (T ) = ag .S .

0, 2.ag .
i
0, 2.ag .

Vertical spectrum

0 T TB

Sv e (T ) = av g .S . 1 +

TB T TC

Sv e (T ) = av g .S ..3, 0

TC T TD
TD T

T
TB

. (.3, 0 1)

Sv d (T ) = av g .S .

T
TB


i
. . 2q,5 32

0, 2.avg .
i
0, 2.avg .

Sv d (T ) = av g .S . 2q,5

Sv e (T ) = av g .S ..3, 0

TC
T

Sv e (T ) = av g .S ..3, 0

T C .T D
T2

.
i

Sv d (T ) = av g .S . 2q,5

TC
T

Sv d (T ) = av g .S . 2q,5

TC TD
T2

3. Seismic load based on EUROCODE

Corresponding linear system

f0
Elasto-plastic system

fy

u
u v u0

um

Fig. 2. Elasto-plastic system and its corresponding linear system.

One of the most often used simplified interpretations of f y or

Ry [2] is the following:

1
1/2
f y = (2 1)
1

1
Ry = (2 1)1/2

Tn < Ta
Tb < Tn < Tc or
Tn > Tc
Tn < Ta
Tb < Tn < Tc
Tn > Tc .

The term behavior factor q, presented by national standards STN

ENV 1998 [10] and ENV 1998 [1] devoted to seismicity, reflects:
failure intensity of the structure depending on its importance, the
subsoil, and ductility and structural regularity classes. Its value
equals 1.0 in the case of linear behavior.
Basic relations dedicated to an estimation of the elastic and
design acceleration response spectrum are compiled in Tables 1
and 2, where Se (T ) denotes the ordinate of the design spectrum, T
vibration period of a linear single degree of freedom system, ag
design ground acceleration in the return period of the occurrence,
Tb , Tc limits of the constant spectral acceleration branch, Ssoil
parameter, qbehavior factor depending on the importance of the
structure, damping correction factor with reference value = 1
for 5% viscous damping.
The recommended values of parameters for spectrum type 1 in
the Table 1 are the following
av g = 0.9ag ,

TB = 0.05s

TC = 0.15s

TD = 1.

(5)

The q factor established in STN ENV 1998 [10] to 3 is expressed

in the following form:
(4)

The factor q described by Eurocode (EC) [1]standards and also

by a national standard denoted STN ENV [10] is comparable to the
decrease factor Ry of the ultimate strength value. Factor Ry is used
by the US technical literature and standards [11].

q = qo kw > 1.5

(6)

qo basic value of the behavior factor, depends on the type of

structure, and the kw factor reflecting the prevailing failure mode
in a structural system with walls.
Eurocode 8 [1] advises for various structures following ranges of
values for the behavior factor: steel structures, concrete structures,
non-bearing structures. In Table 2 there is shown a comparison
of behavior factor values, depending on system of the structure

J. Krlik, J. Krlik Jr / Engineering Structures 31 (2009) 28652872

2867

Table 2
Elasto-plastic system and its corresponding linear system.
Ductility factor q
Standard

Level

STN ENV (2005)

ENV 1998-1-1 (2003)

DCL
DCM
DCH
DCL
DCM
DCH

FEMA 368 (2001)

NORM B4015 (2002)

DCM
DCH

CRACK (nonlinear calculation)

Reinforced concrete
Frames multistory regulatory

Shear walls

Frames & shear walls

2.5
3.75(3.00)
5.00(4.00)
1
3.90(3.12)
5.85(4.68)
36
1.5
3.0

2.00
3.00(2.40)
4.00(3.20)
1
3.00(2.40)
4.40(352)
45
1.5
2.5

2.25
3.38(2.70)
4.50(3.60)
1
3.90(3.12)
5.85(4.68)
5.5
2.9
3.0
2.37/2.84

Fig. 3. Stressstrain concrete diagram.

and on criterion ductility. A wide scatter of the ductility factor q

values contributes to an enhanced demand to verify their accuracy
in respect of the type of structural system and boundary conditions.
The ductility magnitude of the structural system depends mostly
on the load level.

Fig. 4. Kupfers plasticity function.

The stressstrain relation (Fig. 3) is defined following ENV

1992-1-1 following
Loading in compression region

4. Nonlinear analysis of the reinforced concrete wall

Nonlinear analysis of the seismic resistance is determined by
the weakest member of a coupled frameshear wall hybrid bearing
system (Fig. 1). It means that: resistance of the entire structure
depends on the critical members resistance. The failure in critical
cross-sections can evoke collapse of the entire structure [4,1215,
6,7]. In the case of the hospital facility [16], there is a behavior
factor determined by the bearing capacity of the horizontal bracing
system located on the 48th axle, the place of coupling shear walls
and bracing frame.
The numerical model of the coupled frameshear wall system is
based on ervenkas reduction model [8] of the concrete strength
and also on Kupfers biaxial stress plasticity criterion in the plane
of the principal stresses. A new program was developed called
CRACK [9] containing aforementioned techniques. This program
cooperates with the system ANSYS.
The constitutive model presented is a further extension of the
smeared crack model. The smeared crack model, used in this work,
results from the assumption, that the field of more micro cracks
(not one local failure) brought to the concrete element will be
created. A validity of this assumption is determined by size of finite
element, hence its characteristic dimension, where A is the element
area (versus integrated point area of element).
One concrete layer was considered as orthotropic material for
which the direction of a crack is the same as the direction of the
principal strain.
In this model the Kupfers bidimensional failure criterion of
concrete is considered. The concrete compressive stress fc , tensile
concrete stress ft and shear modulus G are reduced after the
cracking of the concrete.

cu < eq < 0,

cef = fcef

k 2

1 + (k 2)
(c =
0.0022, cu =
0.0035).

eq
,
c

(7)

Softening in compression region

cm < < cu ,

eq

ef
c

fcef

eq c
1
cm cu

(8)

Tension region

t < eq < m ,

cef = ft exp(2( eq t )/tm ).

(9)

In the case of plane stress the strength function in tension and in

compression was considered as equivalent values. In the plane of
principal stresses can be defined the relation between the one and
two stress states due to the plasticity function by Kupfer (Fig. 4).
Compressioncompression
fcef =

1 + 3.65a

(1 + a)2

fc ,

a=

c1
.
c2

(10)

Tensioncompression
fcef

= fc rec ,


rec =

1 + 5.3278

c1
fc

rec 0.9.

(11)

Tensiontension
ftef = ft ret ,

ret =

A + (A 1) B
AB

B = Kx + A,

(12)

x = c2 /fc
ret = 1 x = 0,

ret = 0.2 x = 1.

(13)

2868

J. Krlik, J. Krlik Jr / Engineering Structures 31 (2009) 28652872

The shear concrete modulus G was defined for cracking concrete

by Kolmar in the form
G = rg .Go ,

rg =

1
c2


ln

c1 = 7 + 333 (p 0.005) ,

c1

0 < p < 0.02

(14)
(15)

where Go is initial shear modulus of concrete, u is strain in the

normal direction to crack, c1 and c2 are constants dependent on
ratio of reinforcing, p is the ratio of reinforcing transformed to the
plane of the crack.
The limit of damage at a point is controlled by the values of the
so-called crushing or total damage function Fu . This is defined in
the principal strain space in terms of the plastic strains at the point,
the limit equivalent of plastic strain up .
Function of concrete failure (loss of integrity) can be defined in
dependency to the components of principal stresses in the crack
plane of layer lth by the function of failure surface Ful . Thus
Ful = Ful ( p , up , ) = 0,
2
3

u 1

(17)

{cr } = [T ] { } ,

(18)

where [T ], [T ] are transformation matrices for the principal strain

and stress in the direction in the layer.
[T ]
cos2
sin2
sin cos
0
0
2
2

sin

cos

sin cos
0
0

2
sin

cos

2
sin

cos

cos
2

0
0

(19)

0
0
0
cos sin
0
0
0
sin cos

[T ]
cos2
sin2
2 sin cos
2
sin2
cos

2 sin cos

sin

cos

sin

cos

cos 2

0
0
0
0

0
0
0
cos
sin

0
0

0
.
sin
cos

The strainstress relationship in the Cartesian coordinates can

be defined depending on the direction of the crack (in the direction
of principal stress, resp. strain)
[cr ] = [Dcr ] {cr } and therefore
[ ] = [T ]T [Dcr ] [T ] {} = [D] {} .

(21)

where [Tc .cr ], [Ts ] are the transformation matrices for concrete and
reinforcement separately.

Bl E l
x
Bl l E l
xy x

0
 l  0
Dcr =

Bl lxy Exl

Bl Eyl

0
0

0
0

0
Glxy

0
0
Glyz

ks

(22)

Gzx

ks

Eyl
Eyl lxy

ks = 1 + 0.2

For the membrane and bending deformation of the reinforced

concrete shell structure, we have chosen the SHELL91 layered shell
element, on which we propose a plane stress on every single layer.
We propose that the crack in one layer of shell element be
oriented perpendicular to the orientation of principal stresses.
Hence, the membrane stress and strain vector depends on the
direction of principal stress and strain in one layer

n
 T     X
 l T  l   l 
= Tcl . Dlcr Tcl . +
Ts
Ds
Ts

2

Exl

where Exl (or Exl ) is Youngs modulus lth layer in the direction x
(or y), Glxy , Glxy , Glxy are shear moduli lth layer in planes xy, yz and
zx; ks is the coefficient of effective shear area

1 = 2 = 1 1p < 0, 2p < 0,

 p
u,tens
1p > 0, 2p > 0,
1 = 2 =
up,comp
 p

u,tens
1 =
;
2 = 1 1p > 0, 2p < 0,
up,comp
 p

u,tens
1 = 1;
2 =
; 1p < 0, 2p > 0.
up,comp

Bl =
(16)

where

{cr } = [T ] {} ,

 l

j0=1

c2 = 10 167 (p 0.005)

v "
u
 p 2  p 2 #
u
1

l
+ 2
up = 0;
Fu = tu
1
2

The stiffness matrix of reinforced concrete for the layer lth can
be written in the following form

(20)

1.2,

25t 2

(23)

where A is the element area, t is the element thickness.

After cracking has occurred, the elasticity modulus and
Poissons ratio are reduced to zero in the direction perpendicular
to the cracked plane, and a reduced shear modulus is employed.
Taking 1 and 2 the two principal directions in the plane of
the structure, the stressstrain relationship for concrete l-layer
cracked in the 1 direction, is

1
0

2
0

12 = 0

13

23 l
0

0
E
0
0
0

0
0
Gcr
12
0
0

0
0
0
Gcr
13 /ks
0

0
12

13

G23 /ks l 23 l

(24)

where the shear moduli are reduced by parameter rg1 by Kolmar

(14) as follow
Gcr
12 = Go .rg1 ,

Gcr
13 = Go .rg1 ,

Gcr
23 = Go .

When the tensile stress in the 2 direction reaches the value ft0 , a
second cracked plane perpendicular to the first one is assumed to
form, and the stressstrain relationship becomes:

1
0

12 = 0

13

23 l
0

0
0
0
0
0

0
0
Gcr
12 /2
0
0

0
0
0
Gcr
13 /ks
0

0
12 ,

13

Gcr
/
k

s l
23 l
23

(25)

where the shear moduli are reduced by parameter rg1 and rg2 by
Kolmar (14) as follow
Gcr
12 = Go .rg1 ,

Gcr
13 = Go .rg1 ,

Gcr
23 = Go rg2 .

The cracked concrete is anisotropic and these relations must

be transformed to the reference axes xy. The simplified averaging
process is more convenient for finite element formulation than
the singular discrete model. A smeared representation for cracked
concrete implies that cracks are not discrete but distributed across
a region of finite elements.
The smeared crack model, used in this work, results from the
assumption, that the field of more micro cracks (not one local

J. Krlik, J. Krlik Jr / Engineering Structures 31 (2009) 28652872

Plate D1

2869

Plate D4

Force F [kN]

60
50
40
30

Experiment

20

Presented

10

Pressure p [kN/m^2]

70

Experiment
Presented

0
Displacement w [mm]

Displacement w [mm]

Fig. 5. Comparison of experimental and nonlinear numerical analysis of plates.

Fig. 6. The shape of dominant modes of the hospital structural model in X ,Y and Z direction.

failure) brought to the concrete element will be created. A validity

of this assumption is determined by size
of the finite element,
hence its characteristic dimension Lc = A, where A is the element
area (versus integrated point area of element). For expansion of
cracking to be valid the assumption of constant failure energies
Gf = const

n (w) dw = AG .Lc ,

Gf =

w = w .Lc ,

(26)

where w is width of failure, n is the stress in concrete in the

normal direction, AG is the area under the stressstrain diagram of
concrete in tension. The concrete modulus for the descending line
of the stress strain diagram in tension (crushing) can be described
by Oliver depending on the failure energies in the form
Ec ,s =

Ec
1 c

c =

2Gf Ec
Lc .

2
max

(27)

where Ec is initial concrete modulus elasticity, max is the maximal

stress in concrete tension. From the condition of a real solution
of relation (27) it follows, that the characteristic dimension of the
element must satisfy the following condition
Lc

2Gf Ec
2
max

(28)

The characteristic dimension of an element is determined by

the size of the failure energy of the element. A theory of concrete
failure was implied and applied to the 2D layered shell element
SHELL91.
The program CRACK was checked and the results were
compared with the experimental results of Hjek [17,12]. The
reinforced concrete plate D1 was loaded by force F in the center
and plate D4 by pressure p on the area of plate. The comparison of
the experimental and numerical results are presented in Fig. 5.

5. Experimental modal analysis

The seismic resistance of the hospital structure was checked
in the FEM model which was calibrated on the results of the
experimental modal testing. The numerical modal analysis (Fig. 6)
was realized in the program ANSYS. An experimental modal
analysis (Figs. 79) of the hospital structure and subsoil was
performed by Arsenal Research [16]. Stiffness parameters of the
subsoil were experimentally specified (vP = 1261 m/svelocity of
longitudinal waves vS = 600 m/svelocity of transverse (shear)
waves). The subsoil class was defined as B category of Eurocode
scale. Also, then the dynamic shear modulus is following:
Gdyn = vs2 = 792 MPa.

(29)

The numerical modal analysis was realized for various subsoil

rigidities. A comparison of the modal characteristics between the
numerical and experimental analysis is presented in the Table 3.
There are presented the principal frequency characteristics and
effective mass ratio of six calculation models in three directions.
Five models (Nem1 to 5) are considered without the stiffness of
interior brick walls. The model Nem3m includes the interior
wall stiffness. These models have variable stiffnesses of soil
(high, medium and low). The medium soil corresponds to the
median value of soil stiffness obtained from the experimental
measurements. The higher (or lower) soil stiffness is equal to the
upper (or lower) quantile of the soil stiffness using the normal
distribution function, which is considered for the 95% probability
of nonexceedance. The dominant frequencies in the horizontal
direction of the X -axis and Y -axis are in the interval fX =
0.631.88 Hz and fY = 0.692.85 Hz and in vertical direction of
the Z -axis is fZ = 2.007.39 Hz (Table 3). The results from the
experimental measurements show, that the dominant frequencies
are following: fX = 1.88 Hz, fY = 3.12 Hz, fZ = 5.56 Hz.
The calculation model Nem3m with the brick wall stiffness and

2870

J. Krlik, J. Krlik Jr / Engineering Structures 31 (2009) 28652872

Fig. 7. The mode in X direction found experimentally for frequency value equals 1.88 Hz, viscous damping 11.4% [16].

Fig. 8. The mode in Y direction found experimentally for frequency value 3.12 Hz, viscous damping 4.18% [16].

Fig. 9. The mode in Z direction found experimentally for frequency value 5.56 Hz, viscous damping 2.99% [16].
Table 3
Critical modes of hospital building.
Model

Subsoil

Direction X

Vert./Horiz.

Frequency (Hz)

Effect. mass ratio (%)

Direction Y
Frequency (Hz)

Effect. mass ratio (%)

Direction Z
Frequency (Hz)

Effect. mass ratio (%)

Nem1
Nem2
Nem3
Nem4
Nem5

R/R
R/EH
EH/EH
EM/EM
EL/EL

1.12866
1.06114
0.86421
0.69891
0.62927

68.095
71.799
58.735
79.256
79.610

1.52316
1.51682
0.99064
0.78654
0.69355

68.533
69.231
80.775
91.525
92.000

7.39420
7.39416
5.54769
2.77618
2.00706

59.346
59.551
59.372
93.210
98.464

Nem3m

EH/EH

2.85207

69.998

5.54867

47.241

5.56

Model with internal brick wall elements

1.87976

68.022

Experimental measured critical eigen-frequencies

Experiment

1.88

3.12

Notes: RRigid subsoil, EHElastic with the high rigidity, EM Elastic with the medium rigidity, ELElastic with the low rigidity.

high soil stiffness gives us the best results in comparison with

the experimental results (Table 3). However, this case corresponds

to an elastic behavior of the soil and structure. The effect of

the seismic action depends on the principal frequencies in the

J. Krlik, J. Krlik Jr / Engineering Structures 31 (2009) 28652872

3.0

2.5

Elastic
Inelastic q=2.0
Inelastic q=4.7

2.0
1.5
1.0
0.5
0.0

Accel. spectrum Sa [m/s^2]

Accel. spectrum Sa [m/s^2]

3.0

2871

2.5

Elastic
Inelastic q=2.0
Inelastic q=4.7

2.0
1.5
1.0
0.5
0.0

12 15 18 21 24 27 30 33
Frequency [Hz]

9 12 15 18 21 24 27 30 33
Frequency [Hz]

Fig. 10. Horizontal and vertical acceleration spectrum.

Fig. 11. FEM model of frame-wall system in modul 48/A-N (a) Original model (b) Upgraded model.

horizontal and vertical directions (Fig. 10). The peak of the

horizontal (or vertical) acceleration spectrum is in the interval
27 Hz (or 620 Hz). Also, hence the seismic action decreases with
the lower level of the principal frequencies of the structure.
In the case of the earthquake even the soil stiffness is reduced
and the stiffness of the interior walls will be degraded too.
The shear modulus of the soil is reduced depending on the
earthquakes intensity, i.e. G = (0.70.9)Gmax for ag = 0.1 g. The
experience and experimental results from the investigation of the
interaction of interior brick walls and concrete frame establishes
that the thin interior walls and walls without rigid connection
are damaged during earthquake events and their real stiffness is
quickly reduced. The standards (FEMA 368 and Eurocode 8) permit
one to calculate the frame-wall interaction effect in the case that
the connection between them is rigid and the seismic resistance of
the wall is fulfilled.
Also, on the base of numerical analysis of all models the critical
structure element of the hospital building is the frame and wall at
the one side of the building. After the damage of this part of the
structure the building will collapse during the earthquake event.
The stiffness of this building part is lower as from the opposite
side. From that reason the building is rotated about vertical axis
(Figs. 610) and some structure elements between them are more
exposed. Two alternatives of these frame-wall structures (original
and upgraded) was investigated from the point of the capacity
check in elastic and elasto-plastic state (Fig. 11).
6. Nonlinear analysis of the ductility factor
The nonlinear behavior of the hospital building was investigated depending on the ductility capacity of the reinforced
concrete shear wall on the building side. Estimated values of equivalent horizontal forces were calculated by constraint equations

which conveniently match the reference assumption of horizontal

degree of freedom values along the height of the building. Afterwards nonlinear analysis was carried out just for a selected part of
the building structure (Fig. 11). We considered the capacity of the
original and upgraded frame-wall structures.
The ductility factor was established from a ratio (2) of the
critical fragment deflection in a linear behavior to the nonlinear
value (Fig. 12).
The ductility factor for the original model is as follows

nonlin
0.02906
=
= 1.29
lin
0.02246
nonlin
0.10237
=
=
= 2.37
lin
0.04320

|ag =0.64 =
|ag =1.25

(30)

and for the upgraded structure

nonlin
0.02703
=
= 1.33
lin
0.02035
nonlin
0.12808
=
=
= 2.84.
lin
0.04510

|ag =0.64 =
|ag =1.5

(31)

The ductility factor, described by the maximum load equation

(2) according to [2,3] is identical to the behavior factor q for period
interval TB T TC . The value of the ductility factor obtained by
calculations is evidently lower than the limit value defined by the
Euro code standard. But its value is the most similar to an Austrian
national standard called NORM B4015 (2002) [18] recommended
value.
The seismic load described by the acceleration spectrum in
the Eurocode standard is calculated by equations compiled in
Table 1 in accordance to defined behavior factor. A comparison of
acceleration design inelastic spectra on Fig. 10 shows that values

1.6

1.6

1.4

Design accel. ag [m/s^2]

J. Krlik, J. Krlik Jr / Engineering Structures 31 (2009) 28652872

Design accel. ag [m/s^2]

2872

1.4

1.2
1.0
0.8
Nonlinear
Linear

0.6
0.4
0.2
0.0

1.2
1.0
0.8

Nonlinear
Linear

0.6
0.4
0.2
0.0

20

40 60 80 100 120 140 160

Maximum displacement [mm]

20

40 60 80 100 120 140 160

Maximum displacement [mm]

Fig. 12. Maximum displacement and design acceleration dependence (a) Original model (b) Upgraded model.

at the range of their peaks (it means the frequency range between
27 Hz) are significantly reduced, i.e. frequencies which describe
main mode shapes of the structural system.
7. Conclusions
The application of behavior factor with respect to the failure
of the structure significantly affects the design of structures
in seismic regions [16]. Projects of the structures become
economically efficient. In the case of more variable structures, it
means that structures with irregular geometry in the horizontal
as well as the vertical plane, hybrid structures combined with
various bearing systems etc. it is necessary to verify the accuracy of
the behavior factor value. This factor was achieved by a nonlinear
calculation method performed on the weakest element of the
structure [2].
Factors of the behavior, described in recent standards, are in
case of some irregular structures unsuitably defined. The performance of these values would result into incorrect conclusions, as
the article shows. The nonlinear analysis, of the 2D critical substructures (wall, frame, core wall,. . .) subjected to the quasi load,
presents an acceptably accurate view of its resistance.
In this paper was presented the nonlinear analysis of the
concrete structures considered the concrete cracking and crushing,
layered approximation of the shell elements with various material
properties, orthotropic material depending on the direction of the
rotated cracks and the orientation of reinforced steel, modified
Kupfers yield function, degradation of the shear modulus by
Kolmar depending on the properties of the reinforcement [9].
The ductility factor determined from the nonlinear quasi-static
analysis give us the more accurate results.
Acknowledgements
This survey was solved with support of the Ministry of
Education in the Slovak Republic within the grant task VEGA
1/0849/08.

References
[1] ENV 1998-1, Eurocode 8. Design of structures for earthquake resistance. Part1
General rules, seismic actions and rules for buildings. 2003. CEN.
[2] Chopra KA. Dynamics of Structures. Berkeley: Prentice Hall, University of
California; 2001.
[3] Flesch RG. Baudynamik. Praxisgerecht, Band 1. Berechnun-ggrundlagen.
Bauverlag GMBH. Wiesbaden und Berlin. 1993.
[4] Cicio A, Wawrzynek A. Adaptation of the plastic-damage concrete model for
a masonry material subjected to cyclic load. In: VIII international conference
on computational plasticity, COMPLAS VIII. 2005 [on CD].
[5] Krlik J, Krlik Jr J. Probability and sensitivity analysis of soil-structure
interaction of high-rise buildings. Slovak J Civil Eng 2006;3:1832 [Faculty of
Civil Engineering SUT in Bratislava].
[6] Materna A, Salajka V, Brdeko L. Calculation models for the nonlinear solution
of reinforced structures. In: Nonlinear mechanics, III. SlovakPolandCzech
symposium of mechanics. 2000. p. 113116 [Faculty of Civil Engineering SUT
Bratislava].
[7] Wawrzynek A, Cicio A, Kosiski K. Wraliwo odpowiedzi ukadu dynamicznego na wybrane parametry plastyczno-degradacyjnego modelu materiau.
Material XI. Sympozjum, Wpywy Sejsmiczne i Parasejsmiczne na Budowle.
2006.
[8] ervenka V. Constitutive model for cracked reinforced concrete. ACI J 1985;
87782.
[9] Krlik J, Cesnak J. Nonlinear analysis of power plant buildings with the VVER
230 reactor after a loss off coolant accident. Slovak J Civil Engrg 2001;3:1832.
[10] STN ENV 1998. Design of structures for earthquake resistance. STN,
Bratislava. 2005.
[11] FEMA 368. NEHRP recommended provisions for seismic regulations for new
buildings and other structures. Part 1: Provisions, BSSC Washington DC. 2001.
[12] Jerga J, Krima M. Assessment of concrete damage. Building Res J 2006;
54(34):21120.
[13] Juhsov E. et al. Real time testing of reinforced infills. In: Proceedings of
12WCEE. 2000. p. 921/18.
[14] Krlik J, et al. Seismic analysis of reinforced concrete wall and frame
interaction in consideration of ductility. In: Soize C, Schuller GI, editors.
Proc. 6th international conference on structural dynamics. Rotterdsam
(Netherlands): MillPress; 2005. p. 1799804.
[15] Krlik J, Tnes R. Seismic analysis of reinforced concrete coupled systems
considering ductility effects in accordance to Eurocode. In: First European
conference on earthquake engineering and seismology. Abstract book. 2006.
p. 436.
[16] Flesch RG. et al. Seismische analysis. Spitler-Projekt. Leoben. No
2.05.00133.1.0. FPZ Arsenal. Ges mbH. 2003.
[17] Hjek J, Fecko L, Nrnbergerov T. Deformation of reinforced concrete plates
loaded in two directions by long-time forces. Rep. V III-3-4/01.1. STARCH
SAV Bratislava. 1983.
[18] NORM B4015. Belastungsannahmen im Bauwesen-Auergewhnliche
Einwirkungen-Erdbe-beneinwirkungen. Grundlagen und Berechnungsverfahren. NORM, Wien. 2002.