BS 114 VER 2 IR RSI Ver 2 Final

RDSO GUIDELINES FOR CARRYING OUT
RAIL-STRUCTURE INTERACTION STUDIES

ON INDIAN RAILWAYS
BS 114 Version 2
Provisions with Commentary
August 2016
RESEARCH DESIGNS AND STANDARDS ORGANISATION,

LUCKNOW
FOREWORD to First Version
Long Welded Rails on bridges is a dream for track maintenance engineers. Elimination of free joints on
bridges help in following ways:
1. Maintenance is reduced due to lesser vibrations and impact, leading to reduced wear and tear.
2. Noise on bridge is reduced.
3. Safety is enhanced as the settlement on free joints is eliminated, leading to reduced twist in track.
4. The track is less prone to sabotage as there are no fish-bolts that can be opened by unscrupulous
elements.
LWR manual of Indian Railways have allowed to lay LWR on girders with bearings only with rail-free
fastenings so far. This has created problems in field as bridges with even single span ballasted deck girders had to
be isolated by providing SEJs on either approach to isolate the LWR. Efforts were made first by HAG committee
comprising of Director IRICEN, PCEs SE Rly and SC Rly, ED/Track-I/RDSO and ED/B & S/RDSO and then by B & S dte
of RDSO to study the phenomenon of Rail-Structure Interaction. The contribution of Sh A K Goel, the then
Director/IRICEN, Sh Ajay Goyal, then Senior Professor/Bridges-I/IRICEN and all members of HAG committee in this
field is very important. This document is ultimately a compilation of work done on the subject till now. This
document borrows heavily from the “RDSO Guidelines for Carrying out Rail-Structure Interaction studies for metro
systems” Ver 2, issued in January 2015 and the contributors to that report are also acknowledged.
Work done at RDSO by Sh V B Sood, Director/B & S/SB-II and his team comprising Sh B P Singh, DD/SB-II,
Sh Uma Shanker JE/Design, Sh Nilesh Kumar, SSRE and Sh Prakash Kumar Ranjan, JRE/Design who have carried out
RSI studies for over 35 bridges on Indian railways and have done hundreds of trials with different configurations
with an aim to understand the phenomenon and to lay down these guidelines is also commendable.
Feedback/ suggestions/ questions on issues regarding these guidelines may kindly be directed to
directorsteel2@gmail.com.
A K Dadarya
Executive Director/B & S
RDSO
Lucknow,
November 2015
FOREWORD to Version 2
RDSO has made great strides in the field of conducting RSI ever since this BS report was published.
Different types of bridges with girders like PSC, composite etc, bearings like elastomeric, sliding, POT/PTFE etc and
foundations like open, pile, well etc have been done. Till date, 51 cases have been examined and 46 bridges were
found suitable for continuing LWR.
The problems have been noticed in bridges with pile foundations and elastomeric bearings or, in one case,
where the bridge layout was very badly planned. Some errors have been found in provisions of earlier version of
this report, mainly in working out stiffness of cracked RCC piers and well foundations. One issue has been found
that the thermal effects and vehicle effects need to be worked out separately if the bridge supports are
asymmetric such as in case of POT/PTFE bearings or if the spans are dis-similar. Some minor typographical errors
too have been found. These corrections have been made in this report along with the example.
Feedback/ suggestions/ questions on issues regarding these guidelines may kindly be directed to
directorsteel2@gmail.com.
A K Dadarya
Executive Director/B & S
RDSO
Lucknow,
August 2016.
Page 2/57
RDSO Guidelines for carrying Out Rail Structure Interaction studies on Indian Railways. BS 114
CONTENTS
1.0 INTRODUCTION............................................................................................................................ 5
1.1 GENERAL...................................................................................................................................... 5
1.2 ADAPTATION / MODIFICATION TO EXISTING RULES...................................................................... 5
1.3 RELEVANT CODES & STANDARDS ................................................................................................. 5
1.4 IMPORTANT POLICY DOCUMENTS ................................................................................................ 6
2.0 SCOPE .......................................................................................................................................... 6
3.0 GENERAL CONCEPTS .................................................................................................................... 7
3.1 INTERACTION PHENOMENON ...................................................................................................... 7
3.2 PARAMETERS AFFECTING RSI ....................................................................................................... 9
3.2.1 Expansion Length ......................................................................................................................... 9
3.2.2 Span Length ............................................................................................................................... 10
3.2.3 Bending Stiffness of Deck ........................................................................................................... 10
3.2.4 Deck Height & Rotation Distance................................................................................................ 10
3.2.5 Support Stiffness ........................................................................................................................ 11
3.2.6 Track Stiffness ............................................................................................................................ 11
3.2.7 Sectional Properties of Rails ....................................................................................................... 13
3.2.8 Temperature variations .............................................................................................................. 13
3.3 VERIFICATION OF TRACK AND BRIDGE CONFIGURATION THROUGH RSI COMPUTATIONS ........... 13
3.3.1 Additional Stresses in Rails ......................................................................................................... 14
3.3.2 Displacements of Bridge Elements.............................................................................................. 14
3.3.2.1 Longitudinal displacement of Deck due to movement of substructure ...................................... 14
3.3.2.2 Longitudinal displacement of Deck due to rotation of deck ......................................................... 15
3.3.2.3 Relative displacement between rail and deck or between rail and embankment ......................... 15
3.3.2.4 Vertical displacement of upper surface of deck with respect to adjoining structure..................... 15
4.0 STEPS IN CHECKING A STRUCTURE UNDER RSI ............................................................................ 16
4.1 CHOOSE REPRESENTATIVE STRETCH FOR RSI STUDY IN LONG VIADUCTS .................................... 16
4.2 VERTICAL TRAIN LOADS.............................................................................................................. 17
4.3 BRAKING AND TRACTIVE LOADS ................................................................................................. 18
4.4 LOAD FACTORS .......................................................................................................................... 19
4.5 STIFFNESS PARAMETERS OF STRUCTURE .................................................................................... 19
4.5.1 Pre-dimensioning of structure .................................................................................................... 19
4.5.2 Determining stiffness of sub structure ........................................................................................ 19
4.5.3 Determining stiffness of bearings. .............................................................................................. 19
4.6 ANALYSIS METHODOLOGY ......................................................................................................... 20
Page 3/57
4.6.1 Analysis methodology using graphs given in UIC 774-3R ............................................................ 20

4.6.2 Choosing computer program for carrying out RSI ....................................................................... 20
4.6.3 Analysis methodology using FEM ............................................................................................... 20
4.6.3.1 Recommendations for FEM Modeling ......................................................................................... 20
4.6.3.2 Recommendations for FEM Analysis ........................................................................................... 21
4.7 ANALYSIS FOR FRACTURE OR REMOVAL OF RAIL(S) FOR MAINTENANCE .................................... 23
5.0 SPECIAL CASES IN RSI ................................................................................................................. 23
5.1 HORIZONTAL CURVATURE ANALYSIS .......................................................................................... 23
5.1.1 Thermal Analysis ........................................................................................................................ 24
5.1.2 Braking / Tractive Analysis ......................................................................................................... 24
5.1.3 Allowable additional stresses in rails .......................................................................................... 25
5.2 RSI FOR TURNOUTS ON VIADUCT ............................................................................................... 25
5.2.1 Introduction ............................................................................................................................... 25
5.2.2 FEM Modeling of Turnouts ......................................................................................................... 25
6.0 CONTROLLING RSI EFFECTS ........................................................................................................ 26
6.1 MODIFICATION OF BEARING ARRANGEMENT............................................................................. 26
6.2 PROVIDING SWITCH EXPANSION JOINT (SEJ) .............................................................................. 27
6.3 PROVIDING LOW TOE LOAD FASTENERS ..................................................................................... 27
Appendix A: Important Documents which covers the RSI studies ............................................................ 28
Appendix B: Data Required for Carrying out RSI Analysis ........................................................................ 42
Appendix C: Working Out Stiffness of Sub-structures of Different types .................................................. 44
Appendix D: Working Out Stiffness of Different types of Bearings .......................................................... 49
Appendix E: Explanatory example outlining Methodology to be Followed for RSI studies Using FEM
Program ........................................................................................................................................... 51
Page 4/57
PROVISIONS COMMENTARY
1.0 INTRODUCTION
1.1 GENERAL C1.1

The purpose of these guidelines is to define the
methodology for carrying out the Rail Structure
Interaction (RSI) to be considered on bridges of
Indian Railways. The guidelines provide a basis for
carrying out RSI studies and thus to work out the
forces induced in rails and bridge components due to
the interaction effects and to assess if the
arrangement will be safe under the interaction
effects.
Paras 2.8.1(d) and 2.8.2.4.3 of Bridge Rules which UIC Leaflet 774-3R 2001 is the basic code on
were reintroduced/ modified in 2014 (Correction Slip which subsequent codes have evolved. The
no 45 which introduced this para is placed at UIC leaflet is based on earlier research on the
Appendix A). The clause 2.8.2.4.3 of Bridge Rules related phenomena. The leaflet describes
refers to UIC 774-3R for carrying out RSI till the forces methodology to be adopted for carrying out
due to continuation of LWR/CWR on bridges in Indian interaction studies, based upon numerical
conditions are finalized. This document explains methods that idealize the behavior of all the
methodology to be adopted to use UIC 774-3R for elements and actions involved for the
Indian Railways specific scenario. Also, certain computation of stresses and displacements.
aspects of RSI on which UIC 774-3R is silent have Specific clauses of other codes, wherever
been explained with references to provisions used, have been mentioned in the
available in other codes like European codes, Spanish commentary.
National codes or Korean codes etc. Data required for
RSI and example of RSI carried out by RDSO with
explanatory notes has been given at the end.
1.2 ADAPTATION / MODIFICATION TO EXISTING C1.2

RULES
These guidelines are meant to supplement and These specifications cover only one aspect,
explain (but never replace) provisions of Bridge namely RSI, and do not specify the loads to be
Rules and/or other codes/ specifications already in used for design, and other checks required for
vogue for design/ detailing of bridges. This document stability/ safety of the structure. There are
gives some simplified methods to model the other closely linked phenomena like dynamic
structure/ structural behavior but it shall remain the analysis and the vehicle – track – bridge
responsibility of the designer to ensure that the interaction etc which are not covered in these
behavior of actual structure is as per the simplified guidelines. The basic principles of structural
assumptions given in these guidelines. analysis are also required to be adopted to
capture the structure behavior realistically.
1.3 RELEVANT CODES & STANDARDS C1.3

 Rules specifying the loads for design of super-
structure and sub-structure of bridges and for
assessment of the strength of existing bridges
Page 5/57
(Bridge Rules)
 UIC 774-3R October 2001: Track/Bridge
Interaction – Recommendations for calculations.
 IRS Code Of Practice For Plain, Reinforced &
Prestressed Concrete For General Bridge
Construction (IRS Concrete Bridge Code, Reprint
Sept 2014)
 UIC 776 2R: Design requirements for rail bridges
based on interaction phenomena between train,
track and bridge.
 Korean Design Standard: Railway Design Manual
(Volume Track)
 Eurocode 1: Actions on structures — Part 2:
Traffic loads on bridges (EN 1991-2 – 2003)
 TCRP report 155, 2012: Track Design Handbook
for Light Rail Transit, second volume,
1.4 IMPORTANT POLICY DOCUMENTS C1.4
The following documents are important for RSI These policy documents are placed at
studies on bridges: Annexure A. The current policy position
 A & C Slip No 45 to Bridge Rules. regarding LWR on bridges for Indian Railways
 Policy letter no CT/IM/LWR/Part Dated is:
 All ballasted deck bridges shall be
19/25.03.2014 issued by ED/Track-I/RDSO
designed for LWR effects.
regarding providing LWR on bridges on trial basis
for bridges upto 110 m length (individual spans  Parameters for carrying out these studies
upto 24.4 m with fixed-free bearings and 45.7 for bridges on straight have been specified.
with elastomeric bearings) with the approval of For others, in the absence of these
PCE. parameters, the RSI studies cannot be
carried out hence LWR cannot be provided
 Letter no CBS/Project/LWR Dated 05.09.2014
except by using other codes for which
issued by ED/B & S/RDSO which covers the
appropriate approval from railway Board
software to be procured and how the
shall be obtained.
infrastructure shall be created for starting the RSI
 The LWR can be laid on trial for bridges
analysis work on Indian Railways.
upto 110 m length and individual spans
upto 24.4 m with fixed-free bearings and
45.7 with elastomeric bearings with the
approval of PCE.
 For other bridges also, new bridges have to
be checked for RSI effects but provision of
LWR is not yet allowed.
2.0 SCOPE C2.0
Page 6/57
These guidelines explain the interaction The Rail-Structure Interaction phenomenon
phenomenon, parameters affecting RSI, provide originates from the fact that the bridge is
guidance on choosing representative stretches for moveable under thermal and live load effects
conducting RSI, methodology to be adopted for whereas the rails as part of LWR/CWR are not
carrying out RSI, special cases in RSI, use of computer free to move. This difference in movements
programs for carrying out RSI and options available induces stresses in rails/ bridge components
for modification in track if the RSI results indicate which need to be studied.
excessive stresses/ deformations.
These guidelines cover steel/concrete bridges with Different bearings/ bridge forms need to be
simply supported or continuous spans, on straight modelled such as to reflect their actual
alignment, both level as well as those on gradient, behaviour under the RSI phenomenon. In
having any type of bearings on Indian Railways. structures like cable stayed bridges, the
However, these guidelines do not cover the bridges flexibility of deck and non-linear response due
on curves or bridges with long/special spans such as to presence of cables supporting the deck
cable stayed bridges, Bow-string arch girders etc require more complex models that capture all
where the typical structural behaviour of spans phenomenon accurately. RDSO has no
affects RSI phenomenon requiring specialized studies experience presently with these models and
to be carried out or where span arrangement induces so these are beyond the scope of these
excessive movement in track which is beyond the guidelines. Similarly, for large movements in
capacity of a typical Switch Expansion Joint (SEJ) to track, specialized solutions are required to be
accommodate. worked out. These are site-specific solutions,
and track experts are required to study the
site conditions and design these
arrangements.
3.0 GENERAL CONCEPTS C 3.0

General concepts describing the RSI phenomenon The long term phenomena like effect of dead
covering the effect of train loads (vertical as well as loads, deformation of deck under creep/
longitudinal) and the effect of thermal changes are shrinkage etc are considered to be dissipated
given in this section. during various track maintenance operations
and hence are not considered while carrying
out the RSI studies.
3.1 INTERACTION PHENOMENON C 3.1

In jointed track, the analysis of effect of various In jointed track, lots of gaps are available
loads on rails and on bridge is carried out separately. normally in joints and the track/ bridge can
However, this type of analysis is not appropriate move independently to some extent. For rail-
when the continuously welded rails (which restrict free fastenings also, the track and bridge
the free movements) are laid on the structure move independently, so there is no interaction
because then the track-structure interaction shows between these. Hence, for these, there is no
non-negligible effects. requirement for carrying out RSI studies.
The presence of deck under the tracks induces The purpose of RSI analysis is to examine
extra stresses in the rails due to interaction these extra stresses in rails due to the actions
phenomenon and this affects stresses in bridge of temperature change, braking / traction of
components also. The extra stresses in the rail are rolling stock combined with the vertical
induced by thermal expansion/ contraction of bending caused due to live loads. These
bridge deck(s)/tracks, deflection of sub-structure stresses are required to be kept within
Page 7/57
under tractive/ braking forces from the trains and allowable limits so that the track is safe under
the end rotations caused by vertical bending under tension as well as compression, and the bridge
vertical train loads. elements are to be proportioned to take all
The magnitude of these extra stresses in the CWR the loads. If the RSI study indicates extra
mainly depends on the stiffness of various elements stresses in rails beyond permissible limits,
of bridge, resistance offered by the track structure to these can be brought within limits by altering
deformation and the boundary condition of rails (i.e. either the stiffness of the structure and/or the
whether these are continuous or have expansion fixing arrangement of the rails to the bridge
joint(s) in between). The RSI describes the effects, structure and/ or introducing expansion joints.
under various loads, of structural collaboration of
rails and bridge by means of their connection
elements.
Distribution of rail stresses along length of an LWR
Distribution of rail stresses affected by presence of bridge

due to change in length of bridge deck under thermal/ live load
changes
The difference between LWR/CWR on ground vis- The interaction phenomenon can be summed
à-vis LWR/CWR on bridge is that a bridge has lesser up as the interplay of stiffnesses of different
stiffness which results in its deformation under components namely track, girders and
various loads/ thermal effects. The track is supported supports (Bearings, sub-structure and
on the bridge and has to respond to these foundations). The component which is stiffer
movements. But the rails, being continuous, are not will attract more stresses. If the sub-structure
free to move and resist these movements, which is flexible, it will move under loads and the
induces loads in them. These loads cause the track rails will be subjected to higher stresses, which
as a whole to move, which relieves part of the loads, can be unsafe for the train operations. Quite
which are transferred back to the structure. The often the purpose of RSI is to ensure that the
final deformations/ stresses in track and viaduct bridge is stiff enough such that the track is
depend on this interaction, which is basically safe.
governed by the stiffness of track and that of the
bridge. This interaction between track and the bridge
structure is studied as RSI effect.
The RSI phenomenon, as explained above, is non-
linear which can only be solved by an iterative
procedure to get a solution that satisfies all boundary
conditions. There can be no formula to be directly
Page 8/57
used to determine the stresses or deformations etc.
The results can be obtained by two ways: we can use
the charts given in the UIC 774-3R or we can model
the bridge, track and approaches and find out the
results using Finite Element Method based computer
programs which try to solve the non-linear problem
through convergence of results through iterations.
The relative stiffness’s of different elements like
track, deck, sub structure and bearings play
important role in determining the results. The
designer has to change the stiffness’s/ arrangement
to optimize the performance and costs.
3.2 PARAMETERS AFFECTING RSI Before start of RSI study, data on bridge and
LWR shall be available at hand. Complete list
of data required for the same is given in
Appendix B.
3.2.1 Expansion Length C 3.2.1

Expansion length is defined as the distance What shall the expansion length be for
between the thermal center point and the opposite different cases? This question needs to be
end of deck. In simple terms, this means it is the answered if graphs as per UIC 774-3R are
length over which structure is allowed to expand/ being used for RSI computations. This shall be
contract by the supports. Free/ moveable bearings worked out as follows:
allow expansion/ contraction to take place whereas  For simply supported spans with one end
fixed bearings do not allow the same. Expansion having fixed bearing and other end having
length depends on the type of support configuration free bearing, expansion length is the
adopted in a structure. Expansion length is defined
distance between the fixed bearings.
and indicated on different type of structures in para
1.1.3.1 of UIC 774-3R 2001.  For continuous spans having a fixed
bearing somewhere in the middle, there
are two expansion lengths, one on each
side. For succession of decks, the
expansion length at a joint is equal to the
sum of expansion lengths of nearest two
spans.
 If the structure does not have fixed
bearings and arrangement has neoprene
or sliding bearings, the expansion length
has to be worked out between thermal
center-point (i.e. point which will not move
under thermal effects) of the deck and its
extremities.
Para 1.3.1 (Figure 6) of UIC 774-3R 2001 may
be referred to determine expansion length for
Page 9/57
different types of support configurations
commonly adopted in bridges.
3.2.2 Span Length C 3.2.2

The vertical train loads cause rotation at ends of The center to center distance (or effective
decks. Since the rails are not fitted at the neutral axis span) between supports is the span length for
of the deck, the length of fibers at the track level this purpose. As against this, normally the
changes under these loads. This leads to longitudinal expansion length for simply supported spans is
displacement, and thus, stresses in tracks which overall length of girder. To simplify the
depends on the magnitude of load as well as span computations, sometimes, analysis is done
length. The span length is measured as center to considering overall length for both.
center of bearings on supports. Alternately, the span can be modeled and the
actual rotation at ends under live loads can be
determined.
Schematic indication of longitudinal displacement of deck fibres

at rail level under vertical loads (Springs indicate restraints due
to track and supports)
3.2.3 Bending Stiffness of Deck C 3.2.3

The bending stiffness of each deck is required to The stiffness of deck is required to be
calculate the vertical deformation effects on computed to get the deck rotation under
structure under the vertical loads of rolling stock. loads. As explained above, this deformation
leads to change in deck length and tries to
change the length of track which induces the
interaction effect. In case the cross-section
varies along span length, exact computation
may be done. Alternately, to simplify
computations, if the cross-section at center
has lesser section modulus than that at ends,
the cross-section at center may be adopted. If
section at center is higher section modulus
than that at ends, an average section may be
considered without too much error.
3.2.4 Deck Height & Rotation Distance C 3.2.4

The change in length of the deck fibers The longitudinal displacement of deck is
supporting rails is affected by the distance between described in clause 1.3.3 of UIC 774-3R 2001.
deck level supporting track and the neutral axis of the It is evident that if the track is supported
deck which is termed as rotational distance and the nearer to neutral axis of girder, this effect will
total distance from bearing to the top of deck, called be lesser, and vice versa.
deck height.
Page 10/57
3.2.5 Support Stiffness C 3.2.5
A major source of interaction phenomenon is the The effects of longitudinal loads on the
stiffness (or, more correctly, flexibility) of bridge substructure is described in clause 1.3.2.2 of
supports under longitudinal actions (braking/tractive UIC 774-3R 2001. The stiffness of the sub
loads and temperature variations). The longitudinal structure, Kpier = H/Σδi where δi is the
stiffness of sub structure depends on stiffness of deflection of sub structure due to:
individual components viz, foundation, sub structure 1. Displacement due to elastic
and bearing. Stiffness of different parts should be deformation of sub structure.
combined to get total stiffness Ktotal as follows: 2. Displacement due to rotation of
1 1 1 support.
= +
K K K 3. Displacement due to longitudinal
where Kpier is stiffness of each sub-structure movement of foundation.
(pier/abutment) and Kbearing is stiffness of bearing. All the above displacements have to be
Kpier has further components as described in figure worked out at the top of bearing level. While
below: computing stiffness, for sustained
temperature loading analysis, long-term
Young’s modulus shall be used, whereas for
the short-term effects of braking and tractive
loading, instantaneous modulus shall be used.
Young’s modulus should be determined as
per IRS-CBC:2014 and the Young’s modulus
for long term effect is normally taken as half
the Young’s modulus for short term.
a) b) c)
Longitudinal displacement of deck due to a) Elastic deformation
of sub structure, b) Rotation of foundation and c) Longitudinal
movement of foundation.
3.2.6 Track Stiffness C 3.2.6

The track stiffness is a measure of resistance In RSI, the movement of bridge under
offered by the track to longitudinal movement. different loads is considered. The stresses
Stresses are induced in track as a response to all the induced in the track due to these movements
movements in deck/ bridge via interaction through and actual movement of track depends on the
track stiffness. The track stiffness is dependent on interaction effect which is dependent on the
multiple factors like: track stiffness curve. Track stiffness is a very
 Type and condition of track structure. important parameter in RSI and can be
 Load on Track. manipulated by changing the track
 Maintenance condition of Track. characteristics to allow movement at critical
locations thus ensuring that extra stresses on
account of RSI are within limits.
Deformation of track is a bilinear curve as The curves given in para are idealized form of
suggested in clause 1.2.1 of UIC 774-3R. Typical track actual behaviour of track, as indicated below:
stiffness curve is as follows:
Page 11/57
40 kN/m
Longitudinal resistance of track is provided by

fasteners in ballast less track and by ballast in
ballasted track. In idealized curve, resistance is
proportional to the displacement of rail relative to
the supporting deck until a relative displacement of
u0 is reached, which corresponds to elastic limit.
Beyond this point, the fasteners/ballast cannot The elastic limit is 0.5 mm for ballastless
resist any further load and perfectly plastic track (and for frozen ballast) and 2 mm for
behaviour of track is assumed: the resistance force is ballasted track (unfrozen) as per clauses 1.2.1
constant while the movement continues (plastic and 1.2.2 of UIC 774-3R 2001.
shear resistance). The elastic limit is different for
frozen and unfrozen ballasts. Analogous to frictional
behavior, plastic shear resistance of the ballast is
higher when an additional vertical load i.e. train
load is applied to the track.
As per clause 2.8.2.4.3 (d) of Bridge Rules, the The track resistance values for ballastless track
track resistance in ballasted deck bridges for track have not been laid down in IRS Bridge Rules,
structure minimum 52 kg 90 UTS rails and PRC so if such track is used and RSI computations
sleepers at density 1540 nos/KM with elastic are to be done, these values shall be taken
fastenings shall be: from track design engineers.
 25 kN per meter of unloaded track For guidance, limiting plastic track resistance
 50 kN per meter of loaded track given in the clause 1.2.2 of UIC 774-3R 2001
for unballasted deck can be referred
 40 kN/m for unballasted track
(unloaded).
 60 kN/m for unballasted track
(loaded).
The frozen ballast also acts quite like
ballastless track.
A note for capturing this behaviour by computer The implementation of connector elements
programs: The behavior of track under longitudinal representing ballast/ fasteners in the
Page 12/57
forces is quite complex. When the direction of interaction phenomenon causes many other
displacement changes, the ballast behavior becomes complications, including that activation and
elastic again, but the relative displacement from deactivation of elements is a function of the
sliding is not recovered. Any computer program to presence of train load. These cannot be
be used for carrying out RSI studies shall be capable realized in many engineering FEM programs
of capturing this behaviour realistically as per actual commercially available. This aspect needs to
behaviour in field. be examined before an FEM program is
chosen/approved for carrying out RSI analysis.
3.2.7 Sectional Properties of Rails C 3.2.7

The cross sectional area of the rails in track,
Young’s modulus of the rail-steel and other
parameters of rails are required to work out stresses
in rails.
3.2.8 Temperature variations C 3.2.8

The temperature changes induce change in For deck, reference temperature is the
length of deck and/or rails. The decks with bearings temperature at which the rails are fastened to
permit expansion/ contraction with change in the deck. Fastening of rails to deck is done
temperature. The temperature variation is measured either at the time of initial laying or during
with respect to reference temperature. If the subsequent distressing/ other maintenance
deck/rail length changes, the interaction operations. During maintenance, rails shall be
phenomenon described above kicks in. fixed to deck at near about the reference
As per the clause 2.8.2.4.3(f), no checks are temperature for which the RSI computations
required for forces/effects due to continuation of were carried out for the viaduct.
LWR/CWR on integral bridges like Arches, RCC Boxes Temperature variation shall be the difference
and slabs without bearing which are not free to between actual installation temperature of
expand/ contract, and are quite stiff. LWR and the maximum/ minimum
The temperature variations are determined temperature at the location. In the higher of
separately for deck and rails. The rail temperature the two differences, 50C shall be added. This
variation is important only if Switch Expansion Joint 50C will give flexibility of ±50C for fastening
is provided on the bridge or within 100 m on rails in future.
approaches. Otherwise, the LWR does not change in For rails, reference temperature is the stress-
length during thermal variations and the reference free temperature of the rail, which is
temperature of rail/ its temperature variation is not determined on the basis of rail temperature
required to be considered. zone given in LWR manual or actual
temperature records. The stress-free
temperature shall be as decided by track
engineers. It shall be ensured that the stress-
free temperature for LWR on bridge is not
altered without consulting with the bridge
design engineer regarding RSI effects.
3.3 VERIFICATION OF TRACK AND BRIDGE C 3.3

CONFIGURATION THROUGH RSI
COMPUTATIONS
Parameters to be verified during RSI studies are The combined response of track and structure,
Page 13/57
the following: as studied by RSI, can have unfavorable effects
on either the bridge structure or the rails.
Design/ layout/ dimensions of the bridge or
the track configuration may have to be
changed to keep these unfavorable effects
within limits.
3.3.1 Additional Stresses in Rails C 3.3.1

The additional rail stresses due to the various The additional stresses allowed for the
actions should be limited to ensure that no rail interaction effect are within the total rail
fracture takes place due to overstressing and the stresses. The term ‘additional stresses’ has
track structure does not buckle. The additional been used as this component is being verified
stresses permitted for the RSI phenomenon shall be separately to ensure proper functioning of
laid down by track design engineers looking at the rail track under RSI phenomenon. The margin for
stress computations done for the rolling stock, LWR additional stresses is derived from the total
and other effects. stresses by limiting the curvature on bridges
Bridge Rules Clause 2.8.2.4.3 (b) specifies that on while considering the RSI effects, and from the
tangent tracks, the additional stresses in rail as per fact that the track is better maintained on
RSI computations shall not exceed values as given bridge as compared to normal ground.
below: The additional allowable stresses are lower for
Rail Maximum Maximum compression as compared to tension to keep
Section additional additional additional factor of safety towards possibility
stresses in stresses in
compression tension
of buckling. For ballastless track, the
60 Kg 90 possibility of buckling is not there, which shall
UTS Rail
60 N/mm2 75 N/mm2 be considered while determining the
52 Kg 90
50 N/mm2 60 N/mm2 allowable stresses.
UTS Rail For Indian Railways, the additional stresses
for curved track have not been specified in
IRS Bridge Rules. So, if any new bridge is being
planned with curvature, the allowable
additional stresses shall be specified by track
design engineers/RDSO.
3.3.2 Displacements of Bridge Elements C 3.3.2

Too much displacement in the bridge structure The ballast packing can get loose or the entire
can cause instability in track structure. Therefore, as track assembly can get unstable if too much
per clause 2.8.2.4.3 (c) of Bridge Rules, the limits laid displacement is there in bridge elements.
down in UIC 774-3R 2001 on the displacements of These checks also control passenger comfort
bridge elements shall be satisfied. and, indirectly, the additional stresses in rails.
3.3.2.1 Longitudinal displacement of Deck due to C 3.3.2.1

movement of substructure
Due to tractive/braking loads, the displacement The limits are as per clause 1.5.3 of UIC 774-
of rails and deck needs to be limited. The absolute 3R 2001. Excessive movements of decks can
horizontal displacement of the deck, worked out for result in deconsolidation of ballast /
tractive/braking loads through RSI studies, shall not deformation in the track plinth due to which
exceed: proper performance of track cannot be
Page 14/57
- 5 mm in case CWR runs through one or both ensured. This limit also indirectly controls the
ends of the bridge. rail stresses.
- 30 mm in case of bridge with jointed track/ If the deck movement worked out as per RSI
comes to be more than permissible, the
expansion devices.
options are either to discontinue LWR/CWR
on the bridge by providing SEJs on either
approach (and provide jointed track on
bridge), or to provide switch expansion joint at
one or both ends of the girders. As per clause
2.8.2.4.4 of Bridge Rules, the option (Of not
providing LWR on bridge or providing LWR
with one or more SEJ(s)) shall be as per
approval of Principal Chief Engineer of the
zonal railway.
3.3.2.2 Longitudinal displacement of Deck due to C 3.3.2.2

rotation of deck
Due to vertical loads, the longitudinal The limit is as per clause 1.5.4 of UIC 774-3R
displacement of upper surface of deck end from 2001. This check is given to ensure stability of
embankment or between tops of two consecutive ballast under the various deformations. To
decks shall not exceed 8mm. work out the longitudinal displacement of
deck due to rotation of deck, the deck rotation
at ends shall be worked out and same shall be
multiplied by the distance between neutral
axis of girder and rail. For open web/U-girders,
if neutral axis is above the location of deck,
this value can be negative. For plate/
composite girders, this value is always
positive.
3.3.2.3 Relative displacement between rail and deck C 3.3.2.3

or between rail and embankment
The relative displacement between the rail and The limit is as per clause 1.5.4 of UIC 774-3R
deck or between rail and embankment under tractive 2001. This relative displacement determines
/ braking forces shall not exceed 4mm. the stability of track structure. Para 2.1.2.1 of
UIC 774-3R mentions that “….relative rail
displacement is not needed for verifying the
effects of temperature variation and always
lies within the limit value for the effects due to
braking as long as absolute displacement of
the deck stays within the limit value of 5 mm.”
3.3.2.4 Vertical displacement of upper surface of C 3.3.2.4

deck with respect to adjoining structure
The vertical displacement of upper surface of Vertical displacement of the deck is a source
deck in relation to the adjacent structural elements of discomfort for the passengers and this has
also needs to be checked. UIC 774-3R 2001 does not to be limited based on the speed of train
Page 15/57
specify any limits for this and leaves the same to the operations.
concerned authorities to decide. These limits have As per clause 6.5.4.5.2 of EN 1991 – 2003 part
not specified in Bridge Rules and this check may not 2, the maximum vertical displacement shall be
be performed normally for bridges on Indian 3 mm for maximum speeds upto 160 km/h,
Railways. However, where higher speed trains are and 2 mm for maximum speeds greater than
moving, the values as per EN codes may be used. 160 km/h.
4.0 STEPS IN CHECKING A STRUCTURE UNDER RSI C 4.0

4.1 CHOOSE REPRESENTATIVE STRETCH FOR RSI C 4.1
STUDY IN LONG VIADUCTS
Track on very long stretches of viaducts pose the Normally bridges on Indian Railways are not
issue of choosing stretches on which RSI study is to be that long and entire bridge can be taken for
done. RSI studies. But long viaducts can be
Representative stretches for RSI studies shall be planned in hilly or built up areas, so these
identified along the viaduct by studying the following : paras are included.
 Change in Pier Stiffness: This occurs at locations of The results of RSI are dependent on the
o Integral Spans. stiffness of different elements and it is quite
o Change in Bearing Arrangement. clear that any stiff element will attract more
o Extended Pier Caps. force. Choosing representative section is
o Change in soil conditions. very important to get the worst scenario
 Change in Span Stiffness: This occurs at locations possible. The changes in the lateral stiffness
of of span supporting elements and bending
o Change in span length. stiffness of the deck have a major impact on
o Change in girder type. the vertical deck deformation and lateral
o Steel Bridges deck movement.
o Composite Girders.
 Change in Span Arrangement: including:
o Stations on viaduct.
o Cross Over Locations on viaduct.
The RSI studies for a viaduct shall be done for The RSI evaluation for viaducts can be done
stretches. A stretch is defined as viaduct from station for the stretches and in case the stresses/
to station or from its start to next station or complete movements are above the laid down limits,
viaduct. The evaluation shall be done as follows: additional analysis shall be undertaken to
 Small stretches of viaducts upto 20 spans can be identify the trends of stresses/
fully modeled and studied under RSI. deformations and possible remedial
 For longer stretches, RSI studies shall be conducted actions. Based on these analyses, remedial
on representative stretches of the viaduct such as action to alter the viaduct design or track
to cover the worst combinations of long spans/ arrangement shall be determined and RSI
flexible sub structure/ curved alignment/ stations/ studies carried out again.
any other special features of alignment. If a clear
cut representative section is not identifiable, or
there are multiple critical locations, multiple
representative sections shall be chosen for
analysis.
 In addition to complete stretch evaluation, any
special spans (special from span length, sub
Page 16/57
structure height, span type etc considerations)
should be evaluated in a standalone analysis. 100 m track length (whether on viaduct or
 The stretches taken up for RSI study must include embankment) on either side of the viaduct/
minimum 100 m track (whether on viaduct or stretch/span of interest has been specified
embankment) beyond the stretch/span/location since this length is required to anchor the
of interest. rails and to dissipate the longitudinal forces.
As per para 1.7.3 of UIC 774-3R “The model
shall also include a part of the track on the
adjacent embankments over at least
100m.”
4.2 VERTICAL TRAIN LOADS C 4.2

As per clause 2.8.2.4.3 (a), the vertical train load The loading standards specified in IRS
shall be as per design loading or the heaviest trains Bridge Rules consider multiple trains for
actually running on the route, depending on the type each loading, which will require RSI studies
of analysis being done. Train loads shall be enhanced to be done multiple times for different
by the appropriate Coefficient of Dynamic Augment trains. This is likely to be very tedious and
(CDA) specified in the Bridge Rules. The placement of time consuming. The complexity of problem
load shall be done such as to create maximum rotation can be gauged from the following:
at the ends. - EUDL for smaller loaded lengths has more
To avoid running the multiple trains specified in intensity of load, which goes down as the
Bridge Rules for each loading, as an alternate method, loaded length increases.
Uniformly distributed loads may be used as a - Impact factor changes with the loaded
substitute to the actual loads, with the following length.
magnitude: To simplify the computations, the approach
Loading Vertical Load TE intensity (kN/m) BF intensity with uniformly distributed trains has been
↓ intensity with (kN/m)
impact (kN/m) devised. RDSO has done correlation studies
Length of 0-12 12- 0-12 12-40 40- 0- 12-∞ of load specified with the actual maximum
UDL  ∞ ∞ 12
DFC load for spans >10 m, and upto 125 m. The
233.72 136 47 22 0 36 16
Loading best correlation was found if we divide the
HM
Loading
219.04 135 47 25 0 32 14 load into two parts (First 12 m separate and
25 T rest of the load separate). The maximum
Loading 182.95 104 47 22 0 34 13
2008 load is captured with an error less than 5%
MBG
Loading
182.91 91 47 18 0 34 13 on lower side. The error on higher side is
Notes: also reasonable for most spans, going upto
1. For other loadings like BGML/RBG/CC+6+2/CC+8+2 20% for few intermediate span lengths only.
loadings, the UDL for MBG 1987 may be used to be on Extra margin may be available in many
conservative side. cases, so it has been recommended that if a
2. The vertical loads in tables above include full impact bridge fails with uniformly distributed load,
load also. If some bridge has restriction on speed, the it may be rechecked with the actual axle
impact load may be reduced according to IRS Bridge loads.
Note: It is generally not necessary in RSI to
Rules provisions.
capture effects for less than 10 m span as slabs
3. If the bridge fails marginally in RSI with EUDL and other type of structures used for lesser
approach, actual train loads may be used for exact spans are not required to be studied for RSI
analysis. effects.
4. For bridges with uniform spans, the UDL shall be Another benefit of using the UDLs is that
Page 17/57
applied at one end of span and moved by 1/10th of the loads need not be placed at close
span in each increment. If dis-similar spans are there, intervals to capture the peak response. It
the increment shall be worked out such that has been found that placing load at
beginning of load is placed at each end of span (i.e. on beginning of span and increment by 1/10th
each pier and abutment). of span captures the peak behavior. It is
5. If the bridge is symmetric, UDL loads given above important to capture the full span loaded
may be run in one direction only. But if the bridge is case in addition to part span loaded cases.
asymmetric in span configuration or in stiffness of the
bridge elements or in cases of fixed-free bearings such
as POT-PTFE bearings, the UDL loads shall be run in
both the directions.
6. To determine the individual effects viz, thermal and This phenomenon of non-linearity in case of
live load effects, multiple models may have to be asymmetric bridges is there due to
analyzed. For bridges with uniform spans and difference in the location of peak stress due
expansion type bearings at both ends, only one model to thermal effects and that due to live
with thermal effect and live load run from one side loads.
shall suffice. By deducting the thermal stresses from
peak stresses, we can get the effect due to live loads.
However, in asymmetric bridges, such linear deduction
gives erroneous results. In such cases, to work out the
correct peak effects, the live load shall be run with
longitudinal loads in both the directions. To get the live
load effect alone, thermal variation shall be given zero
value. (Total three cases need to be analyzed)
4.3 BRAKING AND TRACTIVE LOADS C 4.3

The braking and tractive (acceleration) forces from The braking forces are applied along the
vehicles are longitudinal forces applied parallel to the direction of train movement and the
path on top of rails, uniformly distributed along the tractive forces are applied reverse to the
train length. To avoid running the multiple trains direction of movement of train. Normal
specified in Bridge Rules for each loading, as an train operations in double line are in
alternate method, Uniformly distributed loads given in opposite directions and the braking forces
tables in para 4.2 above may be used as a substitute to from one track are in the direction of
the actual loads. tractive forces from the other track and
These loads shall be applied concurrently with their effect is additive. However in yards
vertical loads such as to create the most adverse effect and other locations, train movements might
on the structure. For girders at gradients, the live load be occurring in same direction and in this
has a component which is applied as longitudinal load case, simultaneous braking (or tractive,
on the bearings/ sub structure. This load shall be whichever are more critical) forces on both
applied along with the tractive/braking loads, as per tracks shall be considered. The EUDL for
the direction of movement of trains. longitudinal loads shall be considered for
For multiple tracks supported on same girder, the the same lengths as for the vertical loads
tractive and braking loads shall be applied as per specified in para 4.2 above.
normal traffic operations with appropriate mode i.e. Clause 1.4.3 of UIC 774-3R provides that
braking or traction such as to produce worst effect on load from two tracks only need to be
the substructure. For more than two tracks, only two considered. Since the longitudinal loads are
tracks shall be considered loaded when carrying out not always applied at the full level by all the
Page 18/57
RSI analysis. trains, this is reasonable. The same clause
If there are multiple tracks supported on different provides that tractive load on one track and
girders, the sub-structure stiffness may be divided braking load on other shall be considered.
appropriately to reflect the share of each track, and However, if regular operation conditions
analysis may be done for individual tracks. are not like this, the actual loads for these
conditions shall be applied.
4.4 LOAD FACTORS C 4.4

The RSI case is an SLS check, the load factors may This also matches with the load factors
be taken from load combination 3/4 of IRS Concrete given in clause 1.5.1 of UIC 774-3R.
Bridge Code. The load factors for the vertical,
longitudinal and thermal loads shall be taken as 1 for
simply supported as well as continuous spans.
4.5 STIFFNESS PARAMETERS OF STRUCTURE C 4.5

4.5.1 Pre-dimensioning of structure C 4.5.1
To start with, the structure has to be given certain The pre-dimensioning allows the designer
dimensions. These can be assumed through the to assume structural stiffness and run an
experience of the designer, or from other similar RSI before the actual design is taken up.
structures already constructed in the past or guidance After getting the idea of stresses/
can be taken from the pre-dimensioning method displacements for the assumed stiffness,
specified in clause 1.6.1 of UIC 774-3R 2001. the designer can optimize the design and
run RSI again to verify if the results are
acceptable. The procedure is iterative till
the desired level of optimization is
achieved.
4.5.2 Determining stiffness of sub structure C 4.5.2

The stiffness of sub-structure has to be determined Soil behavior under different conditions is
using the principles of structural analysis. The quite complex and working out soil
deflection of foundation mainly depends on soil stiffness/ soil spring stiffness is quite a
stiffness. If computer program is used, soil has to be difficult task and requires understanding of
modeled as springs and for this soil spring stiffness the engineering properties of soils in
needs to be worked out. A suggested methodology for subgrade and their behaviour under loads.
computing support stiffness for different type of
structures is given in Appendix C.
4.5.3 Determining stiffness of bearings. C 4.5.3

The bearings can be fixed/free type or elastic The friction in bearings shall be assumed
bearings. The fixed bearings can be considered as rigid, realistically. Para 2.1.3.5 of UIC 774-3R
permitting no movement and the free end can be states that “The effects of friction on rail
modeled as free, neglecting the friction. A suggested stresses and displacements are always
methodology for considering bearing stiffness for favourable especially when the support
different type of bearings is given in Appendix D. stiffness is low, so that ignoring friction is in
general conservative for safety.”
Accordingly, the roller/ PTFE ends may be
considered to be “free”.
Page 19/57
4.6 ANALYSIS METHODOLOGY C 4.6

4.6.1 Analysis methodology using graphs given in UIC C 4.6.1
774-3R
Annexure A and B of UIC 774-3R 2001 have graphs The graphs are applicable for single span/
which have been plotted for single 60 Kg track bridges single deck only. For succession of spans/
with fixed bearing at one end having single span less decks, computer program has been
than 110 m. These can be modified for multiple track, recommended by UIC 774-3R in para 3.2
different rail section, different temperature variations, even though simplified rules for analysis of
single deck with multiple spans etc. bridges with succession of decks have been
given in para 3.3. These rules are applicable
for succession of decks subject to certain
conditions. It may be noted, however, that
the results obtained using these rules are
generally conservative. With easy
availability of computer programs, it is not
recommended to use graphs for design.
These can, however, be used for initial
dimensioning of the bridge elements.
4.6.2 Choosing computer program for carrying out C 4.6.2

RSI
RSI studies can be done using computer program or The use of an FEM based computer
can be done using graphs given in UIC 774-3R. Due to program for numerical simulation of RSI is
several limitations of the graphs, computer programs allowed as per para 3.4 of UIC 774-3R.
are generally used. Any computer program which has Validation of software with the test cases
the capability to model the actual complex non-linear given in UIC 774-3R, or better still, against
behaviour of the bridge and track elements can be other such software also is required before
used. The computer program shall, however, be starting to rely on the results given by a
validated before being permitted for use. The particular software. The validation of
validation shall be done using the test cases given in software is covered in para 1.7.1 of UIC
Appendix D of UIC 774-3R. A computer program shall 774-3R.
be considered validated when the error on the single
effects as well as on overall effect is less than 10% with
respect to corresponding type of analysis (sum of
effects or global effect). Larger tolerances, upto 20%,
can be accepted if error is on safe side.
4.6.3 Analysis methodology using FEM C 4.6.3

This section describes the approach(es) that can be
followed to obtain prudent results in a numerical
simulation for RSI analysis.
4.6.3.1 Recommendations for FEM Modeling C 4.6.3.1

The study of the track-deck interaction involves the These recommendations are given in para
Page 20/57
implementation of numerical models that captures the 1.7.3 of UIC 774-3R. There are more
actual configuration and properties of the structure recommendations in UIC 774-3R but all of
and the track. The model should be able to adequately these have not been reproduced for the
represent the structural behaviour under different sake of brevity. Actual leaflet may be
loads. Few important aspects of model are as follows: referred to study the complete
a) Normally line modeling is done as we are not recommendations.
interested in detailed stresses. However, if
complete model is prepared for design of elements
like girders, sub structure etc, then the same may
be used for carrying out RSI also.
b) In a model, the elements corresponding to the rails
and deck should be located at the level of
respective centers of gravity. Likewise, the
connections corresponding to support devices
should be placed at the level of their centers of
Elements of a typical model
rotation. This will capture the bending effects
properly.
c) The longitudinal behaviour of the track-deck Capturing the non-linear behaviour of
connection shall be modelled as a bi-linear spring connection between rail elements and deck
which can capture load/displacement relation elements is the most important and
similar to that illustrated in clause 3.2.6 above. involved part of modeling for RSI and
deserves close attention from the design
Separate springs shall be used for loaded and
engineer.
unloaded elements.
d) In some cases, it is possible to replace the
mentioned elements by a connection of equivalent
stiffness to that of the foundation/column/support
group.
e) The maximum element length shall not exceed 2 m.
f) An example of RSI study done by RDSO using
Equivalent model of girder
software LUSAS Bridge Plus explaining all steps is
placed at Appendix E.
4.6.3.2 Recommendations for FEM Analysis C 4.6.3.2
Page 21/57
The analysis using a computer assisted modeling Both type of analyses are allowed as per
can be achieved by two methodologies as follows: para 1.7.3 of UIC 774-3R. The choice of
analysis option is largely dependent on the
 Simplified Analysis: A simplified analysis calls for situation being evaluated.
running thermal and live load actions individually In case of simply supported spans,
and then arithmetically combining them using simplified analysis will provide reasonable
factors. results. For optimization of design and in
 Complete Analysis: A complete analysis calls for case of special spans such as arch bridges,
applying the thermal loads and then, on the cable stayed bridges and truss bridges etc,
deformed stiffnesses of bilinear springs, running use of complete analysis will be required.
an additional live load analysis.
In the simplified analysis, first step is application of
thermal loading. The longitudinal resistance of ballast
is taken from Unloaded stiffness curve and is limited
by the Limit of resistance of unloaded track.
Separately, train loading is applied and analyzed. In
this case, longitudinal resistance of ballast is taken
from Loaded stiffness curve and is limited by the Limit
of resistance of loaded track.
The sub-structure/ foundation behaviour is short
term in either case. However, the rubber bearings
have different stiffnesses for thermal loads which are
slow acting loads and the live loads. To get the exact
results, separate stiffnesses can be used. However, the
sub-structure forces as well as track stresses are
lesser with thermal if the bearing stiffness is lesser.
So, only one analysis can be performed with bearing
stiffness for rapid acting loads. In case of marginal Difference in approach of the two types of
cases, exact analysis can be performed analysis is illustrated graphically as below:
The results are then combined by superposing the
results of train load case on the results from thermal
load case. The error arises in this case because there is
an apparent increase in the resistance of the ballast
due to ignoring the resistance already mobilized by the
track for the thermal loading while considering the
train loading. This results in an assumption of greater
yielding load for track than the actual curve. Results
of this approach give higher stresses in the rails and
slightly lower substructure deflections and reactions.
In complete analysis, first step is application of
thermal loading similar to simplified analysis using the
Unloaded stiffness curve. In second step, however,
train loading is applied on the results obtained from
the first step. In this case, longitudinal resistance of
ballast is taken from Loaded stiffness curve and is
limited by the Limit of resistance of loaded track. In
this case, there is no overestimation of the track
Page 22/57
resistance.
4.7 ANALYSIS FOR FRACTURE OR REMOVAL OF RAIL(S) C 4.7

FOR MAINTENANCE
In ballastless track laid on bridges, analysis of the Most of the track on bridges on Indian Railways
scenarios of fracture on bridge or replacement of rails is yet ballasted and this phenomenon is not
are also important. The effects to be considered important. The gap at fracture in ballasted track
include: is governed by the track resistance and the
a) Gap created in rail in the event of fracture. effect of bridge is not going to be substantial in
b) Stress accumulated in sub-structure after fracture: this case. The stresses transferred to
Proper repairs to rail/weld fractures and proper substructure due to fracture of rail are also
limited by the ability of ballast to transfer the
procedure for removing rails for
load and the effect is dissipated over time
maintenance/renewals using rail tensors to under movement of track.
minimize the stresses locked up in the sub-
structure due to track maintenance is important in
For further guidance on this aspect, “RDSO
this case. guidelines for carrying out Rail-Structure
Interaction studies on metro systems” may be
Therefore, following proper track maintenance referred.
guidelines is very important for LWR on ballastless
bridges.
5.0 SPECIAL CASES IN RSI C 5.0

Following special cases in RSI need more involved These are conditions which occur in typical
studies: railway systems but are not covered by UIC 774-
3R. The parameters for these studies have also
not laid down in IRS Bridge Rules and currently
these studies, if required to be done, will have
to be done using other codes only for which
appropriate approval from RDSO/Railway
Board might be required.
5.1 HORIZONTAL CURVATURE ANALYSIS C 5.1

Due to the horizontal alignment of track on curves, The magnitude of the radial force is a function of
axial forces in the rail and superstructure have an rail temperature, rail size, curve radius, and
Page 23/57
outward component, resulting in radial forces on longitudinal fastener restraint.
bearings and sub-structure. The track structure
interaction analysis in case of horizontal curvature,
consequently, is more involved. For such cases, the
analysis for thermal case and tractive/braking loads
has to be carried out separately.
Following forces are recommended to be
considered in the two cases:
5.1.1 Thermal Analysis C 5.1.1 The radial interaction of the rails in
This analysis shall consider the following effects:
curved portion both for the thermal based
o Temperature Gradient analysis and braking analysis have a very
o Tangential Expansion different response owing to the radial
redistribution of the stresses / forces in the track
The rail forces due to the temperature can be and plinths.
predicted mathematically as follows:
Radial Force Per Length of Rail

EαΔT KL
= +
R n
Where, E = Modulus of Elasticity of Rail Section
α = Coefficent of Thermal Expansion
ΔT = Temperature Gradient
A = Cross − Sectional Area of Rails
K = Fastner Slip value divided by spacing
L = Radial Length
n = number of tracks
R = Radius of Rail Curve
5.1.2 Braking / Tractive Analysis C 5.1.2

This analysis shall consider the following effects:
o Braking / Tractive forces
o Nosing Force on Rail
o Lurching (Vertical Bending)
The effect of tractive/ braking forces has to be

studied through modeling duly considering the effects
of curvature.
The additional stresses as per RSI studies shall be
compared with the permissible additional
compressive/tensile stresses specified by the track
design engineers looking at the curvature, other track
Page 24/57
features and forces on track etc.
5.1.3 Allowable additional stresses in rails C 5.1.3 For information of the reader, allowable
The allowable additional stresses in rails for curved additional stresses in rails for curved alignment
track cannot be the same as those for straight track. have been specified in the Korean Design
These have to be separately studied and specified by Standard: Railway Design Manual (Volume
the track design engineers. No values for allowable Track), enumerated below:
stresses for curved track have been specified for For Ballasted Track: To allow for the lower
Indian Railways so far and hence it is not possible to stability of track on curved alignment which is
carry out the RSI studies for curved ballasted track as subject to lateral loads from trains:
yet. The permissible additional Compressive stresses
on account of RSI shall not exceed:
For R≥1500 : 72 N/mm2
For 1500>R≥700 : 58 N/mm2
For 700>R≥600 : 54 N/mm2
For 600>R≥300 : 27 N/mm2
The permissible additional Tensile stresses on
account of RSI shall not exceed: 92 N/mm2
For Ballastless Track: Since the load is taken by
fasteners, which can be designed for the load
actually coming and there is no problem of
stability, the permissible additional Tensile as
well as Compressive stresses on account of RSI
shall not exceed: 92 N/mm2. However, the
fasteners in this case need to be checked for
additional stresses.
5.2 RSI FOR TURNOUTS ON VIADUCT C 5.2

5.2.1 Introduction C 5.2.1
The presence of turnout in track affects the The designs of turnouts having such anti-creep
distribution of stresses in rails in RSI studies as the arrangement are not yet available for Indian
stiffness turnout structure is much more as compared Railways and the consequently CWR are not
with the normal track. When CWR (continuous welded passed through the turnouts on Indian Railways.
rail) is continued through a turnout on viaduct, RSI
effects can cause movements/thermal stresses which
may cause damage to anti-creep arrangement
between the straight tongue rail and stock rail.
5.2.2 FEM Modeling of Turnouts C 5.2.2

The connecting element system of stock rail/lead The curvature of the stock rail / lead rails can be
rail should be modelled with spring to model the ignored owing to the close spacing of fasteners/
interaction behavior. The track resistance for the sleepers. The close fastener spacing provides
turnout portion shall be modelled as bi-linear curve enough radial restraint to prevent any
similar to the normal track, with appropriate values. instantaneous buckling. The heel joint is
recommended to be modeled using spring with
a linear stiffness curve depicting its longitudinal
resistance.
Page 25/57
A Typical idealized turnout
Modeling of turnout to capture its behaviour
6.0 CONTROLLING RSI EFFECTS C 6.0

The control of RSI effects viz, stresses in rails and/or
deflection of bridge components beyond the limits laid
down in UIC 774-3R is the next question which arises after
the RSI analysis is completed.
The obvious option available is to redesign the bridge Making elements stiffer might not always be
elements to make the girders and/or sub-structure stiffer. good from other considerations like seismic
However, if this is not required from other structural loads. This balance also has to be stuck by the
reasons, making structure stiffer only for RSI effects might bridge designer.
be uneconomical. RSI effects can also be controlled by
adopting any of the following measures individually or in
combination. The decision in this regard shall be taken on
techno-economic considerations, which shall be a joint
decision of track and viaduct design engineers.
6.1 MODIFICATION OF BEARING ARRANGEMENT C 6.1

If additional rail stresses due to RSI exceed the limits, Changing bearing configuration is a much
changing the expansion length can be an option to reduce cheaper option as compared to making the
these. By choosing different locations of fixed bearing in structure stiffer. This is an important
case of continuous spans, the expansion length can be parameter for optimization of design. Using
changed. Changing the bearing type, their stiffnesses and fixed-free bearings in place of sliding/
their locations is another option which helps in controlling neoprene type bearings can help increase the
bridge deflections as well as track stresses. stiffness at minimal cost. Also, on a less stiff
pier, the free bearings can be provided so that
the deflections and stresses are controlled.
Page 26/57
6.2 PROVIDING SWITCH EXPANSION JOINT (SEJ) C 6.2

SEJs are devices provided at the ends of LWR/CWR Generally speaking, bridges with expansion
which permit longitudinal movement of rails and at the lengths of the order of 100m may be
same time maintain correct guidance/ support to the designed without resorting to rail expansion
wheels. Allowing LWR to move will reduce the stresses in devices. Expansion lengths of the order of
rails and can be a solution in locations with longer spans/ 300m to 400m will very probably necessitate
taller piers where the rail-stresses are beyond permissible at least one rail expansion device. Expansion
limits. Due to the SEJ(s), the horizontal deck forces are not lengths greater that this may necessitate at
transferred to approaches but to the fixed bearings, least two expansion devices or different track
alleviating the effects on the rail. These also provide relief in arrangement to cater to the large movements
the desired stiffness of the sub-structure as the allowable of deck end. While deciding the location of
movement of decks for locations where SEJ is provided is SEJ(s), it shall be ensured that these are not
also 30 mm as against 5 mm where LWR/CWR is continued adversely affected by bending effects in the
through. rail due to the close proximity of end of bridge
deck etc.
The decision of providing SEJ in track shall be taken SEJs are generally undesirable from the point
jointly by the bridge designers and the track maintenance of view of track maintenance. These shall be
engineers. The decision shall be taken on techno-economic provided only where unavoidable, and after
considerations. On Indian Railways, this decision shall be consultation with the track design engineers.
taken be approved by Principal Chief Engineer of the zonal
railway as per clause 2.8.2.4.4 of Bridge Rules.
6.3 PROVIDING LOW TOE LOAD FASTENERS C 6.3
Another method to modify the behaviour of LWR on The locations where low toe load fasteners
bridge is to change the behaviour of track under loads. Low are to be provided shall be clearly identifiable
toe load fasteners are special fasteners having reduced slip in field and appropriate maintenance
resistance on rails. These alter the track stiffness curve, thus instructions shall be issued for ensuring that
ensuring more movement of rails. This can relieve stresses these are not disturbed/ replaced by improper
in rails. However, adequate care needs to be exercised that fasteners during maintenance activities.
the reduction in clamping force on rail does not jeopardize
the stability of rail.
These are required to be provided, for example, in long Such long spans may also require more
spans where the thermal stresses built up in the rails or movement of rails than can be accommodated
rotation of deck might induce large stresses, especially near in the conventional SEJs. For this, special
the ends of spans. To remedy this situation, the low toe arrangements which can permit large
load fasteners may be provided in small stretches near the movement in track need to be designed and
supports. The toe load required and lengths upto which provided.
these need to be provided shall be designed taking different
field scenarios of thermal and train loads into account.
-0-0-0-0-
Page 27/57
Appendix A: Important Documents which covers the RSI studies

A & C Slip No 45 to IRS Bridge Rules 1964:
Page 28/57
Page 29/57
ED/Track-I/RDSO’s letter no CT/IM/LWR/Part Dated 19/25.03.2014:
Page 30/57
Page 31/57
Page 32/57
Page 33/57
Page 34/57
Page 35/57
Page 36/57
Page 37/57
Page 38/57
Page 39/57
ED/B & S/RDSO’s letter no CBS/Project/LWR Dated 05.09.2014:
Page 40/57
Page 41/57
Appendix B: Data Required for Carrying out RSI Analysis
1. Bridge Details:
a Bridge Number.
b Location of Bridge
c Section
d Between Stations
2. Span details: Complete drawings shall be available, having the following details
a No. of spans in the Bridge.
b Span Lengths: Clear span, effective Span, overall length of each of the spans.
c Type of Girder: I-type (single I, Double I , four I) , BOX, U - type etc.
d Location of Rails: The location of rails w.r.t. girder.
e C/C of Girder leaves: Center to center distance between girder leaves, no of girder leaves.
f Details of Material: Grade of concrete M45 , M35 , M30 , M25 etc, Grade of Steel etc.
g Properties of Material: Young's Modulus, Poisson’s Ratio ,Coefficient Of thermal expansion etc.
h Details of Different Cross-Sections: The cross-sections of girders at different locations including
height of different members, depth of neutral axis,
Note:
i. In RSI analysis, considering overall length of span in computations will be slightly
conservative. (Actually overall length shall be used for thermal effects and effective span for
bending of deck).
ii. For complete analysis, varying cross-section of girders at different locations can be
considered. For simplification, the cross-section at the middle only can be taken if the cross-
section in center is less than that the ends or an average of the cross-sections at canter/
ends if the cross-section at center is higher than that at ends may be taken.
3. Loading Details
RSI analysis shall be done for either standard loadings like 25T Loading 2008, MBG, RBG & DFC Loading
Or actual loads for which the section/bridge is fit (as per clause 2.8.2.4.3(a) of Bridge Rules) such as GC
Loading, CC+8+2 etc. EUDL charts shall be available which give the bending loads for different spans.
4. Track Details
a Type of track: Ballasted or non-ballasted
b Rail section: 60kg or 52kg etc.
c Rail Material properties: Young's Modulus, poisson’s Ratio, Coeff. Of thermal expansion.
d Track details: Height of sleepers, Depth of ballast cushion etc.
e Fastener Details: Sleeper density, type of fasteners.
f Location of SEJ, if any, within 100 m of either abutment.
g Curvature and gradient in track, if any.
5. Temperature Details
a Maximum rail temperature
b Minimum Rail Temperature
c Stress-Free temperature of LWR (Shall be maintained within ±50C, else RSI studies are required
to be done again).
Page 42/57
d Maximum annual deck temperature

e Minimum annual deck Temperature
f Temperature of deck at the time of laying LWR/CWR (±50C variation shall be allowed later on,
else RSI studies are required again)
6. Bearings Details: Complete drawings shall be available, having the following details
a Type of bearing: Elastomeric, Sliding, POT-PTFE Bearing etc.
b No. of bearings per span: No. of bearing provided in each end of girders etc.
c Details of Sliding Bearings or POT-PTFE bearings: Coefficient of Friction considered in design,
vertical reaction on each bearing in Dead Load + Superimposed dead load condition
d Details of Elastomeric Bearings: Dimensions of bearing( L x B x h where L- length across the
track, B- width along the track, h- total height of bearing), Thickness of elementary layers of
neoprene, cover at top/ sides, Shear modulus of rubber, etc.
7. Sub-structure Details: Complete drawings shall be available, having the following details
a Type of Piers and abutments
b Material of construction: Grade of concrete M45 , M35 , M30 , M25 etc, steel grade etc and
masonry properties.
c Properties of Material: Young's Modulus, Poisson’s Ratio, Coeff. of thermal expansion etc.
d Dimensional details: Heights and length/ width of different cross-sections including pier cap/ pile
cap/ well cap etc.
e Reinforcement Details: Type of reinforcement, spacing of bars etc.
8. Foundation And Soil Details: Complete drawings and soil bore-log shall be available, having the
following details
a Type of Foundation: Open, pile, well etc
b Details of Open Foundation: Heights and length/ width of different cross-sections.
c Details of Pile Foundation: No. of piles per abutment/pier, Diameter of piles, Length of piles,
details of pile like concrete grade, reinforcement details etc.
d Details of Well Foundation: Diameter of well, Length of well.
e Soil details: In open foundations, bearing capacity of soil is required. In pile foundation, the
complete bore-log details including the type of soil, depth of different soil layers and
characteristics of soil in different layers etc are required. For well foundations, normally soil
characteristics are not required.
9. Flood Details:
a Rail Level:
b Bottom of Girder:
c High Flood Level:
d Bed level: .
e Scour Level:
f Level at which rock/ non-erodible strata is there.
-0-0-0-0-
Page 43/57
Appendix C: Working Out Stiffness of Sub-structures of Different types

Stiffness is the ratio of longitudinal load applied and deflection of the sub-structure at the
bearing level, units being kN/mm or t/mm. The different components that support a girder are: a)
Foundation, b) Sub-structure (pier/abutment) and c) Bearings. By algebraically combining deflections of
individual elements, we can get the total deflection at the bearing level (This is not strictly correct, as
the secondary effects are neglected, but the error associated is acceptable for the normal bridges with
upto, say, 30m height from top of bearing to scour level). In this appendix, the stiffness computations of
two components, viz, foundations and sub-structure are explained. The stiffness computations for
bearings are explained in the next appendix.
1. Stiffness of Sub-structure (Pier/abutment): The portion of structure between bearing and the
top of foundation is referred to as sub-structure here. For pile foundations, the top of
foundation is at the top of pile cap while that for well foundation is at the top of well cap. For
open foundations, there is no clear cut demarcation between the sub-structure and foundation.
In this clause, the interaction of the foundation with the soil is considered as stiffness of
foundation and all other effects are considered as stiffness of sub-structure. The effects on sub-
structure have been described in para 3.2.5 above. For modelling, the structure may be
considered fixed at the pile cap level or well cap level or at the base of foundation level and
other effects can be added algebraically. Alternately, the complete sub-structure, foundations
(including piles/ wells) and soil (as appropriate springs) can be modeled so that we can get the
combined effect of all effects.
The sub-structure deflects longitudinally under the bending effect of longitudinal loads. The
bending moment increases from top to bottom as the lever arm for bending increases, and the
deflection effects are also more for the lower portions of sub-structure. There are two usual types of
sub-structure encountered:
a) Gravity structures constructed with mass concrete or brick/stone masonary etc which
usually don’t develop any tension, or permit very little tension.
b) Reinforced concrete structures which can develop tension and reinforcement steel is
provided to take care of the same.
For non-cracked RCC or gravity structures, the deflection is easy to work out and requires the
bending moments and section modulus at different heights to be worked out.
The sub-structures are generally having non-uniform (tapered or stepped or both) sections, and
the deflection can worked out by modeling the sub-structure in some structural analysis software
which can model the tapered and stepped sections.
For cracked RCC structures, the problem is a bit complicated. The cracked depth of an RCC structure
depends on deformation of the structure. But since the deformation of structure depends on the
sectional modulus, which depends on the depth of cracking, the problem is an iterative one. The
deflection can be worked out by modifying the effective E for the RCC structure by multiplying E with
diminishing ratio of (Ieff/Igr) for arriving at effective value of E for cracked RCC structure. As given in
ANNEX C- 2.1 of IS: 456-2000, Ieff can be worked out as follows:
= ; but Ir ≤ Ieff ≤ Igr
.
Where,
Ieff = Effective moment of inertia of cracked RCC section,
Page 44/57
Igr = Gross moment of inertia, in mm4,

Ir = Moment of inertia of cracked section=bx3/3+mAst (d-x) 2, in mm4,
∗
= = Cracked moment of resistance, in N.mm,
f cr = Modulus of rupture of concrete,
yt = Distance from centroidal axis of cross section, neglecting the reinforcement, to extreme fibre
in tension,
M = Maximum moment in the section, in N.mm,
Z = Lever arm = d - x/3, in mm,
X = Balance depth of neutral axis, solved by duly equating moment in tension and compression,
in mm,
d = Effective depth of the section, in mm,
bw = Breadth of the web,and
b = Breadth of compression face.
2. Deflection of Open Foundations: The structures with open foundations deflect in two ways
under the effect of longitudinal loads. The foundation as a whole rotates and the foundation
translates under the effect of longitudinal load, as shown below:
a) b)
Longitudinal displacement of deck due to a) bending of foundation and b) longitudinal movement of
foundation.
The translation of the foundation under the longitudinal loads as mentioned in para 3.2.5 figure
(c) is not appreciable under normal loads and this effect can be neglected without error in
computations.
2.1 Computation of Deflection of open foundation due to rotation of foundation:
The foundation can rotate under the longitudinal loads if the foundation is resting on soil. The
soil in this case behaves like a spring. The soil can be considered to act like a spring as it is
compressed by vertical load and gap will not appear on the side rotating upwards if the rotation is
small (which is the case in railway bridges). The value of spring stiffness to represent the soil is quite
difficult and no definitive values are available. The designer shall take the spring stiffness values
carefully as the soil behavior under different loading conditions is different. The problem is
compounded by the fact that on railway bridges, the soil investigation is not done to determine this
value. The angle through which the foundation will rotate is given by the formula (see step 9 and fig
8.6 (Page 90) in the chapter 8: Distribution of Externally-Applied and Self-induced Horizontal Forces
among Bridge Supports in Straight-Decks, in the book “Concrete Bridge Practice Analysis, Design and
economics” by Dr V K Raina, Third Edition, First Reprint November 2009, published by Shroff
Publishers & Distributors Pvt Ltd) :
Page 45/57
12
tan ∅ =
Where M = Bending moment at the base of open foundation, = 1.00 * h if we consider 1 kN
(or 1 t) load applied at a distance of h meters (distance between top of bearing and
bottom of open foundation);
a = plan dimension of open foundation normal to bridge;
b = plan dimension of open foundation parallel to bridge; and
c = Coefficient of subgrade reaction of soil
Design tip: The open foundations on Indian Railways are mostly allowed on rocks. In such cases, the
spring to be considered shall be very stiff. Even these values are not available. For such cases, RDSO
is using the values of spring stiffness given in the table no 3-2 at page no 409 in the Chapter-
“Special Footings And Beams On Elastic Foundations”, of book “Foundation Analysis And Design”,
by Joseph E. Bowles. For soft rock, the spring stiffness value given in the table for dense sand i.e.
128000 kN/mm is adopted presuming that the actual spring stiffness value will be higher for
rocks. For hard rock, this phenomenon can be considered negligible.
Care to be exercised: The values given above are not related to the soil properties like cohesion and
angle of internal friction etc. These values shall not be taken in case the open foundation is resting
on soil. For these cases, spring stiffness shall be taken from soil investigation carried out to work out
the same. Literature of geotechnical engineering may be referred on methods for working out spring
stiffness of soil.
3. Deflection of Pile Foundations: The pile foundations deflect under longitudinal loads structures
through two actions, as figuratively shown below:
Deflection of pile under longitudinal loads; Axial Deformation of piles causing rotation of pile cap
under bending loads
The combined effect of the two effects is quite complicated and looks something like this:
Page 46/57
However, for simplicity sake, without too much error, the two deflections can be worked out
independently and added up to get the required total deflection. These deflections can be worked
out individually as follows:
3.1 Bending of piles in longitudinal direction: The piles bend under longitudinal loads. If the pile is
socketed in rock, the deflection can take place about this fixed point. In most cases, piles are not resting
on rock, or there is considerable layer of soil between the rock level and scour level, and the bending of
piles takes place about a ‘point of fixity’ which develops by considering the soil as springs which
compress on one side and extend on other side under longitudinal loads, as shown below:
The pile below the point of fixity does not bend or deflect under loads. This point of fixity is
below the point of maximum scour. The point of fixity shall be worked out as per paras C-4.1 and C-4.2
of Annexure C of IS 2911-2010 Part 1, Section 2. The following aspects may be seen:
i. The values for sand and normally loaded clays are very close to each other. The clays may
not be considered pre-loaded for most railway bridges in absence of detailed soil
properties being available.
ii. Railway bridges have group of piles with stiff pile cap on top, hence the piles shall be
considered to be fixed headed.
iii. The point of fixity shall be worked out as per para C-4.1 of the annexure mentioned above.
iv. The total deflection due to lateral deflection of piles shall then be worked out as per type
of soil between scour level and the fixity level.
3.2 Rotation of pile cap due to elastic deformation of piles under differential axial loads: Under
bending effects, the piles are subjected to compressive/ tensile loads depending on location
with respect to the neutral axis. The piles, consequently, axially deform, increasing in length if
subjected to tensile loads and shortening in length if compressive load is there. This change in
length gives rise to rotation of the pile cap which, in turn, rotates the sub-structure giving rise to
deformation at top of sub-structure. Procedure for working out this effect shall be as follows:
i. The ‘free-standing’ length of pile from bottom of pile cap to the point of fixity worked out
above is required to determine the change in length.
ii. Load on individual pile shall be worked out by considering the bending moment, location of
pile and the stiffness of pile group, using method similar to one used to determine the load
on individual rivets/ bolts in a group, as given below:
∗
=
∑
Where M = Bending moment at the pile cap level;
Page 47/57
= Axial load in pile ;

xi = Distance of individual pile from the neutral axis; and
∑ = Summation of square of distances of all i piles from neutral axis.
iii. Rotation of pile cap can then be worked out by dividing the deflection of outermost pile by
the spacing.
4. Structures with Well Foundations: The structures on well foundation are resting on single or
multiple wells. Diameter of wells is generally quite high as compared to the piles. The principles
given below are for single wells (cross-section may be circular or D-shaped or double D-shaped
or multi-cellular etc) which are the most commonly used well-form on Indian Railways. Wells are
rigid and due to large size have appreciable bottom friction. (Refer book “Theory and Practice of
Foundation Design” By N. N. Som & S. C. Das).
Accordingly, it is recommended that well shall be modelled with a fixed support at the
bottom, and supported by point springs (multilinear unsymmetrical) with stiffness equal to
difference between passive earth pressure and active earth pressure. For working out stiffness
of multilinear springs, full pressure may be considered to have been mobilized at the X/H ratio
as given in table 6.4 of “Geotechnical Engineering Handbook” by Braja M Das. The values for
mobilization of passive pressure may be used. The table is extracted below:
Table 6.4: Movement(X) of wall required to activate active and

passive conditions
Type of Backfill Soil X/H for Active state X/H for passive state
Dense sand 0.0005 0.0002
Loose sand 0.002 0.006
Soft clay 0.02 0.04
Stiff clay 0.01 0.02
-0-0-0-0-0-
Page 48/57
Appendix D: Working Out Stiffness of Different types of Bearings

1. Fixed-free bearings such as POT-PTFE Bearings: The structures with fixed-free bearings such as
those with roller-rocker bearings, those with POT-PTFE/Spherical/cylindrical bearings are easy to model.
The fixed end is considered rigid for the loads and has infinite (or very-very high) rigidity in modelling.
The free end offers frictional resistance to the longitudinal loads. This friction need not be considered in
modelling and the rigidity of these bearings is considered zero. There is slight error due to this
assumption but the magnitude of this error is less as the girder is not permitted by the fixed end to
move and only slight movement takes place at the free end due to bending of girder under loads. Para
2.1.3.5 of UIC 774-3R states that “The effects of friction on rail stresses and displacements are always
favourable especially when the support stiffness is low, so that ignoring friction is in general
conservative for safety.” Therefore, generally, considering the fixed end bearing as infinitely rigid and
free end bearing as infinitely flexible is OK. However, if exact analysis is desired, or in marginally failing
cases, the friction of free bearing can be considered.
Special Note: a) Some elastomeric bearing designs have special stoppers which convert one end of
girder to fixed while other end is free to move. The behaviour of these bearings needs to be seen and
appropriate decision about modelling needs to be taken. If the free end neoprene bearing stiffness is
to be considered, the stiffness shall be determined as per procedure given in para 3 below.
b) In case of phosphor bronze bearings, the phosphor bronze plate is provided at one
end of girder and the other side has steel to steel interface which is not greased. The friction in the
ungreased steel to steel interface is quite high and this end is virtually fixed. The other end has lower
coefficient of friction (0.15 as per clause 2.7.1 of Bridge rules). The coefficient of friction of phosphor
bronze end can be neglected, or considered as per procedure given in para 2 below, at the discretion
of designer carrying out the RSI analysis.
2. Expansion Bearings (Including sliding bearings): All structures have moving bearings at one end.
The movement can be permitted by steel to steel interface, or
through rollers, or through stainless steel on PTFE etc. Some
structures like plate girders have no fixed bearings and the girder
can move at both ends over sliding bearings. (The restraint is
provided by high coefficient of friction between the bearing and the
bed plate.)
The clause 2.7.1 of Bridge rules (Reproduced alongside) gives the

coefficient of friction for different sliding arrangements. This friction
opposes the movement of the girder. This resistance divided by the
total movement can be considered as the resistance offered by the
bearing to oppose that movement i.e. the stiffness of that bearing.
Movements (expansion/ contraction) under thermal effects takes
place under permanent loads like dead load (self-weight) and
Superimposed dead loads including load of track, ballast, pathways
etc.
The stiffness of this bearing for thermal load can be computed as
follows:
Page 49/57
Actual Value for Bridge No

20, Km 56.875 in KQR-GRD
S No Item Formula/ Notation Section, EC Railway.
(Composite Girder with
sliding bearing)
No of bearings on each end of girder n 2
W=Total Load/No of
Vertical Load (DL + SIDL) on one bearing (kN) 4650/(2*2) = 1162.5
bearings
Span Length (Overall) (m) L 19.775
If
Coefficient of thermal expansion for deck (per 0C) α 1.17 * 10-5
0.25 (For sliding
Coefficient of Friction at the interface µ
bearing)
0
Total thermal variation of desk from mean ( C) (+/-) Δt ±20
Then Total expansion/ contraction of girder, ΔL (mm) (+/-) 0.5 * L * α * Δt ±0.00231*1000 = ±2.31
And Stiffness of bearing (kN/mm) µ * W / ΔL 125.8
For loaded case, the stiffness will work out higher as live load is also there in addition to dead
load and superimposed dead load already present. Therefore, for exact analysis, the stiffness needs to
be worked out for unloaded case and loaded case separately and the analysis needs to be carried out
twice with the two different values. However, since the movement of bearing under load is not much,
the stiffness worked out for the unloaded case (in which the major amount of movement also takes
place) can be used for both the thermal as well as loaded case.
3. Structures with Neoprene Bearings: The stiffness of neoprene bearings can be worked out if we
know the plan dimensions, thickness of rubber bearings and the number of rubber layers. The stiffness
can be worked out as follows:
Actual Value for Bridge No 374,

Km 123/222 in Bangalore-Hassan
S No Item Formula Section, SW Railway (PSC Girder
with neoprene bearings)
Length of bearing (mm) L 560
Width of bearing (mm) B 360
Side cover (mm) c 6
If No of rubber layers (Nos) n 3
Thickness of each layer (mm) t 12
Load on bearing (kN) W 100
Static Shear Modulus (kg/mm2) G 0.1
2*G*(9.81/1000)*(L-
Then Static stiffness of bearing$ 10.39
2*c)*(B-2*c)/(n*t)
$
The dynamic shear modulus is double the static shear modulus as per UIC 772-2R, Clause 3.1.3
The above stiffness worked out is for loaded case. Under thermal loads, the static shear
modulus has to be considered and for this, the stiffness works out to be half of the above value. The RSI
analysis shall be done for thermal case considering the static shear modulus whereas for train loads,
dynamic shear modulus shall be used.
Note: It may be seen that the stiffness of elastomeric bearings is quite low as compared with the
sliding bearing worked out above. Thus, when the bridges having elastomeric bearings are analyzed for
RSI effects, the displacement under longitudinal loads becomes critical controlling parameter. This
flexibility of supports means that the track disperses higher loads in such cases as compared with the
sliding bearings or fixed-free arrangements.
Page 50/57
Appendix E: Explanatory example outlining Methodology to be Followed for RSI

studies Using FEM Program
1. Bridge Details: Bridge No 531, NW Railway
a Bridge Number: 531.
b Location of Bridge: 204/9-10,CH:207802.16
c Section: UDAIPUR-HIMMATNAGAR, NW Railway.
d Between stations: SES-RDD.
2. Span details:
a No. of spans: 3.
b Span Lengths: Clear span: 16.93 m, effective Span: 17.85 m, overall length: 19.05 m for all spans.
c Type of Girder: PSC 4I-Girder,RDSO/B-10245R.
d Location of Rails: At deck level.
e C/C of Girder leaves: 1.23m; Total Depth: 1.5 m.
f Details of Material: Grade of steel: IS:2062 Grade B; RCC Deck: M40.
g Properties of Material:
Concrete (Short term) Steel
Young’s Modulus 31000 N/mm2 (Cl 5.2.2.1 of CBC) 2.1x105 N/mm2 (cl 3.8 of SBC)
Poisson’s ratio 0.2 0.3
Coeff. of Thermal expansion 1.08x10-5/0C 1.17x10-5/0C
h Details of Different Cross-Sections: Uniform cross-section of girder. Critical Values computed:

X-Section Area:5.69 m2; Iyy:6.79722 m4; Izz:1.48946 m4; J:6.97722 m4; Asy:3.38027 m2; Asz:1.87711
m2; ytop: 0.559303m (All values are worked out for short term loading as the live loads are
transient loads; thermal effects are slightly over-estimated due to this).
Note: To be on conservative side, overall length 19.05 m used for all computations (The effect is very
minor).
3. Loading Details: 25t Loading, Full speed (125 KMPH for goods and 160 KMPH for passenger).
4. Track Details
a Type of track: Ballasted, 350 mm cushion.
b Rail section: 60kg 90 UTS.
c Rail Material properties: Young's Modulus: 2.1e+5 N/mm2, poisson’s Ratio: 0.3, Coeff. Of
thermal expansion: 1.17e-5.
d Track details: Height of sleepers: 0.21 m, Depth of ballast cushion: 0.35 m, Wearing coat: 0.08m.
e Fastener Details: Sleeper density: 1660 nos/KM, type of fasteners: Elastic (ERC).
f Location of SEJ within 100 m of either abutment: NIL.
g Curvature and gradient in track, if any: NIL.
5. Temperature Details
a Maximum rail temperature: 550C
b Minimum Rail Temperature: 70C
c Stress-Free temperature of LWR (Shall be maintained within ±50C, else RSI studies are required
to be done again): 360C.
d Maximum annual deck temperature: 510C
e Minimum annual deck Temperature: 70C
f Temperature of deck at the time of laying LWR/CWR: 290C
Page 51/57
Accordingly, maximum deck temperature variation from mean temperature to be considered for RSI:
340C (Actual variation 290C; ±50C margin for maintenance);
Maximum rail temperature variation from mean temperature to be considered for RSI: NIL as there is
no SEJ within 100 m on either side of the bridge.
6. Bearings Details: Complete drawings shall be available, having the following details
a Type of bearing: Elastomeric Bearing.
b No. of bearings per span: 4 in each end of girder.
d Dimension of Elastomeric Bearings: 550x 400x 78 mm.
7. Sub-structure Details: Complete drawings shall be available, having the following details
a Type of Piers and abutments: PCC piers and abutments.
b Material of construction: Grade of concrete, Piers: M30; Bed Block: M30..
c Properties of Material: Young's Modulus, Poisson’s Ratio, Coeff. of thermal expansion etc.
Concrete M30
Young’s Modulus 28000 N/mm2 (Cl 5.2.2.1 of CBC)
Poisson’s ratio 0.2
Coeff of Thermal expansion 1.08x10-5/0C (Assumed)
d Dimensional details: As per Drg No DyCE(C)137-UDZ/790/D-UDZ-HMT. Abutments are

rectangular/ tapered twin, spill-through type and piers oval in shape. Bed blocks are cuboid in
shape. Piers and abutments have different heights.
8. Foundation And Soil Details: Complete drawings and soil bore-log shall be available, having the
following details
a Type of Foundation: A1, A2, P1, P2: Open foundation.
b Details of Open Foundation: Single Step A1: 9.345m x17.204 m x 0.915 m; P1:11.116 m x 16.116
m x 1.220 m, P2:11.723m x16.272m x0 .920m, A2:10.33 x 17.204x1.220m.
c Soil details: Open Foundation, soft rock; Bearing capacity: 46T/m2.
9. Flood Details:
a Rail Level: 291.033 m.
b Bottom of Girder: 288.843 m.
c High Flood Level: 280.055 m
d Bed level: 277.705m.
e Scour Level: 276.500 m for both abutments and piers .
f Level at which rock/ non-erodible strata is there: 274.106m.
10. Computation For Stiffness

(1)Stiffness of Elastomeric Bearing:
Formula used for calculating Stiffness of elastomeric bearing,
(2 * G * 9.81) [(a  2c) * (b  2c)]

k= * , in kN/mm.
1000 n * hi
Page 52/57
In this bridge Elastomeric bearing used is as per RDSO drawing: RDSO/BA-10245R, Dimensions of the
bearing deduced from above drawing are as follows:
Size of bearing: 550*400*78 mm.
G= static shear modulus= 0.1 kg/mm2,
Length of bearing, a= 550 mm,
Width of bearing, b= 400 mm,
Cover, c= 6mm,
Individual layer thickness, hi= 10 mm
Number of layers of elastomers, n=5
Putting above values in the above equation gives the stiffness of elastomeric bearing as below,
∴ k= (2*0.10*9.81/1000)*[(550-12)*(400-12)/10*5]=8.19 kN/mm
(2) Stiffness of Abutments A1 & A2 with open foundations: To simplify the computations, open
foundations are considered supported vertically with soil springs and the effect of soil on sides of
foundation is neglected. This assumption is on safe side.
The stiffness of soil springs (ks) supporting open foundation vertically is taken as 128000 kN/mᶟ (This soil
Stiffness is for dense sand as on page no: 409 of foundation analysis and design “by Joseph E Bowel’s
Book .Normally open foundations are founded on hard strata/rock hence this assumption is OK or on
safe side. (See Para 2.1 (Design tip) of Appendix-C of these guidelines).
The deflection of sub-structure has two components:

(i) Deflection Due To Rotation of Foundation δ1: This is to be worked out as per Para 2.1 of APPENDIX-C
of these guidelines. Data in this case:
A1: a=9.345m, b=17.204m, h=11.417m, A2: a=10.33m, b=17.204m, h=13.537m. Using above formulae
for given values of a, b and h for Abutments A1& A2 as per GAD of the bridge.
For A1, δ1=2.56807E-07 m.
For A2, δ1=3.26609E-07 m.
(ii) Elastic deformation of Abutments (δ2): Modeling of Abutments A1 & A2 was done by MIDAS CIVIL
software and 98.1kN longitudinal force was applied at bearing level. If Abutments is of mass concrete,
then Ec given in IRS concrete bridge code is to be used. The deflected shape from modeling was as
follows:
Page 53/57
As per modelling, elastic deformation per kN of longitudinal force For A1, δ2=5.83E-07 m, For A2,
δ2=7.72E-07 m.
∴Total deformation at bearing level per kN of longitudinal load:
For A1, δ = δ1 + δ2 = 2.56807E-07 + 5.83E-07 = 8.39E-07m = 0.000839mm;
For A2, δ = 3.26609E-07 + 7.72E-07 = 0.0011 mm.
Finally stiffness for Abutments A1 & A2:
For A1 = 1/δ = 1/0.000839 = 1191.36 kN/mm.
For A2 = 1/δ = 1/0.0011 = 910.35 kN/mm.
(3) Stiffness of piers (P1, P2) with open foundations:
The deflection of sub-structure has two components:

(i) Deflection Due To Rotation of Foundation δ1: This is to be worked out as per Para 2.1 of APPENDIX-C
Of these guidelines. Data in this case:
P1: a = 11.116m, b = 16.116m, h=13.86m, P2: a = 11.273m, b = 16.272m, h = 14.737m. Using above
Formulae for given values of a, b and h for Piers P1& P2 as per GAD of the bridge.
For P1, δ1 = 3.87059E-07 m.
For P2, δ1 = 4.19206E-07 m.
(ii)Elastic deformation of piers (δ2): Modeling of piers P1 &P2 was done by MIDAS CIVIL software and
98.1kN longitudinal force was applied at bearing level. If Pier is of mass concrete, then Ec given in IRS
concrete bridge code is to be used. The deflected shape from modeling was as follows:
Page 54/57
As per modelling, elastic deformation per kN of longitudinal force For P1, δ2=7.18E-07 m, For P2,
δ2=8.60E-07 m
∴Total deformation at bearing level per kN of longitudinal load:
For P1, δ = δ1+δ2 = 3.87059E-07 + 7.18E-07 = 1.11E-06m = 0.00111mm;
For P2, δ = 4.19206E-07 + 8.60E-07 m = 1.28E-06 m = 0.00128 mm.
Finally stiffness for Piers P1&P2:
For P1= 1/δ = 1/0.00111 = 904.81kN/mm.
For P2= 1/δ = 1/0.00128 = 781.65kN/mm.
11.Live load and increments to be considered:

(i) Assuming left and right hand approach lengths = 200m (This shall be of adequate length to
accommodate train to be run.) and left hand abutment is the reference point.
(ii) Total length of deck/bridge = 3*19.05m = 57.15m
(iii) Therefore, Total length of deck/bridge with abutment = 457.15m
(iv) As per Para-4.2 of these guidelines, for 25 t load to be taken as:
Loading Vertical Load intensity

with impact (kN/m) TE intensity (kN/m) BF intensity (kN/m)
Length of UDL 0-12 12-∞ 0-12 12-40 40-∞ 0-12 12-∞
MBG Loading 182.95 104 47 22 0 34 13
(v) Longitudinal load intensity (TE) for (40-∞) = 0.00 kN/m. The span is only 19.775m and TE
will govern single span as well as double span loaded conditions, hence BF case not
required to be run. For larger span both cases needs to be run separately.
(vi) Location Increment for each Analysis = span/10 = 19.05/10 = 1.905m as per note 4 of para
4.2 of these guidelines.
(vii) Number of Track Loading Locations = (total length of bridge/location increment for each
analysis)+1 = 31
Page 55/57
For easy understanding, screenshot of LUSAS spreadsheet with data input is given below.
Amount
For
Deck Temperature 34
Temperature
Number of
Track Loading
Locations 31
Parametri Parametri Starting Finishing
Track Amoun Location
c Starting c End Location of Location
For Selection t (per Loaded Incremen
Loading Type Position Position Loading for of Loading
Rails to be unit Length t for each
for for First for Last
Loaded length) Analysis
Loadings Loadings Analysis Analysis
Vertical1 1 188 200 182.95 12 0 57.15 1.905
Vertical2 1 0 188 104 188 0 57.15 1.905
Acceleration1 1 188 200 -47 12 0 57.15 1.905
Acceleration2 1 160 188 -22 28 0 57.15 1.905
12.Track stiffness to be considered: As per Para 3.2.6 of the guidelines, the plastic resistance for
ballasted track is:
Unloaded condition = 25kN/m/mm.
Loaded condition = 50kN/m/mm.
Limit for elastic deformation = 2 mm.
13.Properties of rails to be considered:
For 60 kg UIC rail, from LUSAS BRIDGE PLUS software modeling done and data collected.
A Iyy Izz J Asy Asz Eccentricity Description
Track with 2
0.015440 0.0000614 0.011832 0.0118263 0.0127098 0.006599 0 UIC 60 kg
Rails
E ν α Description
210000 0.3 1.17E-05 60 kg Rails 90 UTS
14.Eccentricity and other dimensions to be considered:

Eccentricity between rail and slab = (height of NA of rail section+ height of sleeper at mid +ballast
cushion+ wearing coat) = (0.086m+0.21m+0.35m+0.08m) = 0.726m.
Eccentricity between rail and deck section (Distance from Neutral Axis of rail to top of deck + Distance of
neutral axis of girder from top) = 0.726 + 0.559 m = 1.285 m
15.Results: With above data input, the results obtained from LUSAS are summarized as follows. (It is
seen that the bridge passes from RSI considerations.)
Tempe- COMBINE Only Permissi
OK or Reference For
rature D (Temp. Train ble Unit
NOT permissible value
Only + Loading) Loading Value
MAX. Horizontal
Displacement Of UIC 774-3R, Cl.-
1 = 3.44 6.19 2.75 5.00 mm OK
Deck (Under 1.7.2, No SEJ
longitudinal loads
Page 56/57
only)
Max. Relative
Displacement
Between Rail & UIC 774-3R, Cl.-
2 = 3.225 4.175 0.95 4.00 mm OK
Slab (Under 1.7.2
longitudinal loads
only)
Additional Tensile
As Per IRC –
Axial Stress in Rail
3 = 6.08 29.22 75.00 MPa OK Bridge Rules para
due to RSI
2.8.2.4.3, CS-45
Effects(MPa)
Additional
Compressive Axial As Per IRC –
4 Stress in Rail due = 4.43 31.95 60.00 MPa OK Bridge Rules para
to RSI 2.8.2.4.3, CS-45
Effects(MPa)
Max. Deck End 2.021 E-
5 = 1.46E-06 0.0002 m
Rotation 04
Height Of CG
6 = 1.285 m
From Rail Level
Max.
Displacement Of UIC 774-3R, Cl.-
7 = 0.26 8.00 mm OK
Deck End Due To 1.7.3
Rotation
Tempe- COMBINE Only Permissi
OK or Reference For
rature D (Temp. Train ble Unit
NOT permissible value
Only + Loading) Loading Value
UIC 774-3R, Cl.-
1.7.2, Permissible
Value Not Given
Max Vertical
In IRS Bridge
8 Displacement Of = 0.00 mm 3.00 mm OK
Rules. It is 3 mm
Deck End
as per para
3.3.2.4 of RSI
guidelines
Other Results Of Modelling
Max Relative
Displacement
9 = 2.369 3.103 3.10 mm
Between Rail &
Track
Bending Moment
10 = 7457.09 kN.m
In Deck
Peak Longitudinal
11 Reaction On = 187.34 kN
Abutment (N)
Max. Shear Force
12 = 1553.83 kN
In Deck
Max. Shear Force
13 = 71.61 kN
In Track
-0-0-0-0-0-
Page 57/57

BS 114 VER 2 IR RSI Ver 2 Final

Uploaded by

Copyright:

Available Formats

BS 114 VER 2 IR RSI Ver 2 Final

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

BS 114 VER 2 IR RSI Ver 2 Final

Uploaded by

Copyright:

Available Formats

What are the guidelines trying to achieve by allowing long welded rails on bridges?

What are the guidelines trying to achieve by allowing long welded rails on bridges?

What problems were created by only allowing rail-free fastenings on bridges so far?

What problems were created by only allowing rail-free fastenings on bridges so far?

RDSO GUIDELINES FOR CARRYING OUT

RAIL-STRUCTURE INTERACTION STUDIES

RESEARCH DESIGNS AND STANDARDS ORGANISATION,

4.6.1 Analysis methodology using graphs given in UIC 774-3R ............................................................ 20

1.1 GENERAL C1.1

1.2 ADAPTATION / MODIFICATION TO EXISTING C1.2

1.3 RELEVANT CODES & STANDARDS C1.3

3.0 GENERAL CONCEPTS C 3.0

3.1 INTERACTION PHENOMENON C 3.1

Distribution of rail stresses along length of an LWR

Distribution of rail stresses affected by presence of bridge

3.2.1 Expansion Length C 3.2.1

3.2.2 Span Length C 3.2.2

Schematic indication of longitudinal displacement of deck fibres

3.2.3 Bending Stiffness of Deck C 3.2.3

3.2.4 Deck Height & Rotation Distance C 3.2.4

3.2.6 Track Stiffness C 3.2.6

Longitudinal resistance of track is provided by

3.2.7 Sectional Properties of Rails C 3.2.7

3.2.8 Temperature variations C 3.2.8

3.3 VERIFICATION OF TRACK AND BRIDGE C 3.3

3.3.1 Additional Stresses in Rails C 3.3.1

3.3.2 Displacements of Bridge Elements C 3.3.2

3.3.2.1 Longitudinal displacement of Deck due to C 3.3.2.1

3.3.2.2 Longitudinal displacement of Deck due to C 3.3.2.2

3.3.2.3 Relative displacement between rail and deck C 3.3.2.3

3.3.2.4 Vertical displacement of upper surface of C 3.3.2.4

4.0 STEPS IN CHECKING A STRUCTURE UNDER RSI C 4.0

4.2 VERTICAL TRAIN LOADS C 4.2

4.3 BRAKING AND TRACTIVE LOADS C 4.3

4.4 LOAD FACTORS C 4.4

4.5 STIFFNESS PARAMETERS OF STRUCTURE C 4.5

4.5.2 Determining stiffness of sub structure C 4.5.2

4.5.3 Determining stiffness of bearings. C 4.5.3

4.6 ANALYSIS METHODOLOGY C 4.6

4.6.2 Choosing computer program for carrying out C 4.6.2

4.6.3 Analysis methodology using FEM C 4.6.3

4.6.3.1 Recommendations for FEM Modeling C 4.6.3.1

4.6.3.2 Recommendations for FEM Analysis C 4.6.3.2

4.7 ANALYSIS FOR FRACTURE OR REMOVAL OF RAIL(S) C 4.7

5.0 SPECIAL CASES IN RSI C 5.0

5.1 HORIZONTAL CURVATURE ANALYSIS C 5.1

Radial Force Per Length of Rail

5.1.2 Braking / Tractive Analysis C 5.1.2

The effect of tractive/ braking forces has to be

5.2 RSI FOR TURNOUTS ON VIADUCT C 5.2

5.2.2 FEM Modeling of Turnouts C 5.2.2

A Typical idealized turnout

Modeling of turnout to capture its behaviour

6.0 CONTROLLING RSI EFFECTS C 6.0

6.1 MODIFICATION OF BEARING ARRANGEMENT C 6.1

6.2 PROVIDING SWITCH EXPANSION JOINT (SEJ) C 6.2

6.3 PROVIDING LOW TOE LOAD FASTENERS C 6.3

Appendix A: Important Documents which covers the RSI studies

ED/Track-I/RDSO’s letter no CT/IM/LWR/Part Dated 19/25.03.2014: