MARCO ROSIGNOLIromeo piremeranen Bridge Launching Marco Rosignoli Parma, Italy — “L! ThomasTelfordPublished by Thomas Telford Publishing, Thomas Telford Ltd, 1 Heron Quay, London El4 45D. URL: www.thomastelford.com Distributors for Thomas Telford books are USA: ASCE Press, 1801 Alexander Bell Drive, Reston, VA 20191-4400, USA Japan: Maruzen Co. Ltd, Book Department, 3-10 Nihonbashi 2-chome, Chuo-ku, Tokyo 103 “Australia: DA Books and Journals, 648 Whitchorse Road, Mitcham 3132, Victoria First published 2002 ‘Also available from Thomas Telford Books Cable stayed bridges. R. Walther, B. Houriet, W. Isler, P. Mota and J-R, Klein. 2nd edition, 1999. ISBN 0.7271 2773 7. Current and future trends in bridge design, construction and maintenance 2. ICE, Highways Agency and International Association for Bridge Maintenance and Safety, 2002. ISBN 0 7277 3091 6. Integral bridges. G. L. England, N. C. M. Tsang and D. 1, Bush, 2000. ISBN 0 7277 2845 8. Management of highway structures. P. C. Das, 1999. ISBN 0 7277 2775 3, ‘Manual of bridge engineering. M. J. Ryall, G. A. R. Parke and J. E. Harding (eds), 2000. ISBN 0 7277 2774 5. Post-tensioned concrete bridges. Highways Agency, Service D’Etudes Techniques des Routes et Autoroute, Transport Research Laboratory and Laboratoire Central Des Ponts et Chaussées, 1999. ISBN 0 7277 2760 5. Prototype bridge struciures. M. Y. H. Bangash, 1999. ISBN 0 7277 2778 8 ‘Steel bridge strengthening. Highways Agency and W. S. Atkins, 2001. ISBN 0 7277 2881 4. Cover photo: the Steine Bridge (courtesy Greisch) ‘A catalogue record for this book is available from the British Library ISBN: 0 7277 3146 7 © Thomas Telford Limited 2002 All rights, including translation, reserved. Except as permitted by the Copyright, Designs and Patents Act 1988, no part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying or otherwise, without the prior written permission of the Publishing Director, Thomas Telford Publishing, Thomas Telford Ltd, 1 Heron Quay, London El4 4D. This book is published on the understanding that the author is solely responsible for the statements made and opinions expressed in it and that its publication does not necessarily imply that such statements andor opinions are or reflect the views or opinions of the publishers. While every effort has been made to ensure that the statements made and the opinions expressed in this publication provide a safe and accurate guide, no liability or responsibility can be accepted in this respect by the author or publishers, Typeset by Keyword Publishing Services Printed and bound in Great Britain by MPG Books, Bodmin Contents Preface List of symbols 1. Introduction 1.1. Introduction to incremental launching 1.2. Incremental launching of PC bridges 1.3. Incremental launching of composite bridges 1.4. Incremental launching of prestressed composite bridges Incremental launching of bridges 2.1. Structural system of the superstructure 2.2. Static system and seismic design 2.3. Geometric constraints 2.4. Launching techniques 2.4.1. Launching of light superstructures 2.42. Friction launching of heavy superstructures 2.5. Launch bearings and guide devices 2.6. Launch and lock forces 2.7. Correction of launch stresses 2.7.1. Launching nose 2.7.2. Stayed front system 2.7.3. Temporary piers 2.7.4. Deck launching onto arches 2.8. Launch stresses in the piers 2.9. RTM method for analysis of the continuous beam 2.10. Shear and bending moment envelopes 2.11. Verification of design assumptions by monitoring 2.11.1. Blastic modulus of concrete 2.11.2. Thermal correction of results 10 a7 15 16 19 19 27 29 32 33 43 53 55 58 88 94 97 103 110 118 120 aaa 123vi BRIDGE LAUNCHING 3. PC bridges one 3.2. 3.3. 3.4. 3.5. aan 3.7. aaae 2.11.3. Coefficient of thermal expansion of concrete 2.11.4. Shrinkage Presizing of the superstructure Deck segmentation and yard organization Casting phases 3.3.1. Monolithic casting 3.3.2. Two-phase casting in a single formwork 3.3.3. Two-phase casting in a double formwork 3.3.4. Segment extraction 3.3.5. Curing yard Assembly and launching of precast segments Launch bearings and guide devices 3.5.1. Pier-cap arrangement 3.5.2. Effects of vertical misalignment of launch bearings 3.5.3. Local stresses above launching bearings Effects of time-dependent behaviour of conerete 3.6.1. Shrinkage 3.6.2. Creep 3.6.3. Prestressing steel relaxation 3.6.4. Effects on bridge launching Prestressing 3.7.1. Launch prestressing 3.7.2. Dimensioning of launch prestressing 3.7.3. Service prestressing Reinforcement 4. Composite bridges ae 4.5. Conceptual design and deck presizing 4.1.1. Multigirder systems 4.1.2. Two-girder systems 4.1.3. Composite box girders Segmentation of the steel girder Assembly yard organization Launching of the steel girder 4.4.1, Launching bearings 4.4.2. Launching nose 4.4.3. General launch stresses and instability 4.4.4. Local launch stresses and instability In situ casting of the concrete slab 4.5.1. Casting methods and cracking control 4.5.2. Longitudinal prestressing of the concrete slab 4.5.3. Transverse prestressing of the conerete slab 123 123 125 uae 128 131 ae 141 142 146 149 151 153 155 156 165 170 iar 172 173 174 178 179 190 194 206 209 210 ne vies 217 ane ae 229 231 235 236 239 246 247 254 257 CONTENTS vii 4.6. Incremental launching of the concrete slab onto the steel beam 257 4.6.1. Technological features 258 Advantages with respect to in situ casting 265 Stress analysis 267 5. Prestressed composite bridges 271 5.1. Introduction 271 High performance concrete 272 Lightweight concrete 275 5.1.3. Introduction to prestressed composite bridges 276 5.1.4. Cross-sectional efficiency 279 5.2. PCS bridges with stiffened-plate webs 281 5.2.1. | Web-slab lower node 283 5.2.2. Web instability 284 5.2.3. Launching 286 5.3. PCS bridges with corrugated webs 288 5.3.1. State of stress in the corrugated webs 291 5.3.2. Cross-sectional deformability 294 Connection between webs and slabs. 296 Web instability 299 Fabrication of corrugated plate girders 305 Field activities 307 53.7. Launching 308 5.3.8. Case studies 311 References and bibliography 319 Index 333Preface In recent decades, the availability of new means of calculating and of new materials has permitted a rapid evolution of bridge design and construction methods, under pressure from the competitive requirements of the international construction industry. The increase in labour costs has played a fundamental role in the indus- trialization of bridge construction methods. However, industrialization requires investments in terms both of research and of specialized construction equipment, and these investments must be amortized. This requirement has resulted in a marked specialization of operators, and the advent of bridge subcontractors equipped for just one or two construction methods, and often for just one structural material (either prestressed concrete or steel). Contemporaneously, a changing emphasis on the environment has modified demand and led designers to plan infrastructures more and more carefully inserted into their sites. The use of tunnels significantly reduces the impact of new infrastructures on the environment; however, costs apart, this design philosophy has to face the difficulty of passing under existing obstacles. Rivers, channels and, to a lesser extent, existing infrastructures, require construction of bridges. Infrastructures deeply inserted into the environment result in short bridges. In hilly areas, bridges are often directly inserted between tunnel inlets to span the obstacle. In spite of their modest length, however, modern bridges must fulfil several requirements. They must have low maintenance costs and minimum effects on the environment and the obstacle to be overpassed. They must also be easy to build and competitive in terms of construction costs. The length of a bridge has a strong impact on its construction cost. In indus- trialized countries, labour costs can easily exceed half of the total construction cost of the bridge. Reduction of labour costs requires industrialization, with related investments that have to be amortized. If the bridge is short, amortization of equipment costs (not only the initial investment, but also transport, assembly, and final dismantling) can be prohibitive. On the other hand, labour costs resulting from a poorly industrialized yard can be prohibitive as well. The solution is not easy.x BRIDGE LAUNCHING These new demands stimulated research on new structural systems and bridge construction methods that combine low labour costs and low investments for adaptable equipment that can be amortized in several jobs. Finally, a better evaluation of the total cost of a bridge (construction and maintenance during its service life) singled out the intrinsic quality of the bridge and its good adaptability as primary objectives. These issues produced substantial innovations in medium-length bridges. For contractors who operate in the international market, innovation and export capability are two basic and tightly interconnected requirements. The ability to export depends on the ability to propose competitive alternatives, which in turn derives from cultivating research and innovation. In recent years, the contraction of the public works market and the resulting tighter competition has led contractors to reduce overhead expenses, and with them research, and to seek a systematic reuse of available equipment. This, in turn, is prevented by innovative ideas requiring new investments. In spite of that, however, innovation is an inevitable requirement for the international activity of high-tech countries, which cannot compete with emerging countries in labour costs and must therefore exercise their know-how to propose advanced alternatives. Some emerging countries have already reached high technological levels, so, to preserve their export capacity, the most evolved nations cannot but continue the research of new ideas. Incremental launching of bridges is a logical and effective response to most of these requirements. The aim of this book is to merge the structural, organizational and economic aspects of the incremental launching of bridges. These aspects are tightly related and must be considered as a whole in the design of such structures. The author is a freelance bridge consultant who has spent many years as project manager for bridge contractors, leading workgroups composed of designers and field engineers, and verifying each time that a broad view can result in a successful approach to high-tech structures. This book is addressed to bridge contractors, technical and design-office engineers, and owners, to help them to widen their partial knowledge of the subject. Chapter 1 introduces the incremental launching construction method for PC, composite and prestressed composite bridges. Chapter 2 deals with the techno- logical and design aspects common to these three different types of launched bridges. Chapter 3 is devoted to prestressed concrete bridges, while Chapter 4 deals with conventional composite bridges, whose launching is similar to the launch of steel bridges. Finally, Chapter 5 is devoted to prestressed composite bridges. Symbols The symbols listed below have been used in the text. Some passages have required further symbols, described in the respective sections. a distance between web centres on the top slab axis ay base of the fold in a folded plate web ay peak ground acceleration (PGA) dz. peak ground acceleration during construction A cross-sectional area ‘Ay area of the bottom slab or flange, area of the braces of a temporary pier A, cross-sectional area of the concrete slab A; net area of each flange A; cross-sectional area of the nosedeck joint, Ay area of each pillar of a temporary pier A, cross-sectional area of the steel girder Ay cross-sectional area of a web b distance between web centres on the bottom slab axis, breadth of concrete in cross-section, flange width, flange segment length by breadth of the bottom flange br step of the folds in a folded plate web by breadth of the neo-fion plate b breadth of the top flange B total breadth of the top slab c web sloped depth C4 cross-sectional property related to warping (inertial term in the BEF equation) C; friction coefficient Cf; breakaway friction coefficient at bearing i Ci friction coefficient of the contact plates of the friction launcher G constant C, torsional stiffnessxii BRIDGE LAUNCHING theoretical breadth (from web axis) of the side cantilever of the cross- section distance of the bottom slab axis from the cross-sectional centroid distance of the top slab axis from the cross-sectional centroid longitudinal flexural stiffness of a cross-sectional plate transverse flexural stiffness of a unit length of web eccentricity eccentricity of the additional prestressing in the front zone of the superstructure downward eccentricity of the resulting force efficiency of material upward eccentricity of the resulting force elastic modulus elastic modulus of the braces of a temporary pier mean elastic modulus of concrete during the heating period mean elastic modulus of concrete during the cooling period elastic modulus of the launching nose elastic modulus of steel axial deformation modulus of a folded steel plate flexural stiffness of the superstructure flexural stiffness of the launching nose ultimate stress cube compressive strength of concrete at i days of curing 28-day characteristic compressive strength of concrete tensile strength of concrete effective tensile strength of concrete ultimate strength of steel characteristic yield strength of steel general force additional prestressing in the front zone of the superstructure prestressing force in the bottom slab theoretical friction force friction force opposed by the formwork design friction force horizontal force due to the longitudinal gradient of the launch surface hemisymmetrical force breakaway force at support i lower prestressing force in the nosedeck joint section lock force maximum prestressing force minimum launch prestressing parabolic prestressing force force in the stays of the cable-stayed front scheme, longitudinal forces in the slabs SYMBOLS xiii M poe M Mi Mp a Pe Pee Pr prestressing force in the top slab thrust force upper prestressing force in the nosedeck joint section effective shear modulus of steel average percent gradient of the launch surface shear modulus of steel theoretical depth of the cross-section (distance between the centre lines of slabs or flanges) depth of the cross-sectional central core average thickness of concrete depth of the fold height of the pier height of the tower in the cable-stayed front system net height of web plate (measured clear between flanges) total depth of the cross-section (distance between the extreme edges) total depth of the steel girders (distance between the extreme edges) moment of inertia about the centroid of the cross-section moment of inertia of the nosedeck joint cross-section moment of inertia of the launching nose BEF modulus of foundation, coefficient for seismic demand during construction, generic coefficient or constant constants or coefficients length of the span length of the bank span of the continuous beam, length of patch loading along the bottom flange edge critical length of the front cantilever length of the launching nose iength of the neo-fion plate effective length, length of the equivalent hinged rod length of patch loading along the web edge inverse of Poisson’s ratio (1/v) bending moment negative moment in the first support section at the end of launch design bending moment moment in the cross-section above support I moment in the cross-section above support I due to cantilevered masses ultimate cross-sectional flexural strength axial force in a member, compression force minimum prestressing force applied to the anchor blocks uniformly distributed service load probability of exceedance of the design seismic event during service life probability of exceedance of the design seismic event during construction hemisymmetrical component of the unit load uniformly distributed load produced by parabolic prestressingxiv BRIDGE LAUNCHING uniformly distributed superimposed dead load Euler's critical load nominal static ultimate shear strength of headed stud connector nominal static ultimate shear strength of headed stud connector in the presence of tension uniformly distributed dead load uniformly distributed dead load of the launching nose unit weight of the steel structure of a composite bridge dead load of the superstructure radius of gyration of the cross-section about the centroidal axi of curvature, uniform friction load local buckling strength design support reaction hemisymmetrical component of the support reaction reaction at support I ultimate support reaction applied through bottom flange plate symmetrical component of the support reaction radius of curvature of the equivalent tendon support reaction of bearing i support reaction at the friction launcher local yielding strength section modulus modified section modulus state array in section j reduced state array of support section J time, thickness thickness of the bottom slab or flange construction duration return period of the design seismic event return period of the design seismic event during construction thickness of the top slab or flange web thickness torsional moment, temperature transfer matrix between the cross-sections j and k reduced transfer matrix between the support sections J and K ultimate tension force in headed stud connector shear force vertical misplacement of the launching bearings hemisymmetrical component of vertical misplacement of launch bearings symmetrical component of vertical misplacement of launch bearings total plastic work for a complete corrugation distance of the centroid from the lower edge radius SYMBOLS xv distance of the centroid from the upper edge angle, coefficient of thermal expansion, dimensionless progression of launch critical dimensionless progres angle, restraint coefficient difference between two values loss of prestress at time ¢ orthogonal strain creep strain shrinkage strain friction angle, creep coefficient rotation specific weight friction force amplification factor for design of launch devices design safety factor for loads design safety factor for materials design safety factor from uncontrolled sliding vertical deflection, imperfection constant distortional component of the cross-section deformation torsional component of the cross-section deformation BEF characteristic length, slenderness ratio BEF characteristic length for the calculation of distortional edge stresses development ratio of the folded web Poisson’s ratio torsional rotation flexural efficiency of the cross-section structural efficiency of design orthogonal stress compressive stress (positive) critical buckling stress critical vertical stress ideal biaxial stress tensile stress in a prestressing tendon initial tension in a tendon allowable contact pressure on PTFE residual tensile stress (negative) tensile stress (negative) ultimate compressive stress total length of the neo-flon plates on a launching bearing tangential stress critical tangential stress critical tangential stress for global buckling critical tangential stress for interactive buckling critical tangential stress for local buckling ion of launchxvi BRIDGE LAUNCHING Ta Trim v design ultimate shear stress Timoshenko’s critical tangential stress contemporaneity factor for launching bearing breakaway force |. Introduction The first prestressed concrete bridge was built in 1941. From the end of World War II onwards prestressed concrete bridge construction developed quickly, thanks to the pioneers of the new technology: Guyon, Freyssinet, Leonhardt, Magnel, Morandi, Mérsch and Ross, among others. In the past, the design of almost all these new bridges was limited to quasi- isostatic systems, since analysis of statically indeterminate structures was not compatible with the means available for calculation. For many years, the basic criterion in the design of prestressed concrete (PC) bridges has been to assure the possibility of simple static analysis. So, multispan bridges built on falsework used the simple support scheme, the first balanced cantilever bridges in the 1950s were hinged at mid-span, and the first superstructures built with movable shuttering systems in the 1960s were simply supported at the piers or articulated at the counterflexure points. Front the 1960s onwards, the-extraordinary-progress in computing techniqu the technological advances in prestressing, and the improved knowledge of materi- als extended the possibili in the design and, above all, in the analysis of PC. These new possibilities stimulated technological innovation and the definition of new construction methods. In the creative thinking of these 60 years, many construction methods were developed to guarantee adequate margins of competitiveness of PC over steel constructions. At the same time, however, the cost reduction of steel plates, the better organization of workshops, new splice techniques, and new field assembly methods reduced the lower threshold of competitiveness of steel bridges to spans that can be easily reached with PC as well. Finally, new combinations of reinforced concrete slabs and steel girders led to composite sections explicitly addressed to these borderline spans. This further stimulated research in the concrete and steel- work industries, and created new connections between these different technologies. New construction methods took advantage of the most recent technological advances both in the design of new equipment and in widening the field of applica- tion of already familiar techniques. So, although the idea of bridge launching2 BRIDGE LAUNCHING derives from Palaeolithic technology (think of a tree-trunk), its application to PC bridges has been made possible by the availability of launching techniques already tested in the field of steel bridges, and of new materials such as Teflon. For many decades, the light weight of steel structures had permitted their launch by means of winches and lubricated wooden slides or steel rollers, with a friction that, although considerable, did not cause inadmissible stresses in the piers or require excessively expensive launching equipment. The high flexural efficiency of the steel girders and the ability of the material to work both in tension and in compression eased resistance to the temporary launch stresses and avoided the need for overdesign with respect to service requirements. Finally, launching of the steel girder had many advantages over conventional assembly, as fitting and joining of the steel elements was performed on the ground instead of in the air, without affecting the areas below the bridge, and with greater safety and simplicity (Fig. 1.1). However, compared with traditional hoisting techniques for steel girders, costs were higher as a result of the operational slow-down caused by the availability of only one working yard. As a consequence, the adoption of incremental launching ‘was initially limited to bridges high above the ground or in inaccessible areas. In the eyes of the PC bridge pioneers, some of these aspects were less critical, others even promising. In fact, construction duration and yard organization for PC bridges are different from those for steel bridges, and the cost of labour and equipment is so high that every possible alternative must be carefully examined. Front launching of a superstructure built on the ground promised savings in both equipment and labour, but in practice this was limited by the much greater weight of concrete and its low tensile strength. Fig. 1.1. Incremental launching of the steel girder of the Schrotetal Bridge (author) INTRODUCTION 3 These obstacles were gradually overcome. The advances in prestressing tech nology lightened the superstructure, made it elastic and less subject to cracking, permitted joining subsequent segments, and allowed introduction or removal of tendons according to need. The commercial availability of finite-element software for personal computers eased continuous beam analysis in the subsequent support configurations assumed during launching. The introduction of the steel-Teflon contact along sliding surfaces permitted reduction in launch friction. Finally, construction equipment reached the technological level necessary to move huge masses with the required precision. None of this was available in 1959, the year that saw a first attempt to launch precast concrete segments for the bridge over the Ager River [1]. The superstructure was built on a timber falsework, which supported precast segments built on one bank and launched one by one into their final position (Fig. 1.2). The joints between adjacent segments were cast in situ, and prestressing was obtained by means of long external tendons placed within the box cell and tensioned from the two ends of the superstructure. The bridge segments were moved by sliding them along a wooden rail by means of wooden skids lubricated with engine oil. The friction coefficient was high, but without negative effects as the segments were moved one by one. This first trial demonstrated how it would have been simpler to avoid the false- work and to launch the whole superstructure directly on to the piers by taking advantage of its flexural stiffness, This concept was applied not much later for the Rio Caroni Bridge [2,3]. The falsework was eliminated and the whole length Fig. 1.2. Precast segments of the Ager Bridge (courtesy Leonhardt)4 BRIDGE LAUNCHING of the superstructure was built behind an abutment by joining precast segments built in a casting yard and advanced to the assembly position by sliding along a wooden rail, with joints cast in situ. On completion of segment assembly (Fig. 1.3), the superstructure was prestressed with external tendons distributed within the box cell so as to obtain a centroidal force. Their anchorages were arranged in the front end-diaphragm of the super- structure, and at the opposite end they were deviated around a cylindrical concrete block, whose jacking from the rear end of the superstructure produced the required prestressing force. The superstructure was launched with the help of a front steel-truss nose. One temporary pier was built in each span to halve the launching span (Fig. 1.4). The bridge was launched above movable bearings, and at the end of each launch stroke it had to be lifted to relocate the bearings (Fig. 1.5). On completion of launch, axial prestressing was not sufficient for the design loads, so the external tendons were made eccentric by vertically shifting their mid-span and support deviation points. At the same time, the rear deviation 1phase: construction of segments construction : : Of segnents Positioning DO) norereere construction of Final piers 2” phase! axial prestressing of the deck launching nose construction of temporary piers 3" phase: launching Fig. 1.3. Construction scheme of the Rio Caroni Bridge INTRODUCTION 5 Fig. 1.4. Launch of the Rio Caroni Bridge (courtesy Leonhardt) cylinder was released. Tendons were finally drawn close to the webs, fastened to them, and protected by means of prepack concrete (Fig. 1.6). Although conceptually ingenious, the shifting of tensioned tendons was a very complex and expensive operation. This stimulated the idea of introducing undulated tendons on completion of launch to correct a permanent axial Jaunch prestressing, and this concept has characterized most iaunched bridges since then. Also the need for complete assembly of the superstructure behind the launch abutment was quite limiting, and this stimulated the idea of match-casting longer deck segments directly behind the abutment and joining launch tendons every two or three segments. Launching of the superstructure thus became incremental. That was how, in 1965, the incremental launching method (construction of long deck segments behind an abutment alternating with the launch of the whole deck section) was applied for the first time in the construction of a bridge near Kufstein, Germany. Once the new segment has been joined (if precast) or has reached the necessary strength (if cast in situ) and launch prestressing has been introduced, the whole superstructure is launched forwards by a length equal to the segment. This gives free access to the assembly yard or to the formwork, another segment can be built in contact with the rear one, and so on until completion of the superstructure (Fig. 1.7). From one point of view, this construction method simulates a horizontal ‘slip form’, with the difference that the form is fixed and the superstructure itself moves with respect to it, in a process that is somehow similar to a cadenced extrusion.6 BRIDGE LAUNCHING Final bearing > i dock LAUNCHING stainless steel plate LIFTING a BEARING RETURN LOWERING Fig. 1.5. Discontinuous launch with mobile bearings In this first experience of incremental launching, the use of a specific launch prestressing was avoided by reducing launching spans with two temporary piers per span and by accepting edge tensile stresses, temporarily controlled by reinforcement and permanently avoided by parabolic prestressing (Fig. 1.8). In subsequent years, there have been many improvements in this construction technique. Sliding is facilitated by neoprene-Teflon (neo-flon) plates inserted between the superstructure and fixed launch bearings, thus eliminating the need for lifting. The front cantilever support devices range from steel or concrete launch- ing noses to cable-stayed systems. Thrust devices have evolved towards compact friction units that guarantee large safety margins and smooth travel, thanks to the combined use of high-pressure hydraulic and digital controls. Fast continuous INTRODUCTION 7 GUIDE DEVICE steel plates rio spacer prestressing wires or strands concrete wel FINAL ARRANGEMENT reinforcement wire net Injection pipe pre-pack concrete mortar anchor bolt bush-hanmered surface concrete web Fig. 1.6. Final arrangement of prestressing tendons in the Rio Caroni Bridge beam solver algorithms permit easy evaluation of the stresses and deflections of the superstructure in every launching phase as well as of their envelopes. Although the incremental launching construction method was originally con- ceived for bridges some 100 m long, the search for the greatest cost reduction8 BRIDGE LAUNCHING Fig. 1.7. Incremental launching extended its use both to short bridges and to very long bridges. This construction method is now applied to small bridges, where labour cost can be lowered without major investments, and to large bridges, where greater labour cost savings are achieved by massive field industrialization. In both cases, investments for specialized construction equipment can be adapted to the specific demand and can be amortized in more bridges, as equipment is very adaptable [4]. The technological advances of incremental launching permitted experimentation with new types of monolithic handling of the superstructure. So, though incremen- tal launching still represents the most common handling technique and its evolution has influenced the development of others, both the symmetrical approach and INTRODUCTION 9 Fig. 1.8. Launch of the third box girder of the Kufstein Bridge: the rectangular piers are temporary (courtesy Leonhardt) rotation have won a role in critical applications and in many cases they represent the least expensive and most elegant solutions [4]. In the last two decades, there have been many ingenious applications of launch techniques for PC bridges. These range from beams supported by arches, curved superstructures, cable-stayed bridges launched symmetrically from the two banks, or built along one bank and then rotated into their final position, or built by launching the superstructure above temporary piers and then suspending it once it has reached the final position. There has also been research and development in the field of steel constructions. A better knowledge of instability has led to lighter structures, often characterized by thin web plates. This required improvements in the launch bearings, which must follow the rotations of the superstructure with uniform support reactions. This also resulted in lighter noses and specialized systems for transfer of launch thrust. The evolution of launch bearings was particularly significant in conventional composite bridges, where the steel girder is launched first and the deck slab is cast in situ on completion of girder launching. The large deflections and rotations of these flexible girders prior to cross-section completion result in several specific requirements for launch devices. In spite of these advances, the different cost evolution of concrete and steel plates still suggests the use of reinforced concrete for the compression areas of10 BRIDGE LAUNCHING medium-span steel bridges. A concrete deck slab is particularly suitable for simply supported spans; however, the continuous beam offers several advantages, and control of the upper edge tensile stresses in the support areas is therefore an issue common to most composite bridges. The control of longitudinal tensile stresses in the concrete slab requires specific segmental casting sequences for the slab, which cause high costs and long construc- tion times. This suggested incremental launching of the concrete slab onto the steel girder. By using the same thrust devices first for the steel girder and then for the concrete slab, it is possible to build the composite bridge in a very restricted yard and to avoid any interference with the areas below. Experience gained in the launch of PC bridges was invaluable for this new step forward. Research in PC also led to external prestressing. When external tendons are used, the web thickness of the box girder is no longer restrained by the need to contain and to deviate tendons. This permits lightening the webs and enhancing cross-sectional efficiency. However, shear forces dictate the minimum thickness of concrete webs, and steel plate webs are much more efficient for this purpose. This led to the concept of prestressed composite bridges. In these new box girder typologies, two concrete slabs resist the flexural stresses thanks to prestressing, and tendon deviation reduces the shear stresses in the webs down to levels that can be resisted by thinnest steel plates. Compared with the PC box girders, these prestressed composite sections are much more efficient, with related savings in weight, concrete, reinforcement and prestressing. These new sections require only about a quarter of the steelwork of conventional composite sections. Advance in the launch techniques for PC and steel bridges led to the launch of prestressed composite bridges. Incremental launching is probably the most competitive construction method for medium-span box girder bridges. 1.1. Introduction to incremental launching Construction methods that take advantage of the self-resistance of structures are based on a rather intuitive principle. As several structural components are designed for permanent stress levels that are higher than those reached during conventional construction, this temporary overstrength can be exploited to simplify construction and to reduce the carrying demand on construction equipment (and the related costs). The use of structural strength during construction results in significant economic advantages for both PC and steel bridges. However, several aspects should be taken into account to correctly estimate the economic implications of self-resisting construction methods In bridges made of concrete, the structural material is unable to resist significant tensile stresses. The use of prestressing, limits or avoids tensile stresses; however, this stress adjustment is obtained with tendons whose alignment depends on the distribution of the internal actions to be corrected, and can be modified only by expensive operations. In steel bridges, the structural material is inherently able to INTRODUCTION I resist both tensile and compressive stresses, but this simplicity is only apparent, as compressive stresses cause instability. In both cases, therefore, a structural design for the permanent actions may be inadequate for self-resisting construction if the temporary stresses are very different from the final ones. On the other hand, if construction stresses are so significant as to require design adjustments, this might result in a structure that is overdesigned with respect to service requirements. Design for construction stresses depends on the level and distribution of the temporary internal actions, which are governed by the dead load of the superstructure and the temporary structural systems used for construction. When construction takes place with the same static system of the permanent structure, a load increase after completion of construction (for instance, application of superimposed dead load) results in a proportional increase in internal actions. Other types of load (live load, wind) or actions (temperature gradients, imposed displacements, settlements) may result in non-proportional increases or sign inversion of internal actions. However, this second family of effects is always superimposed on the basic effects of dead load, and the resulting variations in internal actions are relatively small. Variations in the static system have a more profound influence, especially when they result in sign inversion of bending moment in areas of the structure that are loaded in a stable manner on completion of construction. A mid-span section crossed by prestressing tendons nearby its lower edge is hardly compatible with negative moments. A support section crossed by prestressing tendons near its upper edge will hardly resist positive moments. As any operation on tensioned prestres- sing tendons is expensive, less eccentric tendons must be used to enhance the structure adaptability to these temporary load conditions, and this results in a less efficient final prestressing scheme. Similar conclusions can be drawn for the flange plates of the steel girders of conventional composite sections. As construction stresses depend on the dead Toad of the superstructure and the static system, both of them must often be adjusted. Dead load should be as low as possible during launching. In composite bridges, the concrete slab is a prime component of the section weight and it is therefore added only after completion of the launch of the steel girder. In prestressed composite bridges and conventional PC box girders this is usually impossible. However, appendages can be applied after launching, temporary support devices such as launching noses and stayed systems can be used to adjust the internal stresses in the front zone of the superstructure, and temporary piers may diffusely decrease launch stresses in the most delicate cases. The economic advantages of self-resisting construction methods often exceed the costs resulting from the design requirements of construction phases. However, a broad vision of the overall design economy is necessary with these types of structures, Segmental bridges are usually erected [5,6] in one of five ways: span-by-span construction with truss, falsework, or movable shuttering system: progressive placement; span-by-span lifting; balanced cantilever construction; or incremental12 BRIDGE LAUNCHING launching and derived methods. The first three methods do not involve self-resistance of the structure during construction. © With span-by-span construction, the PC superstructure is cast in situ in a movable shuttering system that proceeds from one abutment towards the opposite one. Construction joints are usually placed at the counterfiexure points. The form carrier provides factory-type operations transplanted to the job site. It may be supported on the piers, from the edge of the previously completed construction or from the pier. The same principle has been applied in conjunction with precast segments that may be assembled on a steel truss to make a complete span. Prestressing tendons ensure the assembly of the various segments in one span while maintaining full continuity with the preceding span. This construction method is rarely applied to steel girders. In principle, progressive placement is similar to the span-by-span process, because the construction starts at one end of the structure and proceeds continuously to the other end. Its origin, however, is derived from the cantilever concept. Precast segments are placed in successive cantilevers on the same side of the same pier rather than by balanced cantilevers on both sides of the pier. Temporary stays deviated by a steel tower placed at the pier limit the cantilever stresses and deflections. This method is seldom applied to superstructures cast in situ or steel girders. © Span-by-span lifting is mainly used for the steel girders of composite sections to be completed with a concrete slab cast in situ. It is also used for PC 1-girders, while the dead load of PC box girders is excessive for this construction method. @ The balanced cantilever process is primarily suitable for long spans where construction activities for the superstructure can be planned at the deck level and without the use of extensive falsework. Long viaduct structures with relatively short spans are better processed using either span-by-span methods or incremental launching. This procedure cantilevers segments from a pier in a balanced fashion on each side until mid-span is reached and a closure is made with a previous half-span cantilever from the preceding pier. The same erection process is repeated until the structure is completed. For construction cast in situ, the movable formwork is supported from the previously erected segment or from a movable form carrier while the new segment is being formed, cast and stressed. For precast segmental construction of box girders made of steel, reinforced concrete or composite materials, the new segment is usually supported from a movable winch system anchored to the preceding segment. Balanced cantilever construction strictly follows the self-resistance criteria. In a PC bridge, prestressing tendons in the deck slab and compression forces in the bottom slab resist the negative moment generated by the dead load of the cantilever and the weight of the form carrier. The permanent effects of dead load and prestressing are similar to construction stresses, and the effects of superimposed dead and service loads are mostly of the same sign. INTRODUCTION 13 Since construction stresses and permanent stresses are similar, this construction method does not require particularly refined prestressing schemes. Therefore, although it dates back to the beginning of PC bridges, the advances in prestressing in recent decades have not produced substantial changes in balanced cantilever construction, whose optimum field of use is still limited to rather long spans. The use of the balanced cantilever method has now widened to include the stayed-cantilever construction of arch bridges and the progressive erection of cable-stayed bridges. The conceptual solution is the same [5], with temporary stays in arch construction and permanent stays in the cable-stayed bridges. 1.2. Incremental launching of PC bridges Compared with traditional construction techniques, the incremental launching of a PC bridge cast in situ is revolutionary [7] from many points of view: © The standard equipment is simple, being limited to a formwork supported on the ground, a thrust system and special devices (a launching nose that may be integrated with temporary piers or a stayed front system) that reduce the launching stresses. © The formwork, the batching plant (if any) and the steel cage assembly template are located in a small yard, so the stocking and handling of loose materials by means of a tower crane are considerably simplified. e The superstructure is built in a fixed, sheltered location. Each operation, from the steel cage assembly to concrete casting or prestressing, is much simpler and safer than on a falsework or at the tip of a cantilever, and may be organized in parallel rather than in series. © A high level of quality control is indispensable at each stage of construction, since mistakes or irregularities cause launch difficulties. The construction of well-built and durable structures is therefore in the best interest of the contractor. e The construction of the superstructure does not require placement of equipment such as footbridges or falsework between the piers, with clear advantages when the bridge spans a river (Fig. 1.9), highway, railway or inaccessible place, or presents reduced clearances. e Unlike balanced cantilever construction, it is possible and convenient to build long superstructure segments, even more than 30 m long. This reduces the number of construction joints, weak points in any structure. Compared to balanced cantilever construction, launching results in faster construction and requires only one casting bed instead of two or more pairs of form carriers. Although the incremental launching construction method is generally associated with bridges cast in situ, it is perfectly suitable for precast segmental construction as well. Bridges composed of precast segments arranged along temporary support14 BRIDGE LAUNCHING Fig. 1.9. Skye Bridge (reprinted with permission, Dywidag Systems International) rails, connected by closure joints or epoxy, and incrementally launched, offer several advantages. © Construction of most of the superstructure is independent of pier construction, as precast segments can be stored elsewhere. The use of labour and casting equipment is optimized in a series of highly repetitive operations with the best quality control and amortization of invest- ments (precast segments for launched bridges are absolutely conventional in dimensions and handling). ¢ Assembly and launching of the superstructure are fast operations that require little specialized equipment (support/launch rails, a launching nose and a thrust system). Equipment is easily reusable and inexpensive in terms of initial investment and assembly and dismantling costs. Yard operations are limited to the segment arrangement along the support) launch rails, the joint sealing and the introduction of prestressing in the new section of the superstructure. The construction duration can be half of that required for casting in situ, although with additional labour costs because of these additional activities, So far, the incremental launching construction method has been mainly used for bridges cast in situ, as this permits combining the advantages of the industrialized and repetitive work processes with those deriving from a small number of construc- tion joints with continuous reinforcement. In addition, segment length and yard organization can be easily adapted to the specific requirements, thus enhancing the overall competitiveness of the construction method. | 1 i [ | INTRODUCTION 15 Launching of precast segmental bridges [72] can be a very competitive solution too. Bridges whose length does not permit amortization of the costs of a launching truss for segmental construction can be optimum targets, especially when it is necessary to allow unrestricted use of the areas below. Tens of modular highway overpasses may be built with a precasting plant and few sets of inexpensive launch equipment (support/launch rails, a launching nose and a thrust device), with all the advantages of the continuous box girder structural solution. A precasting plant can simultaneously feed production lines with truss assembly and production lines with incremental launching, thus further enhancing amortization conditions and planning flexibility of the project. All the conventional launch techniques (incremental launching from one abutment towards the opposite one, monolithic launching from the opposite abutments with mid-span closure, rotation about the piers adjacent the obstacle to overpass and mid-span closure, launching followed by transverse shifting) may be used as well. 1.3. Incremental launching of composite bridges In the last decades, many factors increased the competitiveness of composite bridges and reduced their lower use threshold to spans of about fifty metres, which previously were the dominion of PC box girders [4,12]. Progress in iron metallurgy led to rolled steel of high and reliable mechanical properties. The development of design codes based on the strength of materials led to a better evaluation of safety. The progress of assembly techniques replaced riveted splices with bolted or welded ones. In the last years, this evolution has been further accelerated by the cost stagnation of steel plates and the parallel general increase in labour cost. The qualitative advantages of composite bridges are numerous. The high tensile and shear efficiency of the steel piates combines with the good compressive efficiency of concrete. Rapidity of construction and the possibility of building most of the structure in a workshop improve planning and risk management. The long durability resulting from different and renewable protective treatments and the possibility to modify the structure with time to adapt it to new service conditions further improve the flexibility of design. Finally, the presence of few structural elements whose function is clearly recognisable enhances the architectural quality of the structure. The optimum field of utilization of a composite section is the simply supported beam. In the negative bending regions of a continuous beam, the bottom flange compressive stresses require thick plates or closely spaced stiffeners to prevent instability, while the tensile stresses in the deck slab require longitudinal rein- forcement to control cracking. The presence of a concrete bottom slab in the negative bending regions for double composite action is competitive only for long spans. In most cases, therefore, the cross-section is open and only aesthetic reasons or particular design requirements justify, in straight bridges, the higher cost of closing the girder with a steel bottom plate.16 BRIDGE LAUNCHING These limitations also influence the launch of composite bridges, as the dispersion of support reactions within the webs and the negative bending stresses affect every cross-section of the superstructure during launching. Since the dead load of the concrete slab is 75 to 85% of the total weight of the superstructure, the steel girder is launched first and the deck slab is placed on to the steel girder only on completion of its launch. This solution reduces the launch stresses in the steel girder and avoids tensile stresses in the concrete slab. On the other hand, however, the deflections of the steel girder are large during launching because the flexural stiffness is much smaller than that of the final composite section. This requires precambering and results in specific launch requirements in terms of both support adaptability and control of local stresses. The concrete slab may be either cast in situ or in turn launched onto the steel girder. Manual form handling for slabs cast in situ is expensive, risky and time consuming, and the final quality is quite poor. Therefore, casting in situ takes place either with form carriers (movable shuttering systems that shift along the steel girder to reach the different casting positions) or on precast planks. The use of form carriers is preferred in most cases although their cost can be adequately amortized only in long bridges. In the case of short or medium-length bridges, precast concrete planks are often used as left-in-place formworks. However, con- crete planks usually do not contribute to the carrying capacity of the composite section and their weight therefore increases the weight of the steel girder, Precast planks are expensive and are to be placed onto the steel girder, with risks for workers and the areas below and additional costs. As an alternative to casting in situ, a continuous concrete slab may be cast in a fixed form placed behind an abutment. The slab is composed of segments match- cast against each other with continuous longitudinal reinforcement. Segment casting is followed by the launch of the whole slab section onto the steel girder, and this sequence (construction of a new slab segment, and launch of the whole strip) is repeated until completion of the slab, which is finally joined to the steel girder to achieve composite action. Incremental slab launching is an interesting construction method. The form system is much less expensive than a movable shuttering system. Slab construction is a continuous process, as the discontinuous in-situ casting sequences aimed at avoiding tensile stresses in the negative bending regions are no longer necessary. Steel cage assembly and concrete handling are extremely simplified. Slab launching also avoids interference with the areas below, reduces shrinkage and creep cracking, eases introduction of longitudinal prestressing, simplifies structural analysis, improves durability and aesthetic quality, and reduces risks for workers. 1.4. Incremental launching of prestressed composite bridges In spite of its numerous advantages, a conventional composite section has some limits when used in a continuous beam. Its construction by incremental launching requires launching of the steel girder alone and placement of the deck slab later, and the difficult casting conditions of a bottom slab discourage the use of double INTRODUCTION 17 composite action. Of course, construction of a concrete bottom slab would be much simpler in a fixed formwork supported on the ground. This led to the concept of box girders composed of two concrete slabs and two steel webs, to be launched with the completed section. The presence of two concrete slabs requires the massive utilization of prestressing, and concrete slabs, light steel webs and prestressing tendons are combined in new typologies of prestressed composite sections. In these sections, longitudinal bending moments are resisted by axial forces in the concrete slabs and the shear forces, reduced by tendon deviation, are mainly resisted by the steel webs. Structural behaviour depends on the shear force distribution between the steel webs and the prestressing tendons. The different material efficiency suggests shaping the tendons so as to balance the shear forces produced by the permanent loads and one-half of the live loads, so that the steel webs resist only the shear fluctuations due to the presence or absence of live loads. It is thus possible to obtain slender cross-sections that guarantee the highest efficiency in service. On the other hand, any operations in which the dead load shear forces are not reduced by shaped tendons are difficult and delicate. Prestressed composite sections actually involve several launch challenges.2. Incremental launching of bridges 2.1. Structural system of the superstructure With the progress of materials technology and calculation methods, the continuous beam structural system has been shown to have several advantages over statically determinate schemes. Continuous beams require less structural materials than either the simply supported beam or Gerber schemes, and guarantee a better control of deflections and fatigue phenomena. Avoidance of internal hinges that interrupt beam continuity in the horizontal plane improves the seismic response of the bridge, and a smaller number of bearings and intermediate joints decreases maintenance costs. Finally, secondary stresses produced by differential settlement of bearings (if any) may be avoided by shimming, and are definitely reduced by creep of concrete. For all of these reasons, the continuous beam scheme has been more and more frequently adopted in recent generations of bridges. The superstructure of a launched bridge is always built as a continuous beam. This makes the bending moments in the mid-span regions and in the support sections similar in value during the launch, although opposite in sign. Centroidal launch prestressing of PC and prestressed composite bridges controls the flexural tensile stresses alternately at both edges without excessive compressive stresses at the opposite edges, and similar considerations apply to the compression flanges of composite bridges. The advantages of a continuous beam suggest permanently maintaining the static system used during launching, at least when the bridge length is compatible with the expansion and contraction produced by thermal and time-dependent phenomena. In this case, on completion of launch it is only necessary to replace the launching saddles with permanent bearings and to complete prestressing. Two longitudinal Static systems of the superstructure may be achieved this way: Fixed bearing at one end and sliding bearings at all the other supports. This scheme permits concentration of the horizontal forces at one abutment, but requires longer sliding bearings and a large expansion joint at the mobile end. Fixed bearing at an internal pier and sliding bearings at all the other supports. This scheme is applicable only in the presence of a short and stiff central pier20 BRIDGE LAUNCHING that can resist the whole horizontal action. Compared with the previous scheme, a central restraint reduces the movement of bearings and expansion joints. Tn a case of flexible piers, two or three central piers may be hinged to the superstructure to better distribute the horizontal forces. The static system used during launching can be modified once the superstructure has reached its final position. © Some design requirements (very long bridges, high temperature differentials, slender piers, etc.) can dictate division of the superstructure into. shorter consecutive continuous beams, or even simply supported spans. e Ina case of variable curvature in plan or profile that makes launching from only one abutment impossible, two sections of the superstructure can be launched from the opposite abutments. As an alternative, if curvature varies in a bridge region not much distant from the ground, the casting yard may be arranged on a temporary embankment raised in this region, so that two deck sections can be launched one after the other in opposite directions. In both cases, the central closure between the two deck sections increases the final degree of redundancy with respect to that used during launching. e The superstructure may be permanently framed to the launch supports to obtain indeterminate arches or portal schemes. This solution is often used with steel and composite bridges. © The superstructure, launched over temporary piers, may be suspended to a tower to obtain a final cable-stayed system. In the 49.16 Lot of the high-speed railway ‘TGV Atlantique’ in France, a PC superstructure composed of five consecutive three-bay continuous beams, each with spans of 29.2, 30.6 and 29.2 m, was launched single 445-m continuous beam. After launching, the temporary prestressing of the permanent joints was released to separate the superstructure into the five final beams, which were moved towards their final position by jacking the internal expansion joints. The same technique was adopted in South Africa for the Olifant’s River Bridge, composed of 23 45-m PC spans for an overall length of 1035 m. The final scheme of this railway bridge consists of two 11-span continuous beams fixed at the respective abutments and a simply supported central span working as an expansion joint. This scheme transfers the braking forces directly to the abutments and permits the use of slender piers. The 23 spans were built as a single continuous beam. On completion of launch, the superstructure was permanently fixed to the launch abutment. After relieving of the launch prestressing of the first internal joint, the free section of the superstructure (a 12-span continuous beam) was launched forwards until the central span was placed in its final position. After opening of the second joint, this operation was repeated on the remaining I1-span front section, fixing it to the second abutment. A similar solution was adopted in France for the highway Oli Bridge, composed of 15 41-m PC spans with a longitudinal 5.4% grade. The presence of piers 60 m tall INCREMENTAL LAUNCHING OF BRIDGES 21 dictated design of two continuous beams fixed at their respective abutments, with a central expansion joint. Since the dead load component along the launch plane was greater than the friction force, launching was obtained with a braked downwards sliding. A braking force of 6 MN was sufficient, compared with the 14-MN thrust force required for launching the superstructure upwards. The same technique was used to open the central expansion joint. After fixing the upper beam to its abutment, the lower beam was allowed to slide downwards by releasing the temporary prestressing tendons of the central joint. The Sinntal Bridge for the high-speed railway ICE in Germany, composed of eight 42.5-m spans, a 51.5-m span and a 30.8-m span, was built by launching as a continuous beam a sequence of single PC spans temporarily coupled to each other by joint prestressing. On completion of launch, the spans were disconnected and detached by jacking the joints until obtaining the final design scheme of 10 simply supported beams. This solution resulted in a significant cost reduction compared to the conventional span-by-span construction with a movable shuttering system. A similar application of temporary prestressing of expansion joints permitted the execution of the Basrah Bridge in Iraq [13], in which the central span of a movable bridge was launched together with the access spans as a single continuous beam. The launch static system may be also modified by joining two sections of the superstructure into a single continuous beam. This solution has been often used for bridges with a straight-curved or curved-countercurved plan alignment, which can be obtained by launching two bridge sections with constant curvature from the opposite abutments, and then connecting them by means of a central closure joint [14]. In the final phases of approach, the launching nose is assembled in a raised position to slide above the front-end diaphragm of the superstructure already in place (Fig. 2.1). A temporary pier below the closure joint improves the casting conditions, but similar results may be obtained with crossed joint bars like those used for mid-span connection of bridges built with balanced cantilever construction. The Schnaittach Bridge was built in Germany by launching a 424-m bridge section with a constant radius of plan curvature of 1000 m downhill from one abutment, and an 864-m section composed of a front clothoid and a rear tangent section uphill from the second abutment. The front clothoid required the progressive rotation of the casting cell during the first segment extractions and the use of shifting launch bearings (Fig. 2.2) placed on temporary extensions of the final piers. On completion of launch, the two bridge sections were connected by casting the joint segment in situ. A similar solution was adopted for the Schrotetal Bridge in Germany. This 495-m composite box girder bridge (Fig. 1.1) was built by launching a 404-m section of the U-girder from an abutment and a 51-m thinner section from the opposite abutment, directly above the railway. A central 40-m varying-depth span was assembled on the ground and lifted by jacking until closing the steel girder. Then, the concrete slab was cast in situ with a movable shuttering system. Incremental launching on 68.1-m spans required the use of a 24.1-m trussed launching nose. Mid-span closure of two bridge sections launched from the opposite abutments also permits overtaking a much longer central span. Temporary stays deviated by22 BRIDGE LAUNCHING Fig. 2.1. During the final approach, the launching nose is raised to slide over the superstructure already in place Fig. 2.2. Launching bearings movable in the transverse direction (author) INCREMENTAL LAUNCHING OF BRIDGES. 23 steel towers placed onto the superstructure have been used to support the two cantilevers of the central span of the Mainflingen Bridge [15]. As an alternative, the superstructure might be directly designed with extradosal prestressing. In this case, Some permanent prestressing tendons deviated by the concrete towers might support the front cantilever. Concrete towers are integral with the superstructure and arrive above the piers of the main spans on completion of launch. In a case of a Jong central span, this construction method might be more competitive than balanced cantilever construction; however, no application of this technique has been reported as yet. Wide superstructures may be attained by joining parallel beams. The 19-m wide, twin box-girder superstructure of the Tanaro River Bridge in Italy was obtained by joining two box girders built in the same casting yard one after the other. On completion of the incremental launching of the first half of the superstructure (Fig. 2.3), this was shifted rightwards by 9 m (Fig. 2.4) to clear the yard alignment for the construction of the second half. Both deck sections were cast on a 50-m continuous support through 10-m segments, and launched in four 50-m strokes, The opposite 5.3% grade at the abutments and a 1900-m radius of vertical curvature required the use of temporary locks to avoid uncontrolled sliding. Temporary piers aligned with the casting yard halved the launch span (from 50 m to 25 m) to further decrease launch prestressing. The total investment for specialized construction equipment was less than €120 000. Fig. 2.3. Launch completion of the first half of the superstructure: the rounded end of the left cantilever is the left edge of the deck slab (author)24 BRIDGE LAUNCHING Fig. 2.4. Transverse shifting of the first half of the superstructure clears the temporary pier alignment for the launch of the second half (author) Continuous beams launched onto arches represent a brilliant solution to the requirements for safety of workers, high structure quality, short construction duration and cost savings. The two parallel arch bridges over the Neckar River in Germany for highway A81 have been built by incrementally launching two 15-m wide, 365-m long continuous PC box-girder superstructures on to two central parailel arches and related access spans (Fig. 2.5). Maximum pier height is 95 m. The central arches have a span of 154.4 m and a rise of 49.9 m. A similar solution has been adopted in France for the arch bridge over the Isére River for roadway CD 22 [16]. A 8.6-m wide, 234-m long continuous PC ribbed- slab has been incrementally launched on to a central arch and related access spans. Maximum pier height is 30 m. The central arch has a span of 134.0 m and a rise of 24.0 m. The arch must have enough flexural stiffness to resist the asymmetrical loads and the horizontal forces produced by launching of the superstructure from one abutment towards the other. This often requires the use of temporary stays that balance launching loads and brace the tallest spandrel columns. Load asymmetry might be avoided by launching two symmetrical superstructures at the same time from the opposite abutments towards mid-span. However, the presence of symmetrical loads on only some of the spandrel columns induces bending stresses in the arch, and the cost of two casting yards rarely makes this scheme competitive. In the Veitshéchheim Bridge for the high-speed railway ICE in Germany (Fig. 2.6), temporary stays were integrated with counterweights to balance load INCREMENTAL LAUNCHING OF BRIDGES 25 Fig. 2.5. Launching of the superstructure above the arch of Neckarburg Bridge (courtesy Leonhardt) Fig. 2.6. Veitshichheim Railway Bridge (courtesy Leonhardt)26 BRIDGE LAUNCHING asymmetry on the arch. The PC box-girder superstructure, 1260-m long and weighing 425 MN, is one of the heaviest beams ever launched [17-21]. The central polygonal arch spans 162.0 m with a rise of 24.0 m, and the access spans have a constant length of 53.5 m So far, no rigid-frame PC bridge has been built by incremental launching, although this structural scheme would permit overtaking spans fitted to arch bridges with lower costs. Many steel portal bridges have been built by launching the steel girder on to inclined legs, as with the Piave River Bridge and the Cadore Bridge in Italy [22]. Compared to the conventional balanced cantilever construction of cable-stayed bridges, bridges launched on temporary piers and then cable-stayed to a tower are often less expensive. The 21.8-m wide, 527-m long PC box-girder of the Wandre Bridge in Belgium (Fig. 2.7) was launched onto temporary piers placed in the Meuse River and a parallel channel. With 18-m-long deck segments, the launch of the 118-MN superstructure required a 35-m-long launching nose. On completion of launch, the superstructure was stayed to a 95.5-m-tall A-shaped tower to attain the final structural system with two stayed spans of 144 and 168 m respectively [23]. With this construction method, the tower can be erected without any restrictions during the incremental launching of the superstructure. Compared with conven- tional stayed-cantilever segmental construction cast in situ, the stays are introduced on completion of launch, with smaller time-dependent effects in both the tower and the superstructure, easier geometry control, and a much smaller number of construction joints in the superstructure. Compared with precast segmental construction of the superstructure, continuous reinforcement through the joints reduces deck prestressing, which can be attained with long external tendons tensioned before removal of the temporary piers. This construction method has Fig. 2.7. Incremental launching of the Wandre Bridge (courtesy. Greisch) INCREMENTAL LAUNCHING OF BRIDGES 27 been proposed for the central cable-stayed span of the Worli-Bandra Sea-link Project in India. The tramway Palizzi Bridge (Fig. 2.8) is a 156-m-long superstructure with overall depth of 1.0 m, twin box-girder cross-section, and a main stayed span of 66 m [24,25]. It was incrementally launched onto six tracks of the FS railways in Milan, Italy, with the use of temporary piers that reduced the launch spans down to about 22 m. Site restrictions required the use of hinged-pendulum temporary piers that were stayed to external restraints. Refined geometry control procedures have been adopted to compensate for the settlements produced by the continuous passage of trains, [26,27]. 2.2. Static system and seismic design On completion of launch, the superstructure is a continuous beam supported on low-friction bearings and transversely restrained by the lateral launch guides. This temporary support configuration has to be modified to achieve the permanent static system of the structure, which will govern its seismic response in the longitudinal and transverse direction. Several solutions are possible in the longitudinal direction ¢ The launch bearings may be replaced with PTFE sliders that transfer small horizontal frictional forces to the piers. This solution simplifies pier cap details and avoids the need for anchor pins between the bearings and the Fig. 2.8. Launch of the first segment of the twin-box-girder superstructure of Palizzi Bridge between the vertical towers (author)28 BRIDGE LAUNCHING superstructure, with related geometric tolerances that are hard to respect with a launched bridge. In this case, the longitudinal forces resulting from the seismic response of the superstructure are mainly transferred to the abutments. Sometimes one abutment is able to resist the whole force; in most cases, however, high peak ground accelerations (PGA) and/or long and heavy PC superstructures require the contribution of both abutments. The seismic demand on abutments may be reduced with dampers activated by sacrificial shear keys. In this case, the significant displacements of the superstructure require large expansion joints, since sacrificial joints might complicate the transit of emergency services after a major earthquake. As an alternative, both abutments may resist longitudinal seismic forces rigidly. This behaviour can be attained by creating an expansion joint in the middle of the bridge or by adding lock-up devices (LUD) to the abutment sliders. LUD are active at the high velocities of the dynamic response of the bridge (not only to seismic demand, but also to braking forces and wind) and practically inert at the lower velocities of thermal expansion. Compared with the use of abutment dampers, longitudinal seismic displacements of the superstructure are much smaller, but the reduction in the seismic demand produced by increased damping is lost. Hinge bearings may be inserted at some piers. This complicates the anchor devices of bearings but permits a better distribution of the seismic response forces between the piers and the abutments. As for the abutments, hinge bearings can be obtained by adding LUD to conventional PTFE sliders so that longitudinal displacements are restrained only during the dynamic response to impulse actions. With the opposite intention, the seismic demand on a hinged superstructure can be reduced by locking the PTFE sliders at some piers with sacrificial shear keys and by equipping them with dampers. Breakage of the shear keys lengthens the vibration periods and the intervention of dampers increases the equivalent system damping, with synergistic effects on the seismic demand. In this case, the shear keys resist static and dynamic service loads. The PC superstructure may be rigidly joined to one or more piers. Vertical reinforcing bars screwed to bar couplers embedded into the pier cap can be embedded in support diaphragms cast in situ that fill the solution of continuity between the pier head and the superstructure. Vertical prestressing bars or tendons may clamp these horizontal joints to locally increase the axial force. This eases control of the hinging sequence of the structure during the design level earthquake (DLE) and decreases the ductility demand at these joints, Of course these interventions are much simpler in the case of composite bridges with steel pier Seismic isolation of the superstructure [28-30] may be attained by replacing the launch bearings with low-damping rubber laminated bearings, lead-plug bearings, high-damping rubber systems, friction pendulum systems and similar devices. Low-damping rubber laminated bearings integrated with abutment INCREMENTAL LAUNCHING OF BRIDGES 29 dampers have been used for the twin launched superstructures of the Megalorema Bridge in Greece. In the transverse direction, it is almost always necessary to anchor the super- structure to the piers. When the standard guides of PTFE sliders are insufficient to resist transverse response forces, sliding shear keys may be anchored to the bottom slab of the superstructure. These shear keys slide into longitudinal guides embedded into the pier caps, so that their transverse action does not restrain longitudinal displacements. These devices have been used along with longitudinal LUD in the Greveniotikos Bridge in Greece [31]. As an alternative, seismic isolation of the superstructure works equally well in all directions; in this case, however, the piers are almost always to be designed for an elastic response to the DLE demand. As regards the analysis methods, a non-linear time-history dynamic analysis is often necessary with isolated bridges [28]. When inelastic structure response is achieved with plastic hinges at the pier bases (and at the pier tops in the case of continuity between piers and superstructure), the use of the force reduction factors provided by codes is often sufficient. However, the adoption of a single-degree-of- freedom model may be excessively approximate for these structures [31], especially in a case of moderate seismicity. In these cases, a non-linear pushover analysis [32] is strongly advisable as it permits: analysis of the non-linear response of the bridge from the first pier hinging to the dynamic equilibrium condition during the DLE and then up to the ultimate limit state © evaluation of the progressive period lengthening and damping increase produced by structural deterioration and a method for taking these parameters into account in determining the seismic demand on the bridge © determination of the onset sequence and progression of plastic hinges and the related pier stiffness distribution to be used for spectral analysis e determination of the force reduction factors to be used in spectral analysis analysis of the non-linear loading conditions of plastic hinges and assessment of their rotation capacity and shear strength © determination of the overall safety factor of the structure from progressive collapse condition. 2.3. Geometric constraints During launching, the superstructure is a continuous beam supported on launching bearings and transversely restrained by lateral guides that prevent drifting movements. Any constraint eccentricity (vertical misplacement of launching bearings or transverse misplacement of lateral guides) with respect to the theoretical launch alignment causes secondary stresses. These can be tolerated only if small, and often with overdesign. Constraint eccentricity also causes launch problems such as excessive wear of neo-fion pads and frequent jacking for removal of damaged30 BRIDGE LAUNCHING bearings, with related higher labour costs. Therefore, it is necessary that the launch bearings and the lateral guides be perfectly aligned with the surfaces of the super- structure they come in contact with during launching. This can be guaranteed in each launch position only in the case of a common geometry. For one solid to slide rigidly inside another, it must be superimposable on itself by rotation, translation, or rotation-translation, Therefore, its alignment must be: (1) tangent in plan and tangent or circular in profile; (2) circular in plan and horizontal in profile (no launch gradient); (3) circular in a plane inclined with respect to the horizontal plane; or (4) curvilinear both in plan and in profile (Fig, 2.9). In cases (3) and (4), the projections of the superstructure on to the horizontal plane are not arcs of a circle but ares of an ellipse. In case (4), this distortion also takes place in the vertical cylindrical surface that contains the axis Fig. 2.9. Incremental launching of the Val Restel Bridge with a 150-m radius of plan curvature (courtesy GeCo) INCREMENTAL LAUNCHING OF BRIDGES 31 of the superstructure. These aspects should not be disregarded when defining the plan layout of the piers and when positioning the launch bearings and the lateral guides, especially in a case of significant curvature [33]. A last case is (5) a circular helix, very rare and substantially comparable with the above even when the radii of curvature are small. These geometric constraints and the uniform distribution of launch stresses generally suggest the launch of constant-depth superstructures, Minor adjust- ments in the vertical alignment of the launch surfaces have often been attained by temporarily attaching shims (hard wood or precast concrete planks) under the bottom surface of webs of PC box girders. This solution generally aims to cover small differences in elevation due to transverse rotations of the bottom slab. Steel plate or trussed extensions have been applied under the bottom flanges of variable-depth steel girders. The adoption of these solutions with PC super- structures is complicated by the higher dead load during launching. Temporary piers are often necessary to decrease the launch stresses in the final mid-span sections, and the adjustable casting cell is more expensive. The cost of the precast concrete shims and of their anchoring and final removal further limits the overall advantages of the solution with respect to conventional balanced cantilever construction. In bridges with shorter bank spans and progressively longer spans towards the centre, varying-depth superstructures have been attained by using a vertical curvature radius for the bottom slab different than that of the deck slab. The Ile Falcon Bridge [34] in Switzerland was launched with a convex curvature radius of 24.900 m for the top slab and 60 000 m for the bottom slab. Over the total 720-m length of the bridge, this permitted increasing the total depth of the PC box girder from 2.1 m at the abutments to 3.7 m in the middle of the central 73-m span. Although some types of launching bearings can be adjusted in the transverse direction and-adapted to varying-widthr cross-sections or varying pian curvatures, when designing a bridge (and a launched bridge in particular) it is always advisable to place the varying curvature zones outside the bridge itself. If a transition curve affects the end section of the superstructure, the lower corners of the cross-section (the only zones of the superstructure that come in contact with launch restraints) can be designed with a constant curvature. Then, the final alignment of the deck slab can be attained by locally adapting the lateral cantilevers, since transition curves begin with radius variations so small that these corrections are generally acceptable. Symmetrical launching from the opposite abutments with central closure permits covering major differences in curvature. As an alternative, the casting yard may sometimes be placed in a central position to launch one section of the superstructure on one side, and eventually a second section on the opposite side with different plan curvatures, with optimum amortization of the casting installation. In fact, incremental launching involves some constraints that the designer must take into account from the early stages of design. On the other hand, technological progress has permitted realizations that were unimaginable just a few years ago,32 BRIDGE LAUNCHING and the technological implications of these high-tech construction methods will continue to influence the design aspects more and more. 2.4. Launching techniques The handling of a PC superstructure involves enormous forces and requires the guidance and control of large volumes. Although lighter, a steel girder basically involves the same launch problems, and lower thrust forces are in this case complicated by much larger deflections and rotations of the support sections. The force necessary to launch the superstructure is proportional to its weight, as both friction resistance and the longitudinal force produced by the gradient of the launching surface are a function of it. In the longest PC bridges this force can easily exceed 10 MN, although in most cases it is only a few MN. This is nearly the same strength as prestressing tendons. Therefore, in the first PC launched bridges it was natural to use the prestressing materials and equipment already available in the yard. From here came the development of drawing devices composed of one or more prestressing jacks anchored to a foundation and acting on strands or bars anchored to the superstructure. With time, some inconveniences of the drawing systems stimulated the develop- ment of thrust devices, some applicable to relatively modest loads, some suitable for higher loads. As a consequence, the launch equipment currently used has mechanical characteristics, power and cost so varying as to substantially define its field of utilization [35]. The use of winches is limited to the lightest loads (short steel structures) and is progressively being abandoned. The less expensive hydraulic devices apply a tow force by means of strands or bars and are suitable for light superstructures (short PC bridges, steel girders with orthotropic-plate deck or for composite superstructures, concrete deck slabs launched onto previously launched steel girders). Intermediate launcher types apply the thrust force by self-clamping to longitudinal continuous restraints and are suitable for prestressed composite bridges and medium-length PC superstructures. Finally, the most expensive devices transfer the launch force by friction and are generally reserved for the movement of large masses in short times, or to solve particular control requirements of the launch forces. Since friction launchers are extremely adaptable, extended amortization conditions may be sought in short PC bridges as well. Second-hand launch equipment of all types is easily available in the EC. Independent of the transmission modality of the launching force to the super- structure, every type of launch equipment requires the presence of an anchor element restrained to the ground. The foundation of the abutment is often suitable for this use, since it is usually dimensioned to resist most of the longitudinal service loads. When the launching force exceeds the capacity of the abutment (a condition not infrequent in the final phases of uphill launching of PC bridges), overdesign may be avoided by joining the foundation of the casting yard with the abutment to create a large concrete bed that cooperates by friction. In case of need, however, specific foundation blocks for the launching jacks can avoid the application of any horizontal forces to the abutment. INCREMENTAL LAUNCHING OF BRIDGES 33 2.4.1. Launching of light superstructure The launch of light superstructures (short PC bridges, medium-length prestressed composite bridges and even long steel girders for non-prestressed composite bridges) is often characterized by the use of inexpensive equipment, because of the low launch forces and the necessity of assembling and dismantling the yard rapidly and with low costs. Drawing by means of prestressing bars or strands is the most common solution, although it presents some disadvantages, especially when launching along inclined surfaces. In fact, the ordinary longitudinal grade of highway bridges falls within the range of variation of the steel-Teflon friction coefficient, 14%, Therefore, when the launch occurs with new or well-greased neo-flon pads, it can be necessary to brake the superstructure during launching or to lock it during the construction of the new segments to prevent uncontrolled sliding. As an alternative to towing, thrust devices are made up of hydraulic pistons that act directly against the rear end of the superstructure by self-clamping to longitudinal kerbs or by locking in racks anchored to similar kerbs. In the case of PC and prestressed composite bridges, these kerbs support the superstructure during segment casting in situ or precast segment assembly. In the case of steel girders, these kerbs are often necessary only for launching, so this solution is expensive and the kerbs are replaced by bars or strands anchored to the abutment, resulting again in a tow system. Back thrust systems have the same weak points as tow systems when launching along inclined planes. None of these devices permit moving the superstructure backwards in a case of excessive advance or launch problems, nor its braking when launching downhill. Therefore, these launch systems should be used with prudence. 24.1.1. Tow systems Coupled prestressing bars and hollow hydraulic jacks anchored to the super- structure (Fig. 2.10) or to the abutment (Fig. 2.11), have been used many times with PC and steel girders to reach launch forces of 0.8 to about 1.5 MN. The short stroke of conventional jacks for prestressing bars, usually about 200 mm, suggests placing two jacks on each bar (Fig. 2.10) so that they alternate. During the stroke of the first jack, the second one returns to idle to be ready to pull the bar when the first jack reaches the end of its stroke, avoiding bar relieving. This scheme requires the use of special guillotine anchorages, and the bar couplers must pass through the central hole of the thrust jacks. The use of single-thrust jacks with bar relieving at the end of each launch stroke should be limited to almost horizontal launch surfaces. The stroke necessary for the new bar stressing is lost in each cycle and this lengthens launch duration, If the casting/assembly yard is long, bar elongation at tensioning may result in excessively short effective strokes of the thrust jacks. In this case, the jacks may be designed for launching, with an effective stroke of about 1m (Fig. 2.11). Their cost is reasonable, although amortization is necessarily limited to launch operations.34 BRIDGE LAUNCHING Fig. 2.10. The twin-jack back-thrust system for the steel U-girder of the Schrotetal Bridge was designed for a total force of 1.3 MN and 210-mm launch strokes (author) In the case of launching uphill, the superstructure must be locked during each return stroke of the tow jacks to avoid uncontrolled backwards sliding. The modest dead load of the superstructure usually allows utilization of the drawbars themselves for this purpose. The use of two jacks on the same bar (Fig. 2.10) avoids any problem, as drawbars are not relieved. In the case of single jacks, the drawbars have sometimes been divided into coupled segments as long as the jack stroke. When the jack reaches the end of its stroke, the bar is locked to an anchorage, the jack is released, its rod is reentered, the superfluous bar element is eliminated and the new bar head is connected again to the jack. This system is effective only in the case of long-stroke launch jacks and requires a lot of bar couplers. When conventional prestressing jacks are used, the less expensive solution is to locate them in cylindrical saddles that permit bar anchoring at their front end and avoidance of bar relieving during jack re-entry. Drawbars are placed under (Fig. 2.11) or next to the superstructure (Fig. 2.10). In the case of high launching forces, drawbars are substituted with tendons of strands. In principle, by using strands the tow force is limited only by the capacity of the jacks, the addition of further strands always being possible. However, it is generally convenient not to exceed the force of 2.5 MN in each tendon in order to use ordinary 19T15 prestressing jacks with adequate safety factors. If needed, it is better to increase the number of tendons, although this causes problems of load distribution that require electronic controls or powerful hydraulic pumps for in-parallel jack feeding with reasonable launch velocities. Strands in each tendon INCREMENTAL LAUNCHING OF BRIDGES 35 Fig. 2.11. Long-stroke bar jack anchored to the abutment (reprinted with permission, Dywidag Systems International) are individually pretensioned before the beginning of launch (isotensioning by ‘combing’). However, when pulling launch tendons with conventional short-stroke prestressing jacks, tens of repetitions of the tensioning cycle compromise the initial isotensioning and require corrections of the force in the strands during launching. Although these operations are generally carried out at the anchor end and do not require dismantling of the launch jacks, they cause significant time losses and involve some risks in the case of steep launch grades. The progressive loss of strand isotensioning also suggests adoption of high safety factors when designing launch tendons. There are numerous possible schemes for applying the thrust force to the super- structure, and physical principles for load transfer. The use of friction between the superstructure and the anchor plates of draw tendons (mobilized by hydraulic local compression as in Fig. 2.12, or by prestressing bars crossing the superstructure as in Fig. 2.13) is limited to short PC bridges launched along horizontal planes, which require minimum draw forces. Higher thrust forces or steel girders require mechanical transmission: transoms anchored to the rear end of the superstructure (Fig. 2.10) or through pins (Fig. 2.14), In the case of a PC bridge, these load-transfer devices induce high36 BRIDGE LAUNCHING anchor plate neoprene. contact surface Jack ol 8008 L return stroke raller Fig. 2.12. Friction transfer of the thrust force by means of hydraulic compression concentrated stresses in regions of the superstructure that often are already loaded by the anchorages of prestressing tendons. Therefore, transfer of launch forces requires careful design and proper local reinforcement. At the end of each construction cycle, the transmission devices must be relocated to the rear end of the superstructure and the draw tendons must be brought back to their initial length. Their weight makes this operation difficult, and in the case of long bridges it may justify the adoption of electrical systems with service trolleys [36]. The problem of superstructure locking in uphill launching can be solved, as with drawbars, by locking strands during the return stroke of the launch jacks. Although few prestressing jacks have two opposite strand clamps to avoid tendon relieving, the beam hangers [>> prestressing bars ROIRUIRTTIRTIIG, launching bean draw tendon Fig. 2.13. Friction transfer of the thrust force by means of prestressing bars INCREMENTAL LAUNCHING OF BRIDGES 37 draw tendon Fig. 2.14. Through pin Jaunch jacks canbe equipped with conventional anchor plates to lock strands at the end of each tensioning cycle. This solution involves minimum costs but does not permit controlled backwards sliding of the superstructure in case of need. As an alternative, it is possible to use two oversized draw cables and to take advantage of the difference between launch and locking forces, by relieving the first tendon while keeping the second one tensioned. In other terms, two tendons are necessary for launching, but they are designed so that one tendon is sufficient to prevent back- wards sliding with the necessary safety factor. In the heaviest bridges it is preferable to avoid anchoring the superstructure by means of the draw tendons and to adopt specific friction-lock devices. These guarantee a higher safety when it is necessary to replace some strands or to re- equilibrate the draw tendons because of strand slippage in the jacks. Finally, friction locks are always necessary during construction of the new segment when rear transoms are used to launch the superstructure. Friction-lock devices usually involve hoisting of the superstructure at the abutment by means of jacks equipped with knurled contact plates able to mobilize a high friction coefficient against the bottom surface of the superstructure. This38 BRIDGE LAUNCHING locking system has a varying effectiveness, locking force being proportional not only to the friction coefficient, but also to the continuous beam support reaction at the abutment. This remains practically constant during most of launching but decreases considerably in the final phases, just when the weight in movement is at its highest and, with it, the sliding force along the launch direction — that is, the force to balance. In some heavy PC superstructures, friction anchoring has been attained with horizontal jacks pressing against the opposite sides of the bottom slab, so that the transverse locking force does not depend on the support reaction of the super- structure. In this case, lock safety depends on contact pressure, i.e. on an artificial force that may vary over time. Therefore, jacks should always be equipped with safety ring nuts and the time-dependent transverse shortening of the bottom slab should be accounted for when determining the minimum transverse force necessary to create a safe lock action. In the case of steep slopes, it can be convenient to launch downhill maintaining the superstructure braked, as the braking force is much smaller than the draw force required for uphill launching, friction working in favour of equilibrium. In reality, however, the possibility of pulling the superstructure backwards should be taken into account when sizing braking tendons, since this facilitates both the release of the temporary locks of the superstructure and the correction of an accidental excessive advance. The draw systems present some limits when the bridge has a curved plan align- ment, as the launch force is applied along the chord subtended by the jacks and the draw tendon anchorages, and not along the local tangent at the bridge alignment. This produces angular discontinuities in the draw strands both at their entry into the jacks and at the anchorages, which must be reduced by deviation devices (Fig. 2.15). In this case, the strands cannot slide inside deviators (friction would be excessive) and it is necessary to place the launch jacks at the rear movable anchorage 2m phase formwork launching bearing 1* phase formwork draw tendon tendon deviator lateral guide continuous foundation draw jack Fig. 2.15. Deviators of draw tendons and guide devices are necessary in the case of curved plan alignment INCREMENTAL LAUNCHING OF BRIDGES 39 (Fig. 2.16). Therefore, tendon deviators must be progressively disassembled to per- mit the jack passing. Even so, the tow force applied along the subsequent chords tends to cause drift phenomena in the superstructure, which require the presence of many lateral guides (Fig. 2.15) in the formwork area. Finally, the drawbars or cables cannot be placed alongside the superstructure (a less expensive location because of its good accessibility) as they touch the bridge along the external side of the curve. To conclude, in spite of the special attention required in particular cases (steep grade, curved alignment) and a certain operative slowness, tow systems remain the least expensive and most efficient solution for the launch of short PC bridges and even long steel girders. They require simple and reusable equipment and are easy to assemble and dismantle. 2.4.1.2. Self-clamping systems As an alternative to draw systems, hydraulic pistons can be anchored directly to the rear end of the superstructure to transfer the thrust force by acting against rigid restraints. Rear thrust requires anchoring the launch pistons at many locations along a continuous restraint. The longitudinal kerbs that support the forming system of bridges cast in situ or the precast segments of segmental bridges allow resisting the thrust reaction by means of friction mobilized by hydraulic compression Fig. 2.16. The rear launch system for the Tiziano Bridge was designed for a 3.2-MN thrust force attained with two launch tendons, one of which was sufficient to avoid uncontrolled backward sliding (author)40 BRIDGE LAUNCHING (Fig. 2.17). As an alternative, a continuous mechanical contrast (Fig. 2.18) is more expensive but more reliable in a case of uphill launching. The use of self-clamping systems is generally limited to light structures launched along a horizontal alignment, as the high thrust forces required for heavy super- structures or steep grades cause several difliculties: © Pulling of the superstructure backwards is practically impossible (as with the tow systems) as this would require bidirectional anchor devices between the thrust pistons and the superstructure in every construction joint. Therefore, recovery of the elastic launchwards deflection of a pier caused by the seizing of a neo-flon pad is an extremely delicate operation, and the wrong insertion of a neo-flon pad requires hoisting of the superstructure. Both these operations increase the launch duration and immobilize labour on the piers. © The presence of rear pistons slows down the formwork set-up and interferes with the distribution of the joint reinforcement and the prestressing tendons. # poopy release, Forward stroke Fig. 2.17. Rear thrust pistons with hydraulic clamps INCREMENTAL LAUNCHING OF BRIDGES 41 | oC Fig. 2.18. Rear thrust pistons with mechanical contrast in a continuous rack e The flexural stresses produced by the inclination of the thrust pistons suggest launching the whole cross-section, and the use of paired forms (bottom slab and webs in the rear formwork and deck slab in the front one) is discouraged. The casting/assembly yard has to be located nearby the abutment to transfer the thrust reaction directly to its foundation. This causes large rotations in the joint sections between the segments of the superstructure, whose reduction often requires the insertion of a temporary pier in the bank span, with additional costs. © At the beginning of each launch, a high thrust force (static friction) is applied far from the abutment. In the foundation system of the casting/assembly yard, hinged at the abutment, the moment produced by this long lever arm may produce rotations that jeopardize the vertical alignment of the whole casting yard, unless expensive indirect foundations are used.42 BRIDGE LAUNCHING In the case of light structures launched along a horizontal alignment, however, the use of rear thrusters offers some advantages over tow systems. Launching is much faster: this is a basic issue because two or three people are necessary at every support point during the launch, and labour saving can soon cover the higher cost of launch equipment. Launch is easier in the case of curved plan alignment as well. Finally, the use of an anchor rack permits hydraulic braking of the super- structure during downhill launching. The braking system of Fig, 2.19, anchored to Fig. 2.19. Rear braking system for steep-slope downhill launching INCREMENTAL LAUNCHING OF BRIDGES 43 the superstructure by two temporary tendons, has been studied for a downhill launch on an 8% slope in Brazil. On the whole, the investment required by these thrust devices, certainly functional for light or medium-weight bridges, increases quickly with the weight of the superstructure and soon becomes uneconomical. Their optimum field of utilization is limited to medium-length PC bridges, long U-girders for composite bridges and prestressed composite bridges. In long PC bridges the rear thrusters may be combined with strand-drawing systems sized for kinetic friction, so that the rear pistons cooperate only in overcoming the static friction at the launch beginning. However, long and heavy PC superstructures suggest adoption of launch devices based on transmission of the thrust force by friction, 2.4.2. Friction launching of heavy superstructures In the heaviest bridges, the thrust force is transferred by friction using one or more pairs of ‘launchers’ placed under the webs of the superstructure. In the simplest version, a launcher is composed of a vertical jack pushed along a lubricated surface by a horizontal piston. The piston acts against a foundation block that supports the superstructure in the pauses between two subsequent launches (Fig. 2.20). In the most advanced version, a friction launcher is a monolithic device composed of a sledge containing the vertical jacks. The sledge slides between Fig. 2.20, Basic configuration of a friction launcher (author)44 BRIDGE LAUNCHING longitudinal guides along a steel~Tefion base under the thrust of one or two pistons pivoted to a rear steel block (Fig. 2.21) (notice the level of surface finishing that can be easily reached with incremental launching [37,38)). Further details of this axial launcher can be found in Fig, 2.22, which illustrates the standard working sequence of friction launchers: © Using the vertical jacks, the superstructure is hoisted a few millimetres from the support block of the launcher (or placed near the launcher). As a consequence, the support reaction of the continuous beam is transferred from the support block to the movable launcher sledge. Using the thrust pistons, the sledge advances by sliding on a low friction surface. The thrust force is transferred to the superstructure by friction, by means of the support reaction of the continuous beam on the sledge and the high friction coefficient between the bottom surface of the superstructure and knurled steel plates placed above the jacks. e When the piston limit stop is reached (after an effective stroke that varies from 0.25 to 1.00 m), the superstructure is lowered on to the support block by relieving the hoisting jacks. © The sledge, unloaded, returns idle to the initial position to start this cycle again. Fig. 2.21. Front view of an axial launcher for fish-belly cross sections. Raising capacity is 8.4 MN and the thrust force is 6.4 MN (author) INCREMENTAL LAUNCHING OF BRIDGES 45 LIFTING THRUST LOWERING RETURN Fig. 2.22. Working cycle of a friction launcher46 BRIDGE LAUNCHING For the superstructure to be trailed by the launcher sledges it is necessary that the ratio of the thrust force Fy necessary to move the superstructure to the vertical force Ry; transferred between the superstructure and the launchers be smaller than the friction coefficient C;,, between the launcher sledges and the bottom surface of the superstructure. Otherwise, the hoisting sledges would slip under the super- structure, damaging its surface, but without succeeding in moving it. Fr =< Q1) Ro The vertical force Ry. is the continuous beam support reaction at the launchers. It depends on the static system of the superstructure, i.e. on the position of the launchers with respect to the adjacent supports. In very long PC bridges, therefore, high thrust forces require a significant distance between the abutment equipped with launchers and the front support of the curing area. This requirement restrains the distribution of the temporary supports in the curing area and the sequence of introduction of the launch prestressing. Ry, is generally calculated by considering aligned supports, as the launchers hoist’ the superstructure only a few millimetres. If it is insufficient, forcing the superstructure upwards can increase this reaction. The related stresses must be checked for the conditions of partial launch prestressing in these segments, and partial curing in the case of casting in situ. Hoisting jacks are generally oversized with respect to the highest theoretical values of Ry, to be able to hoist the superstructure, even when support mis- placements or thermal gradients increase Ry. Thrust pistons are designed for the highest launch thrusts considering the friction in the casting/assembly bed, in the launching bearings and in the launchers themselves, in addition to the average longitudinal gradient of the launch surface. The use of a pair of friction launchers anchored to an abutment involves some specific considerations in the initial and final phases of launch. At the beginning of the launch of a PC bridge, the curing area separates the formwork from the abutment equipped with launchers. The need for a perfect connection between the launching nose and the superstructure is generally solved by match-casting, ic. by assembling the launching nose against the front head of the formwork. This requires the launching nose and the front section of the superstructure to be trailed on to the launchers until a support reaction sufficient to continue the launch by friction is obtained. For this purpose, strands anchored to the launching nose and to the friction launchers permit using the thrust pistons of the latter for these first launch phases as well (Fig. 2.23). e In the final phases of launch, Ry, decreases until the reaction of the continuous beam end support is reached. This vertical force is generally insufficient to satisfy the highest thrust force demand (2.1) by friction, especially when launching takes place uphill. Figure 2.24 describes the evolution of the support INCREMENTAL LAUNCHING OF BRIDGES 47 Fig. 2.23. Initial launch phases of the Aronde Bridge. These friction launchers have been designed for a total raising capacity of 16.8 MN and a thrust force of 12.8 MN (author) reaction on the launcher of Fig. 2.21, as recorded during the final launch phases of the Serio River Bridge. The continuous line describes the theoretical evolution of the continuous beam support reaction, black dots have been recorded at the launcher, the support reaction needed for launching with the recorded friction (C;, = 0.029) is dotted, and the effects of overhoisting (forcing) between i3 m and 4 m from the end of launch are evident. When the lack of vertical force affects long launch phases (uphill launching along a steep gradient), a second pair of friction launchers may be placed on a more advanced support (generally a short pier). However, synchronization of two or more pairs of launchers is a rather delicate task, especially when their distance is significant. When the lack of vertical force affects only the last launch phases, it is normally simpler to use a steel rear-nose integral with the superstructure and connected to the launchers by draw strands (Fig. 2.25). In this case, the draw tendons are designed only for a share of the final thrust force, as the continuous beam has a support reaction to exploit by friction (Fig. 2.24). In addition, when the rear end of the superstructure approaches the abutment, launching is no longer affected by the extraction friction from the formwork and the friction at the temporary bearings of the curing area (which together represent a significant part of the total friction). As an alternative to the rear-nose, front drawing with strands and prestressing jacks permits solving the problem in a less expensive way (Fig. 2.26), although with some interference with the bracing system of the launching nose.48 BRIDGE LAUNCHING support reaction at the launcher (MN) =40 =20 0 progression toward the end of launch (mn) Fig. 2.24. Decrease in the abutment support reaction in the last launch phases of the Serio River Bridge The use of friction launchers (single as in Fig. 2.21, paired as in Fig. 2.23 or distributed in more pairs) controlled by numeric-control boards (Fig. 2.27) may result in enormous synchronized thrust forces. This permits moving huge masses, such as those of the longest railway PC bridges. Fig. 2.25. Rear nose and integrative strand pulling in the last launch phases of the Serio River Bridge (author) INCREMENTAL LAUNCHING OF BRIDGES 49 Fig. 2.26. Final front pulling of the Palizzi Bridge (author) Compared with draw and back-thrust systems, the use of friction launchers offers many advantages: © In draw systems, the breaking of any component of the launch system (bars or strands with related anchorages and couplers, anchorages of the thrust transoms to the superstructure, thrust jacks, hydraulic plant) leaves the superstructure free on iow-friction supports. This requires adoption of high safety factors that involve oversizing of equipment and operative slowness. Friction launchers, on the contrary, guarantee absolute intrinsic safety, as the worst consequence of mechanical or hydraulic faults is halt of the launch and descent of the super- structure on to the temporary supports near the launchers, which are designed as friction constraints. Therefore, equipment can be designed for the launch loads and can be overloaded without excessive worry in case of unforeseen requirements. © The electronic control of the hydraulic plant permits synchronization of the action of launchers and recording in real time of the forces produced. By monitoring Ry, it is possible to verify the accuracy of superstructure geometry (some types of geometric errors modify the support reactions of the continuous beam, as discussed in Section 3.5.2) and to correct these errors prior to their accumulation. By relating Ry, to the raising of the superstructure, it is possible to estimate indirectly the elastic modulus of the young concrete. The possibility of setting an upper limit for the thrust force permits avoidance of pier overloads produced, for instance, by the upside-down insertion of a neo-flon pad. This50 BRIDGE LAUNCHING Fig. 2.27. The numeric-control board of the friction launcher of the Serio River Bridge allowed pre-setting and real-time monitoring of launch parameters (forces and dis- placements) (author ) mistake may seem ridiculous but in fact it is rather frequent and potentially very dangerous, as the friction coefficient increases dramatically when passing from a greased steel-Tefion contact to a steel-neoprene one. In addition, pier workers are often so hopeful of successfully removing the seized pad (in most cases this is practically impossible) that launching is stopped when the pier is so significantly deflected launchwards that it begins vibrating. Without friction launchers, pulling the superstructure backwards is impossible. Therefore, recovery of the launchwards pier deflection requires local raising, and the horizontal force that deflects the pier is transferred to the hoisting jacks. These emergency operations require extremely careful procedures and controls, possibly planned from the very beginning. ¢ Launch cycle automation by means of electronic end-of-stroke switches facilitates the operations even more, which can be supervised by one technician only, and increases the launch speed, which can easily reach 10 m/h. ¢ Labour costs are definitely lower because of the higher launch speed (launching immobilizes two to three people at each support) and the possibility of avoiding tow tendon repositioning and anchoring of the superstructure during these phases. Moreover, after launching, the casting/assembly yard is immediately ready for the construction of the next segment. INCREMENTAL LAUNCHING OF BRIDGES. 51 © The lateral guides of the launcher sledges also act, by friction, as transverse guides of the superstructure, and avoid the need for further local devices and the related labour demand. As an alternative, however, the hoisting jacks can be left free to shift transversally (Fig. 2.28). As with the draw or back-thrust systems, it is thus possible to launch by friction superstructures with varying width. In order to transfer the thrust force without stress concentration and surface damage, the contact plates of the friction launchers are articulated so as to adapt to the superstructure. This is usually achieved by connecting the jacks of the hoist- = launching direction corrugated lifting steel plate | Jack mortar shifting. cast-in cylinder Fixing rail lifting wing nut for lateral move Fig. 2.28. Shifting launchers52 BRIDGE LAUNCHING ing sledge in series and by equipping them with ball-and-socket caps. In the case of PC superstructure, the contact plates should also be relatively ‘soft’ to compensate for surface irregularities of the bridge. Their surface is generally engraved in the opposite direction of the launch (Fig. 2.29), so that the thin plates thus obtained can transfer thrust by bending. The opposite arrangement might easily break the concrete surface because of stress concentration. Similar design concepts can be used for the launch of steel girders; of course, in this case, the bottom flanges of the steel girder will be free from splice plates and bolts in the working areas of the friction launchers. The original vertical-load-based friction launchers have evolved into several different schemes. The launch system used for the Olifant’s River Bridge was composed of longitudinal thrust pistons acting on a movable transom placed under the superstructure. The transverse jacks of the movable transom clamped the bottom slab of the PC superstructure and created the contact pressure necessary for friction transmission. The work cycle, similar to that of a vertical-load launcher, involved clamping of the superstructure, thrust, release, and transom return to the initial position. Not depending on the continuous beam support reaction, this scheme has the advantage of not restraining the design of the casting/assembly yard and not breaking down in the final launch taunching direction contact plate UFting jack Fig. 2.29. Contact plates for PC superstructures INCREMENTAL LAUNCHING OF BRIDGES. 33 phases. This system is perfectly suited for horizontal launching; completed with synchronized lock jacks, it can be used in the case of inclined launching as well (both downhill and uphill). Friction launchers with hydraulic clamps are often used for the launch of steel girders [39], although their transfer devices often damage paint. Flat hydraulic pliers clamp the bottom flange of the steel girder and the launch cycle is similar to that used for the Olifant’s River Bridge. Also in this case, launching along inclined planes requires prudence and experience. 2.5. Launch bearings and guide devices The launch of PC, prestressed composite and conventional composite bridges requires different types of launch bearings that comply with the specific charac- teristics of the webs of the superstructure. Launch bearings are described in specific chapters related to the different types of bridges. Lateral guides are also necessary during launching, to maintain the correct plan alignment of the superstructure and to resist the transverse forces produced by wind and the seismic response during construction. The prevailing function of the lateral guides is related to the depth and mass of the superstructure as well as to the plan alignment of the bridge. Large masses such as those of PC bridges require strong lateral guides during launching, especially when it takes place along a curve. Strong guides are also necessary to resist the seismic response of the structure during the design-level earthquake (DLE) for construction phases; the effects of transverse wind are generally modest. Conversely, design of the lateral guides for the lighter steel girders of composite bridges is often governed by transverse wind. Alll the piers of a tangent bridge are equipped with lateral guides during launch- ing although the correction of alignment of the superstructure occurs in most cases by acting at three points only: © near the front end of the casting/assembly yard to correct the position of the construction joint with the next segment eat the launch abutment (especially with draw systems) © at the front-most pier, to adjust the direction of the launching nose at its contact with the next pier. The design load of lateral guides is often empirical, being based on the force necessary to move the superstructure by acting at the most appropriate points during each launch stage. To account for the need for adjustments in the bridge alignment during launching, some standards for PC bridges (AASHTO, ZTV-K88) require that the calculated horizontal forces acting on the lateral guides be increased by 1% of the local support reaction of the continuous beam. At least 3% of the maximum support reaction should be used when launching the steel girder for a composite bridge [39]. When horizontal or uphill launching takes place along an alignment curved in plan, the external lateral guides are also subjected to radial forces resulting from the54 BRIDGE LAUNCHING local reaction to the thrust force. This system of forces (of elastic nature and therefore proportional to the transverse flexural stiffness of the piers) can be analysed by using a model of the continuous beam placed on elastic supports. The forces thus obtained are added to those estimated for the alignment adjustment of the superstructure. When launching downhill, the superstructure is braked, the axial launch force in the superstructure turns from compression to traction, the direction of the guide forces is reversed, and lateral guides located on the interior face of the girder are required. When the gradient of downhill launching is intermediate between breakaway and kinetic friction, the superstructure must be pushed until detaching, and braked during sliding. Therefore, direction of the radial guide forces inverts during launching [34]. The design of lateral guides for wind action is simpler. The design standards provide criteria for evaluation of the transverse load due to wind. This load is applied to the superstructure with the most unfavourable distribution to determine the transverse forces to be resisted at the piers. Transverse wind generally governs the design of lateral guides for steel girders, but it may often be neglected when launching much heavier PC superstructures. Alignment forces are added to wind reactions determined for the maximum wind speed allowed during launching. Exceptional wind conditions might require interruption of the launch of steel girders. In this case, they may be assessed without alignment forces. The design of the lateral guides for the seismic response of the structure during the DLE for construction phases is more complex. Several seismic design codes determine the seismic demand in terms of peak ground acceleration (PGA) and spectral modification factors that depend on the importance of the structure and on the site classification. The PGA, ay, for the seismic design of the permanent structure often reflects a seismic design event with a return period f,.9 of approxi- mately 475 years. Accordingly, the design event has a probability of exceedance p, of about 10% and 19% for a design life of the siructure of 50 and 100 years, respectively. Evaluation of the seismic demand for construction phases requires determination of the probability of exceedance during construction, p,.., and of the construction duration f,. EC8 [40] suggests that p,,. be not greater than 5%. The return period of the DLE for construction phases, f;,., can be determined as [40] 1 and the related PGA during construction, ag.¢, is [40] a gc = «.(#°) (2.3) The coeflicient k varies between 0.30 and 0.45 according to the reliability of the seismological data. For instance, assuming a construction duration of 1 year (for the superstructure) and accepting a 5% probability of exceedance, the return INCREMENTAL LAUNCHING OF BRIDGES 35 period of the DLE for construction phases is 20 years and a,. varies between 24% and 39% of ag. In regions of high seismicity, these levels of seismic demand may be significant and have different effects in the longitudinal and transverse direction. In the longitudinal direction, the superstructure is supported on low-friction launch bearings. Neglecting friction of the neo-flon pads, the whole longitudinal response of the superstructure must be resisted with specific restraints. If friction launchers are used, they also act as lock-up devices and the problem is solved. If launching takes place by means of tow systems or back pushers, bidirectional restraints must be used to lock the superstructure even in the case of a horizontal launch surface. In both cases, the most logical location of these restraints is at the launch abutment, although this stiff restraint results in short vibration periods of the structure and increased spectral demand. Seismic restraints located at some piers might decrease the system stiffness by taking advantage of pier flexibility; however, locking operations are difficult at the top of the piers and may result in pier overdesign in the case of isolated bridges. In the transverse direction, on completion of launch, the superstructure will be connected to the piers by either transverse shear keys or a seismic isolation system. In the first case, the piers are designed for the full transverse response to a, with the additional mass of the superimposed dead load and characteristic live load; there- fore, these piers are rarely designed for the seismic demand during construction. In the case of isolated bridges, pier design for service conditions depends on the period lengthening and damping increase produced by the isolation system. Since the isolation devices are inserted only on completion of launch, the superstructure is not protected during construction phases. Also in this case, however, ag. is much smaller than ay and this is usually sufficient to avoid overdesign of the piers due to construction phase requirements. In both cases, lateral guides are designed accord- ing to the conventional capacity design principles [28], with adequate protection factors. 2.6. Launch and lock forces The horizontal force necessary to move or to lock the superstructure depends on the longitudinal gradient of the launching surface and on the friction forces that oppose imposed movements (launching) and spontaneous movements (uncontrolled launchwards or backwards sliding). The effect of the longitudinal gradient is easy to evaluate. By assuming that the self-weight of the superstructure Q is concentrated at its centroid, the longitudinal grade of the vertical profile of the bridge at this location defines the average percent launch gradient G,,. The horizontal force Fg to be applied to the superstructure to launch it or to lock it along a frictionless launch surface is Fg = GmQ (2.4) Particular care must be paid to the determination of the average gradient of the superstructure during the individual launch phases. When the vertical curvature56 BRIDGE LAUNCHING radius is small, the launch gradient can be very different from the average grade between the opposite abutments, In the Tanaro River Bridge (Fig. 2.3) the local gradient is + 5.3% at the launch abutment and —5.3% at the opposite abutment, so that the average grade of the bridge is zero. During the launch, however, Ga, was +7.2% at the beginning of the first 50-m launch (with the superstructure entirely behind the launch abutment), +5.8% at the beginning of the second launch (100-m superstructure), +4.4% at the beginning of the third launch (150-m superstructure) and +3.1% at the beginning of the final stroke. A similar situation occurred in the stayed Palizzi Bridge [26,27] (Fig. 2.8). Checking G,, in the individual launch phases is necessary for a safe design of the launch devices and the anchor systems that lock the superstructure during construction of new segments, since uncontrolled sliding can occur backwards as well. Assuming that the deck is horizontal, in a launch bearing i, the minimum tan- gential force F? necessary to produce sliding depends on the local support reaction Ry, of the continuous beam and on the breakaway (static) friction coefficient C?; of the launch bearing. Fo=ChiRy, (2.5) In the case of a PC deck, Jaunching occurs on steel~Teflon contacts. C}; depends on many factors, since the behaviour of PTFE in contact with a polished stainless steel surface is rather variable: © C?; decreases when the average contact stress increases. Recommended design coefficients of friction for bearings with stainless steel sliding on pure PTFE continuously lubricated are 8% for a bearing stress of 5 MPa, 6% for 10 MPa, 4% for 20 MPa and 3% for 30 MPa and over [41]. e Temperature: increases in temperature produce effects similar to increases in load. e Lubrication and surface cleanliness: an imperfect cleaning or lubrication of neo-fion pads and/or of the sliding surfaces, or their excessive wear, increase friction substantially. In the absence of test data, for design purposes, the coefficient of friction for pure unlubricated PTFE on stainless steel may be taken as twice the values previously given before for lubricated contacts [41]. Similar conclusions can be drawn for the rollers or sliding devices used to launch the steel girders. The total friction force F; does not depend only on the breakaway condition of launch bearings: © The breakaway forces (2.5) soon decrease towards the values resulting from the coefficient of kinetic friction, C;;, which is lower. Therefore, the piers should be designed for a breakaway friction of at least 5% (7% is more prudent in the case of tall piers), but friction must be neglected when determining the braking or locking forces (kinetic friction may be very low), as required by AASHTO [42]. In case of need, the braking devices for downhill launching may be helped by INCREMENTAL LAUNCHING OF BRIDGES s7 inserting timber plates instead of neo-fion pads at some bearings; however, these solutions should be limited to emergency interventions. © A contemporaneity factor yj <1 should be used to decrease the breakaway forces at the rear launch bearings, since the elastic shortening of the super- structure under the axial compression produced by the thrust force longitudin- ally staggers the bearing breakaway. This reduction is influential only in long bridges. ¢ The friction force opposed by the formwork, Fj, or the foundation kerbs of the assembly area for precast segments is high because of the low bearing stress and the dirt and wear caused by grout spillage. The breakaway friction coefficient of the casting cell may be about 10% in the case of a continuous stainless steel— Teflon contact, and much higher in a case of different sliding surfaces. Excessive tolerances in the vertical alignment of the support rails further increase extrac- tion friction. However, this high friction coefficient is applied only to the dead load of the deck segment to be extracted from the casting cell. The friction component of the transverse forces applied by the lateral guides is often negligible (even in the case of a curved launch alignment) when using neo-flon pads between the lateral guides and the superstructure. The use of rolls, rollers, or other sliding materials (filled PTFE, CMI, CM2, bronze, Ferrozell), frequent when launching steel girders, may increase guide friction substantially, although the dead load of the superstructure is in this case much smaller. As a consequence, the total friction force Fy does not increase linearly with the number n of launch bearings affected by the superstructure, and follows a law of the type = Fiy + vk? (2.6) 7 When sizing the thrust devices it is necessary to apply a load factor y¢ to the friction force (2.6) to account for events such as insufficient lubrication of neo-flon pads, wear of sliding surfaces, and tendency of PTFE to seize to the launching bearings during the intervals between two subsequent launches. The final friction component Fy of the design thrust force is therefore Fe = yeFt (2.7) and the total thrust force Fy is Fr=Fo+Fr (2.8) On the other hand, no resisting friction should be taken into account when design- ing the lock or brake devices, and uncontrolled sliding will be checked for a force F, = ysFa (2.9) where the load factor for uncontrolled sliding ys will be at least 1.3 [42]58 BRIDGE LAUNCHING To conclude: © During launching it is necessary to apply a system of forces able to produce (thrust) or to control (brake) the movement of the superstructure. © During the pauses of the launch it may be necessary to apply a different system of forces to avoid uncontrolled sliding. In both cases, great care must be taken in the evaluation of these systems of forces. In several bridges, the gradient of the launch surface and the average coefficient of launch friction are similar. When Fg and Fp have the same sign in (2.8) (uphill launching), the sign inversion of Fy during the stops of launching may cause back- wards sliding. When Fg and Fr have opposite sign (downhill launching), great care must also be taken in the determination of the expected range of variation of Fp. After breakaway, new and well-lubricated neo-flon pads may result in friction coefficients much smaller than 2%. In this case, downhill launching may require, at the same time, pushing the superstructure until the breakaway and braking it immediately after. Bridge launching affects huge masses, and no intervention is possible if sliding is uncontrolled. Therefore, launching along inclined surfaces requires prudence and experience, and often justifies an independent design check by specialists. 2.7. Correction of launch stresses The most characteristic aspect in designing launched bridges is the need to absorb the stresses due to the temporary support configurations assumed by the deck during launch. In fact, every cross-section of the superstructure passes cyclically in mid-span and above the piers, and is therefore subjected to the maximum positive moment, the maximum negative moment, and the maximum shear. Thus, although supporting only its own weight, each cross-section has to withstand transitory stresses that are not only very heavy, but also significantly different from the service stresses. Thermal stresses and differential support settings worsen the situation. The situation evolves between the two limit conditions of Fig. 2.30. Considering two successive piers and a typical span of the continuous beam, the first condition is the final position A, with the support diaphragms of the deck located above the piers. The second condition is position B, with the deck advanced by half a span and supported on cross-sections that, on completed launching, will be mid-span sections. In a PC bridge, in both positions, mid-span sections are rarely overloaded. Shear force is low, and the bending moment is lower than the values reached under live loads. In addition, the box girder cross-section is well suited to positive moment, since the wide top slab provides a large compressed area that draws the centroid upwards. Therefore, longitudinal prestressing reducing tensile stress at the bottom edge is generally sufficient. INCREMENTAL LAUNCHING OF BRIDGES 59 ® ASS Fig. 2.30. Limit support conditions for the current zone of the superstructure With regard to the cross-sections temporarily above the piers, in position A, the launch stresses are only a fraction of the permanent stresses, and the deck can be sized for the latter. Longitudinal prestressing is still needed to cover the upper-edge tensile stress. Web reinforcement is designed for the operational shear stresses and assessed for the launch shear stresses without the relieving effect of the parabolic tendons (whose sloping sections reduce the shear force in the webs) since these are installed only on completion of launch. On the contrary, in position B, the cross-sections that on completion of launch will be in mid-span to resist positive moments and slight shear forces are subjected to the maximum negative moment and shear. They often need to be adapted to these transitory stresses, and therefore tend to be oversized with respect to the service requirements. During launching, every cross-section of the deck passes above the piers, and the need to resist the same transitory stresses requires that both the cross-sectional moment of inertia and the web thickness be constant along the superstructure. This prevents lightening mid-span sections (an operation that with other construction methods is more and more effective as span increases, especially in constant-depth bridges) and extends oversizing to the whole deck, causing structural burdening [43,44]. In addition, the cyclic changes of sign in bending moment and shear force prevent the launch prestressing from undulating. This maintains an axial layout and limits its contribution to the uniform compression of all cross-sections. On completion of launch, this axial compression is insufficient to absorb the stresses due to live loads, and it has to be integrated with more efficient undulated tendons. It is detrimental as well, because it increases the stresses60 BRIDGE LAUNCHING in areas already compressed by the bending moment diagram of the continuous beam [45,46]. To conclude, launch stresses require numerous checks and specific types of prestressing that optimize the bridge performance during launch, reducing the need for additional material at those points where oversizing, in service conditions, would be pointless or even damaging. In a prestressed composite bridge the situation is similar. In these structures, however, the cross-sectional behaviour under bending moments and shear forces is more specialized and this results in additional launch requirements. Compared with a PC superstructure, the higher cross-sectional efficiency and the lower weight reduce the flexural launch stresses. In spite of this, however, launch prestressing is always necessary. This is obtained with straight tendons mostly contained into the concrete slabs. Reaching a correct position of the prestressing force is difficult with only internal tendons and some external tendons are often necessary. This results in more efficient final prestressing schemes as the external tendons can easily be relieved and repositioned on completion of launch. As regards shear stresses, the external polygonal tendons of permanent prestressing are usually designed so as to balance the shear forces produced by the permanent loads and 50% of live loads with tendon deviation forces. Therefore, the thin steel webs only resist the fluctuations in the shear forces due to the presence or absence of live loading. During the launch, however, prestressing must be uniform to accommodate the changes in sign of bending moment, and the steel webs have to resist full shear forces. This may result in locally oversized details, especially when the bridge is launched without temporary piers that halve the launch spans. The launch of a conventional (non-prestressed) composite bridge is relatively simpler as these structures are built by launching the steel girder first and then building the concrete slab and the weight of the steel girder is relatively small. Just for a first comparison, let us assume we have to launch a bridge with 50-m spans and a 13-m-wide deck slab. A PC superstructure with internal prestressing may have an average concrete thickness of, say, 0.60 m, which results in a dead load of 195 KN/m. A prestressed composite bridge may have an average concrete thick- ness of 0.35 m, which results in a partial dead load of 114 KN/m; adding the weight of two corrugated plate webs, say, 6 KN/m, the overall dead load is 120 KN/m and the weight saving is 38%. Finally, the average thickness of the concrete slab of a composite box girder bridge may be 0.28 m, which results in a partial dead load of 91 kN/m. Adding the weight of a U-shaped steel gitder, about 25 kN/m, the total dead load is 116 KN/m and weight saving is 41%. With these spans and deck dimensions, a composite box girder is not much lighter than a prestressed com- posite bridge. The narrow bottom slab of the latter does not increase the total conerete quantity much, and the high efficiency of prestressing strand significantly decreases the total weight of the steel webs. These economic aspects will be explored in Chapter 4. At the moment, it is important to note that 22% of the dead load of this composite box girder is due to the U-girder and the remaining 78% to the concrete INCREMENTAL LAUNCHING OF BRIDGES 61 deck slab. Launching the whole cross-section instead of only the U-girder results in launch stresses five times greater, and the related costs (oversizing of the steel girder, expensive launch devices, etc.) are higher than those resulting from casting of the deck slab in situ on completion of launch of the steel girder. Therefore, the steel girder can be presized for conventional design conditions (dead load of the concrete slab and related casting equipment prior to the onset of composite action, and composite resistance to service loading), and minor design adjustments resulting from launch requirements are related to only its own dead load. Following the previous comparison, the dead load of the U-girder during launching is 25 KN/m compared with 195 kN/m for a PC box girder. Support reactions are much smaller during launching, but the open steel girder is very flexible and the large rotations of its support sections require rocking launch bearings. The cross-sectional centroid is often located below the middle of the girder depth, and this may result in oversized top flanges. Finally, the support diaphragms of the steel girder are much heavier than the standard section and the envelope of the launch moments is disturbed by the migration of these concentrated loads. Structures that present so many and different load conditions require a careful presizing, to avoid excessive stresses in one of the many launch or service stages, but without excessive prudence, which would result in an oversized structure. These aspects are discussed for the different types of launched bridges in the relevant chapters; however, the considerations that follow are common to all these types of structures. Most of the launch stresses depend linearly on the distributed dead load q of the superstructure, so its reduction is of fundamental importance. If p is the distributed service load, the bridge cost depends on p +q and the overall structure efficiency can be expressed [46] as ee p+q The reduction of dead load limits the influence of launching stresses and increases the general efficiency of the bridge by creating a reserve available for service loads. Since the cost of both materials and construction equipment depends on dead load (that is, on the area A of the cross-section), and the moment of inertia / is the main parameter in estimating carrying capacity, the efficiency of the cross-section can be evaluated in terms of moment of inertia reached, with self-weight being equal; that is, radius of gyration [45] I Pea (2.11) With area being equal, in a section of depth H! whose centroid is distance z, and 2 respectively from the upper and lower edge, the highest radius of gyration is obtained by concentrating masses at the edges: Fax = Zu (2.12) Ps (2.10)62 BRIDGE LAUNCHING and the cross-sectional efficiency can be expressed in relative terms as flexural efficiency, p¢: maz (2.13) The closer the actual value for p, is to the ideal value of 1, the better the flexural efficiency. Generally, PC box girders have flexural efficiency of about 0.55, and prestressed composite bridges can reach 0.70. Considering a continuous beam composed of many bays of the same span and subjected to a constant dead load, far from the ends, the cross-sections above the piers remain vertical, and the static scheme of each span is that of the perfectly fixed beam. On the contrary, during launching, just before reaching the next pier, the front zone of the deck overhangs the entire span (Fig. 2.31). The negative moment in the deck cross-section passing above the first pier is six times greater than that at the rear supports and the shear doubles. The launch bearings transfer significant forces (the support reactions of the superstructure), and their most logical location is therefore directly below the webs of the cross-section. On the other hand, the launch surface must be regular, and this requires a cylindrical geometry of the solid to be launched (in most cases, a constant depth of the cross-section), With these technological restraints and the stress differences of Fig. 2.31, a sizing carried out for the temporary stresses of the front section would burden the entire deck design, whereas a sizing optimized for the rear section would be inadequate in the front one. This second alternative is clearly less expensive. Therefore, it is necessary to introduce devices that control the stresses in the front section of the superstructure, 250.0 200.0 150.0 100.0 50.0 bending moment (MNm) Figure 2.31. Bending moment envelope in a bridge with 45.5-m spans and 0.207 MN/m dead load, without correction devices in the front span INCREMENTAL LAUNCHING OF BRIDGES 63 beyond that critical cantilever L,, for which the negative moment in the cross- section above the first support is equal to the design negative moment (perfectly fixed beam) of the rear zone: Ww (2.14) and which is L,, = 0.41L. There are three possible solutions: © Limiting the temporary stresses by reducing the spans (ie. increasing the number of supports by inserting temporary piers between the final piers). © Limiting the difference between the temporary stresses in the front and rear parts of the deck by supporting the overhang with a cable-stayed scheme. Reducing the cantilever weight by using a lighter steel launching nose. In the early applications of incremental launching, the final spans were reduced with temporary piers distributed along the entire length of the bridge. In fact, it was observed that the use of just one temporary pier located below the front overhang involved significant time and costs, acceptable only for spans so long as to make launch-stress reduction advantageous in the rear section of the superstructure also. It was also observed that the use of many temporary piers in each span did not actually produce the dramatic stress reduction that might be predicted at first sight, owing to the different elastic and thermal settlings of the supports, the construction tolerances and the flexural stiffness of the superstructure. This resulted in the general use of a single temporary pier per span. Nowadays, the cost of the temporary piers for PC bridges (even if prefabricated and reusable), their foundations, and the additional labour they require is justified only in the case of very long or variable spans, where savings in launch pre- stressing (unless made unnecessary by a higher parabolic prestressing) and reinforcement may be competitive. The use-of temporary piers is also infrequent in launching the steel girders of composite bridges, although in some cases it has permitted the launch of the composite section complete with the concrete slab. These limitations narrow the field of application of the principle of generalized launching stress reduction. Instead, the stress differences along the superstructure are normally limited to make it possible to overcome the full span with a cantilever. In the first of these techniques, the cantilever is supported by means of a stayed scheme comprising a tower integral with the superstructure and strand tendons anchored symmetrically to the tip of the cantilever and two spans behind, thereby creating a sort of external prestressing with great eccentricity. In relation to the position reached by the tower during launch, this scheme requires continuous adjustment of the force in the stays because of the concentrated action exerted by the tower on the deck. Just before the contact with the pier the tower passes nearby the front support and the maximum pull in the stays supports the cantilever and reduces its elastic deflection. However, when the new pier has been reached the pull must be relieved so as not to increase the positive moment in the front span64 BRIDGE LAUNCHING with the concentrated action of the tower. Resort to this technique is discouraged not only by the longer time required for the continuous adjustments of the stay force and the deflection of the superstructure, but also by its intrinsic delicacy and complexity, presenting considerable stress concentrations and serious risks in case of errors. The second technique provides, instead of a support for the deck end, a limitation of its cantilever weight by means of a lighter extension — the launching nose — that anticipates the contact of the deck with the pier. The launching nose is a ‘passive’ solution, perhaps less elegant than the stayed scheme but much simpler; it is safe, fast and economical, so its adoption has become virtually standard in full-span incremental launching of PC and prestressed composite bridges. Even with the use of temporary piers the reduction of the cantilever weight is advantageous, and it is essential in the longest spans. The stayed scheme also takes advantage of it. The launching nose thus characterizes practically all the applications of incremental launching. 2.7.1. Launching nose The role of the launching nose becomes more and more important as the self-weight of the superstructure increases. Composite bridges are generally built by launching the steel girder and then casting the concrete slab. As the self weight of the steel girder is only one-fifth of the final cross-sectional weight, the stiffness-to-weight ratio is favourable during launching and the use of a launching nose may sometimes be superfluous. When necessary, short, light trusses are used, like that in Fig. 1.1. The heavier prestressed composite bridges are in an intermediate situation: their stiffness-to-weight ratio is higher than that of a conventional PC superstructure, and relatively light noses can be used in these cases as well. The most demanding launching noses are those for conventional PC superstructures. In the 30 years during which the incremental launching construction method has been used, launching noses of all types have been used. They have been more or less rigid trusses (Fig. 2.32) or plate girders, made of steel or concrete, with front realignment sledges or hydraulic systems to recover the clastic deflection. These launching noses have sometimes been designed according to customs and experience, but often they were already available and therefore ready to be reused. The most common type of launching nose for PC bridges consists of two steel plate girders connected to the front end of the superstructure with their bottom flanges aligned with the lower edge of the deck. Acting as an extension of the superstructure perfectly integral at the joint, the launching nose affects the stresses in the superstructure (or, more exactly, in the continuous beam comprising the deck and the launching nose itself) and permits their homogenizing. The behaviour of the nose-deck elastic system is governed by three dimensionless parameters that describe its geometrical and mechanical characteristics: INCREMENTAL LAUNCHING OF BRIDGES 65 Fig. 2.32. Trussed launching nose (author) the nose length compared with the span to be overcome, L/L e the unit weight of the launching nose compared with the unit weight of the deck front zone, gn/q e the flexural stiffness of the nose compared with the one of the deck front zone, Eqly/EI. The model developed in the following chapter makes it possible to illustrate the influence of these three parameters and to optimize rapidly the nose-deck inter- action [47], avoiding the trial and error use of sophisticated calculation methods. As this model neglects shear strains, results may be approximate in the case of prestressed composite bridges, which are generally affected by significant shear deformations. 2.7.1.1. Nose—deck interaction model In a PC bridge, a correct presizing of the nose-deck system can start from a nose length L, equivalent to about two-thirds of the typical span: Ly = 0.65L (2.15) and from a first attempt unit dead load q, obtained from the following equation: Qn = kD (2.16)66 BRIDGE LAUNCHING where, in KN and metres, it is 0.012 < k < 0.020 for highway PC bridges and 0.018 < k < 0.030 for railway bridges, with progressively higher values in the case of wide and heavy superstructures. These approximate values for Ly and qo can be used as a starting point for the optimization of the nose-deck interaction. In order to develop a theoretical model of the nose-deck interaction, the follow ing assumptions are made [48,49]: e The nose and the superstructure are of constant stiffness and weight: since they actually vary both in the launching nose and in the deck front zone, it is correct to consider their weighted average. The concentrated load of the end diaphragm of the superstructure, the temperature gradients and the mis- placement of launching bearings, have a significant influence on the launch stresses. These load components have not been included not to excessively complicate the model; however, practical applications need to account for these aspects, which can be easily introduced through the related support rotations. © The number of equal spans behind support C (Fig. 2.33) is assumed to be so large as to assimilate the superstructure to a continuous beam composed of. infinite bays of constant span L. e Launch prestressing is centroidal also in the deck front zone, so as not to introduce additional redundant unknowns. Defining the progression of launch with the distance x of the nose-deck joint section from support B (Fig. 2.33), the cantilever configurations assumed before the nose reaches support A vary from the starting position, x = 0, to the contact one, x = L — Ly, and the dimensionless launching progress (2.17) D c 2 a L L L px, tn Fig. 2.33. Nose-deck system in the first phase of launch INCREMENTAL LAUNCHING OF BRIDGES. 67 At the start of launch, for a = 0, the bending moment in the support section B is that of the cantilever nose, which in a dimensionless form (the moments due to cantilever masses will be marked with an asterisk) is Mg Van (laa wen -2e(Z) ev Once launch starts, this increases (in absolute value) with the dimensionless law Mi dun () , VLn Wiggle eT @.19) The moment Me at the third support C is Mc __ks+ky 1 ky Mi c_ _ks+he | ky Mp (2.20) qe ky+kql ky +k gh? with the coefficients k; of Fig. 2.34. Finally, the support reaction at the front support is R L+a)? daly IL, i Ae Ose (140452) (2.21) Once support A has been reached, the nose tip is hoisted until it realigns with the theoretical profile of the deck in order to permit the subsequent sliding on launching bearings. The recovery of the elastic deflection creates a positive moment that reduces My and the second launch stage starts (Fig. 2.35), in which the nose slides on support A until the concrete deck arrives. This stage is defined by 1-L/L 0.90, Mg does not depend on £,/,/E7. Therefore the flexural stiffness of the nose can be used only to prevent the launching moments after contact with support Aand until stabilization of the EOL value (for 0.20 < a < 0.90) from exceeding the moments at contact and at EOL, since deck oversizing is generally more expensive than stiffening the launching nose. Optimum stiffness, in this case, turns out to be around E,J,/EI = 0.200 (thicker line). Greater stiffness is pointless, since it cannot prevent reaching the EOL moment, and less stiffness increases the negative design moment of the super- structure. Moreover, the maximum cantilever moment is much lower than the EOL one: contact with the pier is uselessly anticipated; i.e. the launching nose is excessively long, and it is convenient to shorten it. On the other hand, in the case of a short nose (equal relative weight but L,/L = 0.50), the progression of My is as given in Fig. 2.37. (Note there that, with a short nose, the maximum cantilever moment is higher than the EOL one, which again is higher than the support moment of the perfectly fixed beam.) Until the contact with support A it is favourably affected by the lower weight of the nose but unfavourably by the greater length of the cantilever concrete segment. On reaching A Mj is reduced more effectively, and at the EOL it stabilizes on the value given by (2.24). The effects of the flexural stiffness are comparable as well, and also in this case optimum stiffness should be around £,/,/EI = 0.200 (thicker line). However, keeping Mg below the EOL value during the intermediate launch70 BRIDGE LAUNCHING 0.16 Ne -0.12 4/12 -0.08 é g bending moment at support B (M/qt”2) 0.00 L 0.00 0.25 0.50 075 a 4.00 Fig. 2.37. Progression of Mg with launch for L/L = 0.50 and qq/q = 0.10 in relation to the relative flexural stiffness, Eyly/ET stages loses significance since this threshold is substantially exceeded in the previous cantilever configuration. In contrast to what occurs for the long nose, in the case of a short nose the maximum cantilever moment is greater than the EOL one, and it is therefore the first phase of launch to affect the design of the superstructure. Contact with the pier occurs too late; i.e. the nose is too short and should be extended. Proceeding by trial and error, the optimum nose length is obtained in Fig. 2.38 with E,/,/EI = 0.200, the cantilever moment at contact is equal to the EOL one, and in the intermediate stages it is lower. 0.16 a Ka 8 112 -0.08 004 bending moment at support B (Miq12) 0.00 0.00 0.25 0.50 O75) fa: = 1.00 Fig. 2.38. With qu/q = 0.10, the cantilever moment at contact is equal to the EOL one for L/L = 0.65 INCREMENTAL LAUNCHING OF BRIDGES nN The diagrams examined so far analyse the behaviour of the nose-deck elastic system for a given relative weight; Fig. 2.39 compares the diagrams of Mg calcu- lated for the optimum parameters obtained for g,/q = 0.10 as a function of this ratio. Mg depends significantly on the relative weight of the nose only during the first phase of launch. In the second phase it is limited to a family of curves contained in 15-20% of the EOL value, which twist because of M4. The different relationship between Mp and q,/q in the two phases of launch means that, once the system has been optimized for a given relative weight, an increase in q,/q modifies the diagram of Mg towards the type shown in Fig. 2.37 (‘short’ nose) and vice versa its reduction modifies the diagram towards the type shown in Fig. 2.36 (‘long’ nose). To force the value of Mj on contact towards the EOL moment it is then necessary to modify the nose length, shortening the canti- lever concrete segment (i.e. lengthening the nose) as the relative weight of the latter increases. Figure 2.40 compares the diagrams of My that restore this condition: starting from qq/q=0.10, as the relative weight increases equalization is obtained with longer noses and for progressively lower values of My, which tend to the perfectly fixed beam moment of the rear zone of the superstructure. Proceeding by trial and error, the curve of Fig. 2.41 is obtained, which defines, as o/q varies and for the optimum flexural stiffness, the values of L,/L that equalize the negative moment on contact with the EOL one. This diagram makes it possible to determine easily the optimum length of the launching nose in relation to its relative weight. In conclusion, a correct design of the nose-deck system can be arrived at starting from first-trial parameters obtained from (2.15) and (2.16). The values of Ly and qq so defined will be refined until the equalization of the contact moment with the EOL one is obtained. Then, the minimum flexural stiffness necessary to contain the 0.16 g 025 z 02 = -0.12 |— g 0.45 5 =1h2 iS eae oA ] z | : 9.65 & -004 0.00 0.00 0.25 0.50 075 a 1.00 Fig. 2.39. Progression of Mg for Ly/L = 0.65 and Eyly/EI = 0.200 in relation to the relative weight q,/q. The thicker line is the optinum diagram of figure 2.3872 BRIDGE LAUNCHING 0.12 J “12 0.08 x bending moment at support B (M/q1A2) 1 3 8 0.25 0.50 075 a 1.00 Fig. 2.40. Obtaining equalisation of bending moment on contact and at the EOL by adjusting the nose length as qn/q changes moment in the intermediate launch stages under this common threshold may be defined. For this purpose, it is necessary to remind that the diagrams in Figs 2.36- 2.43 are only aimed at illustrating the influence of different parameters on the behaviour of the nose-deck elastic system. When presizing a launching nose for a practical application, the concentrated load of the front diaphragm of the super- structure, the expected misplacement of launch bearings, and the temperature gradients stated by the design code, should always be taken into account. This can be easily achieved by inserting additional components of the rotations of the support sections in Equations (2.20) and (2.22). 1.000 = 5 | 8 o760| —— 2 § e 0.500 3.020 0.060 0.100 0.140 0180 relative nose weight Fig. 2.41. Evolution of the relative length L/L that equalises the negative moment as 4n/4 changes INCREMENTAL LAUNCHING OF BRIDGES 23 In Fig. 2.42, the progression of Mc (2.20) shows that, independent of the nose length, the recovery of elastic deflection forces Mc towards a single curve that stabilizes for « = 0.60 on a slightly lower value than the perfectly fixed beam one. From this point onward, except for irrelevant additional fluctuations on contact with the following piers, the support moment is no longer affected by the front elastic system. In short, with regards to the negative moment, it may be concluded that by adopting launching noses of sufficient flexural stiffness and adequate length for the relative weight, the deck zone affected by the nose-deck interaction can be limited to a length of about 1.5L, increasing to 2L in the case of very deformable noses. The maximum positive moment in the front span is Gan) 1 (Ra), 140 (Ln\” (dn q t)+3(¢6) Tq (Z) (* and Fig. 2.43 illustrates its evolution in the case of a long launching nose. In contrast to what happens to Mp, the flexural stiffness of the launching nose affects only the progression of M% towards the maximum value, which is in any event reached for a = 0.90. It then starts to decrease because of the effect of Mi. By shortening the nose, the progression of M*%, is substantially similar. The maximum value is obtained a little further on in the launch and is slightly higher because of the smaller contribution of M4. Therefore, the flexural stiffness of the launching nose does not govern the positive moment behaviour of the system, which depends essentially on the continuous beam static scheme of the superstructure and can be modified only by means of Mj. A synergistic action of increasing the length and the weight of the nose can reduce at the same time the negative equalized moment and the maximum positive moment. Consequently, the choice among available noses can be made by also taking these savings into account. (2.25) 0.120 = 3 : 08 & =12 3 -0080 s = os 5 os g 0.040 0.00 025 0.50 (O75 apo Fig. 2.42. Evolution of Mc for different values of Ly/L14 BRIDGE LAUNCHING 0.08 2 5 5 124 £ + 0.04 = 2 04 005 2 04 68 0.025 $ a < 0% 40 0.60 080 a ~~—«.00 Fig. 2.43. Evolution of the maximum positive moment in the front span for L/L = 0.80 and qq/q = 0.10 in relation 10 Eqly/EL Finally, it should be noted that the consideration that the first bay of the continuous beam is in any event subject to higher positive moments is deceiving, since correct bridge design involves shorter end spans just to compensate these increases, while during launch the front section of the deck has to overcome all the longer intermediate spans. For the B-C span one obtains 1(Mc Msg 1\?, Mp ole +5) +5 (2.26) 2\qPP qh?’ 2) ° ql? MBE ql and the relative diagram, compared with the progression of Mc , confirms that the rear zone of the deck, unaffected by the nose-deck interaction, begins with reasonable approximation at a distance of 1.5L from the front end, To draw some conclusions, the design of a PC bridge or of a prestressed compo- site bridge built by incremental launching depends on the technique adopted to control launching stresses, a technique that in all cases involves the use of a launching nose. The optimization of the nose-deck interaction produces savings in structural materials. It may be easily carried out by means of the described model (modified to include the effects of the front end diaphragm, temperature gradients, bearing settlement, and the cable-stayed front system, if present) and an ordinary spreadsheet. Adopting the most common scheme, which provides for using just a launching nose, the evolution of the negative moment in the superstructure may be adjusted by acting on the parameters governing the nose-deck interaction. Among several combinations equally suited to the negative moment, some of them improve the behaviour at the positive moment as well, with a higher cost of the nose balanced by lower costs of structural materials. INCREMENTAL LAUNCHING OF BRIDGES. 15 2.7.1.2. Sizing of the launching nose The optimum stiffness and dimensions of a launching nose depend on many factors, economic considerations included. The convenience of amortizing the investment over several jobs suggests designing launching noses to be as versatile as possible. Since the incremental launching construction method for PC bridges reaches spans of about 60m, it may be convenient to design the launching nose directly for the longest spans even in the case of initial use on shorter spans, and to adapt it to the lower depth of the superstructure. The probable need to reduce the nose length can be addressed by subdividing it into spliced elements, while its greater stiffness and weight further reduce the launch stresses in the deck front zone. The role of the launching nose, originally aimed at controlling the negative launch moments, can thus extend to the control of the positive moment in the front span. The convenience of the higher investment required for this purpose can be verified based on the amortization expectations and the savings in the front zone of the superstructure. In most cases, a launching nose for a PC bridge is composed of two braced steel plate girders. Launching noses made of PC have been adopted in the case of impossible reuse or difficult transport, to be demolished on completion of launch. Prestressed composite noses (steel plate webs and PC flanges) have been used for the launch of extremely heavy and stiff superstructures, If there are no limitations in the depth, truss noses are very efficient. However, the bottom chords of these noses have to withstand the local stresses generated by the migration of the support reactions and are therefore designed for high shear forces and local moments. Together with a great quantity of manual welding, this makes the use of truss noses infrequent for the launch of PC bridges, and the general trend is towards the use of plate girders composed of transportable elements connected by vertical splices with high-grip bolts (Fig. 2.44). On the contrary, truss noses are often the best solution for the launch of the steel girders of composite bridges. Preloaded pins (see Fig. 2.46) are frequently used to join the flanges since con- ventional splice plates with high-grip bolts require hoisting of the superstructure (to avoid interference with the launch bearings) and consequently the discontinuous launch procedures of Fig. 1.5. The use of preloaded pins specializes the action of connection devices: the bolts along the gusset plate mainly resist shear forces while the pins, located in widened holes to avoid lateral contact, resist only axial forces. Horizontal splices in webs are necessary only in the case of deep girders for railway bridges or long spans (Fig. 2.45) and are generally designed for high-grip bolts. The bracing system of the launching nose is almost always designed specifically for the bridge, since it is convenient to align the nose girders with the bottom corners of the deck cross-section to use the same launch bearings. The use of wide launching bearings might permit the reuse of a bracing system designed for more closely spaced girders; however, the high stresses generated in the front diaphragm of the superstructure by the transverse eccentricity of the forces and the moments transferred by the nose girders make this solution infrequent. When16 BRIDGE LAUNCHING PLAN lap plate horizontal bracing segment 1 segment 2 segnent 3 SIDE VIEW lap plate (- lifting points Jack al | LL | Hi connection system approach device J Fig. 2.44. Typical segmentation of a launching nose buying a second-hand nose, therefore, the cost of new horizontal braces and cross- frames should always be taken into account. The external edge of the bottom flange of the launching nose may be aligned with the external corner of the bottom slab (PC and prestressed composite bridges) or of the bottom flange (steel girders) of the superstructure. This allows use of the same lateral guides for the nose and the superstructure. This solution is particularly advantageous in the launch of steel girders since the lateral guides can be small low- friction plates acting directly against the external flange edge. The guide action on the launching nose should be aimed only at centring the support reaction along the nose webs, since alignment corrections of the super- structure by acting on the nose would cause excessive stresses in its bracing system. Should major alignment adjustments be necessary, they can be generated by shimming one of the lateral guides at the most advanced pier reached by the super- structure, When designed specifically for the bridge, the highest depth of the launching nose is generally equal to the depth of the superstructure and is maintained only for few metres from the joint section. In fact, the evolution of the negative moment in the superstructure is such that the full flexural stiffness of the launching nose is INCREMENTAL LAUNCHING OF BRIDGES 7 Fig. 2.45. Horizontal field splices facilitate transport of deep girders. The bottom flange is strengthened with continuous box stiffeners (author) necessary only in the central launch phases. The front section can therefore be lightened, by decreasing both the web thickness and the girder depth, with savings in costs and cantilever masses. The design stresses should be kept reasonably low to accommodate load differences between the two girders due to construction and assembly tolerances, flatness defecis in the bottom flanges, and vertical misalignment of launching bear- ings. Load differences can be significant in launching noses with closely spaced girders, and diaphragms and cross-frames are generally too flexible to effectively control them, Consequently, each nose girder should prudently be designed for 75-100% of the whole support reaction transferred by the nose. This criterion should also apply for the local prestressing systems that join the nose girders to the superstructure. The design of a launching nose starts from the envelopes of the bending moments and the shear forces resulting from analysis of the continuous beam. The load envelopes are determined in each nose girder, and the depth of the rear nose section is taken equal to the overall depth of the superstructure. The transverse width of the bottom flange may be chosen so as to use the same neo-flon pads used for the superstructure. Defining the transverse width of a neo-flon pad with bye, its launchwards length with Lye, and the total length of the neo-flon pads on a launching bearing with ELy¢ (see Fig. 3.23), the calculation of the contact stress is immediate. The transverse bending moment generated by the contact stress governs the thickness , of the bottom flange. The transverse flexural78 BRIDGE LAUNCHING stress generated by the support reaction Ra is calculated by assuming that the neo-fion pads are centred along the web and the support reaction is equally distributed between the two nose girders _ 3Rabar © 8 ELgp Then, this transverse stress is increased to account for the eccentricity in the support reaction and in its local distribution, and it is finally compared with the axial stress allowed by code. Once the bottom flange area has been determined, the top flange area is calcu- lated with a trial-and-error process based on the expected web thickness, the result- ing location of the cross-sectional centroid, and the comparison of the edge stresses under the maximum and minimum moments with the allowable stress. The use of different types of steel for the opposite flanges of the nose girders is infrequent. The final dimensions of the top flange plate are determined based on the influence of the width-to-thickness ratio on instability of the compression flange. Additional infor- mation can be found in Chapter 4. The web thickness, /,, is determined from the envelope of the shear forces trans- ferred by the nose girders. The design tangential stress is kept low to account for the uncertainties about the distribution of support reactions between the two nose girders, and to decrease the number of vertical web stiffeners, whose spacing can thus be determined based on the bracing requirements of the nose girders. Based on the final web stiffener geometry, the code equations for web panel buckling define the minimum thickness of web plates, which should not be less than 10 or 12 mm to ease plate handling in the workshop and to reach adequate overall stiffness. Finally, local buckling of web panels under the thrust of compression flange is assessed. By assuming a 45° dispersion of the support reaction into the bottom flange and, again, a perfect load distribution between the two nose girders, the vertical com- pressive stress at the bottom edge of the web panel above the launching bearing is = Ra © 2ty(ELng + 2p) Detailed criteria for assessment of the web panels are provided in Chapter 4. If the transverse stresses in the bottom flange (2.27) or the vertical stresses in the web panels (2.28) are excessive, closely spaced flange stiffeners can create additional support lines in the flange and improve dispersion of the support reaction into the webs (Fig. 2.46). The triangular box stiffeners of Fig. 2.45 are less expensive in terms of hand welding but complicate the use of preloaded pins for the vertical field splices due to jacking clearances In proximity of the rear nose end, the stresses in the lower region of the webs are high because of the concomitance of several stress states, (2.27) o o, (2.28) © Longitudinal flexural stresses are generated by the support reaction of the con- tinuous beam and by the front nose portion overhanging beyond the support point. INCREMENTAL LAUNCHING OF BRIDGES 19 Fig. 2.46, Vertical flange stiffeners combined with pre-loaded pins (author) © Vertical compressive stresses are generated by dispersion of the support reaction into the webs. © Local strut-and-tie force systems arise from the application point of the support reaction to the shear keys welded on to the contact plate with the superstruc- ture. When the shear keys are located at the lower edge of the superstructure to transfer shear forces directly to the concrete webs (Fig. 2.47), transmission of shear forces is concentrated in the lower region of the nose girders. © High local stresses are generated by the anchor systems of the prestressing bars that clamp the nose against the front end of the superstructure. These anchor systems also concentrate most of the longitudinal flexural stresses produced by the support reactions. The principal stresses are sometimes calculated in these regions of the nose girders by assuming that the stress-state is a biaxial one. In the most delicate cases, a 3D finite-element analysis of the whole joint region (rear portion of the nose girders, front diaphragm of the superstructure and front portion of the box girder section) is necessary. Clearance requirements of the tensioning jacks of the anchor system between the launching nose and the superstructure rarely permit adequate distribution of bottom flange stiffeners (Fig. 2.46) or adoption of the box stiffeners of Fig. 2.45. In addition, it is often necessary to use four clamping bars on each level (see bar holes in Fig. 2.47) to increase the vertical lever arm of joint prestressing. In this case, anchor boxes welded to the nose girders are often necessary, as in Fig. 2.48. A 3D finite-element analysis facilitates nose detailing and the design of transverse prestressing or reinforcement of the front diaphragm of the superstructure,80 BRIDGE LAUNCHING Fig. 2.47. Gusset plate and concentrated shear key of the launching nose of Figure 2.48 (author) especially when the concrete webs are inclined. Similar considerations govern the design of a launching nose for the steel girder of a composite bridge. In this case, the rear end of the trussed nose is often a plate diaphragm directly welded to the steel girder as in Fig. 2.55. Nose girders are connected to each other by cross-frames and horizontal bracing. Cross-frames reduce the load differences between the two girders due to eccentric support reactions; when the nose is twisted, they also oppose distortion. Cross- frames are designed as crosses or transverse frames with stiff joints between the vertical web stiffeners and horizontal cross-beams placed in the plane of the horizontal bracing (Fig. 2.49), sometimes supplemented by diagonals that further improve stability of the compression flange. In this case, it is advisable to design the cross-beams so that the distance between their bottom flange and the bottom flange of the main girders is at least one-half of the transverse width b; of the latter [11]. The walkway to the front work platform for nose realignment at contact with the pier (see Fig. 2.23) can be placed directly on the top flange of cross-beams. In narrow noses (Fig. 2.48) the longitudinal spacing of cross-frames is often such that the panels of the horizontal bracing are square; the rectangular scheme is more common in wider noses, as in Fig. 2.49. The role of the horizontal bracing is to stiffen the nose girders in resisting the wind loads and the transverse forces produced by the lateral guides that centre the nose webs on the launching bearings. Horizontal bracing is designed with cross INCREMENTAL LAUNCHING OF BRIDGES 81 Fig. 2.48. Anchor boxes for the prestressing bars of the joint system (author) Fig. 2.49. Framed cross-diaphragm (author)82 BRIDGE LAUNCHING or K schemes, where the verticals are the cross-beams and the chords are the main girders 2.7.1.3. Coupling to the superstructure The connection system between the launching nose and the superstructure must guarantee the transfer of the bending moments and the shear forces by providing perfect structural continuity. On completion of launch, the connection system must also permit an easy disassembly of the launching nose. In PC bridges, the most effective solution for the transfer of the positive and negative moments is a local prestressing system composed of alloy steel bars. Strand tendons are less frequently used because it is more difficult to relieve them for maintenance or monitoring purposes. The bar portion embedded into the super- structure is generally 1.52 times as long as the deck depth to obtain an anchorage as long as the whole deck depth both for the clamping bars and for the launch prestressing tendons of the superstructure (Fig. 2.50). The top flanges of the launching nose are sometimes anchored to temporary reinforced concrete blocks cast directly on to the deck slab. This solution may be necessary in the presence of many anchorages of launch prestressing in the top slab region affected by the nose, but inclined deck webs or available second-hand noses deeper than the superstructure are also good reasons for this solution. The positive moment transferred through the nose-deck joint is generally much higher than the negative moment due to the cantilever nose. As a consequence, 5824 4 een — shear keys \ \ temporary concrete segment Fig. 2.50. Side view of the anchor devices of the launching nose INCREMENTAL LAUNCHING OF BRIDGES 83 most of the clamping prestressing is applied at the bottom flange level. In this node, the distribution of the clamping bars is governed by technological restraints such as the nose geometry, the diameter and length of the stressing jack and the diameter of its contact plate alongside the anchor nuts of the tensioned bars. The most practical solution for a satisfactory transmission of longitudinal prestressing between the nose anchor system and the prestressing tendons of the superstructure con: in modifying the layout of the latter so as to avoid the bar anchorages and to introduce the tendons deeply between the bundles of bars to make the stress transmission as direct as possible. Although tendon curvature causes prestress losses, with this scheme the prestressing forces migrate from the clamping bars towards the prestressing tendons through longitudinal compressive stresses. A scheme with bundles of clamping bars and parallel prestressing tendons would involve transfer of prestressing forces through shear stresses and would require the use of transverse prestressing to avoid cracking. Splicing of joint prestressing can be complex also in the upper node, especially when inclined webs move the upper cross-sectional node of the superstructure away from the top flange alignment of the nose, whose webs are vertical (Fig. 2.51). This can require a mild transverse prestressing in the top slab that absorbs the principal tensile stresses resulting from the shear forces caused by this eccentricity. However, most of the joint prestressing is concentrated at the bottom flange level and the splicing forces in the top slab are rarely critical. Starting from the maximum positive and negative moments in the joint and the distance between the gravity axis of the superstructure and the centroid of the upper and lower bundles of bars (e, and ¢; respectively), the calculation of the two prestressing forces (the lower force F, and the upper one F,) is immediate. The clamping bars should be oversized with respect to these minimum requirements to resist the load differences between the nose girders without joint decompression and related bar overloading. When presizing the clamping bars, allowance should be Fig. 2.51. Front view of the anchor systems of Fig. 2.5084 BRIDGE LAUNCHING made for re-entry of anchorages, friction losses, elastic shortening of the compres- sion concrete at tensioning and prestress losses due to time-dependent phenomena (launching of long bridges takes several months). Launch procedures and checklists should also include periodic checking of bar tensioning and retensioning when a predetermined threshold is exceeded. The use of one or two hydraulic load cells is advisable in the case of long bridges for earlier detection of excessive prestress losses in the joint system. When the depth of the superstructure allows comfortable work within the box cell, the clamping bars can be tensioned from the rear anchorage. This eases bar retensioning in case of need; bar geometry is less restrained, and the bundles of bars may be closer to the nose webs because there is no jacking clearance on the nose side. This results in compact joints (Fig. 2.52) and no risks for workers during the operations of bar check and retensioning. The joint surface, of net area A; (after subtraction of bar holes) and moment of inertia J, is designed by assuming a linear distribution of the contact stress along the joint surface. The edge stresses for the maximum positive moment Fut Fi, Minas + Futu = et , A L 4 4h Fat Mites + Faeu = Fei , oy ows 7 f i Fig. 2.52. Bar tensioning from the rear anchorage results in compact gusset plates (author) INCREMENTAL LAUNCHING OF BRIDGES 85 are respectively compared with the allowable compressive stress and the least joint compression stated by the design code. Similar verifications are carried out for the negative moment generated by the cantilever nose and the downwards eccentricity of the clamping prestressing. This last load condition sometimes prevails, as in Fig. 2.53. In PC bridges, the design shear force in the joint, V, is transferred through single shear blocks (Fig. 2.47) or more shear keys distributed along the contact surface. Figures 2.50 and 2.51 illustrate a solution that avoids carving the front end of the superstructure with shear keys. In spite of these shear transfer devices, it is advi- sable that the joint prestressing force guarantees joint stability by friction. With a coefficient of friction of 0.4 for a steel-concrete contact and a safety factor of 2, we obtain <02 (2.30) Fut h~ 2.30) When the launching nose is deeper than the superstructure and/or the deck webs are inclined, the nose can be anchored by means of steel blocks (Fig. 2.54) or concrete blocks directly cast on to the deck slab. In the case of concrete blocks, the solution Fig. 2.53. Contact stresses (N/mm?) in the joint section of the launching nose of Fig. 2.4886 BRIDGE LAUNCHING with vertical reinforcement through the joint requires demolishing the block, cutting the joint reinforcement, and patching the deck surface on completion of launch. In both cases, therefore, the anchor blocks are joined to the superstructure by friction. The horizontal force transferred by the nose to the anchor block acts as a shear force V;, in the joint between the block and the deck slab. The minimum vertical force Nain to be applied to a concrete block by means of vertical pre- stressing bars is determined by applying a safety factor of 2 to a friction coefficient between two concrete surfaces of about 0.6: (2.31) The coefficient of friction can be increased with distributed shear keys obtained by pressing corrugated sheet metal on the fresh conerete of the deck slab. In the case of a steel anchor block, we have again Novia = zy (2.32) The use of some shear keys in the joint surface is always advisable, as in Fig. 2.54. In the particular case of this bridge, the whole vertical shear force between the nose and the superstructure has been resisted with vertical bars. Therefore, the front bundle of vertical bars is particularly powerful. Fig. 2.54. View of the anchor blocks of Fig. 2.46 from inside the casting cell (author) INCREMENTAL LAUNCHING OF BRIDGES 87 Similar considerations govern the design of the joint between the launching nose and the steel girder of a composite bridge. When the bridge has a composite box girder section, the nose is complete with a plate diaphragm that joins the nose trusses with the inclined webs of the U-girder. The nose~deck joint may be designed as a conventional grip-bolt field splice, although a welded joint on a shortest sacrifical front overhang (Fig. 2.55) is often the most efficient solution. 2.7.1.4, Recovery of elastic deflection When approaching the pier, the nose tip is generally distant from the theoretical vertical profile of the launch surface because of the elastic deflection of the nose- deck system, In the case of a regular sequence of spans and of piers with similar vertical stiffiness, the nose tip deflection at contact is always downward. It can easily exceed 10 om in the case of a PC bridge, and several tens of centimetres in the case of the steel girder of a composite bridge (Fig. 2.32). The need to recover this deflection to permit the continuation of launch might be solved with a hoisting wedge; i.e. by inclining the bottom flanges of the launching nose so that its advance on the pier forces its progressive realignment. In reality, this realignment method is used rarely and only for steel girders, as the high horizontal forces generated by the wedge action can overload the piers. In addition, the launch bearings must be anchored to the pier cap to avoid accidental displacements or overturning, and must be articu- lated to accommodate the large rotations of the support section. Fig. 2.55. Joint diaphragm between the truss nose and the U-girder of a composite bridge (author)88 BRIDGE LAUNCHING In most cases, therefore, realignment is obtained with hydraulic devices. For a PC bridge, the simplest solution consists of placing on the pier cap, immediately before the launching bearings, a pair of hydraulic jacks equipped with rounded top plates. When the launching nose has arrived above these jacks, neo-flon pads are inserted between them and the nose. Then, jacking realigns the nose and the neo-flon pads permit the launch to continue until the nose tip arrives on the launching bearings. This scheme requires jack transfer from pier to pier; in addition, the effective stroke of flat jacks is quite short and the complete recovery of nose deflection can require several hoisting cycles, with significant dead times. As a lot of workers are present at the piers, these dead times are expensive. These weak points often make it convenient to use steel brackets hinged to the nose tip and moved by hydraulic pistons (Fig. 2.23): when they reach the launching bearings, these brackets are lowered onto neo-flon pads and forced until the elastic deflection has been completely recovered. When launching a steel girder, these pivoted arms are longer and moved by less powerful long-stroke pistons (Fig. 2.32). 2.7.2. Stayed front system A second solution to the need to reduce the transitory stresses deriving from cantilever configurations consists in supporting the front end of the superstructure. This is attained with a cable-stayed device composed of a tower hinged to the superstructure and strand tendons fixed at the tower head and fan-anchored symmetrically with respect to it, to obtain an adjustable external prestressing system with high eccentricity (Fig. 2.56). In the initial phases of construction, when the first two spans of the super- structure are still in the casting yard (Fig. 2.57), a short launching nose and a steel tower are anchored to the superstructure. The tower is positioned so that when the nose tip reaches a pier, the tower is some metres behind the previous pier. Transversely, the two legs of the tower are supported on to the deck webs by means of long-stroke hydraulic jacks that control the vertical force transferred to the superstructure and permit simultaneous adjustment of the pull of stays. For this device to be effective in all launch phases, it is necessary to modify the pull of stays continuously because of the concentrated load that the tower transfers to the superstructure. When the nose tip approaches the pier the force is the highest and the system is effective (Fig. 2.58), but when the tower overtakes the pier and advances to mid-span the same force would increase the positive moment and must be reduced by loosening stays. In practice, many factors limit the use of this support scheme to the longest spans. of PC bridges, and always in combination with a launching nose to reduce the cantilever weight (Fig. 2.59), and with it the required capacity of the stays, as much as possible. © The need for a long casting yard (about two spans): before launching on the bank span, the superstructure must be long enough to anchor the stays and to INCREMENTAL LAUNCHING OF BRIDGES. 89 concrete seat Fig. 2.56. Cable-stayed front system long stroke Jack dynamonetre. yo=- > 1 deviation tower removable pin90 BRIDGE LAUNCHING Fig. 2.57. Launch phases with the stayed front system. In case of a short nose, each Iaunch phase is as long as the whole span guarantee adequate safety against tilting, unless a temporary pier is added in the bank span. © The complexity of calculations: the tension in the stays depends on the elastic deflections of the stayed system and of the superstructure, on pier settling, on bearing misplacement, and on the thermal strains both in the deck and in the stays. In addition, non-linear stay behaviour reflected in the Ernst modulus [5,27] depends on the stress level. Although the corrections deriving from the apparent variation in the elastic modulus of stays are usually necessary only in the analysis of large cable-stayed bridges, these corrections may be significant in the front-stayed system because of the low tension in the stays in the first phases of launching. © The need to control and to continuously adjust the stresses in the tower and in the stays. e The stress concentration in the deck region under the tower, which in spite of stiffening is often affected by cracking problems. INCREMENTAL LAUNCHING OF BRIDGES a1 Fig. 2.58. Position of maximum pull in the stays of the Boivre Bridge (courtesy Spie Batignolles) unneopanenersis Fig, 2.59. Launch of the Charix Viaduct on 64-m spans (reproduced with permission, Scetaroute-Berenguier)92 BRIDGE LAUNCHING © The intrinsic delicateness of the solution, the risks in case of error, and the need to reach the front pier at the end of each launch to avoid long stops with stressed stays. In the case of short launching noses, this last requirement adds further restraints on the length of the casting yard. © The irreversible deck deformations generated by creep of concrete in the case of launch steps as long as the whole span. In this case, it is advisable to end each launch with a front cantilever able to reduce the positive moment in the first span, or to adopt a relatively long launching nose that permits halving the length of the launch steps. © A longer front deck section requires higher axial prestressing than a conven- tional use of the launching nose. For all these reasons, the combined use of a relatively short launching nose 04< = < 0.6 (2.33) and of a cable-stayed system with a tower of height /i hn 0.40 < i-L, < 0.44 (2.34) is generally limited to wide and heavy superstructures with 60-70-m spans. In this case, it permits the use of launching noses designed for spans 10-40% shorter. Over these spans, the use of temporary piers is probably the best solution. The use of temporary stays is particularly advantageous in the case of a long central span obtained by launching two bridge sections symmetrically from the opposite abutments with mid-span closure. The cable-stayed support of the two cantilevers permits overtaking central spans much longer than 100 m. If the central span is designed with extradosal prestressing, the deviation towers of the external tendons may be used during launching to deviate the temporary stays. Compared with the use of a launching nose only, the action of stays modifies the evolution of the launch stresses in the superstructure. Before the contact of the nose tip with the pier, the stays absorb most of the excess negative moment resulting from the use of a short launching nose (Fig. 2.37). Once the pier has been reached, the stays are relieved and launch continues with the nose only. The action of the stays and that of the launching nose are staggered over time and are not super- imposed on each other. Therefore, the launching nose and the stayed system can be designed according to independent criteria. The length and the unit weight of the nose are determined first. Then, the optimum flexural stiffness is defined according to the criteria examined in Section 2.7.1.1, As already discussed, the launching nose cannot control the end-of-launch moment, so it is convenient to design the superstructure for the EOL moment and the action of stays so as not to exceed this design value in the cantilever configuration. INCREMENTAL LAUNCHING OF BRIDGES 93 At the beginning of launch the stays are unstressed, and they are not necessary until a critical advance L,, that generates a negative moment in the front support section equal to MR°" is reached. In dimensionless form (2.17) axle (ano da (Ln? =a ‘ini =1.268 (2) +0.212 2.35) ql iC L g\L Cw de = Beyond az,, the action of stays should limit the support moment Mj below or at the EOL moment until the contact with the pier; ie. for ag, $ respectively. Defining the vertical component of the load transferred to the pier by the vector OF, its orthogonal component with respect to the launch surface is ON = OV cosa (2.41) and the tangential component is OT = OV sinw (2.42) On applying the thrust force the resultant begins to rotate, and launch starts only when its inclination exceeds the equilibrium angle ROR, = 26. Therefore, the limit equilibrium resultants are cosa cos OR, = OR, = OV (2.43) When launching uphill, to rotate the resultant from OV to OR; it is necessary to apply a force F equal to — OT = -OV (cose tgp + sine) (2.44) F=TT,=0 and the horizontal force transferred to the pier is = Ons OV cose sind +a) (2.45) In the case of downhill launching on the other hand, if a thrust force has to be applied (i.e. for a < ¢ as in Fig. 2.66), the resultant rotates from OV to OR; under a force F: F =TT, = OT, - OT = OV cosa tgp — sina) (2.46) If the local gradient is higher than friction (ic. for @ > @ as in Fig. 2.67), the superstructure slides without any intervention, and to brake it the force F must be directed uphill. In this case, a minimum value of the friction angle must be used for calculations instead of the static friction angle (2.40), and in the most delicate INCREMENTAL LAUNCHING OF BRIDGES 105 Fig. 2.66. Pier cap forces for a local gradient of the launch surface smaller than the static friction (% < $) cases # should be directly set to 0. In both cases, the horizontal force transferred to the pier is H =O, =0V cosa cong ina) (2.47) To conclude, when launching uphill the slope adds to the friction, the thrust force and the horizontal forces applied to the piers are higher, but the movements of the superstructure are better controlled. When launching downhill the forces are lower but controlling them is more difficult and delicate, as the reduction in the coefficient of friction after breakaway may change the direction of the force to be applied. In practice, during the stops for construction of a new segment during a downhill launch the deck is anchored to the launch abutment, and thermal contraction (and, toa lesser extent, creep and shrinkage of concrete) cause the superstructure to move slightly uphill at the pier bearings. Since the slope and the friction act in opposite106 BRIDGE LAUNCHING Fig. 2.67. Pier cap forces for a local gradient of the launch surface greater than the static frietion (4 > ) directions during launching, the launchwards pier deflections are very small, if any. Therefore, deck contraction may easily recover these residual deflections and cause a situation of ‘negative friction’ in which the friction forces and the slope com- ponent act in the same direction. Sometimes it is also necessary to pull the deck backwards (uphill) to recover launch errors. For all of these reasons, the pier should always be designed for the slope added to the friction, independent of the planned launch direction. The horizontal force transferred to the pier may accidentally exceed the theoretical value based on friction and slope. A typical and not infrequent case is the insertion of a neo-fion pad upside down, with Teflon in contact with the super- structure and neoprene in contact with the launch bearing. In the case of tail piers, it is prudent to use real-time monitoring systems (inclinometers) and emergency stops, when the pier is not directly supported with horizontal braces or inclined stays. INCREMENTAL LAUNCHING OF BRIDGES 107 Once the system of forces applied to the pier cap has been determined, the eccentricity of its application point with respect to the pier axis can be evaluated. In theory, the neo-flon pads should assure a uniform distribution of the support reaction along the contact surface, so that Ry can be assumed as centred on the launching bearings. In this case, the launching bearings can be shifted backwards from the pier axis to create a load eccentricity, e, able to reduce the moment generated in the pier-base section by the horizontal component H of the force applied to the cap. Setting /, as the pier height, the backwards eccentricity sufficient to cancel the pier base moment is (2.48) Tn practice, the inevitable tolerances and the imperfect parallelism of the launch bearings and the deck surface make the vertical force transferred to the pier eccentric. This eccentricity is often modest; however, its negligibility must be veri- fied case by case on the basis of the length of the contact surface of the launching bearings compared with the longitudinal pier dimension. Finally, the pier itself has a tolerance in verticality (especially if built with slip forms) and the creep of con- crete increases over time the launchwards deflection generated by the horizontal force applied to the pier cap. The vertical load transferred to the pier is also eccentric in the transverse direction as a result of level tolerances in the launching bearings and geometric imperfections in the superstructure, which together generate load differences between the two bearings. Transverse eccentricity of the support reaction is rarely critical in the permanent piers, which are designed for full service loads; in the temporary piers, when necessary, it may be controlled with jack-bearings able to adjust the torsion in the superstructure and the resulting transverse bending in the pier (Fig. 2.65). For long bridges or steep launch gradients it may be necessary to place friction thrusters at some piers to better distribute the thrust action (Fig. 2.68). Front staying of the permanent piers equipped with thrusters is generally unavoidable. The experience gained with friction thrusters placed on stayed piers led to the development of launch reactors. These hydraulic devices, schematically similar to small friction launchers, apply to the superstructure a thrust force equal to their own internal friction combined with the local grade, so that no resulting horizontal force is transferred to the pier. These devices are directly controlled by pier inclinometers to make their response instantaneous and free from the thrust pulses of the main launchers. Synchronization of the launch reactors with the main thrust systems is delicate, and several launch interruptions are necessary because the reactors do not reach their end of stroke simultaneously. For all of these reasons, these devices are quite expensive and should be reserved for the most critical cases. As regards stress verification, the pier is a cantilever partially fixed at the base and is therefore very sensitive to the couples and the horizontal forces applied to the108 BRIDGE LAUNCHING Fig. 2.68. Pier friction thrusters and rigid stays are often necessary for steep-grade uphill launching top. Consequently, pier instability must be assessed during launch as well. The slenderness ratio of the pier, A, is (2.49) where r is the cross-sectional radius of gyration and Lo is the effective length (double the pier height in the case of a perfectly fixed footing, and progressively longer as the degree of base constraint decreases). When the slenderness ratio is not INCREMENTAL LAUNCHING OF BRIDGES 109 very high it is sufficient to calculate the stresses neglecting the pier deflections and to check the critical cross-sections for combined compressive and bending stresses with the amplification factors prescribed by codes, which are generally based on Euler’s critical load: (2.50) creep deflection of the pier generated by the constant application of a horizontal force to its top. Finally, if the pier is slender and deformable, the stresses must be determined by accounting for the launchwards deflection, ic. with an analysis including non-linear P-delta effects. Similar verifications must be carried out in the transverse direction as well. In this case, however, the low guide forces applied to the superstructure rarely require non- linear analyses and the torsional stiffness of the superstructure often results in negligible rotations of the pier cap. Therefore, the effective pier length mainly depends on the transverse mobility of the superstructure and is < 1g

You might also like

• Fear: Trump in the White House

  

  From Everand

  Fear: Trump in the White House

  Bob Woodward

  Rating: 3.5 out of 5 stars

  3.5/5 (738)
• A Man Called Ove: A Novel

  

  From Everand

  A Man Called Ove: A Novel

  Fredrik Backman

  Rating: 4.5 out of 5 stars

  4.5/5 (4610)
• The Sympathizer: A Novel (Pulitzer Prize for Fiction)

  

  From Everand

  The Sympathizer: A Novel (Pulitzer Prize for Fiction)

  Viet Thanh Nguyen

  Rating: 4.5 out of 5 stars

  4.5/5 (122)
• • Magazines
  • Podcasts
  • Sheet music
• A Heartbreaking Work Of Staggering Genius: A Memoir Based on a True Story

  

  From Everand

  A Heartbreaking Work Of Staggering Genius: A Memoir Based on a True Story

  Dave Eggers

  Rating: 3.5 out of 5 stars

  3.5/5 (231)
• Grit: The Power of Passion and Perseverance

  

  From Everand

  Grit: The Power of Passion and Perseverance

  Angela Duckworth

  Rating: 4 out of 5 stars

  4/5 (590)
• Devil in the Grove: Thurgood Marshall, the Groveland Boys, and the Dawn of a New America

  

  From Everand

  Devil in the Grove: Thurgood Marshall, the Groveland Boys, and the Dawn of a New America

  Gilbert King

  Rating: 4.5 out of 5 stars

  4.5/5 (266)
• Never Split the Difference: Negotiating As If Your Life Depended On It

  

  From Everand

  Never Split the Difference: Negotiating As If Your Life Depended On It

  Chris Voss

  Rating: 4.5 out of 5 stars

  4.5/5 (844)
• Angela's Ashes: A Memoir

  

  From Everand

  Angela's Ashes: A Memoir

  Frank McCourt

  Rating: 4.5 out of 5 stars

  4.5/5 (441)
• The Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good Life

  

  From Everand

  The Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good Life

  Mark Manson

  Rating: 4 out of 5 stars

  4/5 (5810)
• Shoe Dog: A Memoir by the Creator of Nike

  

  From Everand

  Shoe Dog: A Memoir by the Creator of Nike

  Phil Knight

  Rating: 4.5 out of 5 stars

  4.5/5 (540)
• Principles: Life and Work

  

  From Everand

  Principles: Life and Work

  Ray Dalio

  Rating: 4 out of 5 stars

  4/5 (608)
• The World Is Flat 3.0: A Brief History of the Twenty-first Century

  

  From Everand

  The World Is Flat 3.0: A Brief History of the Twenty-first Century

  Thomas L. Friedman

  Rating: 3.5 out of 5 stars

  3.5/5 (2259)
• Her Body and Other Parties: Stories

  

  From Everand

  Her Body and Other Parties: Stories

  Carmen Maria Machado

  Rating: 4 out of 5 stars

  4/5 (822)
• The Little Book of Hygge: Danish Secrets to Happy Living

  

  From Everand

  The Little Book of Hygge: Danish Secrets to Happy Living

  Meik Wiking

  Rating: 3.5 out of 5 stars

  3.5/5 (401)
• A Tree Grows in Brooklyn

  

  From Everand

  A Tree Grows in Brooklyn

  Betty Smith

  Rating: 4.5 out of 5 stars

  4.5/5 (1929)
• The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers

  

  From Everand

  The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers

  Ben Horowitz

  Rating: 4.5 out of 5 stars

  4.5/5 (348)
• The Gifts of Imperfection: Let Go of Who You Think You're Supposed to Be and Embrace Who You Are

  

  From Everand

  The Gifts of Imperfection: Let Go of Who You Think You're Supposed to Be and Embrace Who You Are

  Brené Brown

  Rating: 4 out of 5 stars

  4/5 (1092)
• John Adams

  

  From Everand

  John Adams

  David McCullough

  Rating: 4.5 out of 5 stars

  4.5/5 (2409)
• Yes Please

  

  From Everand

  Yes Please

  Amy Poehler

  Rating: 4 out of 5 stars

  4/5 (1898)
• The Emperor of All Maladies: A Biography of Cancer

  

  From Everand

  The Emperor of All Maladies: A Biography of Cancer

  Siddhartha Mukherjee

  Rating: 4.5 out of 5 stars

  4.5/5 (271)
• The Glass Castle: A Memoir

  

  From Everand

  The Glass Castle: A Memoir

  Jeannette Walls

  Rating: 4.5 out of 5 stars

  4.5/5 (1716)
• Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race

  

  From Everand

  Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race

  Margot Lee Shetterly

  Rating: 4 out of 5 stars

  4/5 (897)
• Team of Rivals: The Political Genius of Abraham Lincoln

  

  From Everand

  Team of Rivals: The Political Genius of Abraham Lincoln

  Doris Kearns Goodwin

  Rating: 4.5 out of 5 stars

  4.5/5 (234)
• Rise of ISIS: A Threat We Can't Ignore

  

  From Everand

  Rise of ISIS: A Threat We Can't Ignore

  Jay Sekulow

  Rating: 3.5 out of 5 stars

  3.5/5 (137)
• The Woman in Cabin 10

  

  From Everand

  The Woman in Cabin 10

  Ruth Ware

  Rating: 3.5 out of 5 stars

  3.5/5 (2521)
• The Art of Racing in the Rain: A Novel

  

  From Everand

  The Art of Racing in the Rain: A Novel

  Garth Stein

  Rating: 4 out of 5 stars

  4/5 (4203)
• Wolf Hall: A Novel

  

  From Everand

  Wolf Hall: A Novel

  Hilary Mantel

  Rating: 4 out of 5 stars

  4/5 (3811)
• The Yellow House: A Memoir (2019 National Book Award Winner)

  

  From Everand

  The Yellow House: A Memoir (2019 National Book Award Winner)

  Sarah M. Broom

  Rating: 4 out of 5 stars

  4/5 (98)
• Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future

  

  From Everand

  Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future

  Ashlee Vance

  Rating: 4.5 out of 5 stars

  4.5/5 (474)
• On Fire: The (Burning) Case for a Green New Deal

  

  From Everand

  On Fire: The (Burning) Case for a Green New Deal

  Naomi Klein

  Rating: 4 out of 5 stars

  4/5 (74)
• The Unwinding: An Inner History of the New America

  

  From Everand

  The Unwinding: An Inner History of the New America

  George Packer

  Rating: 4 out of 5 stars

  4/5 (45)
• The Light Between Oceans: A Novel

  

  From Everand

  The Light Between Oceans: A Novel

  M.L. Stedman

  Rating: 4.5 out of 5 stars

  4.5/5 (789)
• Sing, Unburied, Sing: A Novel

  

  From Everand

  Sing, Unburied, Sing: A Novel

  Jesmyn Ward

  Rating: 4 out of 5 stars

  4/5 (1104)
• The Perks of Being a Wallflower

  

  From Everand

  The Perks of Being a Wallflower

  Stephen Chbosky

  Rating: 4.5 out of 5 stars

  4.5/5 (2104)
• The Constant Gardener: A Novel

  

  From Everand

  The Constant Gardener: A Novel

  John le Carré

  Rating: 3.5 out of 5 stars

  3.5/5 (104)
• The Outsider: A Novel

  

  From Everand

  The Outsider: A Novel

  Stephen King

  Rating: 4 out of 5 stars

  4/5 (1850)
• Brooklyn: A Novel

  

  From Everand

  Brooklyn: A Novel

  Colm Toibin

  Rating: 3.5 out of 5 stars

  3.5/5 (1947)
• Little Women

  

  From Everand

  Little Women

  Louisa May Alcott

  Rating: 4 out of 5 stars

  4/5 (104)
• Steve Jobs

  

  From Everand

  Steve Jobs

  Walter Isaacson

  Rating: 4.5 out of 5 stars

  4.5/5 (807)
• Manhattan Beach: A Novel

  

  From Everand

  Manhattan Beach: A Novel

  Jennifer Egan

  Rating: 3.5 out of 5 stars

  3.5/5 (792)
• Bad Feminist: Essays

  

  From Everand

  Bad Feminist: Essays

  Roxane Gay

  Rating: 4 out of 5 stars

  4/5 (1018)