CABLE STAYED FOOTBRIDGE MADE OF UHPC -...

High Performance Fiber Reinforced Cement Composites (HPFRCC7), Stuttgart, Germany – June 1-3, 2015

449

CABLE STAYED FOOTBRIDGE MADE OF UHPC

J.L. Vitek (1), M. Kalný (2) and R. Coufal (3)

(1) Metrostav a.s. and CTU Prague, Prague, Czech Republic

(2) PONTEX, Ltd., Prague, Czech Republic

(3) TBG Metrostav, Ltd., Prague, Czech Republic

Abstract A cable stayed footbridge over the Labe River in Čelákovice is the first structure in the

Czech Republic, where the ultra high performance concrete (UHPC) was applied in a load carrying structure. The UHPC was developed from local materials. The self-compacting UHPC was delivered to the production plant as ready mixed concrete. Extensive testing of the material and of structural elements provided necessary data for design. The production technology of segments was developed. Large segments (11.3 m long) were assembled by cantilever method using a special launching gantry. The footbridge is a lightweight structure with long durability and low requirements on maintenance.

1. INTRODUCTION

The footbridge over the Labe River is located in the town of Čelákovice about 25 km east of Prague. While the town of Čelákovice is on the left bank of the Labe River, the recreational area with a large forest is on the right bank. The town of Čelákovice decided to build a new footbridge, which would provide a comfortable connection between the town and the recreational area for pedestrians and cyclists. Also light utility vehicles, preferably those used by an integrated rescue system, may use the new footbridge. In the feasibility study two alternatives were assumed – a cable stayed bridge and a suspension bridge. Finally the client (Town of Čelákovice) preferred to build a cable stayed structure with the main span crossing the river 156 m long. The client required to build a bridge without supports in the river, which created a landmark of the town.

The design of the footbridge developed since initial stages in 2004. The hybrid steel concrete deck was considered in the tender design in 2012 [2]. Metrostav was developing the ultra high performance concrete (UHPC) at that time. The construction of the footbridge became an excellent opportunity for its application. Metrostav as a principle contractor proposed the change of the original project and a cable stayed footbridge with a segmental precast bridge deck completely made of UHPC was designed.


450

2. STRUCTURAL DESIGN

The footbridge has three spans (43 + 156 + 43 m) and it is 3 m wide. The two steel pylons (37 m high) of the A shape are located on the river banks. The longitudinal section is plotted in Fig. 1. The deck has two edge beams which are prestressed by prestressing bars (during construction) and by two prestressing tendons composed of 15 strands 15.7 mm in diameter (after completion).

Figure 1: Longitudinal section of the footbridge

The slab is 60 mm thick and it is reinforced by transversal ribs in the distance of approximately 1 m. The ribs are reinforced by two bars 16 mm in diameter (Fig. 2). The segments were produced from UHPC. Because of the first application of UHPC, the design was made under the assumption of the concrete class C110/130. The conservative approach was accepted because of little experience with a new material. However, the measurements on the site showed that the real average cylinder strength varied in the range 145 to 150 MPa.

Figure 2: Cross-section of the footbridge deck

The foundation blocks under the pylons and abutments are supported by piles. Abutments are massive reinforced concrete blocks which form a counterweight to uplift reactions of the superstructure. The legs of the pylons are at the top part connected by struts in which the stays


451

are anchored. Under the bridge deck, there is also one strut connecting the legs of the pylons, which is used for lateral guiding of the bridge deck without carrying any vertical load.

The bridge deck is supported by stays in the distance of 11.3 m. This is a rather long distance for segmental bridge deck. This arrangement came from earlier design stages and the number of stays could not be increased due to tender conditions. Therefore the segments 11.3 m long had to be produced and assembled on the site. The joints were glued by epoxy resin as it is usual at segmental bridges. The stresses in joints proved to be critical for design, especially in construction stages, where decompression had to be avoided with a compression stress reserve of 1 MPa.

The stays (delivered by Redaelli) were made of the fully locked rope stays. Their surface is closed by “Z” wires with galvanization corrosion protection. The diameter of the rope varied from 20 to 36 mm in dependence on the required load carrying capacity. At the top end, the stays were connected to the pylon using a fork and pin joint. At the bottom end the stays were equipped with a threaded bar which formed an adjustable connection to the steel anchorage elements embedded in the edge beams of the bridge deck.

3. DEVELOPMENT AND TESTING OF UHPC

The UHPC was developed by the team of Metrostav and TBG Metrostav for both, precast and cast in situ structures. The aim of the research was to develop UHPC from local constituents [1]. The compression strength should exceed 150 MPa (measured on standard cylinders) and the flexural strength was required at least 15 MPa. In order to avoid brittle failure the high strength steel fibres were added in a rather high amount of 160 kg/m3 of UHPC. The developed UHPC was transported to the production plant, where segments were cast; therefore workability about 90 minutes was required. For casting of segments the self-compacting UHPC was considered as the best alternative. Beside mechanical parameters also the development of heat of hydration, autogenous shrinkage, drying shrinkage and creep was measured on laboratory specimens [3].

Additional experiments were focused on verification of the basic elements of the footbridge. The small dimensions of the edge beams of the bridge deck did not allow for application of standard anchor plates of prestressing bars. Smaller anchor plates had to be used. The resistance of UHPC elements loaded by prestressing force applied on a smaller anchor plate was investigated experimentally using the ETAG 13 recommendations. It was observed that the smaller anchor plate in completely sufficient, even if the reinforcement was significantly reduced, since the fibre reinforcement of UHPC is able to carry the stresses under the anchors. The two alternatives were tested. i) No bar reinforcement was used under the anchors and ii) small longitudinal bars and stirrups were in the specimens, since such reinforcement was used in edge beams of the bridge deck. The cracks did not appear in the specimens until the load reached the level 1.4 – 1.7 of the characteristic prestressing force Fpk. The tests were executed at the age of concrete of 5 days.

The transversal ribs of the footbridge are reinforced by two bars of 16 mm in diameter. The slab is not reinforced by any reinforcement with exception of fibres. The load carrying capacity of the deck element in transversal direction was tested. Initially the two point loads were applied (representing the axial load of a light vehicle). The maximum axial load is about 25 kN. When the load of 80 kN was achieved and no cracks appeared, the test was rearranged and the only one point load located in the middle of the slab between the transversal ribs was


452

increased up to 110 kN. The slab failed in bending of the transversal ribs (Fig. 3). The slab itself remained without any damage (with exception of small cracks) although the point load was located between the ribs.

Figure 3: Failure of the transversal ribs (failure load 110 kN)

In order to verify the slab (60 mm thick) performance exposed to punching, the test of the complete segment was prepared. The ribs were supported in the middle of their length and the slab was loaded by a point load acting on the circular area 200 mm in diameter. The load was increased until failure. The first cracks appeared at the level of loading of about 150 to 200 kN, the failure loads varied in the range of 320 to 370 kN. (The test was repeated 4 times). The obtained failure loads were different because of the slightly different structural arrangement of the test. The smaller failure load was measured at the end of the segment (the end span of the slab between ribs) and in the vicinity of the drainage opening. The scatter of the measured failure loads was reasonably small. The tests proved that the load carrying capacity of the slab in much higher than it was required in the design. It was accepted as a conservative design, taking into account the first application of the UHPC.

Additional extensive testing was focused on the bond of mild steel and UHPC and of the prestressing steel and UHPC. A number of pullout tests were carried out. Very favourable results were obtained. The anchorage length of the mild steel bar longer than 3 diameters resulted in failure of the reinforcing bar, while the anchorage zone remained without any damage. The bond of prestressing strands and UHPC was also significantly better than their bond with normal strength concrete. The anchorage lengths could be considerably reduced in UHPC. The larger set of tests will be required, since the statistical evaluation would be required before specification of design recommendations.

4. PRODUCTION OF SEGMENTS

The segments have a complex shape. The top surface is inclined and the top level of the edge beams is well above the slab. The application of self-compacting concrete required to apply the top formwork. Casting of segments in other positions would not lead to easier casting process.


453

Figure 4: Failure of the slab in punching (failure load 370 kN)

A typical segment had the length of 11.3 m. Such long segment would be very difficult to cast in one stage. The casting process is rather complex using a self-compacting concrete. It was very important to keep the homogeneity of UHPC especially the uniform distribution of fibres. Therefore it was decided to cast the segments in two parts, each 5.65 m long. The segments had one joint fixed (to form a large segment) and one joint free, i.e. the joint which was glued on site. The fixed joint was a classical working joint which was reinforced with mild reinforcement. The segments were cast in a steel mould, which was designed and produced specifically for this project. The mould was 5.65 m long; it had a fixed bottom part, longitudinal tiltable sideboards and a top cover. The mould had only one face, since a match casting was assumed, and the second face was created by the segment which was cast earlier. Each second joint was a working joint (reinforced) and the other joints were the match cast joints without reinforcement, later glued on the site during assembly of the bridge deck. The self-compacting concrete was used, which normally creates a horizontal top surface, but it was not required in this case.

The mould was equipped with reinforcement, ducts for prestressing bars and tendons and with steel elements for the anchorage of stays. Then the mould was filled with self-compacting concrete from two truck mixers that flowed into the openings in the middle of the longitudinal edge beams (Fig. 5). The concrete filled the complete mould space. The air went out by the openings in the top cover, which were subsequently closed and which served also for inspection of the casting process. The concrete was transported from the mixing plant of TBG Metrostav, which was about 25 km from the precasting plant, which is situated at the river harbour, so that the segments could be shipped to the site. The hardening of concrete had to be slowed down because of the transport of concrete to a long distance. In order to accelerate the production, the segments were heated up to the temperature of 60°C which allowed for demoulding after 7 to 8 hours. The segments could be produced in 2 – 3 days intervals, depending on the weather conditions. During hardening, the temperatures in concrete were monitored and the hardening process was carefully observed, so that the earliest possible time for demoulding was specified. After demoulding the segment was moved horizontally and formed a face for the next segment.


454

Figure 5: Filling of the mould from two truck mixers

5. ASSEMBLY OF THE FOOTBRIDGE

After construction of the foundations, the steel pylons were transported to the site in three parts, which were welded on the site. Then the complete pylon was lifted by two cranes and anchored to the concrete foundation. At the ends of the footbridge the abutments were cast.

The bridge deck was assembled from the segments, which were initially connected by prestressed bars located in edge beams. The side spans were assembled on the light fixed scaffolding delivered by PERI. The first segment was assembled under the pylon and the others were laid on the scaffolding, then connected to the previous segment and prestressed by prestressing bars. The remaining part at the abutment was cast in situ using ordinary concrete.

The main span was assembled by cantilever method. The first segment, adjacent to that under the pylon, was laid on the shore and lifted by a launching gantry. The following segments were transported by pontoons and also lifted by the launching gantry (Fig. 6). The launching gantry was designed for the footbridge. There were two identical products which were used for assembly of the main span symmetrically from both sides of the river. The launching gantry had two main longitudinal steel beams which were anchored to the top of the already completed footbridge and it formed a cantilever approximately 12 m long. The new segment was lifted from the pontoon (or from the shore in the case of the first segment) using four bars operated by hydraulic jacks. The new segment was lifted about 0.5 m from the last segment. Then the prestressing bars (32 and 36 mm in diameter) were connected. Then the new segment was moved longitudinally to the previous segment. The joint was provided with epoxy glue and it was closed by moving the frame and by prestressing of the bars.

The cantilever of the launching gantry had an auxiliary stay suspended on the top of the pylon, which carried a substantial weight of the lifted segment. The reaction at the pylon was transferred to the abutment by back auxiliary stays. After prestressing of the new segment, the definitive stays were installed. Then it was possible to release the auxiliary stay on the launching gantry and to move the gantry into the next position.

After assembly of 7 segments on each side of the river, the middle short segment was lifted and two closing joints were cast in situ. For this operation only one launching gantry was


455

used. It provided a stiff connection between the two cantilevers. The formwork for the two closing joints was suspended on the launching gantry. When the closing joints were cast and the concrete hardened, the longitudinal tendons (2 x 15 strands of 15.7 mm in diameter) could be prestressed and grouted.

Figure 6: Assembly of the segments over the river using a launching gantry

During assembly of the segments, their positions and the forces in stays were carefully monitored and adjusted if necessary. After finishing of the superstructure, the sprayed waterproofing which is used also as a pavement was installed together with a light steel railing. During the static and dynamic loading test the deflections and natural frequencies were measured and a good agreement with the assumptions of the structural analysis was found. The bridge was completed in spring 2014 (Fig. 7).

6. CONCLUSIONS

The project proved that the development of UHPC from local constituents was successful. A lot of experience was collected, but many problems remain for future solution. Many decisions were based on experimental verification. Recently various recommendations for design were published which will be later modified into design codes. Contemporary design recommendations, which have to be taken into account, should be updated. Higher concrete strength and its resistance against environmental effects require also modification of the prescribed concrete cover, detailing rules, etc. Without such changes the structures made of UHPC would suffer in competition with other structures.

During construction of the footbridge, it appeared that the design is rather conservative. It was a correct approach at the time of the design and planning, if the footbridge was designed now, additional savings could be found. However, the existing footbridge is lighter and more durable than that which would be built without application of UHPC according to the previous proposals.

The footbridge in Čelákovice is the first bridge in the Czech Republic, which has a superstructure made of UHPC. At the moment, it is also a cable stayed bridge with the largest span in the country. Main participants of the construction are listed in the Table 1.


456

Figure 7: A completed footbridge

Table 1: Main participants of construction

Client The town of Čelákovice Bridge Designer Pontex Consulting Engineers, Ltd. Design Supervision Stráský, Hustý and Partners, Ltd. Main Contractor Metrostav, a.s Major Subcontractors TBG Metrostav, VSL Systemy CZ, PERI, Freyssinet CS, OK-BE Construction time 01/2013 – 04/2014

ACKNOWLEDGEMENTS

The authors acknowledge the contribution of J. Komanec and V. Kvasnička (Pontex, Ltd.), who were involved in the design, the work of J. Brož, P. Koukolík and O. Koliba, who were responsible for construction (Metrostav, a.s.). The UHPC research was partly supported by the Ministry of Industry and Trade of the Czech Republic (Project no. FR-TI3/531) and some measurements during the construction of the footbridge were supported by Technological Agency of the Czech Republic (CESTI – TE01020168)

REFERENCES

[1] Vítek, J.L., Coufal, R., Čítek, D., UHPC – Development and Testing on Structural Elements. Elsevier, Procedia Engineering 65, 218-223

[2] Kalný, M., Kvasnička, V., Komanec, J., Cable stayed footbridge with UHPC deck. Proc. of the 1st International Concrete Innovation Conference, Oslo, Norway, June 11/13, 2014

[3] Vítek, J.L., Coufal, R., Brož, R., Footbridge segments made of UHPC. Proc. of the 9th International Conference on Short and Medium Span Bridges, Calgary, Alberta, Canada, July 15-18, 2014

CABLE STAYED FOOTBRIDGE MADE OF UHPC -...

Documents

Transcript of CABLE STAYED FOOTBRIDGE MADE OF UHPC -...