Service life of repaired concrete bridge decks

Pre-mature deterioration of reinforced concrete causesproblems world-wide and the search for reliable cost effectiveand environmental friendly repairs is intense. The service lifeof the repaired structure is important in order to establishthe annual cost and environmental impact for differentalternatives. This doctoral thesis focuses on the service lifeof concrete bridge decks repaired with bonded concreteoverlays. The approach was broad and based on the performanceof a number of bridge decks after approximately ten years inservice. The main aims were to establish the overall servicelife and to improve the design where improvements were needed.Efforts have been made to make the repairs as robust aspossible and the crucial production stage is always considered.The following parameters were investigated: Wear and rutting Bond strength and extent and cause of cracks Relative humidity and temperature in the deck Freeze-thaw resistance and number of freeze-thawcycles Concrete cover and service life criterion with respect tochloride initiated corrosion Service life predictions using moving boundarytheory Chloride concentrations in de-icing water and in thesurfaces of concrete overlays Absorption of water and chloride ions due to capillarysuction of de-icing water Chloride ingress in solid and cracked parts of bondedconcrete overlays The wear and rutting were found to be limited. The bondbetween new and old concrete was unaltered or increased afterten years in service. The cracking on the overlays wasgenerally limited. The main cracking causes were insufficientcuring and reflective cracking due to construction joints (coldcasting joints) in the old bridge deck. The cracks can beavoided or made non-hazardous with respect to corrosion andfreeze-thaw damages with simple measures. The chloride ingress was low in solid and limited in crackedparts of the overlay. Samples of concrete, snow, slush andwater were taken from the overlay during two years to quantifythe chloride concentrations and their variations. The chlorideconcentration interacted with the weather, the local climateand the salt-spreading procedures. The chloride content in theoverlays was affected to at least 10 mm from the surface. The relative humidity in the overlay was stable and low atapproximately 20 mm from the wear surface. The number of(salt)-freeze-thaw cycles was high. The thickness of theoverlays was found to be fairly normal distributed and this wasused in the service life predictions. The moving boundary method is suitable to predict theingress of a chloride threshold value. The method is easy touse with a probabilistic approach and simple and cheap toupdate and no parameters have to be guessed. Numerical methodswere also used and the influence of various surfaceconcentrations was investigated. The transport of chloride ionsinto cracked concrete was treated with a two dimensional modeland the influence of crack depth was found to be limited. Both conventional capillary absorption tests and cyclicwetting and drying test out-doors were conducted on matureconcrete cast and aged in-situ. The conclusion was thatcapillary suction of de-icing water could be the dominatingfactor in the chloride absorption process. Key words: Repairs, bridge decks, bonded concreteoverlays, service life, chloride ingress.