Structural bearings and expansion joints for bridges ( PDFDrive )

Table of Contents 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

Structural Engineering Documents

Gunter Ramberger

Structural Bearings

for Bridges

International Association for Bridge and Structural Engineering

Association lnternationale des Ponts et Charpentes

lnternationale Vereinigung fur Bruckenbau und Hochbau

IABSE AIPC IVBH

Copyright 0 2002 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

Table of Contents

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

2.2 The role of expansion joints

2.3 Calculation of movements of expansion joints

2.4 Construction of expansion joints

2.5 Materials for expansion joints

2.6 Analysis and design of expansion joints

2.7 Installation of expansion joints

2.8 Inspection and maintenance

2.9 Replacement of expansion joints

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 Minneso- ta; 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 Engi- neering Document

1 Bearings

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 of- ten 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 develop- ment of ageing-, ozone- and UV-radiation-resistant elastomers and plastics, new ma- terials 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 compres- sion 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

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

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 com- bined

-

Symbol Function Construction

Point rocker bearing Pot bearing; Fixed elastomeric bearing;

Free spherical sliding bearing; Link bearing with universal joints (tension and compression)

Constant line rocker sliding bearing Free rocker sliding bearing; Free roller bearing; Free link bearing

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 fol- lowing: 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 re- strained More than three vertical supports of the superstructure lead to statically in- determinate 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 sec- tions) 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 distri- bute 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

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)

I

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

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

Case a) has the advantage that only vertical bearing reactions and no permanent hori- zontal 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 super- structure 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

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 gird- ers 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 dif- ferences cause horizontal restraint forces, provided that bearings have no clearance on the abutments

The author prefers the statically determinate system with only one lengthwise re- strained bearing at the abutment concerned because the actual clearance of a bearing

is not determinable in reality (fig 1 ‘4.2-2)

Fig 1.4.2-1: “Classical” layout

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

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 reac- tion force For all bearing constructions it is easier to transfer horizontal forces in com- bination 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 sub- and superstructure (fig 1.4.2-4)

I

I

Fig 1.4.2-4: Inclination of the resultant force

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

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 bear- ing 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 tem-

pe ra tu re)

Fig 1.4.2-9: Geometrical situation

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

It is important to note that the layout of the bearings has a great influence on the struc- tural system The above mentioned arrangements of bearings are typical for average bridges The following examples show some special effects which have to be consid- ered for the design of bridges and bearings These examples do not lay claim to com- pleteness

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

Special bearing conditions, advice etc

16 I Bearings

I

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 f o r 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

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

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

- 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, es- pecially 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 rail- ways, are not

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

Fig 1.5.2-2: ZnJuence area for 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 de- fined 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 consi- dered

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 con- sist 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 di- rections by elastic displacements or rotations Under vertical loads the elastic block

bulges, leading to vertical displacements A solution to this problem was found by re-

inforcing 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 su- perstructure 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 ap- propriate 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)

I

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 sur- face 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

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 188 1, the physicist Heinrich Hertz found the solution of this problem: caused by the elastic deformation the theo- retical 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

(fig 1.6.2-1 ; fig.1.6.2-4)

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

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

These bearings were invented in the 1950s They combine the two desirable proper- ties: 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 unrein- forced 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 struc- tural steel and lubrication grease PTFE is a plastic with high mechanical and chemi- cal 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-

1 Bearings

24

tomeric pad should not exceed 4.0 kN/cm2 The allowable compression for the PTFE

is 3 kN/cm2 for permanent loads and 4.5 kN/cm2 for short term loads (traffic, wind etc.) Pot bearings have the advantage of a very high vertical stiffness (nearly incompres- sible 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

Fig.1.6.3-1: Function of a pot bearing

astomere disc Lid

Sealing Pot - wall Pot - bottom

Fig 1.6.3-2: Fixed pot bearing

Fig 1.6.3-3: Pot bearing constraint in one direction

Fig.1.6.3-4: Members of a pot bearing

Anchoring plate Sliding plate Polished stainless steel PTFE (Polytetrafluorethylen) Lid

Pot -wall Sealing Elastomere disc Pot - bottom

Fig.1.6.3-5: Free pot bearing

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 dis- placements 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 displace- ments 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

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

or one upper leaf with foot plate and pin holes, connected by a pin Leaf bearings al- low 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,

, with R equal to the distance between the

in the perpendicular direction 6, =

2 R 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 func- tioning 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 con- structed by additional sliding elements and (if necessary) guiding systems

base plate

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

Next Page

1.7 Materials for bearings 29

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 101 13

- Quenched and tempered steels according to EN I0082

Eurocode 3 may be used for the design of all bearing components made from struc- tural steel according to EN 10025 and EN 101 13 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

So we have a Poisson’s ratio v = 0.5 and a Young’s modulus of elasticity E =

2 ( 1 +v) G 3 G The fracture strain of rubber lies between 250 % and 500 % Rub- ber 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 approxi- mation for E, co,npr is given by

with o = -, F: compression force

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 same as under compression, but with a + ~ a we obtain E, hendlng = el 1 Lo,npr Actually,

the maximum (3 is not in the middle of one half but nearer the outer side; thus we finally obtain: a + < - , E l c o m p r This is described very well by the following approximate formula:

Fig 1.7.2-3: Rotution - restraining moment

Fig 1.7.2-4: Displacement - restruining~forces

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 pres- sure 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 sur- face, 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-

on the orientation of the large polymer molecules; during movement they are orientat-

ed 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 val-

ues of the friction coefficient pLd between PTFE and stainless steel, depending on the compression force (EN 1337-2)

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)

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 move- ments but they occur at high velocity Wearing is mostly caused by the second case

For the stainless steel plate, austenitic steel X6CrNiMo17122 according to EU- RONORM 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

For the design of bearings the following problems should be addressed: compression between two spherical bodies, compression between a spherical and a flat body, com- pression between two cylindrical bodies, compression between a cylindrical and a flat body along a generator line As already mentioned, Heinrich Hertz obtained the solu- tion 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

Hertz found the following maximum compression stresses max (T and widths b on the contact areas:

r,, r2 radii of the bodies in contact

V

max (3

b

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

Ov = d0i2 + O2 + Oj3 - (3~(32 - 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

- O3Oi and yielding begins when

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

f,

R radius of the cylinder

tensile strength of the material

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

(sphere against plane surface) Fz,d 5 170 R 2 f"

Ed Compared to Hertz's formula with

Pin and pin plate for leaf and link bearings

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

Fig 1.8.2-2: Shear force

Fig 1.8.2-3: Bending moment

For normal bridge bearings we have: c = 0, a = ~

The design values for the resistances are

b

2

d 2 n

4 Shear: F,,, = 0.6 A f u p / Y M p = 0.6 ~ fup / Y M p = 0.47 1 d’f,, / Y M p

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

fUp tensile 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):

a) Depending on the pin plate thickness t:

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 of a bearing

Generally, bearings should not be built directly on the construction beneath To guar- antee that the area below a bearing is fully sealed a layer of mortar or of a similar prod- uct 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 clear- ance 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 bear- ings 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 im- portant 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 pro- tected 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 dif- ferent 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 hard- ens 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

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 su-

b) correct positioning of the bearings themselves and of parts of the bearing relative to

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:

perstructure

each other

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 ex- ample for such a report

For maintenance the bearings should be cleaned, lubricated (if necessary and pos- sible) and coated with paint Small defects should be repaired as far as possible

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 bear- ings must be carried out under full traffic, restricted traffic or without traffic, depend- ing 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 con-

reinforcement against splitting tension

Fig 1 I I - I : Stiffened areas f o r hydraulic jacks

To replace a bearing, the bridge has to be lifted by one or more hydraulic jacks For hy- draulic jacks, adequately stiffened areas to transmit the hydraulic jack forces to the sub- and superstructure are required Concrete parts must be reinforced against split- ting 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

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:

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 ap- plied 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 se- cured 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 some- times can occur (fig.l.11-3 and tig.l.11-2)

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 suffi- cient 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

United Kingdom: BS 5400

Teil 1 bis 14

Steel, Concrete and Composite Bridges

Section 9.1 Code of Practice for design of bridge bearings Section 9.2 Specification of materials, manufacturing and installa-

tion 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

EN 1337- 10 Inspection and maintenance

EN 1337- 1 1 Transport, storage and installation

Spherical and cylindrical PTFE bearings

Guided bearings and Restrained bearings

A recommendable American Standards about bearings is the following:

AASHO-LRFD: American Association of State Highway Officials ( I 994)

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

Beyer, E und Wintergerst, L.: Neue Briickenlager, neue Pfeilerform Der Bau- ingenieur 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