The Charming History of the Incremental Launching Method

The charming history of the incremental launching method by Marco Rosignoli

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, Mörsch and Ross, among others.

The design of most of these new bridges was limited to isostatic or poorly redundant systems, since analysis of hyperstatic systems was not compatible with the calculation means available at that time. For many years, the basic criterion in the design of prestressed concrete (PC) bridges has been to assure the possibility of simple structural analysis. So, multispan bridges cast on falsework used the simple support scheme, the first balanced cantilever bridges in the 1950s were hinged in midspan, and the first PC decks built with movable scaffolding systems (MSS) in the 1960s were supported at the piers or articulated at the counterflexure points. Most of the PC bridges of the post-war period were also designed to minimize the quantities of structural materials.

From the 1960s onwards, the extraordinary progress in computing techniques extended the analysis capability of hyperstatic structures. The technological maturity of prestressing led to light, ductile bridges with enhanced capability of resisting temporary construction stresses. Enhanced knowledge of materials stimulated technological innovation, and materials with higher performance found intelligent use and adequate processing methods. Increasing labor costs amplified the labor component of the construction cost of new bridges, and new construction methods were developed to rationalize production, regularize quality, increase the erection rate, and assure safety of workers and public.

In the creative thinking of these decades, several construction methods were developed to assure competitiveness of PC over steel construction. At the same time, however, the lower cost of high-quality hot-rolled steel plates, better workshop organization, new splicing techniques, and new field assembly methods extended the use of steel-concrete composite bridges to spans that previously were the domain of PC bridges. New types of prestressed composite box girders combining PC slabs and steel corrugated-plate webs were also developed to address these borderline spans. This further stimulated research and created trans-disciplinary connections between different technologies.

New construction methods took advantage of recent technology advances to widen the field of application of already familiar techniques and to diminish the labor component of the construction cost of the bridge. Structures can be moved, but this requires a different way of thinking about them, and the availability of adequate technology. So, although the idea of launching a bridge is not new (think of a tree-trunk in Palaeolithic technology) and numerous steel bridges have been launched in the 19th century, launching PC bridges was made possible one century later by the availability of a new low-friction material such as PTFE (Teflon®).

For many decades, the light weight of steel structures had permitted their launch by means of winches and lubricated wooden skids, with frictional loads that, although considerable, did not cause excessive stresses in the piers or require expensive launch equipment. The flexural efficiency of the steel girders and the ability of the material to indifferently resist tensile and compressive stresses facilitated launching and avoided overdesign. Launching does offer several advantages over in-air erection; however, compared with ground crane assembly, costs were often higher due to the availability of only one working point, and launching was initially limited to bridges high above the ground or in inaccessible areas.

In the eyes of the PC bridge pioneers, some of the disadvantages of launch technology were less critical, and others even promising. Construction duration and yard organization for a PC bridge are different from those for a steel bridge, and the cost of labor and equipment is so high that every possible alternative must be examined. Launching of a PC deck built on the ground promised savings in both labor and equipment, but its practical application was limited by the weight of the deck and the low tensile strength of concrete.

These obstacles were gradually overcome. Advances in prestressing technology lightened the deck, made it more flexible and ductile and less subject to cracking, simplified splicing of tendons, and allowed introduction and removal of prestressing according to need. The commercial availability of personal computers and structural analysis programs simplified analysis of the continuous beam in the multiple support configurations of launching. The development of steel-Teflon skids offered a substantial reduction in launch friction, and technological advances in electro-hydraulic equipment offered the possibility of moving huge masses with due precision.

Nothing of this was available in the year 1950, when a small 3-span PC bridge was built in France on the two banks of a river and launched on a full-length falsework over the river for midspan closure. The 56m Vaux-sur-Seine Bridge was the first launched bridge, and the first bridge to use external continuity prestressing.

The year 1959 saw a first attempt to launch precast concrete segments for a 280m-long, 4-span bridge over the Ager River in Germany. The deck was assembled on a full-length timber falsework. Precast segments cast behind the abutment were skidded into position along the falsework by means of a wooden rail lubricated with engine oil. The joints between the 9.5m segments were cast in-place, and long external tendons within the box cell were tensioned from the end diaphragms to provide post-tensioning.

The Ager Bridge experience suggested avoiding the falsework and launching the completed deck over the piers. This concept was applied in the year 1961 for the Rio Caroni Bridge in Venezuela. The 480m deck was assembled full-length behind the abutment by joining precast segments with wet joints. On completion of assembly, the deck was post-tensioned with a large external U-tendon within the box cell to obtain a centroidal force. The tendon was anchored to the front end diaphragm and deviated around a semi-cylindrical concrete block at the rear deck end. The rear deviation block was jacked away from the end diaphragm to generate the required prestressing force.

A steel truss extension was applied to the front end of the deck to control negative self-weight bending in the front cantilever, and one temporary pier was erected in every span to halve the launch span. The deck was launched with movable bearings sliding on low-friction surfaces applied to the pier-caps; at the end of each launch stroke, therefore, the deck had to be jacked to relocate the bearings.

On launch completion, the external tendon was made eccentric by lowering the midspan deviation points and by lifting the pier deviation points. During these operations, the jacks at the rear deviation block were progressively retracted to keep the pull in the tendon constant. The tendon was finally attached to the inner face of the webs and protected with pre-pack concrete.

Adjusting the geometry of the loaded tendon was a complex and expensive operation, and this led to the idea of combining permanent axial tendons designed for the launch stresses with integrative draped tendons applied at the end of launching. This concept characterizes most launched bridges since then. Full-length assembly of a precast segmental deck behind the abutment also showed limitations, and this led to the idea of match-casting longer deck segments behind the abutment and progressively launching the deck. Launching thus became incremental.

The incremental launching method was applied for the first time in 1965 for the construction of a PC bridge over the River Inn in Kufstein, Germany. The deck was cast segmentally behind the abutment, on a platform located along the projection of the bridge axis. After curing the new segment, the deck was pushed forward to clear the casting cell. Another segment was match-cast against the rear deck end, and the process (adding a new segment and launching the entire deck) was repeated until deck completion. The use of multiple temporary piers avoided the need for launch prestressing in the Kufstein Bridge. The launch stresses were controlled with reinforcement, and parabolic prestressing was applied at the end of launching.

In the following years, incremental launching construction was improved in multiple aspects. Low-friction neoprene-Teflon (neo-flon) pads are now inserted between the deck and fixed launch bearings to avoid deck lifting. Modular launch noses control shear and negative self-weight bending in the front deck region in combination with specialized schemes of launch prestressing. The thrust devices offer reliable and smooth electro-hydraulic operations synchronized by programmable logic controllers (PLC). Fast algorithms have been developed to analyze the launch stresses in the continuous beam and to provide time-dependent stress envelopes that simplify structural design.

After a few years required to reach technological maturity, incremental launching construction began to compete with balanced cantilever construction of long PC viaducts. For different reasons, the advantages of rapid deck construction in a fixed location found interest in bridge owners and contractors. The launched bridges can be rapidly built over active highways and railways without impact on traffic, are perfectly compatible with urban environment and sensitive areas, and assure safety for workers and public. Fixed logistics diminishes the construction cost, enhances quality, and allows continuous operations in bad weather.

Although the incremental launching method was originally conceived for highway bridges of a few hundred meters of length, it has eventually been applied to shorter and longer highway and railway bridges, in ingenious applications all over the world. The first launched bridges of the French TGV high-speed railway lines date back to 1997. Incremental launching thus became one of the most competitive construction methods for PC bridges of medium length and simple geometry. The number of PC bridges launched in the 20th century exceeds 2500, and their total deck surface exceeds 3 millions of square meters.

The last decades have seen ingenious applications of the launch techniques: curved decks with tight radius, decks launched over arches, cable-stayed decks launched symmetrically from the opposite abutments for midspan closure or cast on temporary supports and rotated into position, and continuous decks launched over temporary piers and suspended from pylons or tied arches on launch completion.

The technological maturity of incremental launching opened new perspectives for monolithic handling of heavy and ultra-heavy structures. Although incremental launching is still the most common of these construction methods and its evolution has influenced the development of the others, monolithic launching, symmetrical launching, rotation and jacking and skidding have acquired their own role for specific applications. From the 1970s on, hundreds of bridge decks have been lifted, lowered, rotated, launched or skidded all over the world.

In the field of steel constructions, a better knowledge of instability and the commercial availability of high-grade steels with reliable mechanical properties have led to lighter steel girders that are easier to launch. Non-stiffened webs are used to maximize robotized welding with profile-tracking equipment, and control of web buckling required improvements in the launch bearings. Light launch noses and specialized systems for transfer of the thrust force are also available.

In spite of the advances in launch technology, the different cost of reinforced-concrete slabs and orthotropic steel decking systems still suggests the use of reinforced concrete for the deck slab of medium-span bridges. In a continuous beam, controlling deck slab cracking in the negative bending regions requires oversized longitudinal reinforcement and discontinuous segmental casting sequences that increase costs and construction duration of the deck slab. This lead to the idea of incrementally launching a continuous concrete slab over pre-launched steel girders.

Research in PC also led to external prestressing. When external tendons are used, the web thickness can be designed for the principal compressive stress allowed by the design standards for the serviceability limit state (SLS). Lighter webs diminish the weight of the cross-section and enhance its flexural efficiency. When shear dictates the minimum web thickness, steel webs offer a higher strength-to-density ratio than concrete webs. This led to the concept of prestressed composite box girders with steel corrugated-plate webs, where two reinforced concrete slabs resist bending, and deviation of external tendons reduces the shear force in the webs down to levels that can be resisted by corrugated steel plates. Compared with a PC box girder, a prestressed composite box girder with steel corrugated-plate webs offers savings in weight, concrete, reinforcement and prestressing. Compared to a non-prestressed composite deck, these new sections require only a quarter of the steelwork. Finally, advance in the launch technologies for PC and steel bridges led to the incremental launching construction of the first prestressed composite bridges.