1st Croatian Conference on Earthquake Engineering

1CroCEE 22-24 March 2021 Zagreb, Croatia

DOI: https://doi.org/10.5592/CO/1CroCEE.2021.19

Experiences in seismic design of structural bearings and

expansion joints for RC bridges according to Eurocodes

Viktor Hristovski1, Marija Vitanova2, Nikola Hristovski3

1
 rofessor, Ss. Cyril and Methodius University in Skopje, R.N. Macedonia, Institute of Earthquake Engineering
P
and Engineering Seismology - IZIIS, viktor@iziis.ukim.edu.mk
2
Assistant Professor, Ss. Cyril and Methodius University in Skopje, R.N. Macedonia, Institute of Earthquake
Engineering and Engineering Seismology - IZIIS, marijaj@iziis.ukim.edu.mk
3
M.Sc. student, Ss. Cyril and Methodius University in Skopje, R.N. Macedonia, Institute of Earthquake
Engineering and Engineering Seismology - IZIIS, nikola_hr@yahoo.com

Abstract
The proper seismic design of bridge structural bearings and expansion joints enables avoiding
high repair costs after earthquakes. The paper discusses some experiences of the authors in
design of bridge bearings and expansion joints according to Eurocodes. Within the scope of the
detailed design of several new road overpasses, railway underpasses and railway bridges in the
Republic of Macedonia, a special project on design of structural bearings and expansion joints
has been realized. Based on the performed analyses, several issues are discussed. First, it has
become evident that the geometry and the layout of the bridge structures have an influence on
the layout and the fixation of the bearings (i.e., fixed, guided in one direction or bearings movable
in all directions). The bridge shape in plane, the number of spans, the alignment disposition, etc.
have also affected the adoption of bearing types and dimensions. Finally, the dynamic parameters
and seismic behaviour of the bridges have appeared to be crucial in the design of the bearings.
The paper also includes a comparison between design of bearings according to the Eurocodes
and the Macedonian codes. For the purpose of comparative investigation of different layout
solutions for the bearings’ fixation, apart from static analyses, dynamic non-linear time-history
analyses, using time-history ground acceleration records that correspond to the Eurocode 8
design spectra, have also been performed. From the analyses, the most superior layouts in terms
of fixation, position and number of bearings have been obtained for each treated bridge structure.
Precious experience has been gained during realization of the project. Since the Eurocodes’
design philosophy has already moved toward development of new technologies for bearings,
it is inevitable to make changes in the Macedonian practice. This philosophy has brought more
freedom to the design process, especially in making decisions in obtaining optimum solutions
depending on predominant actions, especially seismic ones.

Key words: s tructural bearing, elastomer, Eurocodes, EN 1337, bridge structure, thermal action,
seismic action, FEM, expansion joints

1st Croatian Conference on Earthquake Engineering - 1CroCEE 873

 1 Introduction
The paper is focused on problems in seismic design of structural bearings and expan-
sion joints for RC bridges according to Eurocodes and EN 1337 based on the authors’
experiences. Although it mainly deals with seismic actions that have an influence on the
bearing design, nondynamic actions, especially thermal and breaking ones, have also
been included in the discussion.
The detailed design of several new bridge structures (road overpasses, railway under-
passes and railway bridges) along Corridor VIII–eastern section (Kumanovo–Beljakovce
section) in the Republic of Macedonia, performed according to the Macedonian codes,
has served as a basis for additional design and comparative analysis of structural bear-
ings according to Eurocodes. Since the bearings and expansion joints for the bridges
have been adopted based on the result of this additional design, the project appeared
to be among the first ones in Macedonia realized by implementation of the Eurocodes
and EN 1337. During the design process several issues have been considered. These
have mainly been related to the influence of the layout of the bridge structures and the
fixation of the bearings discussed in the section 2 of this paper. However, the dynamic
parameters and the seismic behaviour of the bridges have appeared to be crucial in
the design of the bearings. To evaluate the effect of the bearing fixation, comparative
dynamic non-linear time-history analyses have been performed for different layout so-
lutions for the bearings, as discussed in section 3 of the paper. During the realization
of the project, it has become clear that the present Macedonian practice that is mainly
based on design of movable elastomer bearings disregarding seismic actions should
inevitably be abandoned. The Eurocodes’ bearing design philosophy offers wider con-
cepts for fixation of bearings - fixed, guided in one direction or bearings movable in
all directions, so that new horizons in obtaining optimum solutions for bridge bearings
have been opened.

2 Detailed design of six road overpasses, five railway underpasses

and one railway bridge
In the paper, the results from the performed analyses and detailed design of structural
bearings and expansion joints for the structures along the “Corridor VIII – Eastern Sec-
tion”, Kumanovo – Beljakovce section, Republic of Macedonia, are discussed. The fol-
lowing structures have been analyzed: 1. Road overpasses OP30 (km 3+653), OP31 (km
7+555), OP32 (km 9+060), OP33 (km 12+060), OP34 (km 17+758), OP35 (km 26+582);
2. Railway underpasses UP45 (km 14+951), UP46 (km 16+211), UP47 (km 20+220),
UP48 (km 22+631), UP49 (km 24+768) and 3. Railway bridge BR54 (km 24+460).
Within the scope of the detailed design the authors of the paper have performed static
and seismic analyses of the mentioned bridge structures and have computed the ac-
tions according to EN 1990, EN 1991-1-5, EN 1991-2, EN 1992-2, EN 1998-1, and EN

874 SEISMIC PERFORMANCE OF STRUCTURES

1st Croatian Conference on Earthquake Engineering - 1CroCEE
1998-2 (see [1-6]), needed for proportioning of the elastomeric bearings. The design
actions on elastomeric bearings, defined as appropriate Ultimate Limit State (ULS) and
Serviceability Limit State (SLS) combinations, have been calculated and the so called
“Typical bridge bearing schedule” has been prepared according to EN 1337-1, Annex B,
Table B.1 and EN 1337-3, Annex E, Table E.1 (see [7, 8, 9]) for all bearings with speci-
fied different types of fixation for all structures. It is important to note that, in absence
of a National Annex of the Republic of Macedonia to Eurocodes, recommended values
of the parameters have been adopted in the analyses, unless otherwise specified in the
design text.

2.1 Numerical models for finite glement analysis and applied actions for bea-
ring design
In order to evaluate the design values of the effects of actions Ed on elastomeric bear-
ings for all considered bridge structures, 3D numerical models have been developed for
all designed structures. These numerical models have been used for calculation of the
natural mode shapes and periods of vibrations, as well as for calculation of the static
reactions of each structural bearing, obtained due to all permanent, variable and seis-
mic actions. The analyses have been performed using the Finite Element Method (FEM)
based, general purpose software package Tower Radimpex 7.0 [10]. The structures
have been appropriately modeled using the generated finite element mesh: Girders
and columns have been modeled using beam finite elements. Slabs have been mod-
eled using plate elements and bearings have been modeled using link elements. Since
the bearings were the target of the analyses, the structures have been analyzed with-
out consideration of the abutments. For structures with more than one span, middle
columns have been included in the models, resulting in optimal design of elastomeric
bearings. The 3D design models have appeared especially superior as they allowed ap-
propriate consideration of the horizontal seismic actions.
The elastomeric bearings have been modeled by link elements with stiffness calculated
according to specifications given in EN 1337-3 (see [8] and [11]), so that stiffness for all
6 degrees of freedom (three translations and three rotations) has been calculated and
prescribed. Also, the boundary conditions of the link elements have been prescribed, ac-
cording to the designed type of bearing fixation. The shear modulus of the elastomeric
bearings has been adopted as Gg = 0,9 MPa (EN 1337-3, 4.3.1.1, see [8]). The bulk
modulus Eb has been adopted as Eb = 2000 MPa (1337-3, Eq. (20), Note 1, see [8]).
During analysis, the initial values of bearing dimensions and types have been adopted
as in the original design. However, in the design process (according to EN 1337-3), the
bearing dimensions and types have been changed. The boundary conditions (fixation)
have remained unchanged.
The considered action on structures has been classified into three groups (see [1], i.e.
EN 1990, 4.1.1): permanent actions (G), variable actions (Q), and seismic actions (A).
Permanent actions for railway bridges include self-weight of structure, weight of rails,

SEISMIC PERFORMANCE OF STRUCTURES 875

1st Croatian Conference on Earthquake Engineering - 1CroCEE
 sleepers and ballast variable actions. For road bridges, permanent actions include self-
weight of the reinforced concrete structural elements (girders, slabs, columns), weight
due to asphalt, hydro-isolation, RC parapets, noise barriers and/or solid safety barriers
on the cantilever parts, RC conduits elements, etc. Variable actions for railway bridges
include railway traffic actions (EN1991-2, 6.1., 6.2, 6.3), horizontal forces (centrifugal
force (EN 1991-2, 6.5.1), noising force (EN 1991-2, 6.5.2), traction and braking (EN
1991-2, 6.5.3), thermal actions (EN 1991-1-5, 6.1, 6.1.3)), wind actions (EN 1991-1-
4, 8.3.2) and actions due to shrinkage and creep of concrete. Variable actions for road
bridges include road traffic actions (EN1991-2, 4.3 and 4.4), thermal actions (EN 1991-
1-5, 6.1, 6.1.3)), wind actions (EN 1991-1-4, 8.3.2) and actions due to shrinkage and
creep of concrete. Characteristic values of actions have been defined according to EN
1990, 4.1.2. Design values of actions have been defined according to EN 1990, 6.3.1
using partial factors g and probability factors yi. The final adopted ULS combinations
needed for calculation of the design values Ed of the effects of actions which are neces-
sary for design of bearings for both railway and road bridges have been adopted accord-
ing to EN 1990, 6.4.3.2, Eq. 6.10b for STR for persistent or transient design situation,
and according to EN 1990, 6.4.3.4, Eq 6.12b for seismic design situation. Also, the final
SLS combinations have been adopted according to EN 1990, 6.5.3a, Eq 6.14a for a char-
acteristic situation and according to EN 1990, 6.5.3a), Eq 6.15b for a frequent situation.

2.2 Applied seismic actions for bearings and expansion joints design according
to Eurocodes
Seismic actions have been included in the analyses according to EN 1998-1 and
EN1998-2. Mass has been calculated using the characteristic value of permanent loads
Gk and a part of the characteristic traffic loads Qk,1, using factor y2,1 equal to 0.3 for rail-
way bridges and equal to 0.2 for road bridges (EN 1998-2, 4.1.2(4)P):

M = Gk + y2,1 Qk,1. 1)

The damping ratio has been adopted to have the value of 0,05 (EN 1998-2, 4.1.3). The
behavior factor q for bridges with fixed bearings (or partially fixed bearings) according
to EN 1998-2, 4.1.6 (10) should be adopted equal to 1,0 if the first horizontal period of
vibration T < 0,03 sec, and if T > 0,03 sec, q should be adopted to have the value of q =
1,5 (EN 1998-2, 4.1.6 (9), (10), (11)). However, in absence of the Macedonian National
Annex and having in mind EN 1998-2, 6.6.2.3 (1), (2) and (4), the behavior factor q for
railway bridges in this study has been adopted to have the value of 1,5 for bridges with
middle columns, and value between 1,0 to 1,5 for bridges with one span. The behavior
factor q for all road bridges with flexible elastomeric bearings (free at all supports) has
been adopted equal to 1,0 (EN 1998-2, 4.1.6 (11P), EN 1998-2, 6.6.2.3 (1) c. and (4),EN
1998-2, 7.4.1 (1)P). A multi-mode spectrum seismic analysis has been carried out for
all cases. The design spectrum Type 1 for elastic analysis in both horizontal directions

876 SEISMIC PERFORMANCE OF STRUCTURES

1st Croatian Conference on Earthquake Engineering - 1CroCEE
has been used (EN 1998-1, 3.2.2.5) for M > 5,5 (EN 1998-1, 3.2.2.2 (2)P, Table 3.2 and
Figure 3.2). Class of soil B has been adopted for all analyses, according to the map of
soils of the Republic of Macedonia and Eurocode 8 classification. The design ground ac-
celeration has been adopted as ag = gI agR (EN 1998-1, 3.2.2.2), where importance class
factor gI has been adopted equal to 1,4 for railway bridges and 1,0 for road bridges (EN
1998-2, 2.1 (6)). agR is the peak ground acceleration (PGA) of the site for a return period
of 475 years. As to the choice of PGA, in absence of a map and Macedonian National
Annex, an approximate map of PGA for the Republic of Macedonia for a return period of
475 years from one research has been used. Considering the site where the structures
are to be built, for all analyses the value of agR = 0.20g has been adopted. This value also
correlates with the positive technical regulative which is in force in Republic of Mac-
edonia (i.e., Regulative 1986, see [11]). According to EN 1998-2 4.1.7 (3)P the effects
of the vertical seismic component on bearings in the upward direction has also been
considered in the analyses.

2.3 Design solutions for bearings

The final adopted bearing dimensions are shown comparatively in Table 1 for design
performed according to Regulative 1986 (see [11]) and Eurocodes (see [1-9]). It should
be noted that, for the underpass structures UP45, UP48 and UP49 and the bridge struc-
ture BR54 the horizontal seismic actions on bearings computed according to Regulative
1986 were comparatively low compared to those computed according to Eurocodes.
That was the reason why Regulative 1986 design was done with no complications re-
garding the bearing fixation elements that sustain horizontal forces, unlike the Euroc-
ode design where special measures (installation of steel plates) had to be undertaken
because of the much higher value of the seismic actions (see the discussion about the
difference in seismic actions when using both codes in the subsequent section 2.3). So,
Table 1 shows comparatively only the designed dimensions of the rubber, without ad-
ditional steel plates and bolts needed in the Eurocode design.
For bridge BR54 (see the model in Figure 1), the bearings based on Eurocode design
have been adopted of all types of fixities: fixed bearings type B with lower and upper
steel plates; guided bearing in longitudinal direction, type D with upper vulcanized PTFE,
upper sliding plate with welded stainless steel sheet and lower steel plate; guided bear-
ing in transverse direction, type D with upper vulcanized PTFE, upper sliding plate with
welded stainless steel sheet and lower steel plate; and free bearings type C with inter-
nal steel plates. The layout of the designed bearings for this bridge is shown in Figure 2,
denoted as “layout 1”.

SEISMIC PERFORMANCE OF STRUCTURES 877

1st Croatian Conference on Earthquake Engineering - 1CroCEE
 Table 1. 
Comparison between bearing dimensions designed according to the Macedonian code and
Eurocode
Bearing dimensions [mm]
Bridge Regulative 1986 Eurocodes (only rubber dimensions)
structure Guided Guided
All types Fixed Free
x-x y-y
BR54 NB 300/500/63 250/400/63 200/500/79.5 250/500/79.5 500/400/177
OP30 NB 150/200/28 / / / 350/450/132
OP31 NB 150/200/28 / / / 350/450/132
OP32 NB 150/200/28 / / / 350/450/132
OP34 NB 150/200/28 / / / 350/450/116
UP45 NB 200/300/41 180/200/35 190/250/25.5 150/250/25.5 150/250/49
UP48 NB 200/300/41 180/200/35 190/250/25.5 150/250/25.5 150/250/49
UP49 NB 200/300/41 150/250/35 150/350/32.5 150/300/25.5 150/250/56

Figure 1. Model of the four-span curved bridge BR54, using the FELISA/3M software package

On the other hand, the overpass structures OP30, OP31, OP32 and OP34 have been
designed with free (moveable in both directions) neoprene elastomeric bearings type
C. As to the displacement requirements this fixity layout did not result in high values
of horizontal displacements in Regulative 1986 design, however it caused problems
in Eurocode design because of much higher level of seismic actions. So, the bearings
have been designed to have higher dimensions that has resulted in low level of contact
stresses, namely less than 3 N/mm2 (see [8], EN 1337-3, section 5.3.3.6, eq. (16)).
Finally, the problem was solved using special thixotropic epoxy adhesive for gluing the
steel plates to concrete (with adhesive strength > 4 N/mm2) on both surfaces in order
to avoid separation and/or slippage of the bearings.

2.4 Design solutions for expansion Joints

The expansion joints have been dimensioned according to the obtained horizontal de-
sign displacements dEd in longitudinal direction of the structures, using the following two
basic combinations:

878 SEISMIC PERFORMANCE OF STRUCTURES

1st Croatian Conference on Earthquake Engineering - 1CroCEE
-- Ultimate Limit States (ULS) - Seismic combination:

dEd = 1,0·dEG±0,4·dESx±0,5·dET (2)

-- Serviceability Limit States (SLS), Characteristic combination:

dEd = dEG+dEQrt±dEHtb±0,6·dET+0,6·dEQsh (3)

In the combinations dEG denotes displacements due to all permanent actions as self-
weight of the reinforced concrete structural elements (girders, slabs, columns), asphalt
and isolation in road bridges, the RC parapets, noise barriers and/or solid safety barriers
on the cantilever parts, the RC conduits elements, etc. Horizontal displacements due to
variable actions are denoted as follows: dEQrt denotes the horizontal displacements due to
vertical road traffic for road bridges, dEHtb denotes horizontal traction and braking displace-
ments; dET denotes thermal displacements, dEQsh denotes concrete creep and shrinkage
displacements, and dESx denotes seismic displacement. Note that the ULS combination 1
is adopted according to clauses and recommended values in EN 1998-2:2005, Eurocode
8 – Design of structures for earthquake resistance – Part 2: Bridges, 2.3.6.3 (5), Note 1.
According to the clause, the adopted values of design displacements will provide normal
work of the expansion joints with no damage during the serviceability period, including
medium intensity earthquakes that are expected to occur during the serviceability life of
the structures (that is the meaning of the coefficient 0.4 for earthquake displacement in
the above ULS combination). According to above mentioned clause in Eurocode, for max-
imum intensity expected earthquakes (with return period of 475 years), the expansion
joints are allowed to suffer damage. However, this damage mechanism would provide en-
ergy dissipation that will increase the ductility, preventing global collapse of the structure
(superstructure falling off). The transverse displacements due to seismic motion have also
been reported in order to design the clearance between the girders and side parapet walls
in transverse direction on the abutment supports.

2.5 Discussion about differences in design according to Macedonian codes

and Eurocodes
The apparently considerable difference between Eurocode design and design of bear-
ings according to the Macedonian code, i.e., Regulative 1986 has arisen as a result of
several factors: 1. Difference in the philosophies of actions on structures, especially
treatment of thermal ones; 2. Different definition of calculation of mass contributing
to seismic actions, i.e., Eurocodes include the contribution of traffic actions while Mac-
edonian codes do not; 3. Difference in ductility factors, i.e. reduction factors of seismic
action, which are lower in Eurocode; 4. Much higher value of braking force in Eurocode;
5. Remnants of old practice in Republic of Macedonia including very often neglect of the
seismic forces in design of bearings, etc.

SEISMIC PERFORMANCE OF STRUCTURES 879

1st Croatian Conference on Earthquake Engineering - 1CroCEE
 As to the first issue, i.e., differences in philosophies of actions, in Regulative 1986, ther-
mal actions are treated as “additional” forces that are never combined with the seis-
mic ones. However, according to Eurocode, thermal actions should be combined with
seismic actions (see [6], i.e., EN 1998-2-2005, 5.5. (2)P). Also, the elastic seismic value
of displacements (see [6], i.e., EN 1998-2-2005, 2.3.6.3) should be enlarged by the be-
havior factor q (see [6], i.e., EN 1998-2-2005, 2.3.6.1 (6)P, (7) and (8)P). In such a way,
the obtained design displacements and forces of bearings are much higher than those
treated according to the Macedonian code.
Further, unlike Regulative 1986, according to the Eurocodes, the mass contribution due
to traffic should be considered in calculation of the total mass for dynamic analysis with
20 % for road bridges and with 30 % for railway bridges (see [6], i.e., EN 1998-2-2005,
4.1.2 (4)P). This results in greater mass and bigger seismic actions applied to structures
when using Eurocodes.
Ductility in Regulative 1986 (see [11], article 23, Table 5) is taken to have the value of 4.0
for “Z1” design action and 5.0 for “Z2” design action, including a special reduction factor
for the total seismic force of 0.6. If we consider the relation between seismic intensity
and design peak acceleration ([11], article 20, Table 2), the total seismic reduction factor
will range between the values of 3.33 and 4.16. However, according to Eurocode (see
[6], i.e., EN 1998-2-2005, all stipulations in section 4.1.6 including Table 4.1), reduc-
tion factor (called in Eurocode “behavior factor”) for the analyzed bridges should be 1.0
for bridges with all movable elastomeric bearings and with maximum value of 1.5 for
bridges with different fixation layouts. This results in a considerable difference of cal-
culated seismic force which could be 2.22 to 4.16 times bigger when implementing the
Eurocodes, compared to the Regulative 1986. In addition, given that seismic actions
used to be often neglected for smaller structures and having in mind the lower values
of braking forces in the Macedonian code, the significance of bearing design was usually
underestimated in practice, resulting in poor solutions. As a result, considerable dam-
age to bearings of bridge structures along Macedonian roads and railway lines has been
observed due to serviceability actions let alone seismic events.
The National Annex to the Eurocodes was recently adopted in the Republic of Mac-
edonia (September 9, 2020). However, despite many deficiencies in the Macedonian
seismic codes [11], if properly implemented with efficiently defined bearing layout, they
could often result in an acceptable design level. On the other hand, the Eurocodes could
seem very rigorous and their direct implementation could cause many problems to de-
signers. Hence, clever adoption of parameters in the National Annex in its next version
should be an imperative in order to overcome these difficulties and inconsistencies.

880 SEISMIC PERFORMANCE OF STRUCTURES

1st Croatian Conference on Earthquake Engineering - 1CroCEE
3 Comparative dynamic non-linear analyses of BR54 railway Bridge
structure
For the final design, the bearings for all treated structures have been analyzed and pro-
portioned according to Eurocodes and EN 1337, as discussed in the preceding chapter.
However, due to deadline restrictions, during the design phase, it was not possible to
perform any parametric optimization of the bearings related to their layout consider-
ing all the influencing parameters. In the post-design activities, we decided to investi-
gate the efficiency of the optimization process that could lead the design phase, using a
model of the bridge structure BR54 (Figure 1) generated in FELISA/3M software system
(see [12]). Namely, as discussed in the introduction, the idea was to investigate the
influence of the fixation (i.e., the boundary conditions) of the bearings on the seismic
behavior of the bridge using five different fixation bearing layouts. The first four layouts
are given in Figure 2, while the fifth one that is not shown in the figure, is defined with
all eight bearings free in both directions, longitudinal (L) and transverse (T). “Layout 1”
represents the actual design solution. Time-history analyses have been performed us-
ing original El-Centro ground acceleration records in x and y directions so that, for both
directions, 0.2g has been prescribed for PGA. These acceleration records fall approxi-
mately into the design spectrum Type 1, as used in the design.
The first three layouts of bearing fixation practically simulated a situation of a simple
beam with a free support on the left side and a fixed support of the right side for all
spans. The first period of vibration for all three cases was 0.4367 sec. with the first
mode shape directed longitudinally in respect to the bridge. Consequently, this resulted
in very similar results for all three layouts as to the obtained internal bearing forces and
displacements due to seismic actions (see Table 2 for “Layout 3”, span 2,). In the table,
the fixed bearings are denoted by “F”. Also, letters “T” and “L” denote the free moveable
directions of the bearings. From the table, it can be seen that the maximum internal
horizontal force due to seismic actions in the fixed bearings riches the value of 762 kN in
longitudinal direction and 409 kN in transverse direction. On the other hand, the maxi-
mum obtained horizontal displacements in the bearings in the case of all three layouts
are about 18.2 mm in longitudinal direction (obtained in the first span, not presented
here) and 0 in transverse direction.

SEISMIC PERFORMANCE OF STRUCTURES 881

1st Croatian Conference on Earthquake Engineering - 1CroCEE
 Figure 2. The analyzed bearing layouts. Note that the fifth one is defined with all eight bearings that are
movable in both directions (not presented in the figure)

Table 2. Internal bearing forces and displacements for span 2, Layout 3

Force Displacement Bearing Force Displacement

Bearing no. Direction
[kN] [mm] no. [kN] [mm]

3 2.5 687 0 Longitudinal

12 L 16 F
243 0 409 0 Transverse

3 2.5 687 0 Longitudinal

11 L 15 T
210 0 0 0 Transverse

3 2.5 762 0 Longitudinal

10 L 14 T
189 0 0 0 Transverse

3 2.5 682 0 Longitudinal

9L 13 F
127 0 393 0 Transverse

882 SEISMIC PERFORMANCE OF STRUCTURES

1st Croatian Conference on Earthquake Engineering - 1CroCEE
Layout 4 of bearing fixation partially restricted the motion in both directions, resulting in
stiffer response, i.e., in the first period of vibration, with a value of 0.2682 sec in longi-
tudinal direction. The maximum developed displacements were 0.7 mm in longitudinal
direction and 0 mm in transverse direction. The shift of the period of vibration in stiff-
er regions of the design spectrum (from 0.4367 sec. to 0.2682 sec.) resulted in much
higher induced seismic forces in the bearings. Namely, the internal horizontal force due
to seismic actions in guided bearings in transverse direction for span 4 reached a high
value of 2338 kN in longitudinal direction. Probably this bearing layout will not behave
well under thermal actions, too. Finally, layout 5 has been proposed just to simulate the
limit case opposite to the previous one (layout 4) where all bearings are designed free in
both directions. As expected, the induced horizontal forces were very low (the maximum
values were 43 kN in longitudinal and 85 kN in transverse direction), while on the other
hand, the obtained horizontal displacements for span 2 were very high (58,4 mm in lon-
gitudinal and 206.6 mm in transverse direction). The obtained first period of vibration of
1.61 sec. resulted in low seismic forces in the bearings for span 1. Unlike the previous
case, this layout will not have problems with thermal actions, however, the braking ac-
tion will be certainly critical, resulting in high displacements like those obtained due to
seismic actions.

4 Conclusions
From the above discussion, it is obvious that bearings, as important parts of bridge
structures, also deserve special attention and proper seismic design. The conclusions
from the investigations can be summarized as follows.
Within the scope of the detailed design of several bridge structures along Kumanovo–
Beljakovce section, a special project on design of structural bearings has been realized,
being one of the first projects realized by implementation of the Eurocodes and EN
1337 in Macedonia. The gained results and experiences from the project have empha-
sized the differences between national and Eurocodes’ design philosophies. Namely,
the structural bearings have been designed according to both codes whereat a consid-
erable difference in the adopted bearing dimensions has been obtained between the
two design solutions. From the performed design, several parameters have appeared to
be responsible for the differences, as discussed in the paper. Especially, the difference in
the computed level of seismic actions according to both codes was one of the most cru-
cial parameters. Further, in order to discuss the influence of the bearing fixation layout
on the seismic response of bridge structures, dynamic non-linear time-history analyses
have been performed for five different layouts. The fourth layout represented the stiff-
est case, while the fifth layout represented the most flexible case. Layouts 1, 2 and 3
have been investigated as more realistic options in practice, defined between the two
limit cases. The investigation has shown that the period shifting due to change of bear-
ing fixation results in different overall seismic behavior of the bridge structures and es-

SEISMIC PERFORMANCE OF STRUCTURES 883

1st Croatian Conference on Earthquake Engineering - 1CroCEE
 pecially different behavior of the bearings. In practice, this means that one should seek
an optimal layout solution that will lead to optimal balance between induced bearing
horizontal forces and horizontal displacements due to seismic actions.
This research is expected to have an impact on the decisions within the frames of the
activities for estimation of the National Annex parameters governing the problems of
seismic design of bearings. These have not yet been specified and standardized in our
country.

References
[1] EN 1990 + A1, Eurocode – Basis of Structural Design (2002, 2005)

[2] EN 1991-1-5, Eurocode 1: Actions on Structures – Part 1-5: General Actions – Thermal Actions
(2003)

[3] EN 1991-2, Eurocode 1: Actions on Structures – Part 2: Traffic Loads on Bridges (2003)

[4] EN 1992-2, Eurocode 2- Design of Concrete Structures – Concrete Bridges (2005)

[5] EN 1998-1, Eurocode 8: Design of Structures for Earthquake Resistance – Part 1: General Rules
(2004)

[6] EN 1998-2, Eurocode 8: Design of Structures for Earthquake Resistance – Part 2: Bridges (2005)

[7] MKC EN 1337-1; 2009, ICS:91.010.30, EN 1337-1:2000, IDT, Structural Bearings – Part 1: General
Design Rules (2009)

[8] MKC EN 1337-3; 2006, ICS:91.010.30, EN 1337-3:2005, IDT, Structural Bearings – Part 3:
Elastomeric Bearings (2006)

[9] MKC EN 1337-8; 2009, ICS:91.010.30, EN 1337-8:2007, IDT, Structural Bearings – Part 8: Guide
Bearings and Restraint Bearings (2009)

[10] Tower 7, An Instruction Manual for Operating with the Program, (2019), Radimpex, Beograd

[11] Regulative for Technical Normative for Design and Calculation of Engineering Structures in Seismic
Prone Areas, draft, Federal Standardization Office, (1986), Beograd

[12] Hristovski, V. (2007), Software Package FELISA/3M, Input Data Manual, IZIIS, Skopje, R. Macedonia

884 SEISMIC PERFORMANCE OF STRUCTURES

1st Croatian Conference on Earthquake Engineering - 1CroCEE