Bridgeweb
March 12, 2010
Email Updates
Concrete evidence
Fatigue cracks are a major problem for orthotropic steel bridge decks, and cracks in the steel deck plate are of particular concern given their effect on traffic safety. Peter Buitelaar, René Braam and Niek Kaptijn report on a
Published:  September 02, 2005

Two bridge deck overlay replacement projects that have been under way in the Netherlands this summer represent the culmination of a major research effort into solving problems with fatigue damage to steel deck bridges.

Projects to rehabilitate the Hagenstein bridges on the A27 near Vianen and the Moerdijk Bridge on the A16 near Breda, which started this summer, are the first large-scale application of a new concrete overlay system that is intended to solve problems with fatigue on such structures.

Serious damage to the bascule of the Van Brienenoord Bridge on one of the main highways in the Netherlands several years ago resulted in a special task force being formed within the civil engineering division of the Dutch Ministry of Transport, Public Works & Water Management. The aim was to investigate the cause, to understand and control the fatigue mechanism for the 80 steel fixed and movable bridges in the Netherlands and to develop practical solutions for their cost-effective rehabilitation and renovation.

A major research project has been carried out over the last six years, including a pilot project in 2003, to try and develop a new high strength concrete wearing course for orthotropic steel bridges which can also extend the service life of the total construction by solving fatigue problems in specific deck details. The resulting solution is very promising since it turns the deck plate into a much more rigid construction with a higher 'plate factor' due the monolithic composite interaction between the reinforced high performance concrete overlay and the steel deck plate. The overlay, with a minimum thickness of 50mm, has been found to reduce stress by a factor of four to five in the deck plate and three to four in the trough wall, extending the service life of the orthotropic bridge deck by a matter of decades without additional maintenance.

Developments in concrete over the last few decades have resulted in high performance concrete of 100-155MPa and even ultra high performance concrete, which can be higher than 400MPa. Much higher strengths can be achieved by a further densification of the cement matrix combined with an additional pressure and heat treatment during setting and hardening.

But HPC and UHPC are also very brittle, so it is necessary to use a large amount of aggregate in the matrix and, if possible, reinforcement with fibres and rebar. This composite or hybrid material is known under the acronym 'compact reinforced composite' and the CRC principle makes it possible to very accurately predict the behaviour of buildings of all sizes under different loadings, especially when scaling up from small models. Heavily-reinforced ultra high performance concrete seems to have extremely good fatigue resistance even under continuous high loads.

One of the first large applications of reinforced high performance concrete overlay was as a white topping on damaged pavements and industrial floors and in cargo ships. The properties of the overlay make it possible to place the overlay as an 'independent' topping or wearing course on a cracked and/ or polluted sub-base or even on an under-dimensioned sub-base made from different materials like asphalt concrete, concrete, wood, ceramics or steel. One or more layers of welded mesh reinforcement are included, and the concrete mixture contains both steel fibres and acrylic fibres based on a special composite of pre-blended materials.

The standard setting time of the mixture is more or less equal of that of a traditional concrete mixture and likewise depends on temperature and relative humidity, although accelerators can shorten the setting time. After curing for approximately 24 hours at 20°C, the overlay is ready for use. Due to the large amount of welded mesh reinforcement and steel fibres, the hardened overlay is able to withstand a certain amount of restrained deformations from the base without the occurrence of surface cracks.

RHPC overlay is a combination of an HPC strength class C110 (based on special pre-blended materials and reinforced with both steel fibres and acrylic fibres) and welded mesh reinforcement. The mesh reinforcement is placed on an 8mm diameter rebar used as a spacer. Thus, the total amount of reinforcement is approximately 24kg/m2 of traditional reinforcement and 5kg/m2 of steel fibres. The total thickness of the RHPC overlay is in this specific case 50mm and the concrete cover on the reinforcement is thus only 18mm. If the thickness of the layer were to be increased, the reinforcement could be adjusted if necessary. To replace the existing wearing course with a RHPC overlay, the bonding between the steel deck plate (thickness 10 to 12mm) and the overlay is of crucial importance to secure total deck rigidity and a uniform monolithic behaviour under all circumstances. For that reason, initial research focused on creating a bonding zone that met all requirements - one that could easily be created by connecting the mesh reinforcement and the steel deck plate by welds, but this might result in undesirable local peak stresses. The best bonding method turned out to be the use of a two-component epoxy-based adhesive with broadcast bauxite aggregate. After hardening of the epoxy, the overlay is cast and the surface shot blasted. No additional wearing course is applied.

As well as research on relatively small samples it was also necessary to perform tests on full-scale structural elements under different loading conditions. Several associated projects at different institutes such as the civil engineering division of Contec, Delft University of Technology and TNO Building & Construction Research, were carried out to investigate and document the material's properties and behaviour, and this research is still going on. The behaviour of an orthotropic bridge deck with bonded RHPC overlay is completely different from an orthotropic bridge deck with a traditional surfacing, due to the much higher stiffness, so more investigation is still required, including detailed finite element calculations.

Tests proved that the intended application of RHPC overlay was a very promising solution for rehabilitation of orthotropic steel bridge decks - both durability and strength were found to be adequate. In 2003 a pilot project was carried out on the Caland Bridge to test the logistics of the process. The pilot project was on two traffic lanes with a width of 6.7m and a length of 80m. The whole project had to be carried out in just six days including rerouting of the traffic, removal of the asphalt wearing course, inspection and repair of the deck plate and the application, hardening, curing and shot blasting of the RHPC overlay. After the removal of the wearing course, a tent was placed to protect the whole area from the weather. 'Time of flight diffraction' inspections were carried out to investigate if there were critical cracks in the deck plate that had to be repaired. This is a very reliable but time-consuming technique and must be done on a bare steel deck. When the inspection was finished at the edges, prefabricated steel L-profiles with dowels welded on were placed in a two-component epoxy paste adhesive with fillers. These L-profiles were necessary to avoid 'curling' of the RHPC due to traffic loads, shrinkage, temperature loading or local debonding. Once all the inspection was finished, the whole deck plate had to be shot-blasted again to remove the corrosion film. The two-component epoxy paste adhesive was placed on the deck plate and calcinated bauxite 3-6mm was sprinkled on. Due to the very low temperature and the fact that no fillers were used in the two-component epoxy paste adhesive there was an uneven distribution of the layer thickness of the epoxy paste adhesive and thus also of the calcinated bauxite. Despite detailed instructions being provided, the welded mesh reinforcement was not placed well and instead of welded connections between the rebars, steel wire was used. This resulted in the reinforcement bending upwards in the middle of the two traffic lanes. Because of the shortage of time, the engineers decided to leave it and to compensate for it by applying an additional 20-30mm of concrete in the middle, making the cover approximately 40-50mm at some locations instead of the maximum allowed 20mm. An accelerator was added to speed up the reaction of the HPC meaning both the finishing and the shot blasting procedure could be carried out earlier than planned.

A compressive strength of minimum 30 MPa was reached within 24 hours of casting, the burlap sheets were removed and the surface was shot blasted to obtain the required skid resistance of at least 64SRT. Immediately after the shot blasting procedure details such as the joints with the adjacent lanes were finished, special trucks were used to remove the tent and the barriers were placed and connected. A special perforated water hose system was placed under the barrier to cure the HPC overlay for seven days without disturbing the traffic. The entire process took less than 120 hours.

Strain measurements on the resurfaced Caland Bridge show a stress reduction by a factor of four or five in the fatigue critical structural details; this equals the reduction factor measured on the small test samples. Further investigations on part of an orthotropic steel deck plate with fatigue cracks and a RHPC overlay also showed a significant stress reduction in the trough wall, suggesting it may be possible to leave some fatigue cracks unrepaired when placing the overlay, saving time and money. The pilot project demonstrated that it was possible to place the RHPC overlay on an orthotropic steel bridge deck even when traffic, including heavily-loaded freight trains, is allowed to use part of the bridge deck.

The problems with the reinforcement resulted at several places in a thicker and unreinforced cover (40 - 50mm) on the reinforcement near the dividing line of the two traffic lanes of the bridge deck. Very fine transversal cracks up to 0.1mm at the surface, and concentrated in the thicker applied middle part, are visible 30-100cm apart, maximum crack dept is the thickness of the concrete cover. There is no shrinkage visible along the steel profiles at the edges in the longitudinal direction.

Further investigations will be carried out, but one possible reason is a combination of the following: thicker concrete cover, less reinforcement in the longitudinal direction, partially restrained shrinkage in the longitudinal direction (by deck plate and troughs), insufficient curing during the first seven days, movements and stresses due to traffic in the orthotropic steel deck plate and the completely different behaviour of the resurfaced deck compared to both a traditionally-surfaced steel deck or the RHPC overlay. Small cracks will not have influence on the durability of the RHPC overlay but crack distance, crack width and crack depth must be minimised. Stresses during the setting and hardening of the RHPC overlay must be reduced as much as possible by using adequate curing starting directly after the compaction of the overlay. The use of special internal working curing compounds mixed with the pre-blended materials will be further investigated as will the possibilities for the placing the overlay using traditional slipform pavers for time savings.

This summer, the Dutch ministry started rehabilitation of two major bridges, with the first phase of the new decks being installed this year and the second phase due to be installed next year.

The Moerdijk Bridge on the A16 near Breda between Rotterdam and Antwerp is believed to have the most intense traffic spectrum in Western Europe with more than 100,000 vehicles crossing per day, and some 2.5 million heavy goods vehicles per lane each year.

A total area of 16,000m2 is being rehabilitated from May - November this year, and the same again next summer. Engineering and supervision is being carried out by the civil engineering division of the Dutch ministry and the main contractor is TBI Beton-en Waterbouw Haverkort.

Subcontractors Bruil-Ede and Cobeton are responsible for placing the heavily-reinforced high performance concrete, which is being supplied by Contec.

The bridge is being rehabilitated because it is suffering from cracks which have been caused by fatigue in the deck plate and the weld between the deck plate and the through wall. The last rehabilitation was completed as recently as 2001 when several thousands of metres of cracks were repaired by welding and local replacement of the deck plate.

But the deck plate on this bridge is uneven, resulting in a thickness of overlay between 47-100mm, this makes it necessary to place the reinforcement on supports and some extra reinforcement is necessary in the locations where the deck is thickest. Traffic is still using the other lanes while the rehabilitation takes place.

Time restrictions are the issue for the other contract - the Hagenstein bridges on the A27 near Vianen, on the motorway between Breda and Almere. A total area of 4,200m2 had to be rehabilitated this year during just 14 days in July. The second phase, a similar total area, will be rehabilitated next July. Client, engineering and supervision are the same on this project as on the Hagenstein Bridges, but the contractor was Strukton Betonbouw. Again the high performance concrete was supplied by Contec.

The use of RHPC as bridge deck surfacing is a good alternative for a conventional deck structure.

Peter Buitelaar is technical manager of Contec, René Braam is research engineer at Delft University of Technology Concrete Structures Group and Niek Kaptijn is senior research engineer in the civil engineering division of the Ministry of Transport, Public Works & Water Management

Keywords: refurbishment, strengthening, bridge decks