Structural Engineering Documents

Gunter Ramberger

Structural Bearings and Expansion Joints for Bridges

I

International Association for Bridge and Structural Engineering Association lnternationale des Ponts et Charpentes lnternationale Vereinigung fur Bruckenbau und Hochbau

IABSE AIPC IVBH

Copyright 02002 by International Association for Bridge and Structural Engineering All rights reserved. No part of this book may be reproduced in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system, without permission in writing from the publisher. ISBN 3-85748-105-6 Printed in Switzerland Publisher: IABSE-AIPC-IVBH ETH Honggerberg CH-8093 Zurich, Switzerland Phone: Int. + 41-1-633 2647 Fax: Int. + 41-1-633 1241 E-mail: secretariat@iabse.ethz.ch Web: http://www.iabse.ethz.ch

Dedicated to the commemoration of the late Prof. Dr. techn. Ferdinand Tschemmernegg, University of Innsbruck.

Preface
It is my hope that this treatise will serve as a textbook for students and as information for civil engineers involved in bridge construction. My intent was to give a short guideline on bearings and expansion joints for bridge designers and not to mention all the requirements for the manufacturers of such products. These requirements are usually covered by product guidelines, which vary between different countries. Not all the references are related to the content of this document. They are more or less a collection of relevant papers sometimes dealing with special problems. I express many thanks to Prof. Dr.-Ing. Ulrike Kuhlmann, University of Stuttgart, chairperson of Working Commission 2 of IABSE, who gave the impetus for this work; to her predecessor of the IABSE Commission, Prof. Dr. David A. Nethercot, Imperial College of Science, Technology and Medicine, London, for reviewing the manuscript, and Prof. Dr. Manfred Hirt, Swiss Federal Institute of Technology, Lausanne, for his contributions and comments.

I wish to thank J. S. Leendertz, Rijkswaterstaat, Zoetermeer; Eugen Briihwiler, Swiss Federal Institute of Technology, Lausanne; Prof. R. J. Dexter, University of Minnesota; G. Wolff, Reissner & Wolff, Wels; 0. Schimetta t, Amt der 00 Landesregierung, Linz; Prof. B. Johannsson, LuleA Tekniska Universitet, for amendments, corrections, remarks and comments. I thank also my assistant Dip1.-Ing. Jorgen Robra for his valuable contributions to the paper, especially for the sketches and drawings, and my secretaries Ulla Samm and Barbara Bastian for their expert typing of the manuscript. Finally, I would like to thank the IABSE for the publication of this Structural Engineering Document.

Vienna, April 2002

Gunter Ramberger

Table of Contents
1.
Bearings
1.1 Introduction 1.2 The role of bearings 1.3 General types of bearings and their movements 1.4 The layout of bearings 1.5 Calculation of bearing reactions and bearing movements 1.6 Construction of bearings 1.7 Materials for bearings 1.8 Analysis and design of bearings 1 .9 Installation of bearings 1.10 Inspection and maintenance 1. I 1 Replacement of bearings 1. I 2 Codes and standards 1.13 References

7 7 7 9 16 19 29 33 37 38 39 41 42

2.

Expansion Joints
2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10
Introduction The role of expansion joints Calculation of movements of expansion joints Construction of expansion joints Materials for expansion joints Analysis and design of expansion joints Installation of expansion joints Inspection and maintenance Replacement of expansion joints References

51 51 51 58 70 72 84 86 87 88

1
1.1

Bearings
Introduction

All bridges are subjected to movements due to temperature expansion and elastic strains induced by various forces, especially due to traffic loads. In former times our bridges were built of stones, bricks or timber. Obviously, elongation and shortening occurred in those bridges, but the temperature gradients were small due to the high mass of the stone bridges. Timber bridges were small or had natural joints, so that the full elongation values were subdivided into the elongation of each part. On the other hand, the elongation and shortening of timber bridges due to change of moisture is often higher than that due to thermal actions. With the use of constructional steel and, later on, of reinforced and prestressed concrete, bridge bearings had to be used. The first bearings were rocker and roller bearings made of steel. Numerous rocker and roller bearings have operated effectively for more than a century. With the development of ageing-, ozone- and UV-radiation-resistant elastomers and plastics, new materials for bearings became available. Various types of bearings were developed with the advantage of an area load transmission in contrast to steel bearings with linear or point load transmission, where elastic analysis leads theoretically to infinite compression stresses. For the bearings the problems of motion in every direction and of load transmission were solved, but the problem of insufficient durability still exists. Whilst it is reasonable to assume the life of steel bearings to be the same as that of the bridge, the life of a bearing with elastomer or plastic parts can be shorter.

1.2

The role of bearings

The role of bearings is to transfer the bearing reaction from the superstructure to the substructure, fulfilling the design requirements concerning forces, displacements and rotations. The bearings should allow the displacements and rotations as required by the structural analysis with very low resistance during the whole lifetime. Thus, the bearings should withstand all external forces, thermal actions, air moisture changes and weather conditions of the region.

1.3

General types of bearings and their movements

Normally, reaction forces and the corresponding movements follow a dual principle a non zero bearing force corresponds to a zero movement and vice versa. An exception is given only by friction forces which are nearly constant during the movement, and by elastic restraint forces which are generally proportional to the displacement. Usually, the bearing forces are divided into vertical and horizontal components. Bearings for vertical forces normally allow rotations in one direction, some types in all directions. If they also transmit horizontal forces, usually vertical forces are combined.

1. Bearings

A special type of bearing transmits only horizontal forces, while allowing vertical displacements. The following table (Table 1.3-1) shows the common types of bearings, including the possible bearing forces and displacements. Friction and elastic restraint forces are not considered.

Symbol Function All translation fixed Rotation all round Horizontal movement in one direction Rotation all around Construction Point rocker bearing Pot bearing; Fixed elastomeric bearing; Spherical bearing Constr. point rocker sliding bearing; Constr. pot sliding bearing; Const. elastomeric bearing; Constr. spherical sliding bearing Free point rocker bearing; Free pot sliding bearing; Free elastomeric bearing; Free spherical sliding bearing; Link bearing with universal joints (tension and compression)

Horizontal movement in all directions Rotation all round

All translation Line rocker bearing Leaf bearing fixed (tension and Rotation about one axis compression) Horizontal movement in one direction Rotation about one axis Horizontal movement in all direction Rotation about one axis
~

Roller bearing; Link bearing (tension and compression); Constant line rocker sliding bearing Free rocker sliding bearing; Free roller bearing; Free link bearing HoriLontal force bearing

All horizontal tranal. fixed Rotation all round Horizontal movement in one direction Rotation all round

Guide bearing

1.4 The layout of bearings

Tuble 1.3-1

8 . 2

1.4 The layout of bearings

1.4.1 General Bearings can be arranged at abutments and piers (fig. 1.4.1-1 ; fig. 1.4.1-2) under the webs of the main girders, under diaphragms (fig. 1.4.1-3), and under the nodes of truss bracings. The webs and the diaphragms of concrete bridges have to be properly reinforced against tensile splitting; steel bridges need stiffeners in the direction of the bearing reactions to transfer the concentrated bearing loads to the superstructure and the substructure. Abutments and piers also have to be properly reinforced under the bearings against tensile splitting.

-77
Fig. I .4. I - I : Bearings at an abutment

I -

~-

Fig. 1.4.1-2: Bearings at u pier

Fig. 1.4.1-3: Bearing at a single pier

I7

10

1. Bearings

The layout of the bearings should correspond to the structural analysis of the whole structure (super- and substructure together!). If the settlement and the deflection of the substructure can be neglected the structural analysis of the superstructure, including the bearings, can be separated from that of the substructure. Sometimes the model for the analysis, especially of the superstructure, will be simplified by assuming the following: bearings are situated directly on the neutral axis of the girder (fig. 1.4.1-6),the motion of the bearings occurs without restraint, bearings have no clearance, etc. In this case we must consider the correct system (fig. 1.4.1-5) at least for the design of the bearings and take into account the influence of the simplifications on the structure.

&
Fig. I .4.1-4: Reality

A
Fig. I .4.1-5: Correct system

On the abutments or separating piers it is normal to use at least two vertical bearings to avoid torsional rotations. At intermediate piers one or more vertical bearings may be used. If more than one bearing is used the rotational displacement at the pier is restrained. More than three vertical supports of the superstructure lead to statically indeterminate bearing conditions, but even the simplest bridge has at least four vertical bearings. If the torsional stiffness of the superstructure is low (e.g. open cross sections) it may be neglected and the layout with four bearings becomes isostatic. If the torsional stiffness is not negligible (e.g. box girders) we have to take it into account for the structural analysis, especially for skewed and curved bridges. On a bridge with n > 3 vertical supports, n - 3 bearing reactions can be chosen freely within a reasonable bandwidth. This possibility can be used to prestress the superstructure and to distribute the bearing reactions as desired. If the bearings are situated (nearly) in a plane we need at least one horizontally fixed and one horizontally moveable bearing. The moving direction must not be orthogonal

I .4 The layout of bearings

11

to the polar line from the fixed to the moveable bearing. If more than two bearings in the horizontal direction are necessary, the basic principle should be that an overall uniform extension, caused by temperature or shrinkage, shall be possible without restraint.

In general, there are two possibilities for the arrangement of the bearings: a) arrangement in a horizontal position (fig. 1.4.1-7) b) arrangement in a position parallel to the road or rail surface (fig. 1.4.1-8).

---_---,--a

Fig. 1.4.1 -7: Horizontal arrangement of the bearings (case a)

( I

f=--

,,displaced

bridge

Fig. 1.4.1-8: Inclined arrangement ofthe bearings (case b)

Case a) has the advantage that only vertical bearing reactions and no permanent horizontal reactions result from vertical loads, but it has the disadvantage that bridges with inclined gradients require a step at the expansion joint due to movements in the superstructure. The greater the elongation or shortening, the greater the step required. Case b) has the advantage that the slope of the expansion joint is independent of the movement of the bridge. The inclination of the surface of support gives the direction of the normal force. Besides vertical reaction forces, also horizontal reaction forces result from vertical loads. Permanent horizontal actions can lead to a displacement by creep of the concrete and the soil and, thus, to crooked piers.

12

1. Bearings

1.4.2 The layout for different types of bridges For single span girders the layout of the bearings is straightforward. One fixed and one moveable bearing is provided on each abutment, all other bearings are just vertical supports, moveable in any horizontal direction. For wide bridges the horizontally fixed bearings are located in or near the bridge axis.
Formerly, the classical arrangement of the bearings for a bridge with two main girders consisted of one fixed and one lengthwise moveable bearing at one abutment and one lengthwise moveable and one free bearing at the other abutment (fig. 1.4.2-1). This layout has the advantage that longitudinal horizontal forces (braking and traction forces) can be distributed into the two bearings at the abutment, but it has the disadvantage that horizontal forces in the cross direction (wind) and temperature differences cause horizontal restraint forces, provided that bearings have no clearance on the abutments. The author prefers the statically determinate system with only one lengthwise restrained bearing at the abutment concerned because the actual clearance of a bearing is not determinable in reality (fig. 1.4.2-2).

++11,
%I,
I

LA: .
_-

- -- -

;c

Fig. 1.4.2-1: Classical layout

Fig. 1.4.2-2: Horizontally statically determinate system (better than classical layout)

- _ - _ _ - - - -------- --- .

Fig. 1.4.2-3: System with separated vertical and horizontal bearings (statically determinate system)

1.4 The layout of bearings

13

For skewed or horizontally curved single span bridges we have to decide whether the horizontal force should be combined with the higher or with the lower vertical reaction force. For all bearing constructions it is easier to transfer horizontal forces in combination with a high vertical force. In this case the resultant force stays nearer to the centre, its angle to the vertical is smaller and leads to smaller bending moments in suband superstructure (fig. 1.4.2-4).

!
I I HI

Fig. 1.4.2-4: Inclination of the resultantforce

Thus, the horizontally constrained bearings for skewed bridges should be placed at the obtuse corners of the bridge, for curved bridges at the outer side (fig. 1.4.2-5).

Fig. 1.4.2-5: Skewed bridge

Fig.1.4.2-6: Layoutfor continuous girders

14

1. Bearings

For straight continuous girders normally two bearings are used at every abutment and pier. If the torsional stiffness is high (box girder) the intermediate piers can be reduced to a round column with one bearing on the axis under the diaphragm. Constrained bearings in the cross direction are the rule at all piers. If the horizontal bending stiffness is very high we can transfer the horizontal forces only at the abutments. The same considerations are suitable also for skewed and curved bridges (fig. 1.4.2-6). Bearings for horizontal forces and guide bearings which transfer only horizontal forces may be used in combination with leaf or link bearings which cannot transmit horizontal forces. The movement of an expansion joint must be linked by a guide like a constraint bearing. The main movement of an expansion joint should be in the axis of the traffic way. Generally, this direction does not coincide with the direction of the polar line from the fixed bearing to the moveable bearing at the abutment (fig.1.4.2-7). If all other bearings have the same angle between the polar line and the moving direction there results a layout of the bearings with no restraints on uniform elongation or shortening (e.g. caused by thermal actions or shrinkage), as shown below (fig.1.4.2-8).

Fig.1.4.2-7: Layout for curved bridges

Fig. 1.4.2-8: Layout for curved continuous girders (no constraint under overall tempe ra ture)

Fig. 1.4.2-9: Geometrical situation

I .4 The layout of bearings

15

The elongation is A,, = k . r, << r,

A- = k . r << r

proportional elongation

The rotation is

A;tana r+A,

A;tana r

k.r.tana =k.tana r

For Cp, = Cp, the bridge simply rotates as a rigid body without constraint. One special case of this general rule is well known: the bearings are moveable in the direction of the polar lines with a = 0 (fig.1.4.2-10). However, this layout has the disadvantage that generally the main movement of the joint does not coincide with the movement of the bearing.

Fig.1.4.2-10: Special case with a = 0

1.4.3 Special bearing conditions, advice etc. It is important to note that the layout of the bearings has a great influence on the structural system. The above mentioned arrangements of bearings are typical for average bridges. The following examples show some special effects which have to be considered for the design of bridges and bearings. These examples do not lay claim to completeness. a) The already mentioned bearing layout, consisting of one bearing fixed in all sliding directions and one fixed lengthwise at one abutment, leads to high constraint forces not only under horizontal but also under eccentric vertical loading (fig.] .4.3-1). It is interesting that this eccentric loading has no prying effect if the bearings are situated directly on the neutral axis of the girder. This effect results only from the (small) eccentricity of the bearing under the lower flange.

16

I . Bearings

A +
Fig. 1.4.3-1: Prying effect due to a eccentric loading
b) A similar situation occurs for a continuous girder with chequer pattern loading.

~ ~

Fig. 1.4.3-2: Prying effect due to chequer pattern loading

c) It is not generally known that a skewed bridge with horizontally fixed bearings only in one line exhibits the same effect under vertical loading, as the following figure shows:

Fig. 1.4.3-3: Prying forces for a skewed bridge with vertical loading
Similar effects can occur for curved bridges. For the correct analysis of the bearing reactions it is always necessary to model the bearings at the very point where they are actually situated, and in combination with the substructure. The deflection of the substructure can influence the constraint bearing reactions significantly.

1.5

Calculation of bearing reactions and bearing movements

1.5.1 Actions According to Eurocode 1 (ENV 1991) the actions can be subdivided into: - permanent actions, - variable actions, - extraordinary actions.

1.5 Calculation of bearing reactions and bearing movements

17

The bridge should take up the desired shape under all permanent loads, at the average temperature (+lo" C in most of the European countries) and, if time-dependant displacements occur, at the time t = 00, at which time all moveable bearings should be in the zero adjustment (null position). Variable actions and extraordinary actions lead to deviation from this form. Variable actions to consider are: - traffic loads, considering the applicable dynamic coefficients - loads due to traffic loads, i.e. nosing forces centrifugal forces braking forces traction forces - wind loads wind on construction wind on traffic loads - settlements of abutments and piers - thermal actions uniform temperature vertical temperature gradient horizontal temperature gradient temperature differences between individual parts of the bridge (e.g. stay cables, pylon and stiffening girder) - creep and shrinkage of concrete Extraordinary actions to consider are: - earthquake actions - vehicle impact - derailment - rupture of the conductor line others
'

1.5.2 Bearing reactions For permanent actions such as self-weight of the construction, dead load and prestressing, the bearing reactions can be calculated as one load case. For the analysis of the bearings it is necessary to consider different combinations of the bearing reactions: - maximum vertical force and the adjacent horizontal force, - minimum vertical force and the adjacent maximum horizontal force, - maximum horizontal force and the adjacent maximum vertical force, - maximum horizontal force and the adjacent minimum vertical force. The simplest way to obtain these combinations is to calculate the variable actions, especially the traffic load, according to the influence line. One should bear in mind that horizontal actions such as centrifugal forces or braking forces are proportional to the vertical traffic load, but other loads, such as wind or traffic or traction forces for railways, are not.

18

1. Bearings

To obtain the extreme bearing reaction it is necessary to consider that all bridges are three-dimensional and not merely plane systems. The influence lines (influence surfaces) of the bearing reactions can be found as the displacement curves (displacement surfaces) of the system, due to unit displacements F = 1 or cp = 1, acting at the position and in the direction of the required force. If these analyses are performed on a three dimensional model, the definitive influence area will result directly (fig.1 S.2-1; fig.1 S.2-2). If plane models are used for the analyses, special care is necessary, particularly with continuous girders with open or box sections. The following examples demonstrate the difference:

Fig.1.5.2-1: Influence area for the verticul bearing reaction A, box section.

Fig.1.5.2-2: ZnJuence areafor the vertical bearing reaction A, open section.

1.5.3 Bearing displacements As already mentioned, the zero adjustment (null position) of every bearing has to be defined. The displacements are measured from that position. Thus, for concrete and composite bridges it is usual to consider displacements under time-dependent actions such as creep and shrinkage from the time of installation of the bearing to the time defined for the null position (normally t = w), from which position the displacements due to variable actions are measured. To obtain the maximum displacements and rotations, again we can use influence lines. The influence line of a displacement can be calculated as the displacement curve due to the corresponding unit force P = I. To take into account the imperfections due to installation, the temperature difference for the calculation of bearing displacements should be assumed higher than for the structural analysis of the bridge, or some additional displacement should be considered.

I .6 Construction of bearings

19

1.6

Construction of bearings

Fig. 1.6-1 gives un overview for the most common bearings.

20

1. Bearings

1.6.1 Elastomeric bearings Elastomeric bearings are the simplest types of bearings. In the basic mode they consist merely of an elastomeric block (usually rectangular or round). The elastomeric works as a soft part between sub- and superstructure and allows movements in all directions by elastic displacements or rotations. Under vertical loads the elastic block bulges, leading to vertical displacements. A solution to this problem was found by reinforcing the elastic block by thin horizontal steel plates, vulcanized to the elastomer (fig. 1.6.1- 1). The reinforcing plates prevent the block from bulging, thus leading to very small vertical displacements, but they do not hinder horizontal displacements in every direction and also allow small rotations in all directions. Every displacement and rotation leads to restraining forces and moments which have to be taken into account on the whole structure.
These restraining forces are possible if the friction between bearing and sub- and superstructure is sufficient. The friction forces F depend on the compressive force C and the friction coefficient p, with F = C . p. If displacements take place under a small compressive force, sliding between bearing and sub- or superstructure can occur. To avoid this it is necessary to use elastomeric bearings with resistance to sliding. This can be achieved by applying vulcanized plates on the bottom and on the top of the bearing, which can be connected to the sub- and superstructure by bolts, pins or appropriate shapes (fig. 1.6.1-2).

Fig.1.6.1-1: Elastomeric hearing (unanchored)

Smaller, short time, horizontal forces can be transmitted by the restraining forces. If these forces are higher or if they are permanent loads a restraining steel construction is required. In these case the elastomeric bearing transmits the vertical force and allows rotations, while horizontal forces in one or two directions are transmitted by the steel construction (fig. 1.6.1-3 ; fig. 1.6.1-4).

Fig. 1.6.1-2: Elastomeric bearing (anchored)

1.6 Construction of bearings

21

Fig. 1.6.1-3: Elastomeric bearing constraint

Combination: elastomeric bearing and steel construction fixed in one direction.

Fig. 1.6.1-4: Fixed elastomeric bearing

Combination: elastomeric bearing and steel construction fixed in two directions.

1.6.2 Steel bearings Steel bearings are the oldest type of bearings. They have been used for more than 100 years. The principle is simple: a flat plate rolls on another steel plate with a curved surface. If this surface is part of a sphere, theoretically we obtain a point tangency. If this surface is part of a cylinder, theoretically we obtain a linear tangency. In the first case we speak of point rocker bearings, in the second case of line rocker bearings. These bearings allow rotations in all or in one direction, but they do not allow displacements (fig. 1.6.2-1 ; fig.1.6.2-4). Under minimal vertical reactions in combination with horizontal loads point rocker bearings and line rocker bearings can exhibit damage of their connections, because of tension. In combination with sliding elements these bearings are very sensitive to this phenomenon, and it causes partial uplift and excessive wear as a result. Linear tangencies can be found also in roller bearings consisting of a roll and a lower and an upper plate (fig. 1.6.2-5). These bearings allow rotations in one direction and displacements in one direction. The problem with these bearings is a point or linear concentration of the bearing force, which theoretically leads to infinite stresses. In 1881, the physicist Heinrich Hertz found the solution of this problem: caused by the elastic deformation the theoretical point of tangency yields to a circle, the theoretical line of tangency yields to a rectangle. The infinite stresses decrease to high but finite stresses, the so called Hertz compression stresses over a very small contact zone. If the radius of the sphere or of the cylinder decreases the Hertz stresses increase. From the local stress concentration the stresses have to be distributed to the contact zones between bearing and sub- and superstructure. Therefore, steel bearings normally need thicker plates for the stress distribution than other types of bearings which transfer the bearing reactions over an area.

22

1. Bearings

Point rocker bearings are used for bearing reactions in the range 500 and 2500 kN, line rocker bearings and roller bearings for loads in the range 200 and 20 000 kN.

Fig.1.6.2-I: Fixed point rocker bearing

Fig. 1.6.2-2: Point rocker bearing constraint in one direction

Fig. 1.6.2-3: Free point rocker bearing

Fig. 1.6.2-4: Line rocker bearing

I .6 Construction of bearings

23

i
Fig. 1.6.2-5: Roller bearing (left side without guide rail; right side with guide rail)
The contact zones of steel bearings cannot be protected against corrosion. Therefore corrosion-resistant layers of high alloyed steel should be used for the contact areas. This can be done by building up a surface by forging or by welding. Between the mild steel and the hardened high alloyed steel of the surface there should be a welded or forged tough buffer zone. The thickness (in mm) of the hardened layer both on the roller (radius R in mm) and of the plate should be t 2 0,14 . R - 2.

1.6.3 Pot bearings These bearings were invented in the 1950s. They combine the two desirable properties: rotation capacity with a very small resistance and transmission of the bearing reaction over a defined area.
The pot bearing consists of a steel pot, filled with an elastomeric disc and a lid or a piston to the top (fig. 1.6.3-1). When subjected to high compression forces, the unreinforced elastomeric disc behaves similarly to a liquid. Rotations can occur due to the nearly constant volume of the elastomer (v = 0,5). Of great importance is the sealing between the elastomeric pad and the lid: if this sealing has a defect the elastomeric pad escapes like a viscous liquid. The standard type of pot bearing allows only rotation (fig. 1.6.3-2). Vertical forces are transmitted to the pad, horizontal forces from the lid to the pot. To release one sliding direction, an additional construction becomes necessary (fig. 1.6.3-3 and fig. 1.6.3-5). This sliding construction consists of three components: a polytetrafluorethylene (PTFE) disc, a surface of polished stainless steel connected to a sliding plate of structural steel and lubrication grease. PTFE is a plastic with high mechanical and chemical resistance, great toughness and very small friction when combined with polished stainless steel. The PTFE disc is 5 to 6 mm thick, where half a thickness is enclosed by the lid. This disc has small round pockets on the surface for the lubrication grease (normally silicon grease) to reduce friction and wearing. To constrain the movement in one direction an additional guide is used for the lid. This guiding device allows movements in only one direction (fig. 1.6.3-3). Pot bearings are used for vertical bearing forces from 1000 kN up to 100000 kN. Depending on the standard applied the allowable compression between lid and elas-

24

1. Bearings

tomeric pad should not exceed 4.0 kN/cm2. The allowable compression for the PTFE is 3 kN/cm2for permanent loads and 4.5 kN/cm2for short term loads (traffic, wind etc.). Pot bearings have the advantage of a very high vertical stiffness (nearly incompressible elastomeric part). It is comparatively independent of the size of bearing and the applied load. This characteristic is important for the bearing of high velocity railway bridges. Bearings with low vertical stiffness can lead to damage of the rails.

astomere disc Lid Sealing Pot - wall Pot - bottom

Fig.1.6.3-1: Function of a pot bearing

Fig. 1.6.3-2: Fixed pot bearing

Fig. 1.6.3-3: Pot bearing constraint in one direction

1.6 Construction of bearings

Anchoring plate Sliding plate Polished stainless steel PTFE (Polytetrafluorethylen) Lid Pot -wall Sealing Elastomere disc Pot - bottom

25

Fig.1.6.3-4: Members of a pot bearing

Fig.1.6.3-5: Free pot bearing

1.6.4 Spherical bearings The basic type of spherical bearing consists of three main parts: the pan, the part of a sphere and the upper plate made of constructional steel (fig.1.6.4-1). To allow displacements between the parts, sliding surfaces are necessary. The pan has a PTFE plate on the upper surface, the part of the sphere has a chrome-plated polished surface on the underface and a PTFE plate also on the upper surface, and the upper plate has a polished stainless steel plate on the underface. The PTFE plates are chambered over half the thickness and have lubrication pockets with silicon grease, like the sliding plates for pot bearings. The friction resistance of the sliding parts causes reaction moments due to rotations. They must be taken into account to consider additional design stresses of the bearing material.
The vertical bearing reaction is transferred over the compressed areas of the PTFE. The basic model is a moveable bearing (fig. 1.6.4-4). To constrain horizontal displacements an additional construction to connect the upper plate with the pan becomes necessary (fig.1.6.4-2; fig.1.6.4-3). British and Italian bearings have one sliding plane only and a deeper concave part to take over horizontal forces (fig. 1.6.4-5). The construction must be checked for uplift and exceeding the stresses in the contact area. In the bearings with two sliding planes the centre of rotation is between the contact areas of the sliding surfaces, whereas in Italian and British bearings it is somewhere in the bridge structure or in the pier or the abutment.

26

1. Bearings

Like pot bearings, spherical bearings are used for vertical forces in the range of 1000 to 100 000 kN.

Polished

Sliding plate

hb
Part of sphere PTFE Chrome plated polished surface

Fig.1.6.4-1: Members o f a spherical bearing

Fig. 1.6.4-2: Fix spherical bearing

I
I

Fig. 1.6.4-3: Spherical hearing constraint in one direction

I
Fig. 1.6.4-4: Free spherical bearing

1.6 Construction of bearings

27

Fig. 1.6.4-5: Italian and British spherical bearing (one sliding s u f a c e )

1.6.5 Leaf and link bearings All the above mentioned bearings are able to transfer compression forces. If tensile forces as well as compressive forces must be transferred, leaf and link bearings are used. These bearings can only transmit forces in the direction of the leaf. To transfer forces in the crosswise direction, separate bearings must be used.
A leaf bearing consists of a foot plate, one or two lower leafs with pin holes and two or one upper leaf with foot plate and pin holes, connected by a pin. Leaf bearings allow free rotation in one direction. Pin and pin holes must have a fit less than 0.3 mm, as in cases of greater slackness and changing forces the pin will punch the hole. Pin plate and pin should be of different types of steel to avoid seizure. Pin plates are made of structural steel, pins often of tempered steel. For link bearings a pendulum is linked to the foot leaf and to the upper leaf by pins. Link bearings allow rotation and displacement in one direction. For pin holes and pins the same rules apply as given for leaf bearings. Link bearings with universal (Cardan) joints are used only in special cases. They allow rotation and displacement in all directions. Displacements 6 of link bearings are always combined with a small displacement 6,

in the perpendicular direction. 6, = __ , with R equal to the distance between the 2R axes of the pins. Therefore this distance should not be too small.

62

1.6.6 Disc bearings Disc bearings were introduced in the late 1960s. The vertical loads are transferred by an elastomeric disc made of polyether-urethane polymer. In contrast to a pot bearing a transverse extension of the elastomeric disc is possible. Bearing capacity and functioning is comparable with an elastomeric bearing. Rotations around the horizontal axis are transferred by differential deflection of the disc. The rotations cause a shift of the axis of the load from the centre of bearing, which must be considered in the design. Horizontal forces are transferred by a shear-restriction device which allows vertical deformation and rotation. The basic type is a fixed bearing. Free bearings are constructed by additional sliding elements and (if necessary) guiding systems.

Next Page

28

1. Bearings

Fig. I . 6.6-I :Fixed bearing

Fig. 1.6.6-2: Uni-directional guided

-top plate

bearing assembly

base plate

Fig. I . 6.6-3: Multi-directional non-guided

Previous Page
1.7 Materials for bearings

29

1.7

Materials for bearings

1.7.1 Steel Structural steel Structural steel is used for all parts of bearings which are not under extraordinary local stress or do not require special properties against corrosion. Structural steel for bearings can be: - Non-alloy structural steels according to EN 10025 - Fine-grained structural steels according to EN 10113 - Quenched and tempered steels according to EN I0082
Eurocode 3 may be used for the design of all bearing components made from structural steel according to EN 10025 and EN 10113 and for all connections (bolts, welds etc.). Quenched and tempered steels are used mostly for non-welded parts under high pressure (parts with Hertz compression, bolts of leaf and link bearings). In contact areas with Hertz compression layers of corrosion-resistant hard steel can be applied by forging or by welding. In the case of hard-surface welding a tough intermediate (puffer) layer must be welded between the steel and the hard-surface.

Stainless steel Stainless steel according to EURONORM 88-2 or I S 0 683 can also be used for bearings. For design one should use EC 3 , part 1-4. Concerning stainless steel for sliding plates see 1.7.3.

1.7.2 Elastomeric parts Elastomeric parts of bearings consist normally of natural or artificial (chloropren) rubber (NR or CR, respectively). Artificial rubber has the same good properties as natural rubber, and in addition it has a higher resistance against ozone, ultraviolet radiation and ageing and is more rigid. The characteristic mechanical property is the shear modulus G between 0.7 and 1.15 N/mm2 at room temperature, decreasing with increasing temperature. When undergoing stress changes the volume of rubber is nearly constant. So we have a Poissons ratio v = 0.5 and a Youngs modulus of elasticity E = 2 . ( 1 +v) . G -- 3 . G. The fracture strain of rubber lies between 250 % and 500 %. Rubber creeps under stress by up to 50 % of the elastic strain, but creeping ends within some days or weeks. Rubber does not break under compression, it can only break under tensile or shear stresses. Compressing a rubber pad changes its shape. The changing of the shape depends on the possibility of displacement at the compressed areas. If the compressed areas are fixed to a rigid surface, the displacement remains small. Thus we obtain the inequality A, > A , > A3 (fig.1.7.2-1).

Fig.1.7.2-I : Vertical displacements depending on the lateral expansion

30

1. Bearings

Fig. 1.7.2-2: Stress distribution

If the surface of the rubber is fixed to a rigid body shear stresses develop between the two surfaces under compression (fig. 1.7.2-2). Under compression we obtain a virtual modulus of elasticity E, Lllmpr which depends not only on the shear modulus G but also on the thickness of the part between two plates. For rectangular parts a good approximation for E, co,npr is given by
=G

1'

conipr

(: )

. (1

0,6

g)

for b 2 a

The maximum stresses under compression between two rigid bodies are

with o = -, F: compression force. ab For bending, the effective modulus of elasticity E, bcndlng is lower than E, i<,,,,pr because we obtain a compression in two half waves under a constant rotation angle a.If both halves develop a constant displacement, the virtual modulus of elasticity would be the a 1 same as under compression, but with a + we obtain E, hendlng = el Lo,npr. Actually, 2 4
~

1.7 Materials for bearings

31

is not in the middle of one half but nearer the outer side; thus we 1 a -E l compr. This is described very well by the finally obtain: a + < , El 2 following approximate formula: the maximum
(3
-

for b 2 a Under the rotation a we obtain a curvature p = and a restraining moment d

~

a=

Mi?
bending
'

with I =

_____

a'b 12

Fig. 1.7.2-3: Rotution - restraining moment

Fig. 1.7.2-4: Displacement - restruining~forces

1.7.3 Sliding elements For sliding elements in constructional bearings it is normal to use PTFE, also known by the registered trade names Teflon and Hostaflon. PTFE is a so called thermoplast. For bearings it is used in the original (virgin) condition, i. e. not sintered and without fillers. As a counterpart to this rather soft material polished stainless steel plates are normally used, and sometimes acetal resin plates or hardened chromium-plated steel plates. Chromium-plated steel plates are not resistant to fluorine ions and are rather prone to corrosion than stainless steel plates. They are allowed for convex elements only. The combination of a soft and a hard part has the advantage that there is no danger of cold welding which can occur on polished metal or plastic surfaces under high pressure. To minimise the friction silicon grease should be used to provide lubrication. To keep this grease between the two surfaces the PTFE has lubricant pockets on its surface, so that a permanent lubrication takes place over several years. The PTFE plates for bearings are normally 5 to 6 mm thick, the depth of lubricant pockets is 2 mm. Un-

32

1. Bearings

der pressure the PTFE yields. To keep the PTFE in the desired shape it is necessary to keep about half the thickness in a <<chamber>> with sharp edges. Over the sharp edges we obtain a small bulge. It is also possible to glue PTFE to a steel surface. In this case the PTFE is about 2.5 mm thick. The friction coefficient increases with decreasing temperature and with decreasing compression. The static friction coefficient (first movement) is higher than the dynamic coefficient. After movement has taken place the dynamic friction coefficient remains at this value and returns to the static value after a few hours. This might depend on the orientation of the large polymer molecules; during movement they are orientated into the direction of motion within a very thin surface layer. When the motion is stopped, the orientation is lost within a few hours. Fig. 1.7.3-1 shows the design values of the friction coefficient pLd between PTFE and stainless steel, depending on the compression force (EN 1337-2).

0.00

I I

I
I

I I
I

I I I

0.0

0.5

1.o

p [kNicm']

1.5

2.0

2.5

3.0

Fig. 1.7.3-I : Friction coejficient depending on the compressionforce

The design value of the ultimate compression load is

f , = 6 , 5 (1 - 0,02. [6 - 30'C)) kN/cm2 for 6 2 30'C , 6 : maximum temperature of the bearing.

The wearing of the PTFE depends on a) the product of compression and velocity of the displacement b) the total amount of sliding during the life-time c) the lubrication of the surface (a loss of lubrication leads to extremely high wearing) d) the roughness and the hardness of the stainless steel surface e) the contact pressure near the edge of PTFE (ironing effect)

I .8 Analysis and design of bearings

33

For slow movements caused by thermal actions we obtain long sliding movements but at a low velocity. Quick movements caused by traffic loads have short sliding movements but they occur at high velocity. Wearing is mostly caused by the second case. For the stainless steel plate, austenitic steel X6CrNiMo17122 according to EURONORM 88-2, surface n (IIIc), should be used. The stainless steel plate must cover the PTFE plate completely in all situations. The thickness of the plate should be at least of 1 .5 mm. The connection to the carrying plate of mild steel can be welded or glued. For 2.5 mm thick plates the connection can be riveted or bolted.

1.8

Analysis and design of bearings

1.8.1 Hertz compression For the design of bearings the following problems should be addressed: compression between two spherical bodies, compression between a spherical and a flat body, compression between two cylindrical bodies, compression between a cylindrical and a flat body along a generator line. As already mentioned, Heinrich Hertz obtained the solution under the following assumptions (1881): 1. The two bodies consist of isotropic, homogeneous and infinitely elastic materials. 2. Only normal stresses (no shear stresses) occur at the contact areas. 3. The radius (width) of the contact areas is small compared with the radii of the involved bodies. Hertz found the following maximum compression stresses max (T and widths b on the contact areas:
Spherical body on spherical body

b=
7

3F(I-v2) .

1
-f-

2E

= 1,109

3 7

-*-

Cylindrical body on cylindrical body

34
with

1. Bearings

~-~

rl

r 2

Fig. I .8. I - I b: Arrangement o f the radii

F 1
r,, r2 E
V

max (3 b

bearing reaction length of the cylinder radii of the bodies in contact Young's modulus Fig. 1.8.1-2: Stress distribution Poisson's ratio (v = 0.3 for steel) maximum normal stress at the contact area half the width of the contact zone

For the usual rocker or roller bearings the max (3 beneath the vertical bearing reaction greatly exceeds the material yield strength (fig. 1.8.1-2). However, at the contact zone we have not only vertical but also horizontal compression stresses. According to the von Mises criterion the comparison stress
- O3Oi and yielding begins when reaches the material yield strength f,. In the present three-dimensional compression regime, (3" will be less than (3, and yielding will not begin until o1 = f,. On the other hand, the maximum strain does not occur at the surface in the middle of the compression zone, so that the hardness of the surface is not the only criterion for the assessment of Hertz compression.
Ov =

d0i2 + O2

+ Oj3 - (3~(32 -

EN 1337-4 - roller bearings

gives for the design line load pd of a roller bearing

(cylindrical body on flat surface): pd 5 18. R .

f 2
E d

with

tensile strength of the material R radius of the cylinder Eddesign value of the modulus of elasticity
f,

1.8 Analysis and design of bearings

35

Compared to Hertz's formula with maxo, =0.418. we find maxo, 1 0 . 4 1 8 . f i . f " = 1,77.fu=oRd .

EN 1337-6 - rocker bearings - gives for the design load Fz,d of a point rocker bearing
5 170. R 2 . (sphere against plane surface) Fz,d
f"
Ed

Compared to Hertz's formula with

we find m a x o , 10.388..1/170.f, = 2,15f, =oRd. For cylindrical rocker bearings the same formulae as for roller bearings are used.

1.8.2 Pin and pin plate for leaf and link bearings A special problem of all leaf and link bearings concerns the design of the pin and the pin plate. Eurocode 3, part 1- 1, gives simple but satisfactory design rules. The design values of the shear force and the bending moment for the pin can be found using the simple model of distributing the force of each pin plate uniformly over the pin.

c ,

In the case of fig. 1.8.2-1 we obtain the shear force and the bending moment according to fig. 1 3.2-2 and fig. 1.8.2-3.

36

1. Bearings

c w
Fig. 1.8.2-2: Shearforce

Fig. 1.8.2-3: Bending moment

For normal bridge bearings we have: c = 0, a = The design values for the resistances are Shear: F,,,
= 0.6. A . f u p /YMp = 0.6.
~

b . 2

d2n
~

.fup /YMp = 0.47 1. df,,

/YMp

The combination of shear and bending has to fulfil the inequality

In this inequality, the central pin plate is controlling. The bearing resistance of plate (thickness t and yield strength f,) and pin is:

F,,,, = 1.5.t . d . f y/YM,

f,, field strength of the pin fUptensile strength of the pin yMp = 1.25 according to EC 3- 1- 1 The bearing capacity of the pin plate at the hole is achieved under one of the following conditions (EC 3- 1 - 1 gives two possibilities):

1.9 Installation of bearings

37

a) Depending on the pin plate thickness t:

t = min (2a, b), e >-FSd Y M p 7 - 2t ' f y
'

e, 2

FSd

'

YMp

2 t . f,

d = e 2 +3 3

b) Depending on the geometry of the pin plate:

1.9

Installation of bearings

Concerning the installation of bearings, the need for a later simple replacement must be taken into account. So it should be common practice to put every bearing between a lower and an upper steel cover plate. These cover plates are anchored or connected both with the substructure and the superstructure. These cover plates are connected to the bearings during the installation but remain fixed to the structure while the bearings are replaced (fig. 1.9-1). Thus, the connection between bearing and cover plates should be constructed in order to allow a simple release. Bolted connections are often used but after many years often the bolts can hardly be unscrewed. According to the author's experience, fastening the bearings with small fillet welds that can be ground off and remade during the replacement process is simpler.

Fig. 1.9-1: Fixing o f a bearing

Generally, bearings should not be built directly on the construction beneath. To guarantee that the area below a bearing is fully sealed a layer of mortar or of a similar product is used. So the height of the bridge at the abutments or piers can be adapted easily and very exactly. It is useful to fix the bearing to the bridge so that there is no clearance at the upper plate and to adjust the bridge by hydraulic jacks. In this situation the

38

1. Bearings

bearings should be adjusted exactly. Thus, the lower plate will get exactly the desired inclination (horizontal or parallel to the gradient, see fig. 1.9-1) and all moveable bearings will have the desired pre-adjustment, which depends on the temperature of the bridge and the expected shrinkage and creep. The installation of the bearings should be done early in the morning when the bridge has a (nearly) constant temperature. The designer has to provide a table with the pre-adjustment of every bearing depending on the measured bridge temperature. For good functioning, careful handling of the bearings during installation is very important. The bearings must be kept free of dirt, mortar, water and dust, especially from all moving parts. Many bearings, such as pot bearings and spherical bearings, are protected against dust by rubber bulges, but others are not protected at all. These have to be cleaned to remove mortar and sand after the installation. The gap between the lower plate of the bearing and the substructure is normally 3 to 5 cm thick and must be completely filled with a mortar bedding. This can be done in different ways: - by a fresh mortar bedding, chambered in the centre where the bearing is set. The excess of mortar will come out on all sides and must be removed. - by a special joint filling mortar which must be mixed in a pan type concrete mixer with a precise quantity of water. This mortar is liquid at first and should be poured in a formwork around the bearing only from one side, so that the air can escape on the other side. The special mortar fills the gap without air bubbles, it sets and hardens very quickly so that after one day the mortar bedding can be fully loaded and the formwork removed. If the gap is less than 1 cm a two-component epoxy resin should be used instead of mortar. Initially this resin is a lighter fluid than mortar, thus completely filling even very small gaps. - by boxing up earth-damp mortar in the gap with a wooden stick also from one side to avoid air bubbles. This method will be difficult for the lower plates with a short side larger than half a metre. All mortars should be non-shrinking.

1.10 Inspection and maintenance

Visual tests of all bearings should be done by qualified personnel at regular intervals. The following properties of the bearings have to be checked: a) sufficient ability to allow movement, taking into account the temperature of the superstructure b) correct positioning of the bearings themselves and of parts of the bearing relative to each other c) uncontrolled movement of the bearing d) fracture, cracks and deformations of parts of the bearings e) cracks in the bedding or in adjacent parts of sub- and superstructure f) condition of the anchorage g) condition of sliding or rolling surfaces h) condition of the anticorrosive protection, against dust, and of the sealings. For the different types of bearings the following checks are of importance:

1.1 1 Replacement of bearings

39

Elastomeric bearings: Displacements and rotations, cracks in the elastomer. Roller and rocker bearings: Displacements and rotations, adjustment of the guiding device, no gap in the contact line. Pot bearings: Sufficient mesh of the lid in the pot, tight sealing of the elastomer in the pot (if the sealing has a defect, the elastomer comes out like a pancake!) Sliding devices - PTFE and stainless steel: Thickness of the PTFE, clean surface of the stainless steel. The result of an inspection should be recorded in a report. EN 1337-10 gives an example for such a report. For maintenance the bearings should be cleaned, lubricated (if necessary and possible) and coated with paint. Small defects should be repaired as far as possible.

1.11 Replacement of bearings

The replacement of bearings is a normal maintenance operation for bridges. Thus, a bridge designer has to provide measures so that a replacement can be carried out easily. The owner of a bridge has to define in the tender if the replacement of the bearings must be carried out under full traffic, restricted traffic or without traffic, depending on the importance of the bridge and the possibility of a traffic ban or a traflk diversion. In case of a replacement under traffic the jacking equipment should allow the same movements as the bearing. To allow rotations the jacks around one bearing should be connected to a single hydraulic circle. That means that the security devices must have a sufficient clearance. Translations are possible by means of additional sliding constructions.

i
_ m _

reinforcement against splitting tension

Fig. 1.I I - I : Stiffened areasf o r hydraulic jacks

To replace a bearing, the bridge has to be lifted by one or more hydraulic jacks. For hydraulic jacks, adequately stiffened areas to transmit the hydraulic jack forces to the sub- and superstructure are required. Concrete parts must be reinforced against splitting tension, steel parts need stiffeners (fig. 1.11-2). Thus, the construction drawings must show in which areas or at which points hydraulic jacks can be set, what are the maximum lifting forces and up to which level the bridge may safely be lifted. This

40

1. Bearings

data is of particular importance if the bridge is supported in a statically indeterminate way at one abutment or pier, in which case the lifting force depends on the height of lift. High stresses can be induced in the cross girder or diaphragm by the lifting device. In such cases it may be necessary to lift the whole cross section uniformly with two or more hydraulic jacks even for exchanging only one bearing. If more than one jack is used the forces can be controlled by hydraulic connection of some or of all jacks: all connected jacks have the same pressure. Hydraulic jacks need some clearance for the installation. For lifting by a few millimetres up to two centimetres flat piston jacks can be used. The following table gives a guide for the required clearances:

Force kN 500 1000 2000 SO00

Required clearance Normal hydraulicjack mm 300 360 450 600

Required clearance Flat piston jack mm 150 180 200 250

Table 1.11-1: Required clearance for hydraulic jacks

There are flat jacks with a height of 80 mm and a lifting force up to SO00 kN. But their stroke is only 20 mm and there is no security device. This kind of jack should be applied for special cases only. New bridges should be constructed for normal hydraulic jacks. In all situations, during the replacement of a bearing the hydraulic jack should be secured by a mechanical device such as an adjusting nut for the piston or lining plates to avoid dropping in case of pipe rupture or rupture of the piston sealing which sometimes can occur (fig.l.11-3 and tig.l.11-2).

!!

pipe or

t-------I
L

----____

c=

Fig. 1.1 1-2: Hydraulic jack with lining plates

1.12 Codes and standards

41

Fig.1.I 1-3: Hydraulic jack with thread and nu1

If the replacement of a bearing takes a long time so that displacements of moveable bearings will occur, the hydraulic jacks have to be equipped with a sliding device, normally PTFE plus a sliding plate of stainless steel. Particular care is required when replacing bearings which transmit horizontal forces: if the friction between the jack and the surface of sub- and superstructure is not sufficient it is necessary to restrain the movement of the bridge by appropriate devices. If the replacement is done under traffic, in most cases, and especially for railway bridges, these devices have to transmit all horizontal forces due to a possible loss of friction.

1.12 Codes and standards

The first attempts to standardize bearings in national codes were made decades ago. In Europe several codes and national standards are available. The best known national standards in Europe on this topic are Germany: DIN 4141 Lager im Bauwesen (structural bearings), Teil 1 bis 14. United Kingdom: BS 5400 Steel, Concrete and Composite Bridges. Section 9.1 Code of Practice for design of bridge bearings Section 9.2 Specification of materials, manufacturing and installation of bridge bearings New European Standards about bearings are the following EN 1337 Structural bearings with the parts EN 1337- 1 General design rules EN 1337-2 Sliding elements EN 1337-3 Elastomeric bearings EN 1337-4 Roller bearings

42 EN EN EN EN EN EN EN 1337-5 1337-6 1337-7 1337-8 1337-9 1337-10 1337-1 1 Pot bearings Rocker bearings Spherical and cylindrical PTFE bearings Guided bearings and Restrained bearings Protection Inspection and maintenance Transport, storage and installation

1. Bearings

A recommendable American Standards about bearings is the following: AASHO-LRFD: American Association of State Highway Officials ( I 994).

1.13 References
Books and special chapters about bearings for bridges: Eggert H., J. Grote, W. Kauschke: Lager im Bauwesen. Verlag von Wilhelm Ernst & Sohn, Berlin, Munchen, Dusseldorf 1974. Lee D.J.: Bridge Bearings and Expansion Joints. Second edition by E & FN Spon, London, Glasgow, New York, Tokyo, Melbourne, Madras 1994. Eggert H., W. Kauschke: Lager im Bauwesen. 2. Auflage, Ernst & Sohn, Berlin 1995. Rahlwes K., R. Maurer: Lagerung und Lager von Bauwerken in: Beton-Kalender 1995, Teil2, Ernst & Sohn, Berlin.

Papers: Albrecht, R.: Zur Anwendung und Berechnung von Gummilagern. Der Deutsche Baumeister 1969, Heft 4, Seite 326, und Heft 6, Seite 563. Andra, Beyer, Wintergerst: Versuche und Erfahrungen mit neuen Kipp- und Gleitlagern. Der Bauingenieur 5 (1962). Andra, W. und Leonhardt, F.: Neue Entwicklungen fur Lager von Bauwerken, Gummi- und Gummitopflager. Die Bautechnik 39 (1969), Heft 2, Seite 37 bis 50. Bayer, K.: Auflager und Fahrbahnubergange fur Hoch- und Bruckenbauten aus Kunststoff. Verein Deutscher Ingenieure VDI im Bildungswerk BV 1956 (Vortragsveroffentlichung). Beyer, E. und Wintergerst, L.: Neue Briickenlager, neue Pfeilerform. Der Bauingenieur 35 (1960), Heft 6, Seite 227 bis 230. Eggert, H.: Briickenlager. Die Bautechnik 50 (1973), S. 143/144. Bub, H.: Das neue Institut fur Bautechnik. Strasse und Autobahn, Band 20 (1 969), Seite 189. Burkhardt, E.: Gepanzerte Betonwalzgelenke, Pendel- und Rollenlager. Die Bautechnik 17 (1939), Seite 230. Cardillo, R. und Kruse, D.: Paper (61/WA-335) ASME (1961). Cichocki, F.: Bremsableitung bei Briicken. Der Bauingenieur 36 (1961), Seite 304 bis 305.

1.13 References

43

Clark, E. und Moutrop, K.: Load Deformation Characteristics of Elastomer Bridge Bearing Pads. University of Rhode Island, May 1962. Desmonsablon, Philippe: Le calcul des piles ddformables avec appuis en caoutchouc. Annales des Ponts et Chaussdes, Paris 4/1960. Eggert, H.: Bauwerksicherheit bei Verwendung von Rollen- und Gleitlagern. Strasse Brucke Tunnel 1971, Heft 3, Seite 71. Eggert, H.: Die baurechtliche Situation bei Lagern fur Briicken und Hochbauten. Der Stahlbau 39 (1970), Heft 6, Seite 189. Einsfeld, U.: Erlauterungen zu den Richtlinien von unbewehrten Elastomerlagern. Mitteilungen Institut fur Bautechnik 6/1972. Franz: Gummilager fur Brucken. VDI-Zeitschrift, Bd. 101/1959, Nr. 12, Seite 47 1 bis 478. Gent, A.: Rubber Bearings for Bridges. Rubber Journal and International Plastics 1959. Grote, J.: Neoprenelager - einige grundsatzliche Erwagungen. Kunststoffe im Bau 7/1968. Grote, J.: Unbewehrte Elastomerlager. Der Bauingenieur 44 (l969), Seite 121. Grote, J.: Vermeidung von Rissen und Dehnungsschaden durch gummielastische Lagerungen. Kunststoffe im Bau 11/1968. Hakenjos, V.: Untersuchungen uber die Rollreibung bei Stahl im elastisch-plastischen Zustand. Technisch-wissenschaftlicheBerichte der Staatlichen Materialpriifungsanstalt an der Technischen Hochschule Stuttgart 1967, Heft 67/05. Heesen: Gepanzerte Betonwalzgelenke, Pendel- und Rollenlager. Die Bautechnik, Jahrgang 25 (1 948), Seite 26 1. Hutten, P.: Beitrag zur Berechnung der Lagerverschiebungen gekrummter, durchlaufender Spannbeton-Balkenbriicken. Dissertation TH Aachen 1970. Jorn, R.: Gummi im Bauwesen. Elastische Lagerung einer Pumpenstation. Der Bauingenieur 36 (1961), Heft 4, Seite 1371138. Keen: Creep of Neoprene in Shear Under Static Conditions, Ten Years, Transactions of the ASME, Juli 1953. Leonhardt und Andra: Stutzungsprobleme der Hochstrassenbriicken. Betonund Stahlbetonbau 55 (1960), Heft 6. Leonhardt, F. und Reimann, H.: Betongelenke, Versuchsbericht, Vorschlage zur Bemessung und konstruktiven Ausbildung. DAfStb, Heft 175. Berlin: Verlag Ernst & Sohn 1966, und Leonhardt, F. und Reimann, H.: Betongelenke. Der Bauingenieur 41 (1966), Seite 49. Leonhardt, F. und Wintergerst, L.: Uber die Brauchbarkeit von Bleigelenken. Beton- und Stahlbetonbau 1961, Heft 5, Seite 123 bis 131. Maguire, C. und Assoc.: Elastomeric Bridge Bearings Pads 1959. Massonnet: Zuschrift zu B. Topaloff, Gummilager fur Briicken. Der Bauingenieur 39 (1964), Seite 428. Monnig, E. und Netzel, D.: Zur Bemessung von Betongelenken. Der Bauingenieur 44 (1969), Seite 433 bis 439. Morton, M.: Rubber Technology. Reinhold Publishing Co. 1959. Mullins, L.: Softening of Rubber by Deformation. Rubber Chemistry and Technology (Feb. 1969).

44

1. Bearings

[351 [361 [371

[431 [441

[491

[531

Nordlin, E., Stoker, S. and Trinble, R.: Laboratory and Field Performance of Elastomeric Bridge Bearing Pads, Highway Research Board (1968). Pare u. Keiner: Elastomeric Bridge Bearings. Highway Research Board Bull 242, 1960. Payne u. Scott: Engineering Design with Rubber Rejcha, C.: Design of Elastomer Bearings. Journal of Prestressed Concrete Institute Oct. 1964, Vol. 9, Nr. 5. Resinger, F.: Langszwangungen - eine Ursache von Bruckenlagerschaden. Der Bauingenieur 46 (1971), Seite 334. Rieckmann, H.-P.: Einfluss der Lagerkonstruktion auf die Knicklange von Pfeilern. Strasse Briicke Tunnel 1970, Seite 36 bis 42 und Seite 270 bis 272. Sasse, H.-R. und Schorn, H.: Bewehrte Elastomerlager - Stand der Entwicklung. Plastik-Konstruktion 1971, Heft 5 , Seite 209 bis 227. Schonhofer: Neugestaltungen auf dem Gebiet des Auflagerbaues und auf verwandten Gebieten. Werner-Verlag, Dusseldorf 1952. Sedyter: Uber die Wirkungsweise von Bleigelenken. Beton und Eisen 1926, Seite 29. Shen, M. K.: Uber die Losung des Balkens mit unverschieblichen Auflagern. Der Bauingenieur 39 (1964), Seite 100. Suess, K. und Grote, J.: Einige Versuche an Neoprenelagern. Der Bauingenieur 38 (1963), Heft 4, Seite 152 bis 157. Thielker, E.: Elastomeric Bearing Pads and Their Application in Structures, Paper 207 of Leap Conference (1964). Thul, H.: Bruckenlager. Der Stahlbau 38 (1969), Seite 353. Topaloff, B.: Gummilager fur Briicken - Berechnung und Anwendung. Der Bauingenieur 39 (1964), Seite 50 bis 64. Topaloff, B.: Gummilager fur Brucken. Beton- und Stahlbetonbau 54 (1959), Heft 9. Uetz, H. und Breckel, H.: Reibungs- und Verschleissversuche mit Teflon. Sonderheft der Staatl. Materialprufungsanstalt an der TH Stuttgart, 7.12.1964, Seite 61/76. Uetz, H. und Hakenjos, V.: Reibungsuntersuchungen mit Polytetrafluorathylen bei hin- und hergehender Bewegung. Die Bautechnik 44 (1967), Heft 5, Seite 159 bis 166. Uetz, H. und Hakenjos, V.: Gleitreibungs- und Gleitverschleissversuche an Kunststoffen. Kunststoffe, 59. Jahrgang 1969, Heft 3, Seite 161 bis 168. Weiprecht, M.: Auflagerung von Briicken. Elsners Taschenbuch fur den Bautechnischen Eisenbahndienst, 1967, Seite 23 1 bis 277, Abschnitt E Bruckenund Ingenieurhochbau. Zies, K.-W.: Stabilitat von Stutzen mit Rollenlagern. Beton- und Stahlbetonbau 65 ( 1 970), Seite 297. AASHO-LRFD: American Association of State Highway Officials (1994). Dupont de Nemours Co.: Design of Neoprene Bridge Bearing Pads, Wilmington ( 1959). CNR-UNI 10018-68 (Italian Standards for rubber bearings).

1.13 References

45

Ministry of Transport: Provisional Rules for the Use of Rubber Bearings in Highway Bridges, Memo. 802, London (1962). Mitteilungen, Institut fur Bautechnik, 1970, Heft 2 und 4, und 1971, Heft 4 und 6. Ohne Verfasser. Auflager aus Teflon. Ausziige aus dem Journal of Teflon 1964, 1965 und 1966, Druckschrift der Du Pont de Nemours International S.A. Geneva, Switzerland. Ohne Verfasser. Bruckenlager. Beratungsstelle fur Stahlverwendung, Dusseldorf, Merkblatt 339,2. Auflage 1968. ORE Office de Recherches et dEssais: Verwendung von Gummi fur Bruckenlager, Frage D 60, Utrecht (1962, 1964, 1965). Wiedemann, L.: Zusatzliche Richtlinien fur Lager im Brucken- und Hochbau. Mitteilungen Institut fur Bautechnik 3/1973, S. 73. Verlag Ernst & Sohn. Eggert: Vorlesungen uber Lager im Bauwesen. Wilhelm Ernst & Sohn 1980/1981. Kauschke, W.: Entwicklungsstand der Gleitlagertechnik fur Briickenbauwerke in der Bundesrepublik Deutschland. Bauingenieur 64 (1989), Seite 109 bis 120. BattermandKohler: Elastomere Federung, Elastische Lagerungen. W. Ernst & Sohn, Berlin, Munchen 1982. Gerb: Schwingungsisolierungen. Berlin, 9. Auflage 1992, Eigenverlag (gegen Schutzgebuhr erhaltlich). Grote, J. und Kreuzinger, H.: Pendelstutzen mit Elastomerlagern. Der Bauingenieur 53 (1978), Seite 63/64. Kanning, W.: Elastomer-Lager fur Pendelstutzen - Einfluss der Lager auf die Beanspruchung der Stutzen. Der Bauingenieur 55 ( 1 980), Seite 455. MauredRahlwes: Lagerung und Lager von Bauwerken. Betonkalender 1995, Ernst & Sohn, Teil 11. Weihermuller, H. und Knoppler, K.: Lagerreibung beim Stabilitatsnachweis von Bruckenpfeilern. Bauingenieur 55 (1980), Seite 285 bis 288. Andra, W.: Der heutige Entwicklungsstand des Topflagers und seine Weiterentwicklung zum Hublager. Bautechnik (1984), Seite 222 bis 230. Eggert, H.: 7 Grundsatze bei der Lagerung von Brucken. 9. IVBH-Kongress Amsterdam 1972, Schlussbericht. Internationale Vereinigung fur Briickenbau und Hochbau, Zurich, Schweiz. Deinhard, J.M., Kordina, K., Mozahn, R., Storkebaum, K.-H.: Der Schadensfall an der Mainbrucke bei Hochheim. Beton - Stahlbetonbau, 72 (1977), Seite 1 bis 7. Eggert, H. und Wiedemann, L.: Nutzungsgerechte Lagerung von Stahl- und Verbundbrucken und unterhaltungsgerechte Konstruktion von Bruckenlagern. IVBH Symposium Dresden 1975. Vorbericht. Eggert, H.: Lager fur Brucken und Hochbauten. Bauingenieur 53 (1978), Seite 161 bis 168, und Zuschrift 54 (1979), Seite 200. Konig, G. et. al.: Spannbeton: Bewahrung im Bruckenbau. Analyse von Bauwerksdaten, Schaden und Erhaltungskosten. Springer-Verlag Berlin, Heidelberg, New York, London, Paris, Tokio 1986.

46

1. Bearings

Pfohl, H.: Reaktionskraft am Festpunkt von Briicken aus Bremslast und Bewegungswiderstanden der Lager. Bauingenieur 58 (1983), Seite 453 bis 457. Eggert, H. und Hakenjos, V.: Die Wirkungsweise von Kalottenlagern. Der Bauingenieur 49 (1974), Heft 3 , Seite 93/94. Lehmann, Dieter: Beitrage zur Berechnung der Elastomerlager. Die Bautechnik I (1978), Seite 19 bis 22, I1 (1978), Seite 99 bis 102, I11 (1978), Seite 190 bis 198, IV (l979), Seite 163 bis 169. Kordina, K. und Nolting, D.: Zur Auflagerung von Stahlbetonteilen mittels unbewehrter Elastomerlager. Der Bauingenieur 56 (1981), Seite 41 bis 44. Kordina, K. und Osterath, H.-H.: Zur Auflagerung von Stahlbetonteilen mittels unbewehrter und bewehrter Elastomerlager. Der Bauingenieur 59 (1 984), Seite 461 bis 466. Kessler, E. und Schwerm, D.: Unebenheiten und Schiefwinkligkeiten der Auflagerflachen fur Elastomerlager bei Stahlbetonfertigteilen. Fertigteilbauforum 13/83, Seite 1 bis 5 (Betonwerk + Fertigteil-Technik). Kessler, E.: Die Anwendung unbewehrter Elastomerlager. Betonwerk + Fertigteil-Technik, Heft 6 (1987), Seite 419 bis 429. Bundesminister fur Verkehr: Schlden an Brucken und anderen Ingenieurbauwerken. Dokumentation 1982. Verkehrsblatt-Verlag, Dortmund. Bundesminister fur Verkehr: Bericht uber Schaden an Bauwerken der Bundesverkehrswege. Januar 1984. Eigenverlag BMV. Beyer, E. und Eisermann, G.: Nachstellbare Bruckenlager. Erfahrungen beim Bauvorhaben Dusseldorf-Hauptbahnhof. Beton 5/1983. Dickerhoff, K.J.: Bemessung von Bruckenlagern unter Gebrauchslast. Dissertation Universitat Karlsruhe 1985. Petersen, Chr.: Zur Beanspruchung moderner Briickenlager. Festschrift J. Scheer, Marz 1987. Hehn, K.-H.: Priifeinrichtung zur Untersuchung von Lagern. VDI-Z 118 (1976), Seite 1 14 bis 118. N.N., Sanierung der Kolnbreinsperre, Projektierung und Ausfuhrung. 1. Auflage Mai 1991. Herausgeber: Osterreichische Donaukraftwerke AG. Hakenjos, V. und Richter, K.: Dauergleitreibungsverhalten der Gleitpaarung PTFE weiss/Austenitischer Stahl fur Lager im Briickenbau. Strasse, Briicke, Tunnel 1 1 (1979, Seite 294 bis 297. Imbimbo M. und Kelly J.M.: Influence of Material Stiffening on Stability of Elastomeric Bearings at Large Displacements. Journal of Engineering Mechanics. Sept. 1998. Zederbaum, J. (1966): The frame action of a bridge deck supported on elastic bearings. Civil Engineering and Public Works Review 61(7 14), 67-72. Leonhardt, F. und Andra, W. (1 960): Stutzprobleme der Hochstrassenbrucken. Beton- und Stahlbetonbau, 55(6), 121-32. Tanaka, R., Natsukawa, K. and Ohira, T. (1984): Thermal behaviour of multispan viaduct in frame. In International Association of Bridge and Structural Engineering, 12th Congress, Vancouver, Canada, 3-7 September. Building Research Establishment (1979) Estimation of thermal and moisture movements and stresses; Part 2, Digest 228, Watford.

1.13 References

47

[96] Emerson M. (1977): Temperature differences in bridges: basis of design requirements. TRRL Laboratory Report 765. Transport and Road Research Laboratory, Crowthorne. [97] Emerson M. (1968): Bridge temperatures and movements in the British Isles. RRL Report LR 228, pp.38. Road Research Laboratory, Crowthorne. [98] Emerson M. (1973): The calculation of the distribution of temperature in bridges. TRRL Report LR 561. Transport and Road Research Laboratory, Crowthorne. [99] Emerson M. (1976): Bridge temperatures estimated from the shade temperature. TRRL Report LR 696. Transport and Road Research Laboratory, Crowthorne. [ 1001 Stephenson, D.A. (1961): Effects of differential temperature on tall slender columns. Concrete and Constructional Engineering, 56(5), 175-8: 56( 1l), 401-3. [ 1011 Garrett, R.J. (1985): The distribution of temperature in bridges. The Journal of the Hong Kong Institution of Engineers, May, 35-8. [ 1021 ComitC Euro-International du BCton (1984). Design manual on structural effects of time-dependent behaviour of concrete (Bulletin No. 142). George Publishing Company. [ 1031 ComitC Euro-International du BCton (1985). Manual of Cracking and Deformations. Bulletin 158E, Lausanne. [ 1041 Neville, A.M., Dilger, W.H. and Brooks, J.J. (1983): Creep of Plain and Structural Concrete. Construction Press, London and New York. [ 1051 Mattock A.H. (1961): Precast-prestressed concrete bridge 5.Creep and shrinkage studies. Journal of the Portland Cement Association Research and Development Laboratories, May. [ I061 Institution of Geological Sciences: National Environmental Research Council (1 976), Atlas of Seismic Activity 1909-1968. Seismological Bulletin No.5. [ 1071 Dollar, A.T.J., Abedi, S.M.H., Lilwall, R.C. und Willmore, R.L. (1975): Earthquake risk in the UK. Proceedings of the Institution of Civil Engineers, 58, 123-4. [ 1081 ICE and SECED (1 985): Earthquake engineering in Britain. Proceedings of Conference of the Institution of Civil Engineers and the Society of Earthquake and Civil Engineering Dynamics, University of East Anglia, April. [ 1091 Lee, D.J. (197 1): The Theory and Practice of Bearings and Expanison Joints for Bridges, Cement and Concrete Association. [ 1 101 Buchler, W. (1987): Design of Pot Bearings, American Concrete Institute Publication, SP-94, V01.2, pp. 882-915. [ 1 1 11 Black, W. (1971): Notes on bridge bearings, RRL Report LR 382, Transport and Road Research Laboratory, Crowthorne. [ I 121 Kauschke, W. and Baignet, M. (1987) Improvements in the Long Term Durability of Bearings in Bridges, American Concrete Institute Publication SP-94, V01.2,577-612. [ 1 131 Taylor, M.E. (1970): PTFE in highway bridges. TRRL Report LR 491, Transport and Road Research Laboratory, Crowthorne. [ 1141 Eggert, H., Kauschke, W.: Lager im Bauwesen, Ernst & Sohn, Berlin 1996.

48

1. Bearings

[ 1 151 Hakenjos, V.: Lager im Bauwesen mit Komponenten aus Kunststoff verdran-

[ 1 161 [ 1 171

[ I 181

[ 1 191

[I201
[ 1211

[ 1221

11231 11241

[ 1251 [ 1261

[ 1271

11281

[ 1291

[ 1301

[ 1311

gen hochbeanspruchbare stahlerne Rollenlager. 13th H.F. Mark-Symposium on 19- 10-94 in Vienna. Marioni, A.: Apparecchi di appoggio per ponti e strutture. ITEC, Milano 1983 Campbell, T. I. and Kong, W. L.: TFE Sliding Surfaces In Bridge Bearings. Report ME-87-06, Ontario Ministry of Transportation and Communications, Downsview, Ontario, 1987. Crozier, W. F., Stoker, J. R., Martin, V. C. and Nordlin, E. F.: A Laboratory Evaluation of Full-Size Elastomeric Bridge Bearing Pads. Research Report CA DOT, TL-6574- 1-74-26, Highway Research Report, June 1979. Gent, A. N.: Elastic Stability of Rubber Compression Springs. ASME, Journal of Mech. Engr. Science, Vol. 6, No. 4, 1964. Jacobsen, F. K. and Taylor R. K.: TFE Expansion Bearings for Highway Bridges. Report No. RDR-3 1, Illinois DOT, June 1971. McEwen, E. E. and Spencer, G. D.: Finite Element Analysis and Experimental Results Concerning Distribution of Stress Under Pot Bearings. Proceedings of 1 st World Congress on Bearings and Sealants, ACI Publication SP-70, Niagara Falls, 198 1 . Nordlin, E. F., Boss, J. F. and Trimble, R. R.: Tetrafluorethylene (TFE) as a Bridge Bearing Material. Research Report, M & R 64642-2, California DOT, Sacramento, CA, June 1970. Roark. R. J. and Young, W. C.: Formulas for Stress and Strain. 5th Ed., McGraw Hill, New York, 1976. Roeder, C. W., Stanton, J. F. and Taylor, A. W.: Performance of Elastomeric Bearings. NCHRP Report 298, TRB, National Research Council, Washington, D. C., October 1987. Roeder, C. W. and Stanton, J. F.: State of the Art Elastomeric Bridge Bearing Design. ACI Journal, 1991. Roeder, C. W., Stanton, J. F. and Feller, T.: Low Temperature Performance of Elastomers. ASCE, Journal of Cold Regions, Vol. 4, No. 3, September 1990, pp 113-132. Roeder, C. W. and Stanton, J. F.: Failure Modes of Elastomeric Bearings and lnfluence of Manufacturing Methods. Proceedings of 2nd World Congress on Bearings and Sealants, ACl Publication SP-94, Vol. 1, San Antonio, Texas, 1986. Roeder, C. W., Stanton, J. F. and Taylor, A. W.: Fatigue of Steel-Reinforced Elastomeric Bearings. ASCE, Journal of Structural Division, Vol. 116, No. 2, February 1990. Roeder, C. W., and Stanton, J. F.: Elastomeric Bearings: A State of the Art. ASCE, Journal of the Structural Division, No. 12, Vol. 109, December 1983. Saxena, A. and McEwen, E. E.: Behaviour of Masonry Bearing Plates in Highway Bridges. Proceedings of 2nd World Congress on Bearings and Sealants, ACI Publication SP-94, San Antonio, 1986. Stanton, J. F. and Roeder, C. W.: Elastomeric Bearings Design, Construction, and Materials. NCHRP Report 248, TRB, National Research Council, Washington, D. C., August 1982.

1.13 References

49

[132] Stanton, J. F., Scroggins, G., Taylor, A. W. and Roeder, C. W.: Stability of Laminated Elastomeric Bearings. ASCE, Journal of Engineering Mechanics, Vol. 116, No. 6, June 1990, pp 1351-1371. [ 1331 Structural Bearing Specification. FHWA Region 3 Structural Committee for Economical Fabrication, Subcommittee for High Load Multi-Rotational Bearings (HLMRB), October 1991.

51

2
2.1

Expansion Joints
Introduction

As mentioned in chapter 1.1, movements in old stone and timber bridges were small and no additional devices were necessary to close the gaps between bridges and abutments due to bridge movements. The first expansion joints were built for steel railway bridges because their movements were not negligible. With the increase of road traffic and of its speed, closing the gaps became necessary for safety reasons, especially at the moveable bearings. Initially, cover plates were used for expansion joints. For longer bridges these cover plates were not sufficient, so that finger joints and sliding plate joints were used. All these types of expansion joints were not watertight and so the water ran down to the bearings and to the abutments. The first watertight expansion joints were built using steel rails between rubber tubes to absorb the movements. This principle led to a lot of different multisealed expansion joints which differed in the means of supporting the steel rails, in the rubber profiles and in controlling the gap widths. Another type of watertight expansion joint is the cushion joint, consisting of a rubber cushion with vulcanised steel plates which transfer the traffic loads. In spite of continuous amendments of all constructions for expansion joints, these still remain wearing parts, especially in bridges with high traffic density and high traffic loads. The following chapters give a short survey of expansion joints for different movements used in the construction of bridges.

2.2

The role of expansion joints

The role of expansion joints is to carry loads and to provide safety to the traffic over the gap between bridge and abutment or between two bridges in a way that all bridge displacements can take place with very low resistance or with no resistance at all. A further requirement is a low noise level especially in an urban environment. The expansion joints should provide a smooth transition from the bridge to the adjacent areas. The replacement of an expansion joint is always combined with a traffic interruption - at least of the affected lane. Therefore expansion joints should be robust and suitable for all loads and local actions under all weather conditions, moisture and deicing agents. The replacement of all wearing parts should be possible in a simple way.

2.3

Calculation of movements of expansion joints

Movements of expansion joints depend on the size of the bridge and the arrangement of the bearings. Normally the form of construction depends on the horizontal translation orthogonal to the joint. But it is necessary to consider all translations and rotations to ensure that the displacements will not reach the limits of the joint construction. To describe the movements of an expansion joint in detail we have to consider three translations and three rotations (fig. 2.3- 1).

52
/ /

2. Expansion Joints

Fig.2.3-1:Possible movements
These movements result from temperature, displacements due to external loads, and creep and shrinkage in concrete and composite bridges. We may obtain the movements (displacements and rotations) from the structural analysis of the system. Movements due to loads depend on the location of the loads. The controlling deformations can be determined with influence lines (fig. 2.3-2 and fig. 2.3-3). The influence line of a deflection is the bending line due to a unit load acting in the direction of the considered movement.
1

.-

Fig.2.3-2:Influence line for a translation

I"
Fig.2.3-3:Influence line for a rotation
To obtain the displacement caused by a rotation it is also possible to calculate the rotations; the displacements can be determined from the known rotations.

2.3.1 Horizontal translation in the direction of the bridge axis u, A change of the environment temperature, creep under normal force and shrinkage lead to a uniform extension or shortening of the bridge (fig. 2.3.1 -1). The thermal expansion coefficients of steel and concrete have approximately the same / K ). A uniform change of temperature about the cross value ( a , = 1,0...1,2. section causes only a horizontal translation of the joint. This applies to composite bridges, too.

2.3 Calculation of movements of expansion joints

53

Fig.2.3.1- I : Uniformly extension or shortening

n

Temperature:

UXt.= UT C
i=l

l i ATi

Creep and shrinkage of concrete bridges

N,,
n

Permanent normal force (compression > 0) Shrinkage coefficient

Shrinkage: u,,., = -EcbW

li
i=l

E,,

A possible problem is the change of the location of the fixing point or the unknown location of the fixing point. On arch bridges the superstructure is usually fixed at the crown of the arch. The fixing point is moved by the deformation of the arch due to the asymmetrical load. Buried expansion joints are often used for short bridges (Chapter 2.4). If the fixing point is situated on longer piers, it acts as a horizontal spring bearing. Due to a movement in the joint a plastic deformation of the asphalt layer occurs and the construction has a certain rigidity. A different rigidity of the expansion joints on the right and left abutment and a possible longitudinal deformation can lead to the cracking of the asphalt layer at one abutment. As the rigidity of this joint is higher than the rigidity of the piers the new fixing point is situated near the undamaged expansion joint (fig. 2.3.1-2).
Cracking of the asphalt layer of the buried expansion joint Fixing point after cracking

Fig.2.3.1-2: Change o f the fixing point

54

2. Expansion Joints

In the case of an elastic fixing point there are additional movements at expansion joints due to acceleration and braking forces. The actual rigidity of piers can differ from the planned rigidity. Moreover, if the bridge is fixed on more than one pier, the position of the fixing point can differ from the planned position.
Creep and shrinkage in composite bridges (acting in the concrete parts of crosssection only) mainly lead to deflections which result in rotations above the y-axis (fig. 2.3.1-4). Creep can be considered using a reduced section area and a reduced moment of inertia, shrinkage by a substitute tensile force Nshacting on the free shrinking concrete. N\,, is a compression force acting on the composite cross-section.

-1

-I-

E,,

Shrinkage coefficient

A, Area of concrete

E, Reduced modulus of elasticity of concrete to consider creep

Fig.2.3.1-3: Equivalent shrinking force

Fig.2.3.1-4: Deflection under load

Horizontal movements of expansion joints can also be caused by vertical movements of the abutments. They are caused by foundation settlements or by replacement of bearings (fig. 2.3.1-5). Statically indeterminate steel and composite bridges can be prestressed by intentional lifting and/or lowering at the bearings.
yr+

-+
u x

positive definition:

cp

2.3 Calculation of movements of expansion joints

'xd 1
(bn

55

Tn

F Y I
Fig.2.3.1-5: Displacement of bearings
UXd1
Uxd n

=4 4 ' (e" +e,>

= $1

'

e, + @" .e ,

If a fixing point is located on a high pier the additional movements due to pier deformation must be considered in the structural analysis. The movements can result from acceleration, braking forces, uniform and non-uniform temperature actions.

2.3.2 Horizontal translation in direction of the cross-section u, A horizontal translation in the crosswise direction results if the angle between the joint and the moving direction of the bearing is not 90 O (e. g. in skew bridges). The magnitude of the movement depends on the magnitude of the movement in the direction of the bridge axis and on this angle (fig. 2.3.2-1 and fig. 2.3.2-2).

u, = sincp. ueff uy = C0Scp~Ueff

Fig.2.3.2-1: Skewed bridge

56

2. Expansion Joints

Fig.2.3.2-2: Skewed bearing conditions

2.3.3 Vertical translation u, Vertical translations u, can be caused by the replacement of bearings (fig. 2.3.3-3) and the geometrical conditions on the abutment (fig. 2.3.3-1 and fig. 2.3.3-2).

u, = u x .tan)

Fig.2.3.3-1: Sloping bridge with horizontal bearings

Fig.2.3.3-2: Bridge with short cantilever on the abutmen2

2.3 Calculation of movements of expansion joints

SN+

57

...............

7 Hydraulic jack

.............

Fig.2.3.3-3: Vertical displacement of bearings (due to bearing replacement)

2.3.4 Rotation around the bridge axis cpx In the case of a replacement of one single bearing at one side a rotation cpx occurs (fig. 2.3.4-1). However, it is possible to avoid this movement by uniform lifting over the cross-section.

Tr- ........

........

Hydraulic jack Fig.2.3.4-1: Lijting on one side

Rotation around the y-axis cpr This deformation is caused by vertical loading and non-uniform temperature. The controlling load positions of the traffic loads can be determined with influence lines.

2 . 3 . 5

Fig.2.3.5-1: Rotation due to deflections 2.3.6 Rotation around the z-axis cpz The deformation cpz is caused by non-uniform temperature action in the horizontal direction, and by wind loads (fig. 2.3.6-1).

58

2. Expansion Joints

'PZ
....~~..........~....

Fig.2.3.6-I: Non-uniform temperature action

2.4

Construction of expansion joints

2.4.1 General The construction of expansion joints has to fulfil the following requirements: - movement capacity - bearing capacity for static and dynamic loading, - watertightness to save bearings, substructure and possible linkage of expansion joints from deterioration, - low noise emission, - traffic safety. To fulfil the last two requirements a limitation of gap widths is essential. Additionally, it is recommended to avoid slopes exceeding about 3 % and vertical steps between joined surfaces exceeding 8 mm (fig. 2.4.1- 1).

Fig.2.4.I-I: Recommended safety requirements

Expansion joints are exposed to pollution. The sealing should not be damaged by inclusions of bigger external bodies. If the gap width is reduced due to a movement of the superstructure the joint must be able to expel grit and silt to the carriageway surface.

2.4 Construction of expansion joints

59

In particular, all elastomeric components must be readily accessible and easily replaceable. 2.4.2 Small movements (up to 25 mm) For movements up to 15 mm it is possible to construct a continuous asphaltic carriageway pavement with a supporting element covering the gap of the superstructure. This kind of joint is also called a buried expansion joint (fig. 2.4.2-1). Up to 10 mm a flat metal plate is sufficient; for movements above 10 mm an elastomeric pad is necessary to avoid pavement cracks at the edges of the supporting plate. An additional reinforcement of the pavement is advisable to provide a uniform strain distribution. The thickness of the pavement should be at least 80 mm and should be equal to the thickness of the corresponding parts of the superstructure and the abutment. To fulfil this requirement the cover of the gap is usually extended into a niche. The asphaltic pavement does not provide sufficient watertightness. An additional sealing is recommended to protect bearings and substructure from deterioration.
Flexible filler

Fig.2.4.2-I: Buried expansion joint

There are covering elements fulfilling the requirements of support, strain distribution and watertightness without additional sealing, e.g. the following kind of joint construction (fig. 2.4.2-2 and fig. 2.4.2-3).
Flexible filler

Fig.2.4.2-2: Buried expansion joint sealed by a rubber profile

60

2. Expansion Joints
Flexible filler Reinforcement

//

Fig.2.4.2-3: Buried expansion joint with continuous sealing and additional rubber projile
For movements between 15 and 25 mm the asphaltic material above the joint can be replaced by a specially modified asphaltic material. Constructions of this kind are called asphaltic plug joints (fig. 2.4.2-4 and fig. 2.4.2-5).The thickness should be at least 80 mm, while the length should not exceed 700 mm. Though movements exceeding 25 mm could be managed in laboratory tests the influence of temperature and of deformation velocity is not known adequately. Incorrect placement of material results in tearing of the adjacent carriageway pavement. Further problems are yielding of asphaltic material under the wheels of standing vehicles, brake and acceleration forces combined with high environment temperatures, and the development of rutting. Because of their low lifetime (though combined with low relative costs) asphaltic plug joints are recommended for temporary purposes.
,Flexible matenal

Fig.2.4.2-4: Asphaltic plug joint

2.4 Construction of expansion joints

61

Fig.2.4.2-5: Asphaltic plug joint additional sealed by a rubber profile

2.4.3 Medium movements (over 25 mm, up to 80 mm) The absorption of medium movements requires an elastic expansion element or an expansion gap across the carriageway surface. For traffic safety, gaps below 5 mm or over 65 mm are not recommended. Thus, the expansion movement of a simple gap construction is limited to 60 mm. Expansion joints for medium movements consist of a sealing element, edge elements, and fixing elements. The sealing element can be replaced by a cushion element that absorbs movements caused by shear deformation (fig. 2.4.3-1).
Sealing element /Edge element ( steel laws

Fixing element

Cushion element

Fig.2.4.3-I: Construction methods of expansionjoints for medium movements

Seals of expansion gaps can be constructed as V-shaped sealing strips (fig. 2.4.3-2) or hollow sections (fig. 2.4.3-4). Movements are absorbed by the folding of these elements. There are special seals for pavements and cyclist areas to decrease the width of the gap to avoid accidents (fig. 2.4.3-3). Traditional cover-plates are prone to rattling and corrosion and hinder the accessibility of possible seals, but they provide the best comfort for pedestrians with high heel shoes (fig. 2.4.3-6).

62

2. Expansion Joints

Fig.2.4.3-2: V-shaped sealing

Fig.2.4.3-3: Special sealing for sidewalks

El

Fig.2.4.3-4: Hollow section

Fig.2.4.3-5: Expansion joint with V-shaped sealing

Fig.2.4.3-6: Expansion joint with cover plate

Fig.2.4.3-7: Expansion joint for sidewalks

2.4 Construction of expansion joints

63

The use of seals made from cellular neoprene extrusion has the advantage of a closed carriageway surface. In addition to the function as sealing, they are able to transfer traffic loads. Movements up to 80 mm can be accommodated (fig. 2.4.3-8).

Fig.2.4.3-8:Seals made from cellular neoprene extrusion

Elastomeric cushion joints (fig. 2.4.3-9) are made from neoprene reinforced with steel plates. Thus, traffic loads can be transferred without significant deflections. The movements are absorbed by increasing and decreasing of the widths of the two gaps on the upper side. The maximum movement is limited by the gap width. The rubber cover of the bearing plate can wear away under traffic or can be damaged (e.g. by snow ploughs) which lowers the skid resistance.

I
I

Fig.2.4.3-9:Elastomeric cushion joint

Especially when using elastomeric cushions and neoprene extrusion seals, the restraining actions can exceed 20 kN/m which in some cases is not negligible.
2.4.4 Large movements (over 80 mm) For large movements, sealing elements and rail elements are coupled. Additionally to the components of a single gap construction, intermediate elements (also called rails), supporting elements and linkage elements are needed (fig. 2.4.4- 1). Linkage elements cause equal gap widths saving the seals from overextending. They must be able to sustain acceleration and braking forces.

64
Intermediate e l e y t S ,,e a n i lg Fixing element element

2. Expansion Joints

Edge,element ( steel jam

Linkage element

Support element

Fig.2.4.4-1: Construction method o f expansion joints for large movements

The following figure shows the coupling of cushion elements. In this case a special linkage mechanism is not necessary due to the high deformation resistance of the single elements which actually act as a spring linkage.

Fig.2.4.4-2: Coupled elastomeric cushion joint

A typical construction is the coupling of V-shaped and hollow section sealing elements. It is called multiple seal expansion joint. These expansion joints can be classified by the kind of supporting and linkage. The folding trellis linkages (fig. 2.4.4-3) satisfy a l l supporting and linkage purposes.

2.4 Construction of expansion joints

65

-7
I

- Fig.2.4.4-3: Rails supported by folding trellis linkage

I I

An additional linkage is needed if the rails are supported by parallel beams. One possibility is the spring linkage (fig. 2.4.4-4). Springs are made of an elastic material. The portion of the resisting force resulting from friction depends on the number of rails and supporting beams whereas the portion of spring force is independent at the number of springs because of the series connection. A disadvantage of this kind of linkage is that acceleration and braking forces cause non-uniform spring deformations. If the gaps are opened near to the maximum value the seals can be overextended. Another possibility of linkage of parallel supporting beams is the use of horizontal parallel linkages (fig. 2.4.4-5).

66

2. Expansion Joints

Fig.2.4.4-4: Rails supported by beams, spacing controlled by springs

Fig.2.4.4-5: Rails supported by beams. Spacing controlled by horizontal parallel linkages

Fig.2.4.4-6: Rails supported by hinged arranged beams (Swivel System)

Fig.2.4.4- 7: Linkage

2.4 Construction of expansion joints

67

If the supporting beams are skew (Swivel System, fig. 2.4.4-6) they control the gap width by means of the kinematic characteristic of the mechanism (fig. 2.4.4-7). The number of supporting beams does not depend on the number of rails. The higher the number of rails the more economical becomes the application of hinged supporting beams. As an alternative to the application of multiple seal expansion joints, special nonwatertight constructions like cantilever-toothed joints or rolling leaf joints (also called roller shutter plate expansion joint) are used. Both the cantilever-toothed joint and the rolling leaf joint are as a rule not watertight, so that an additional drainage system is necessary.
The cantilever-toothed joint (fig. 2.4.4-8), also called finger joint, is a very robust construction but with several disadvantages. The deformation capacity in the crosswise direction is severely limited and vertical deformations of the joint can prejudice traffic safety. To accommodate small vertical deformations without hazard the free finger ends should be rounded. Finger joints with supported fingers (fig. 2.4.4-9) have proved to be not as good as with cantilever fingers. The rolling leaf joint (fig. 2.4.4-10) consists of a tongue plate, a rocker plate, and sliding plates. The acceptable movement depends on the size and number of sliding plates. Rolling leaf joints can exhibit the following disadvantages: - broken hinges (falling shutter plates cause gaps in the motorway), - wear of the bearing surface, - breaking of the restraining spring elements. Some manufactures have carried out important improvements by: - stronger hinges, - use of specially designed bearings for the shutter plates, - stronger restraining elements with elastomeric springs, - rubber seals between the plates (it makes the joint watertight to a great extent).

Fig.2.4.4-8: Cantilever-toothedjoint or finger joint

68
Steel fingers with an elastomeric encasing \

2. Expansion Joints

Fixing Fig.2.4.4-9: Finger joint with supportedfingers

Fig.2.4.4-10: Rolling leaf joint

2.4.5 Expansion joints for railway bridges For the expansion joints for railway bridges it is necessary to consider two elements: - the rails themselves, - the bridge. Nowadays, a continuous track without expansion joints is preferred, due to the comfort of the passengers. Therefore, many modern railway bridges have no expansion devices for the track. Eurocode 1, part 3 (ENV 1991-3), gives rules for the maximum expansion lengths for continuous tracks. The expansion length (i.e. the distance between the thermal centre and the opposite end of the deck) should not exceed 60 m for steel structures with a ballast bed and 90 m for concrete and composite structures, again with a ballast bed. If the expansion length exceeds these values expansion devices should be used.

2.4 Construction of expansion joints

69

Two different types of railway expansion joints are in use by the railway authorities. The first type consists of a parallel joint in the rail and works according to fig. 2.4.5-1.

Fig.2.4.5-I: Paralleljoint in the rail

The second type is normally used for high speed railways (fig. 2.4.5-2). It consists of an ending rail with a slope of 1 : r, with r = 70 to 100, and a tapered rail which is machined in the same slope. This expansion joint has the advantage that there is no gap between the rails and that the wheel load is carried by a full profile section, but the disadvantage that the rail gauge will be widened by the expansion, according to the slope:

AS = A1 .

2 r

E. g., for an expansion of A1 = 200 mm and with r = 70 m the gauge is widened by As = 6 mm. The expansion joints of the rails should not be located directly over the gap because of the rotation angle (py of the bridge. It is better to adjust the expansion joint on the embankment, at a short distance from the bridge.

Fig.2.4.5-2: Feathered joint

70

2. Expansion Joints

For bridges without a ballast bed the gap between the bridge and the abutment normally stays uncovered. For bridges with a ballast bed there are two possibilities: a) to enclose the ballast within the bridge and on the embankment, bridging the gap only by the rails. This construction causes problems to the automatic track ballast tamping machine. b) to build a continuous ballast bed by means of elastic rubber (neoprene) joints or by sliding cover plates. In this case there is no interruption of the ballast bed and no problem for the tamping machine. On the other hand the compactness of the ballast increases and decreases with the expansion in the area of the expansion joint of the rails.

A completely different design philosophy is to install no special expansion devices on bridges with a continuous ballast bed as described above (and, sometimes, with normal bolted rails joints). If high forces caused by temperature-induced expansion or shortening of the bridge occur in the rail, the track will move in the longitudinal direction and will become settled by itself, especially under the vibrations of the passing trains.
The forces between the track and the bridge have to be considered especially for the design of the longitudinally fixed bearing(s) of the bridge. ENV 1991-3 gives the longitudinal action per track FT, = f 8 [kN/m] (LTI-LT*), where LTi are the expansion lengths from the fixed bearing.

2.5

Materials for expansion joints

2.5.1 Steel parts Normally, the supporting members such as edge elements, rails and cross beams are made of mild steel protected by coating or of corrosion-resistant steel. The stirrups of the fixing are curved reinforcing bars. Stainless steel is used for moveable parts like the bolts of a folding trellis linkage and sliding plates connected with PTFE. Members that are difficult to access, e.g. niches for linkage elements, also are made of stainless steel or corrosion-resistant steel. Steel parts embedded in concrete outside of the zone of carbonation, corrosion-resistant steel and stainless steel do not need any protection against corrosion. Parts made of mild steel must be protected. Coatings must have a sufficiently high resistance against mechanical stress, temperature actions, oils, and de-icing salt. The coating should be chosen in accordance with the appropriate national standards. However, a coating consisting of a two-component epoxy priming coat with zinc dust and a twocomponent epoxy final coat with micaceous iron ore is recommended. Steel parts embedded within the zone of carbonation need only a priming coat. A protection against corrosion by means of an elastomeric sheathing is possible if the elastomeric material satisfies the requirements of resistance and durability. In the case of protection by galvanising, hot-dip galvanising is the normal case. Spray galvanising is expensive but also possible.

2.5 Materials for expansion joints

71

2.5.2 Elastomeric parts Elastomeric parts must be resistant to environmental influences, de-icing salt, alkaline and acidic water. They are classified in two categories (load transferring and non-load transferring). Load transferring elements (e.g. cushion elements or elastic bearings of the rails) are made from polychloroprene or from natural caoutchouc. The material must be ageresisting, despite the presence of de-icing salt. Non-load transferring elements (e.g. sealings) are made from polychloroprene or from ethylene-propylene-caoutchoucwith high resistance to tearing and to crack propagation. The thickness should not be below 4 mm. The following table gives the recommended characteristics of applied elastomers.
Characteristic

I Resistance to tearing Tearing strain Resistance to crack propagation Behaviour after a temperature stress (1 4 d; 70 "C) Change of hardness Change of resistance to tearing Change of tearing strain Resistance against potassium chloride (solution: 4 %; 14 d; 23 " C ) Change of volume Change of hardness Resistance against hot asphaltic bitumen (30 minutes; 220 "C) Change of resistance to tearing Change of tearing strain Bond with steel

min. 350 % min. 1 0 N / m m

min. 400 % min. 15 N / m m

max. +7 Shore A max. -20 % max. -20 %

max. +5 Shore A max. -15 % max. -20 %

max. +10 % max. -5 Shore A

max. +10 % max. -5 Shore A

max. -20 % max. -20 %

-

max. -20 % max. -20 % Failure within the elastomeric material

Tah.2.5.2-I : Recommended characteristics qf the elastomeric parts The springs of spring-linked multiple seal expansion joints are made of polyurethane with a high resistance to crack propagation. The material is able to withstand high strains. It can be compressed down to 20 % of the original length. A further advantage is the good damping characteristics.

72

2. Expansion Joints

Asphaltic plug joints are made of a special modified asphaltic material. This must have a sufficient flexibility to absorb the movements of the gap, combined with a sufficient load bearing capacity. The exact composition of the material depends on the producer. However, the binder material usually consists of bitumens modified with plasticizers and polymers. The aggregates, usually, belong to the basalt group.

2.6

Analysis and design of expansion joints

2.6.1 Buried expansion joints and asphaltic plug joints Expansion joints have to satisfy the requirements of ultimate limit state and fatigue strength design. A buried expansion joint or an asphaltic plug joint must only fulfil the construction requirements given in chapter 2.4. The most important rules are: - The thickness of the asphaltic layer should be at least 80 mm. - The asphaltic layer over the supporting construction must have the same thickness as over the superstructure and over the abutment. - The length of asphaltic plug joints shall not exceed 700 mm. Thin cover plates should be verified by a calculation. The spread of the load can be considered by an angle of 45 O (fig. 2.6.1-1).

Fig.2.6.1 -I: Loud spread under a wheel 2.6.2 Single seal and multiple seal expansion joints In most cases the ultimate limit state of a single seal and of a multiple seal expansion joint is analysed correctly, while the fatigue was only considered empirically. However, damage is usually caused by fatigue. Therefore a correct analysis is essential [ 18; 191. The loading acts for a very short time. The probability that the axles of two vehicles are at the expansion joint at the same time is relatively small and only one axle need be considered. As a rule, standards contain a design load of the following type to analyse single members of a bridge.

2.6 Analysis and design of expansion joints

73

q * Rv k

stat

LR Contact length wheel - carriageway surface R Static load

cp Dynamic factor

Fig.2.6.2-I: Design wheel loud

One rail of an expansion joint carries only the portion Fv.k.,tat of the load, depending on the rail width b, the gap width s and the contact length LR(fig. 2.6.2-4).
a,[-I

I no 0 Y5

n yo
0 85 0 80 0 75 n 70 0 65 0 60 n 55 n 50

Yn

80

70

50

40

?n a [ D E G ]

Fig.2.6.2-2: Factor a,

Fig.2.6.2-3: Arrangement of the wheel loads

L,

= L , - ~ S ,
i=l

Fig.2.6.2-4: Loadper rail

2. Expansion Joints

Effective contact length Gap width Number of gaps within the contact length Rail width Portion of wheel load Factor of the influence of the angle between expansion joint and driving direction (fig. 2.6.2-2) If a f 90 O the two wheels of the axle do not cause the maximum loading on the rail at the same time. This fact can be considered by reducing the influence of both wheel loads by the factor a,. Horizontal wheel loads result from rolling friction, acceleration and braking forces, and from the slope of the bridge. Accelerating and braking of a lorry at the expansion joint cause maximum loads but this is a comparatively rare case and, thus, is considered only for the ultimate limit state analysis. Horizontal forces due to rolling friction act at each overrunning and exert an influence on the fatigue of the material.

Ultimate limit state The ultimate limit state is analysed with the single wheel loads of an axle and considering the dynamic factors given in the relevant standards.

The acceleration and braking force are determined from the vertical loading. Edge profiles and their fixing are designed for a horizontal force due to the full wheel load. Intermediate profile:

Fh,k,dyn = 5 . (P,, . Fv.k,stat

Coefficient of static friction of the standard

cpV, ( P ~ Vertical and horizontal dynamic factor Contrary to the fatigue analysis, for ULS verifications a horizontally and vertically fixed continuous girder is a suitable model of the rails. Rails and support beams can be calculated with the E-P or P-P method because actually no yielding occurs due to the high applicable design loads. The ultimate limit state is analysed using the semiprobabilistic safety concept as follows:

2.6 Analysis and design of expansion joints

75

Fatigue design Failure due to fatigue is the main reason for the observed damage. Three types of fatigue fractures have been observed (fig. 2.6.2-5):
1) Failure of the welded joint between rail and support beam 2) Failure of the support beam 3 ) Failure of the rail

Fig.2.6.2-5: Possible cracks due to fatigue

For the fatigue design, the stress range is of interest. At first it is determined by using the loads given in the standards. The horizontal forces due to rolling friction, slope of bridge and acceleration or deceleration must be considered. However, they are smaller than the horizontal force due to acceleration and braking. The factor 5 consists of three parts:

5 = 5s + 5 R -k 6 E

gs gR gE

Factor due to slope Factor due to rolling friction Factor due to locomotive acceleration/deceleration

against driving dmcilon

again~t driving direction

Q
- 5 E [%I

Fig.2.6.2-6: Determination of the factors

tS,kRand cE

accelerat,O"

76

2. Expansion Joints

The vertical load acting on an intermediate or edge profile is Fv.k.star . The horizontal loads are determined as follows: Intermediate profile:

Fh,k.stat = 5. Fv,k,stat

Fig.2.6.2-7: Dynamic loading of a rail

The contact time t, of the wheel depends on the contact length LR,the velocity v and the width of the profile b. b+L, t, =V

T half period). The circular frequency is: The impact load is sine-shaped (t, = -; 2

The impact causes a damped sinusoidal vibration (fig. 2.6.2-8). For the ultimate limit state analysis the response in the fundamental mode of the system is of interest. It is considered by the dynamic value given in the applicable standards. Fatigue of material is caused by the stress range. Normally, only the first and second amplitude of Fv,k,dyn exceed the constant amplitude fatigue limit.

2.6 Analysis and design of expansion joints

77

Fig.2.6.2-8: Dynamic loading and response of system

Fig.2.6.2-9: Dynamic model

The static bending moments in the vertical direction can be determined on the supported continuous beam. It depends on the stiffness of the springs if it has to be taken into account or if the springs can be assumed to be rigid. In the horizontal direction the consideration of the elastic fixing is essential (fig. 2.6.2- 10).

Fig.2.6.2-10: Vertical and horizontal static system

78

2. Expansion Joints

It is important to use the dynamic stiffness of the springs because it differs from the static value. Both the spring stiffness and the damping coefficient are determined by overrun-tests. The frequency fh and the damping coefficient can be determined from the recorded time-deformation curve. The spring stiffness Ch,dyn in the model is varied until the lowest natural frequency according to the experiments is observed. The logarithmic decrement D of the damping coefficient of a spring-linked expansion joint amounts to approximately 10 %. Further possibilities to determine the lowest natural frequency are an analysis by FEM or approximate methods. The following method leads to satisfactory solutions. The fundamental vibration mode shape of the vertical direction can be described by the static bending line of a continuous girder.

A sinusoidal loading causes the following bending deflection curve:

The following formula leads to the stiffness of the spring:

The application of the formulae of the frequency and the rotational frequency leads to the natural frequency of the vertical system: 2= -c (Jj m

With known chdyn and equal span widths the frequency fh of the horizontal direction can be determined in the same way. But the system is an elastically-supported continuous girder. The following figures show some calculated results.

2.6 Analysis and design of expansion joints

fh

79

F
lh=lO@10 m
-8
4

fh

450
400

450

400
350

350
300
250

300

250

1,80 rn
200 150 100 1

2.00 rn

200
150 100

2,20rn

1011

1 2 3 4 5 6 7 8 9 1 0 7 1

c 1 , . ~ ~ - 1[N/ml 0~

Ch,dyn

l o6 /mI

4501
400

L
1,40 m

L
450

1 0m

Ih= 300.108 m>1m45m

, 1 . 6 0

rn

4oo
350

Ih=40@10 m

1 5m

Om
I0 rn I0 m

1,80 rn

2.00 rn

300

2.20 rn
200 150 100

1011

1011

C h.dyn

/mI

C h.dyn 1

o6

[N/m]

Fig.2.6.2- I I : Lowest natural frequencies of an elastically supported continuous girder

m Massofrail [kg/m] I h Moment of inertia [m4] fh Lowest natural frequency [Hz] The dynamic values cp, and cp2 of the first and second modes of the system are added to the value Acp. With an assumed logarithmic damping coefficient of 10 %, the following diagrams give directly the impact factors Acp (fig. 2.6.2-12). Either the first or second figure can be used. They are suitable for the vertical and horizontal direction. L
Ch.dyn

Single span [m] Dynamic stiffness of spring [N/m]

80
Distance of frequency

2. Expansion Joints

The horizontal axis of the diagram (b) contains the natural frequency of the system. This version shows the frequency of resonance as the maximum of the graph of the design velocity. The values Acp of the resonance frequency are comparatively high. Natural system frequencies near the resonance must be avoided at least for the vertical bending. The recommended distance from the resonance frequency is also indicated in the diagram. With a known design velocity a maximum span of the rails can be determined. Longer spans cause higher values Acp, leading to a higher stress range. Another disadvantage is an increasing number of stress cycles exceeding the cut-off limit, which means that more than two modes of the system must be considered. With the values Acpv and AT,, the dynamic difference moments can be calculated.
Mv.k.dyn

= 'Vv

. Mv.k.stat

Mh.k.dyn

= '(Ph

'

Mh.k.stat

The stress range is determined as follows: W V wh The design load of an axle is higher than the actual load. The nominal stresses should be reduced by the factor fred to get the actual design loads. The value of the factor depends on the ratio between design load and loading due to the real traffic situation. The determination of the actual traffic situation requires extensive data for the real loads and their frequency (fig. 2.6.2-13). Infrequent high loads exert an advantageous influence on the fatigue behaviour (overloading effect). The maximum load for fatigue design must be determined considering the real frequency of the actual traffic loads (e.g. there may be load components occurring only in one of a thousand cases). Instead of the nominal stress also the design load could be reduced.
'Ok.rnan.dyn

k .dyn

AMh.k.dyn

2.6 Analysis and design of expansion joints

81

In

0.2

0.4

0.6

0.8

7,O

A 5 i 1 Amax

Fig.2.6.2-13: Example o f a typical loading sequence The stress ranges up to the chosen limit are used to determine a constant amplitude stress range that causes the same damage (fig. 2.6.2-15).
log 4 U R
log Ao

+= 2 MIO

,
No= 5 Mi0
NL= i W MID

D
N = , 2 MIO

log N

No-

5 Mi0

NL=100 MIO Ndarn*

log N

Fig. 2.6.2-14: Fatigue strength curve

Fig. 2.6.2-15: Constant amplitude stress range

This value when compared with the stress range A q m a dyn x provides the factor that allows the fatigue analysis with design loads given in the standards to be used. For in= 0.75 for the conditions of traffic in Germany, stance, [20] recommends the factor fred to be applied to the loads of German Standard DIN 1072. A maximum stress determined in this way is exceeded in only one of a thousand cases. The fatigue design has to fulfil the following equation:

YFt

Partial safety factor of the fatigue loading (yFf= I .O) Partial safety factor of fatigue strength (yMf = 1.15)

?/Mf

A o ~ 1110 , = a,,,,,. A

~ Constant . ~ ~amplitude ~ ~ stress ~ ~ range for 100 million cycles

82

2. Expansion Joints

Can be ascertained by the analyses of the real sequence using the Palmgren-Miner summation (aloe = 0.4).
AOL

Fatigue strength for 100 million cycles

The construction members of the expansion joint are three-dimensional and compact. The fatigue strength A q can be taken from the standard used if it contains a suitable detail category, otherwise tests become necessary. The following testing arrangements were recently used with success (fig. 2.6.2-16). The required number of tests is normally indicated by the standards.

Fig.2.6.216: Recommended arrangement of the tests

The lifetime of a construction can be calculated as a statistical value. It is only applicable for the evaluation of that type of construction.

Ndarnage

365. DTLV . DAAL . p Yd Ndamage DTLV DAAL = 4 P Design life - time in years The number of cycles exceeding the cut-off limit The average of daily lorry traffic in one direction The average number of axles of each lorry The distribution of the DTLV on several lanes p = 1.O in case of one lane p = 0.85 in case of two lanes p = 0.80 in case of three or more lanes

2.6.3 Elastomeric cushionjoint The loads for the ultimate limit state analysis and the reduced loads for the fatigue analysis are determined in the same way as for the seal expansions joints. In the vertical direction the analysed element transfers a portion of the wheel load, depending on the zone of influence. Horizontal loads are determined from the vertical loads using the factor 5.
Intermediate profile: Edge profile:
Fh.k.stat

Fh.k.stat

Fv.k.~tat

5.

Rv.k.stat

2.6 Analysis and design of expansion joints

83

The horizontal loading of edge profiles and their fixings are analysed considering the complete wheel load. Edge profiles and fixings can be analysed in the same way as for multiple seal joints. A possible intermediate profile can be treated as a single span beam (fig. 2.6.3-1).

,213 1

,113

,I13
4 4

213

Fig.2.6.3-1: Calculation of the intermediate profile

The elastomeric parts of elastomeric cushion joints have to withstand stresses and stress ranges due to traffic loads. Their strength can be ascertained by tests. The following testing arrangement is recommended.

Fig.2.6.3-2: Recommended arrangement of the test

The specimen is of the same character as the planned construction and has a length of at least 1200 mm. The loads are applied through an elastomeric disk of 50 mm thickness which is situated in the middle of the cushion element. L R and B R are the dimensions of the load area according to the applicable standard. If the width of sample is smaller than LR, only a reduced load acts on the joint construction. It can be considered by a smaller disk and a force than P. The inclination of P depends on the factor 5. It considers the sliding friction or the roller friction, the slope of the bridge and the locomotives acceleration and is different for the ultimate limit and fatigue tests. The applied force P has the following value for the ultimate limit test:

84

2. Expansion Joints

Fvk

5tdf

Wheel load of the standard

For the fatigue test the loads are reduced by the factor fred.
'red

= fie, P

The construction is applicable if experiments prove that the full load P can be supported as a static load, the reduced load Pred for 2 millions of cycles.
2.6.4 Cantilever-toothedjoint and rolling leaf joint The Bernoulli-Euler theory of bending gives correct results provided that the height to length ratio of a beam is at least 1/5. Fingers of cantilever-toothed joints are often not within this range. If this requirement is satisfied the ultimate load can be calculated easily. Otherwise tests become essential. The fatigue behaviour must be determined by tests anyway because of the three dimensional character of the connection cantilever / edge element. The testing arrangement and the applied loads are the same as for cushion joints (fig. 2.6.4-1). Maximum stresses are caused when the joint expansion is maximum.

Fig.2.6.41 : Recommended arrangement of the test

The behaviour of a rolling leaf joint should be checked in the same way. In most cases neither the application of the Bernoulli-Euler theory of bending is possible nor do the standards contain suitable detail categories for the fatigue design. The loads must be placed in the most disadvantageous position.

2.7

Installation of expansion joints

The design of an expansion joint is performed by determination of the extreme values of the expected movements and the position of installation. The installation data depends on the planned construction sequence. The expansion joint is adjusted by means of an auxiliary construction. For a spring linkage prestressing is necessary (fig. 2.7-1). It is recommended to instal the expansion joint in the early morning when the temperature is distributed almost uniformly over the whole bridge.

2.7 Installation of expansion joints

85

Immediately before the installation the actual temperature of the bridge is measured. If it is not within the considered tolerance the adjustment must be corrected. After that the expansion joint is flushed and fixed temporarily. In the case of a steel bridge it is provisionally bolted or tack-welded. The auxiliary construction must be removed immediately. After carrying out the final fixing, the protection against corrosion is completed. In concrete bridges the expansion joints are provisionally fixed by welding together reinforcement and anchoring. The concrete pour should be at least of the same strength as the adjacent material of the superstructure. While pouring the concrete the joint construction should be protected by a cover.

......
Adjustable auxiliary construction

;a/

Reinforcement

Reinfdcement

Fig.2.7-1: Possible auxiliary constructionfor the installation

In the case of a steel bridge the date of installing the expansion joints has no influence on the expected range of movement. In the case of a concrete bridge or a composite bridge, single unidirectional movements (shortening due to creep and shrinkage) occur. These movements begin with the erecting of the construction and stop within some weeks / months / years. Creep is caused by compressive stresses, especially due to prestressing. The movement due to prestessing forces occurs during the prestressing work. The joint construction has to accommodate the movements which occur after the installation. Therefore, the dimension and, by this, the costs of a joint construction can be reduced by a late installation. The variation of creep and shrinkage is shown in the following figures by means of the coefficient of creep cp(-,t,) and the shrinkage value E,, . In various standards, t = 5 years (= 1800 days) to t = 20 years is set equal t o t = m.

86

2. Expansion Joints

500

1000

1500

2000

500

1000

1500

2000

Time [dl

Time Id1

Fig.2.7-3: Variation o f creep

Fig.2.7-4: Variation o f shrinkage

The maximum increments of shrinkage and creep occur immediately after completion or after prestressing. For example after 100 days (about 3 months), about 50 % of the expected creep deformations and 25 % of the shrinkage deformation have taken place.

2.8

Inspection and maintenance

Expansion joints should be checked regularly by means of visual inspection. The frequency depends on the sensitivity of the construction. Before the inspection the joint is cleaned, and cover-plates may need to be removed. The check should involve the following items: - Damage of the anticorrosive protection. This should be repaired before advanced rust formations appear. The new coating must be compatible with the existing one. - Visible cracks due to fatigue in the steel members. - Damages to the seals. The soiled water of the carriageway can lead to the deterioration and corrosion of the bearings, the substructure and possible the linkages. - Workability of the linkage. If it does not fulfil its function, damage of the seals may result. - Obstruction or damage of the drainage system. The adjacent carriageway pavement should also be checked. A jutting joint construction due to wheelers enhances the impact loading. If it is not possible to repair the entire pavement, asphalt ramps should be erected to protect the joints. Service-free expansion joints are often demanded by the manufacturers. Nevertheless, it is recommended to clean the gaps from grit and silt to protect seals and linkage. The drainage should also be cleaned regularly.

2.9 Replacement of expansion joints

87

2.9

Replacement of expansion joints

The lifetime of an expansion joint should be the same as the lifetime of the carriageway pavement. A complete replacement becomes necessary if the steel parts exhibit advanced fatigue damage. On steel bridges only the bolted or welded connections are removed. A replacement on concrete bridges is more expensive. More frequent is the replacement of single members, especially of the elastomer components. Seals should be replaceable from the carriageway site. Manufacturers offer different systems for easy replacement (fig. 2.9-1).
Edge or intermediate beam Grooved dowel Edge or intermediate beam

Clamping strip

,-/'

ic

Polychloroprene sealing element

Polychloroprene sealing element

Fig.2.9-I: Possiblefixings to the seal

The gap width must be opened to at least 25 mm. In the case of an elastic linkage, smaller widths are possible because the rails can be displaced. On the other hand the seals must not be stretched fully. Expansion joints for large movements should be accessible from the underside to change members of the linkage like elastomeric springs. In the case of a road with several lanes it is desirable to change the seals of the expansion joint in sections. It is possible to join the seals by vulcanization on site. If a replacement of the rails becomes necessary they can also be joined on site. However, the joints should be situated in zones with minimal stress range and must be welded very carefully because of the high fatigue loads.

88

2. Expansion Joints

2.10 References
Books about expansion joints for bridges: Lee D.J.: Bridge Bearings and Expansion Joints. Second edition by E & FN Spon, London, Glasgow, New York, Tokyo, Melbourne, Madras 1994.

Papers: Price, A.R. (1982): The service performance of fifty buried type expansionjoints. TRRL Report SR 740, Transport and Road Research Laboratory, Crowthorne. Price, A.R. ( I 983): The performance of nosing type bridge deck expansion joints. TRRL Report LR 1071, Transport and Road Research Laboratory Crowthorne. Price, A.R. (1984): The performance in service of bridge expansionjoints. TRRL Report LR 1 104, Transport and Road Research Laboratory, Crowthorne. Department of Transport (1989): Expansion joints for use in highway bridge decks. Departmental Standard BD 33/88. Department of Transport (1989): Expansion joints for use in highway bridge decks. Departmental Advice Note BA 26/88. Koster W. (1969): Expansion Joints in Bridges and Concrete Roads. Maclaren and Sons. Busch, G.A. (1986): A review of design practice and performance of fingerjoints. Paper presented to the 2nd World Congress on Joint Sealing and Bearing Systems for Concrete Structures, San Antonio, Texas, September. Watson, S.C. (1972):A review of past performance and some new considerations in the bridge expansion joint scene. Paper presented to regional meetings of the AASHO Committee on Bridges and Structures, Spring. Koster W. (1986): The principle of elasticity for expansion joints. Paper presented to 2nd World Congress on Joint Sealing and Bearing Systems for Concrete Structures, San Antonio, Texas, September. [ 101 Lee, D.J. (1971): The Theory and Practice of Bearings and Expansion Joints for Bridges, Cement and Concrete Association. [ I I ] Demers, C.E. and Fisher, J.W., Fatigue Cracking of Steel Bridge Structures, Volume I : A Survey of Localized Cracking in Steel Bridges - 1981 to 1988, FHWA Publication No. FHWA-RD-89- 166, McLean, VA, 1990 [ 121 Standard Specifications For Highway Bridges. 15th edition, American Association of State Highway and Transportation Officials, Washington, D.C., 1992 [ 131 Tschemmernegg, F., The Design of Modular Expansion Joints, Proceedings of the 3rd World Congress on Joint Sealing and Bearing Systems for Concrete Structures, Toronto, 1991. [14] Dexter, R.J., Kaczinski, M.R., and Fisher, J.W.; Fatigue Testing of Modular Expansion Joints for Bridges, Proceeding of the 1995 IABSE Symposium, Volume 7312, San Francisco, CA, 1995. [ 151 TL/TP-FU 92, Technische Liefer- und Priifvorschriften fur wasserundurchlassige Fahrbahnubergange von Strassen- und Wegbriicken. Bonn: Bundesministerium fur Verkehr, Ausg. 1992

2.10 References

89

- RVS 15.45, Briickenausriistung - Ubergangskonstruktion. Wien: Forschungsgesellschaft fur das Verkehrs- und Strassenwesen, Arbeitsgruppe c<Briickenbau>>, Arbeitsausschuss <cBruckenausriistung>>, Ausg. Januar 1995. [ 171 Braun, Chr.: Verkehrslastbeanspruchungvon Ubergangskonstruktionen in Strassenbriicken. Bauingenieur 67 (l992), P. 229-237. [ 181 Tschemmernegg, F. (a.0.): Ermudungsnachweis von Fahrbahnubergangen nach ENV-1993-1. Stahlbau (1995), P. 202-210. [ 191 Pattis, A.: Dynamische Bemessung von wasserdichten FahrbahnubergangenModulsysteme (Dynamic Design of Waterproof Modular Expansion Joints). Ph.D. dissertation. Department of Civil Engineering and Architecture, University of Innsbruck, Austria (Dec. 1993). [20] Herleitung eines Lastmodells fur den Betriebsfestigkeitsnachweis von StraBenbrucken. Forschung Strassenbau und Strassenverkehrstechnik Heft 430, 1984. [2 11 Ramberger, G.: Bearings, expansion joints and hydraulic equipment for bridges, IABSE, IS. Kongress-Bericht Copenhagen, 1996. [22] Fisher, J.W., Kaczinski, M.R. and Dexter, R.J.. Field and Laboratory Experience with Expansion Joints. IABSE, 15. Kongress-Bericht Copenhagen, 1996. [23] Braun, C.: The Design of Modular Joints for Movements up to 2000 mm. IABSE, 15. Kongress-Bericht Copenhagen, 1996. [24] Nielsen, H.B.: The Storebaelt West Bridge. Railway Expansion Joints. IABSE, 15. Kongress-Bericht Copenhagen, 1996. [25] Crocetti, Roberto: Modular Bridge Expansion Joints - Loads, Dynamic Behaviour and Fatigue Performance. Thesis for the degree of Licentiate of Engineering. Department of Structural Engineering, Division of Steel and Timber Structures. Chalmers University of Technology, 1998. [26] Barnard, C.P., Cuninghame, J.R.: Practical guide to the use of bridge expansion joints. Application guide 29, Transport research laboratory, UK 1997.
[ 161 Richtlinie