Content uploaded by Helga Papp
Author content
All content in this area was uploaded by Helga Papp on Jan 07, 2016
Content may be subject to copyright.
Ŕ Periodica Polytechnica
Civil Engineering
59(4), pp. 495–502, 2015
DOI: 10.3311/PPci.8169
Creative Commons Attribution
RESEARCH ARTICLE
Solutions of Omitting Rail Expansion
Joints in Case of Steel Railway Bridges
with Wooden Sleepers
Nándor Liegner, Gyula Kormos, Helga Papp
Received 22-04-2015, revised 19-06-2015, accepted 19-06-2015
Abstract
The Technical Specifications of D.12/H. of Hungarian State
Railways (MÁV) specifies that a continuously welded rail (CWR)
track can be constructed through a bridge without being inter-
rupted if the expansion length of the bridge is not longer than
40m. If the expansion length of a bridge is greater than 40 m,
the continuously welded rail should normally be interrupted;
a rail expansion joint has to be constructed. The goal of this
research is to provide technical solutions of track structures
on bridges so a continuously welded rail can be constructed
through the bridge from an earthwork without interruption, so
rail expansion joints can be omitted.
Keywords
rail expansion joint ·heat expansion ·rail ·steel bridge ·
wooden sleepers ·rail restraint
Nándor Liegner
Department of Highway and Railway Engineering Faculty of Civil Engineering,
Budapest University of Technology and Economics, H-1521 Budapest, P.O.B.
91, Hungary
e-mail: liegner@uvt.bme.hu
Gyula Kormos
Department of Highway and Railway Engineering Faculty of Civil Engineering,
Budapest University of Technology and Economics, H-1521 Budapest, P.O.B.
91, Hungary
e-mail: kormos@uvt.bme.hu
Helga Papp
Department of Highway and Railway Engineering Faculty of Civil Engineering,
Budapest University of Technology and Economics, H-1521 Budapest, P.O.B.
91, Hungary
e-mail: papphelg@gmail.com
1 Introduction
A finite-element (FEM) model has been developed to deter-
mine the normal, axial forces in the rail, bridge structure and the
bearing in case of a two-span-bridge with an expansion length
of 40 m, where forces occur from the change of rail temperature
and braking and acceleration of trains. Following this, the model
has been converted into bridges with 70m and 100m expansion
lengths with the purpose to find technical solutions, with their
application the resultant normal forces in the rail, bridge and the
bearing do not exceed — or exceed to a lesser extent — those
values resulting in bridges with expansion length of 40m. By
the application of these solutions, the CWR track can be con-
structed through the bridge without interruption, rail expansion
joints can be omitted.
Only the joining of CWR tracks from earthworks to steel
bridges with wooden sleepers are discussed in this research.
There are technical solutions in bridges where the continu-
ously welded rail is constructed through a bridge without inter-
ruption, and longitudinal beams of the bridge can move indepen-
dently from the rails, within certain boundaries. These solutions
are not part of this article.
2 Laboratory testings of longitudinal rail restraint of
rail fastenings
Test series have been carried out in the Laboratory of the De-
partment of Highway and Railway Engineering, Budapest Uni-
versity of Technology and Economics, in order to determine the
longitudinal stiffness and the longitudinal rail restraint of differ-
ent rail fastenings to model the interaction of the rail and bridges
precisely.
The tests were carried out according to standard EN 13146-
1:2012 [3]. The test arrangement is shown in Figs. 1 - 2.
The concrete sleeper, the rail and the fastening assembly were
fixed to a horizontal base. A tensile load at a constant rate of
10kN/min was applied to one end of the rail, while the load and
the displacement were measured. When the rail slipped in the
fastening, the load was reduced to zero rapidly and the rail dis-
placement was measured for two minutes. Without removing
or adjusting the fastening, the cycle was repeated further three
Solutions of Omitting Rail Expansion Joints in Case of Steel Railway Bridges 4952015 59 4
times with three minute intervals in the unloaded condition be-
tween each cycle.
The rail displacement was measured with inductive trans-
ducer of type Hottinger Baldwin Messtechnik (HBM) WA
20 mm, and the load was measured with force transducer of type
HBM C9B 50kN. The data acquisition unit and measuring am-
plifier was HBM Quantum MX 840, evaluation software was
Catman AP. The sampling rate frequency was 10Hz.
Fig. 1. Longitudinal rail restraint test (1)
Fig. 2. Longitudinal rail restraint test (2)
The maximum load to produce an initial elastic displacement
was determined in each cycle. The value of the first cycle was
discarded. The average of the second, third and fourth cy-
cles was calculated and considered to be the longitudinal rail
restraint. The fastening assembly is unable to take on higher
forces, the rail will slip in the fastening longitudinally.
The longitudinal stiffness of the fastening is defined as ratio
of the force producing the initial elastic displacement and the
elastic displacement.
The load – displacement diagram measured on the K(Geo)
fastening with Fe6 washer tensioned with a torque of 250mm
is illustrated in Fig. 3 as an example. In this case there was
no railpad under the railfoot. The longitudinal rail restraint is
obtained to be 20,52kN, and the longitudinal stiffness has been
found to be 40000N/mm.
The tests were carried out on K(Geo) fastening, and on Voss-
loh KS (Skl-12) and W14 fastenings. The results are summa-
rized in Table 1.
Fig. 3. Load – displacement diagram of K(GEO) fastening with Fe6 washer
3 Structure of the FEM model
The finite-element software of AxisVM 12 was used for
model. Two different types of beams are possible to be defined
in the software. One of them is the Euler-Bernoulli beam that
assumes the cross-sections are perpendicular to the longitudinal
axis of the loaded beam. The other one is the Timoshenko beam
that takes into effect the shear deformations, therefore result-
ing in a softer structure. Our model comprises two dimensional
Euler-Bernoulli beams.
The model structures consist of one rail of section 60E1 and
half of the cross-sectional area of the bridge. For interest of the
comparability of different models, each model has got the same
material and cross-sectional properties.
3.1 Bridge structure
The beam modelling the half-cross-sectional area of the
bridge are the following:
•cross-sectional area: 1000 cm2
•elasticity modulus: 210000 N/mm2
•linear heat expansion modulus: 1,20 ·10−51/°C
The static model of the bridge is illustrated in Fig. 4. A fix
support is located at the left hand-side and there are moving
supports at mid-span and at the right hand end, therefore the
expansion length of the bridge is equal to its structural length.
Fig. 4. The static model of the railway bridge
Period. Polytech. Civil Eng.496 Nándor Liegner, Gyula Kormos, Helga Papp
Tab. 1. Longitudinal rail restraint and stiffness of rail fastenings
Type of fastening Longitudinal rail restraint [kN] Longitudinal stiffness [N/mm]
K (Geo) with flat EVA railpad under the rail 26,51 51 400
K (Geo) without any railpad 20,52 40 000
KS, Skl-12 with flat EVA railpad under the rail 16,58 36 000
KS, Skl-12 without any railpad 10,47 14 000
W 14, with flat EVA railpad 11,79 28 000
Tab. 2. Maximum values of normal forces and relative displacements in case
of bridges with expansion length of D=40 m, without any rail expansion joints
(longitudinal ballast resistance in joining track sections p=5 N/mm/rail)
Structure Season K (GEO)
restraint
30,0kN
K (GEO)
restraint
20,52kN
KS Skl-12
restraint
10,47kN
Fixed bearing winter 1581 1375 701
Maximum summer -1581 -1375 -701
normal Bridge structure winter 1581 1375 701
force summer -1581 -1375 -701
[kN] CWR track winter 2009 1966 1930
summer -1761 -1720 -1684
Relative displacement of bridge winter 17,3 23,4 47,5
and rail (at sliding bearing) [mm] summer -17,3 -23,4 -47,5
3.2 Modelling CWR ballasted tracks
It has been assumed in the model that a ballasted track with
continuously welded rail (CWR) joins the bridge at its both
ends. The ballasted CWR tracks are modelled with continuously
elastically supported beams, whose properties are equal to those
of the rail section of 60E1:
•area of cross section: 7670 mm2
•elasticity modulus: 215000 N/mm2
•linear coefficient of thermal expansion: 1,15 ·10−51/°C
The longitudinal resistance of a consolidated and well main-
tained ballast can be 8 to 10 N/mm, whereas that of a newly laid
ballast can be considered to be 5 N/mm in respect of one rail. In
accordance with this, the limiting longitudinal load of the con-
tinuous support of the ballasted track has been assumed to be
9N/mm for the consolidated ballast and 5N/mm for the newly
laid ballast. The model does not take into consideration that the
longitudinal ballast resistance of the track increases under the
load of a train. It is suggested to deal with the case of the loaded
track in another article in the future.
3.3 Modelling the track – bridge interaction
The spacing between the wooden sleepers on the bridge is
0.60m, therefore the beams substituting the rail and that mod-
elling the bridge are connected with non-linear springs every
0.60m. Due to the non-linear behaviour, it is necessary to carry
out second-rank theory computations. The properties of the
springs are defined on the basis of the laboratory tests defined
in chapter 2 and their results summarized in Table 1.
Major in ref. [4] and Birk and Ruge in ref [5] also apply non-
linear elastic relationship between the displacement difference
in the track - bridge interaction and the longitudinal restoring
force.
Fig. 5. Normal force generated by braking in the rail
3.4 Load cases and combinations
The neutral temperature and the linear coefficient of thermal
expansion of the bridge and the rail are different, therefore rel-
ative displacement is generated between the rail and the bridge.
According to the Technical Specifications of D.12/H. of Hun-
garian State Railways (MÁV), the neutral temperature of the rail
is 20+8
−5°C. The temperature of the rail can reach even 60°C in
the summer due to direct sunshine, and as low as -30°C in the
winter. The neutral temperature of the steel bridge is 10°C that
can be changed by ±40°C under extreme weather conditions [6].
European Standard EN 1991-2 require that the braking effect
of the trains onto the rails be substituted by a longitudinally uni-
formly distributed load of 20kN/m per two rails that is 10kN/m
per one rail through a total length of 300m. It has a maximum
value of 6000kN on the bridge. The acceleration of the trains is
to be taken into consideration by an evenly distributed longitu-
dinal load of 33kN/m with a total value of 1000kN [7]. Of the
two effects, it is the braking that produces higher force, therefore
this is critical.
The normal forces generated in the rail by a braking effect is
illustrated in Fig. 5. The braking takes place over the distance
and in the direction indicated by the arrow [8].
Solutions of Omitting Rail Expansion Joints in Case of Steel Railway Bridges 4972015 59 4
Tab. 3. Maximum values of normal forces and relative displacements in case
of bridges with expansion length of D=40 m, without any rail expansion joints
(longitudinal ballast resistance in joining track sections p=9 N/mm/rail)
Struc-ture Season K (GEO)
restraint
30,0kN
K (GEO)
restraint
20,52kN
KS Skl-12
restraint
10,47kN
Maximum Fixed bearing winter 1219 1064 689
normal summer -1224 -1065 -689
Bridge structure winter 1440 1169 689
force summer -1441 -1170 -689
[kN] CWR track winter 1734 1626 1557
summer -1487 -1379 -1311
Relative displacement of bridge winter 12,6 23,4 47,5
and rail (at sliding bearing) [mm] summer -12,6 -23,4 -47,5
Fig. 6. Special positions of braking load
In case of critical load combination the position of maximum
values of normal forces generated by the change of temperature
and by braking should coincide.
The combination of loads comprise of the kinematic load of
change of temperature in winter, that in summer and the braking
effect over a distance of 300m. In order to determine the posi-
tion of loads generating the greatest normal force in the struc-
tures, the braking force has been moved from the position indi-
cated in Fig. 6a gradually in steps of 10m through the positions
in Figs. 6b and 6c to the position shown in Fig. 6d. Braking to
the right and to the left are mirrors of each other. Each braking
load position has been combined with kinematic load of change
of temperature both in summer and in winter.
If the rail temperature is lower than the neutral temperature,
tensile force will arise in the rail that may result in fracture of
the rail and if it is higher than the neutral temperature then com-
pressive force will be induced that may lead to buckling of the
track. The latter is more dangerous in respect of traffic safety.
4 Determination of normal forces in bridges with
expansion length of D=40m without rail expansion
joints
As it has already been mentioned in the introduction, accord-
ing to Technical Specifications of D.12/H. of MÁV, continuously
welded rail track can be joined to the bridge structure without a
rail expansion joint if the expansion length of the bridge is equal
or less than 40m, therefore the normal forces generated in the
structural elements are permitted. As a consequence, as first step
of the research we have determined the normal forces induced
in the rail, bridge structure and the bearing.
The model of the bridge and the continuously welded rail
track has been built in the way described in Chapter 3. The
computations have been carried out in the following cases:
•K(Geo) rail fastening with longitudinal rail restraint of
30,0kN,
•K(Geo) rail fastening with longitudinal rail restraint of
20,52kN,
•KS Skl-12 rail fastening with longitudinal rail restraint of
10,47kN.
The longitudinal stiffness of the fastenings is summarized in Ta-
ble 1.
The normal forces resulting from the load combinations are
summarized in Table 2 in case of a longitudinal ballast resistance
of 5N/mm/rail and in Table 3 in case of 9 N/mm/rail.
The normal internal force diagrams are illustrated in Figs. 7 -
10. They indicate the cases when the greatest normal forces are
generated in the bridge, the fixed bearing and the rail. The 300 m
long section with the uniformly distributed load of braking is in-
dicated. They arise with the application of a rail fastening with
a longitudinal rail restraint of 30,0kN and a longitudinal ballast
resistance of 5kN/m/rail. The force diagram in red colour in-
Period. Polytech. Civil Eng.498 Nándor Liegner, Gyula Kormos, Helga Papp
dicates the normal force in the bridge and that in blue colour
indicates the normal force in the rail.
It has been obtained that the lower the rail restraint is the
lower the normal forces are in the rail and the bridge.
The direction of the maximum normal force is irrelevant in
respect of the bridge and the bearing, the one with the higher
absolute value is considered to be critical.
The technical specifications do not limit the maximum value
of the longitudinal rail restraint and stiffness of the fastening. If
an EVA railpad is inserted under the railfoot in the K(Geo) fas-
tening and the nut is pulled by slightly higher torque than spec-
ified, the longitudinal rail restraint of this fastening can reach a
value of 30kN. The longitudinal stiffness is 51400kN/mm. In
this case the maximum longitudinal force in the bridge and the
fixed bearing is 1581kN both in compression and tension. The
maximum value of the tensile force in the rail is 2009kN and
that of the compressive force is 1761kN. Taking these values
into consideration and the maximum limit values of 3000kN of
braking force per one rail, Table 4 summarizes the maximum
permissible normal forces.
Fig. 7. Normal internal force diagram, when the greatest force is generated
in the fixed bearing and the bridge (D=40m)
Fig. 8. Normal internal force diagram, when the greatest force is generated
in the fixed bearing and the bridge (zoom of Fig. 7)
Fig. 9. Normal internal force diagram, when the greatest tensile force is gen-
erated in the rail (D=40 m)
5 Analysis of bridges with expansion length of greater
than D=40m without rail expansion joints
According to present regulations an expansion joint has to be
constructed between the ballasted CWR track and the bridge if
the expansion length of the bridge is greater than D=40m. As
Fig. 10. Normal internal force diagram, when the greatest compressive force
is generated in the rail (D=40 m)
Tab. 4. Maximum permissible normal forces
Structure Maximum permissible normal force
Fixed bearing 3000kN -3000kN
Bridge structure 3000kN -3000 kN
CWR track 2009kN -1761 kN
a consequence the bridge can change its length due to change
of temperature, however the longitudinal forces resulting from
braking of the trains whose maximum value is 3000kN on one
rail according to standard of Eurocode 1991-2 have to be taken
on by the fixed bearing of the bridge.
We have carried out analyses to determine the normal forces
generated in the continuously welded rail, the bridge structure
and the fixed bearing in cases of bridges with expansion length
of 70m and 100m.
5.1 Bridges with expansion length of 70m
The results of our computations carried out on bridges with
expansion length of 70m without any rail expansion joints are
summarized in Table 5 that contains the longitudinal normal
forces in the rail, the bridge structure and in the fixed bearing
and the maximum relative displacements between the bridge and
the rail, in function of the longitudinal rail restraint of the fas-
tening and the ballast resistance.
In case of application of the same rail fastening, on bridges
with longer expansion length higher normal internal forces are
generated from the same loads.
Rail fastenings with lower longitudinal rail restraint will al-
low higher relative displacements between the bridge and the
rail and will convey lower longitudinal forces from the rail onto
the bridge and vice versa. The rail restraint of Skl-12 fasten-
ing is much less than that of the K(Geo) fastening, therefore
much lower longitudinal internal forces will be generated with
its application.
By comparing the data of Tables 4 and 5, it is obtained that
the internal normal forces generated in the rail in case of a bridge
with an expansion length of 70 m and fastening assembly of Skl-
12 will not exceed the normal forces generated in the rail in case
of an expansion length of 40 m and K(Geo) fastening assembly,
even if an EVA railpad is inserted under the rail and the nut is
drawn slightly higher than specified in the assembly regulations.
In case of a bridge with an expansion length of 70m and a
fastening assembly with a longitudinal rail restraint of 13kN,
similar longitudinal normal forces are generated in the bridge
structure and the bearing as in case of 40m of expansion length
Solutions of Omitting Rail Expansion Joints in Case of Steel Railway Bridges 4992015 59 4
Tab. 5. Maximum values of normal forces and relative displacements in case of bridges with expansion length of D=70 m, without any rail expansion joints
Structure Season Geo rail restraint
of 20,52kN
Skl-12 rail
restraint of
16,58kN
Skl-12 rail
restraint of
12,56kN
Skl-12 rail
restraint of
10,47kN
longitudinal ballast resistance in joining track sections p= 5N/mm/rail
Maximum Fixed bearing winter 1881 1787 1469 1225
normal summer -1881 -1787 -1469 -1225
force Bridge structure winter 1890 1787 1469 1225
[kN] summer -1890 -1787 -1469 -1225
CWR track winter 2075 2009 1963 1947
summer -1829 -1762 -1716 -1700
Relative displacement of bridge winter 27,4 31,6 40,7 46,5
and rail (at sliding bearing) [mm] summer -27,4 -31,6 -40,7 -46,5
longitudinal ballast resistance in joining track sections p= 9N/mm/rail
Maximum Fixed bearing winter 1615 1500 1306 1159
normal summer -1619 -1502 -1306 -1159
force Bridge structure winter 1823 1653 1345 1170
[kN] summer -1823 -1654 -1345 -1170
CWR track winter 1890 1811 1746 1733
summer -1644 -1565 -1500 -1486
Relative displacement of bridge winter 23,2 26,1 29,1 30,5
and rail (at sliding bearing) [mm] summer -23,2 -26,1 -29,1 -30,5
and overdrawn K(Geo) fastening assembly. For a 70m expan-
sion length and 16kN rail restraint, the internal normal forces
in the bridge and in the bearing are higher than in case of 40m
expansion length, however they are much lower than 3000kN,
and still lower than 2000kN.
The application of K(Geo) fastening is not suggested on
bridges with an expansion length greater than 40m.
Continuously welded rail track can be constructed
through bridges with expansion length of D=70m without
rail expansion joints if the rail fastening has got a maximum
longitudinal rail restraint of 15 to 16kN, supposing that bal-
lasted CWR track is joined at both ends of the bridge. In
these cases rail expansion joints can be omitted.
5.2 Bridges with expansion length of 100m
The results of our calculations carried out on bridges with
expansion length of 100m and without any rail expansion joints
are summarized in Table 6. It can be determined that with the
application of a rail fastening with a rail restraint of 10,5kN,
the normal internal force in the rail will not exceed the value
generated in the rail on a 40 m expansion length (2009 kN). If the
rail restraint of the fastening assembly is greater than this value,
the normal internal force in the rail will be higher, especially in
case of a K(Geo) fastening.
The longitudinal internal forces in the bridge and in the bear-
ing do not exceed the limit of 3000kN. In case of a rail restraint
of 10,5kN the longitudinal internal forces are approximately
10% higher than in case of a 40m expansion length.
The maximum relative displacement between the rail and the
bridge is +/-55mm (Table 6). This is +/- 40 mm on a 40 m ex-
pansion length. This difference is negligible regarding the fa-
tigue strength of the rail clip, because the rail starts slipping
in the fastening after an initial elastic displacement of 0,5 to
1,5mm.
Based on our analysis, the continuously welded rail can
be constructed through a bridge with an expansion length
of D=100m without any rail expansion joints, if the rail
restraint is maximum of 11kN, if ballasted CWR track is
joined at both ends of the bridge. Special attention has to be
paid to correct construction of the fastening, if it has a screw or
nut it may not be overtensioned. In case of the construction of a
rail fastening with a rail restraint of greater than 11kN, a more
detailed analysis is necessary.
It can be concluded that the longitudinal rail restraint of
fastening assembly has a dominant influence on the inter-
action of the bridge and the rail in respect of the normal
internal forces.
6 Bridges with expansion length of 100m and with rail
expansion joints
The major goal of our publication, as it has already been men-
tioned at the beginning of this paper, is to provide technical so-
lutions with their application a continuously welded rail track
can be constructed through a bridge without interruption, with-
out any rail expansion joints. This case in this chapter has been
modelled to compare these results with those obtained without
expansion joints.
We have built models also for the cases where there are rail
expansion joints at both ends of the bridge. The models were
built in a similar method discussed in previous chapters. In or-
der to simulate expansion joints, non-linear springs have been
inserted in the model at the ends of the bridge, altogether 10
Period. Polytech. Civil Eng.500 Nándor Liegner, Gyula Kormos, Helga Papp
Tab. 6. Maximum values of normal forces and relative displacements in case of bridges with expansion length of D=100 m, without any rail expansion joints
Structure Season Geo rail restraint
of 20,52kN
Skl-12 rail
restraint of
16,58kN
Skl-12 rail
restraint of
10,47kN
rail restraint of
7,0kN
longitudinal ballast resistance in joining track sections p= 5N/mm/rail
Maximum Fixed bearing winter 2243 2151 1748 1169
normal summer -2249 -2151 -1748 -1169
force Bridge structure winter 2304 2182 1748 1169
[kN] summer -2304 -2182 -1748 -1169
CWR track winter 2220 2148 1977 1929
summer -1974 -1902 -1731 -1682
Relative displacement of bridge winter 33,2 38,1 54,5 71,6
and rail (at sliding bearing) [mm] summer -33,2 -38,1 -54,5 -71,6
longitudinal ballast resistance in joining track sections p= 9N/mm/rail
Maximum Fixed bearing winter 2092 1970 1602 1169
normal summer -2094 -1972 -1602 -1169
force Bridge structure winter 2291 2157 1644 1169
[kN] summer -2291 -2157 -1644 -1169
CWR track winter 2099 2015 1884 1855
summer -1853 -1768 -1637 -1608
Relative displacement of bridge winter 30,0 34,1 41,9 46,0
and rail (at sliding bearing) [mm] summer -30,0 -34,1 -41,9 -46,0
Tab. 7. Maximum values of normal forces in units of kN’s in case of bridges
with expansion length of 100 m, with expansion joints at both ends of the bridge,
longitudinal ballast resistance in joining track sections p=5 N/mm/rail
Structure Season Geo rail restraint
20,52kN Skl-12 rail restraint
10,47kN
Fixed bearing winter 876 876
summer -876 -876
Bridge structure winter -966 -925
summer 895 895
CWR track winter 1504 1504
summer -1543 -1543
of them with longitudinal stiffness of 5kN/mm and a limiting
load of 1,9kN [9]. Above a total horizontal load 19kN the
springs will slide longitudinally, they are not able to take on
higher forces.
Only rail fastening assemblies of K(Geo) with rail restraint
of 20,52kN and Skl-12 with 10,47kN have been modelled in
case of a ballast resistance of p=5 N/mm/rail. The results are
summarized in Table 7. Comparing the values of Tables 6 and
7, it can be concluded that much higher normal internal forces
are generated in the rail, bridge structure and the bearing if rail
expansion joints are omitted at both ends of the bridge. This has
to be taken into consideration during the design, also at consid-
ering the stability of the CWR track against buckling at the joint
of the bridge and the ballasted track.
7 Conclusions
Research has been carried out with the purpose to find tech-
nical solutions to construct continuously welded rail through
bridges with expansion length of greater than 40m without in-
terruption that joins ballasted CWR tracks at both ends. In these
cases rail expansion joints can be omitted. Conclusions are the
followings:
•In case of expansion length of D>40m, the normal internal
forces in the bridge structure, the bearing and the rail will be
higher than in case of expansion length of D=40m. With
increasing expansion length of bridge, the normal internal
forces will increase. This has to be taken into consideration
during the design, also at considering the stability of the CWR
track against buckling at the joint of the bridge and the bal-
lasted track.
•The normal internal forces in the bridge structure, the bearing
and the rail can be decreased by reducing the longitudinal rail
restraint of the fastening assembly. It can however result in
excessive opening of a gap in case of rail fracture in winter.
•The continuously welded rail can be constructed through a
steel bridge with an expansion length of 70 m without any rail
expansion joint if a fastening assembly with a longitudinal
rail restraint of maximum of 15 to 16kN is applied. In these
cases the normal internal forces in the rail will not exceed
those generated in case of an expansion length of 40m with
K(Geo) fastening of 30kN of rail restraint. Normal internal
Solutions of Omitting Rail Expansion Joints in Case of Steel Railway Bridges 5012015 59 4
forces in the bridge and bearing will be approx. 10% higher
than in case of 40m expansion length.
•The continuously welded rail can be constructed through a
steel bridge with an expansion length of 100m without any
rail expansion devices if a fastening assembly with a longitu-
dinal rail restraint of maximum of 11kN is applied. In these
cases the normal internal forces in the bridge and the bearing
will slightly exceed those generated in a bridge with an ex-
pansion length of 40m with K(Geo) fastening, however the
normal force in the rail will be less than those in case of 40 m
of expansion length.
•It can be concluded that the longitudinal rail restraint of fas-
tening assembly has a dominant influence on the resultant nor-
mal internal forces in the bridge and the rail. The less the rail
restraint is, the lower internal forces will be generated in the
structural elements.
•The consolidation of the ballast, that is higher ballast resis-
tance value will serve in favour of safety. In case of higher
ballast resistance less internal forces will be generated in the
rail and the bridge.
References
1MÁV Zrt. D. 12/H. Technical Specifications, Construction and maintenance
of continuously welded rail tracks; Budapest, 2009.
2MÁV D.54. Technical Specification of track construction and maintenance;
Budapest, 1986. KÖZDOK.
3EN 13146-1:2012:E European Standard, Railway applications – Track – Test
methods for fastening systems – Part 1, Determination of longitudinal rail
restraint; Brussels. ICS 93.100.
4Major Z,Special problems of interaction between railway track and bridge,
Pollack Periodica, 8(2), (2013), 97–106, DOI 10.1556/Pollack.8.2013.2.11.
5Birk C, Ruge P,Longitudinal track-structure interaction on railway bridges,
Proceedings in Applied Mathematics Mechanics, 6(1), (2006), 209–210, DOI
10.1002/pamm.200610085.
6MSZ-07-2306/2-90T Static design of railway bridges, forces acting on
bridges, 1990.
7MSZ EN 1991-2:2006 European Standard, Eurocode 1: Actions on bridges,
Part 2: Traffic loads on bridges; Brussels. ICS 91.010.30; 93.040.
8Kormos Gy,Behaviour of continuously welded rail track in mirror of finite
element calculations, (2014). (private communications).
9Kormos Gy,Construction and maintenance aspects of expansion devices
with internal friction with special attention to opening tables, In: Interna-
tional Scientific Conference of Construction; Csíksomlyó, 2002, pp. 142-152.
ÉPKO.
Period. Polytech. Civil Eng.502 Nándor Liegner, Gyula Kormos, Helga Papp
