JoAnn Browning and David Darwin, University of Kansas and Kenneth F. Hurst, Kansas Department of
Transportation
R esearch over the past several decades has addressed the causes of cracking in cast-in-place reinforced concrete bridge decks in North America(1-3) including three detailed studies by the University of Kansas (KU).(4-6) Results of these studies have provided specific guidance on needed modifications in materials and construction techniques to reduce the amount of cracking in reinforced concrete bridge decks. This guidance has been put to use during the first phase of a pooled-fund study under the direction of the Kansas Department of Transportation (DOT) in conjunction with 14 other state DOTs and the Federal Highway Administration. New specifications have been developed for use in the construction of 20 low-cracking, high-performance concrete (LC-HPC) bridge decks (15 in Kansas and 5 in partner states), with an equal number of conventional control decks to evaluate the relative performance and cost.
Specifications
Material specifications were developed using crack survey results in conjunction with construction diaries and laboratory work at KU. It is well established that settlement cracks can be reduced with increased concrete cover, smaller bar sizes, and lower concrete slump. Shrinkage cracks can be reduced by decreasing the volume of water and cement and maintaining an air content above 6 percent. Concrete specified for LC-HPC bridge decks has a maximum cement content of 535 lb/yd3 (317 kg/m3), a maximum water-cement ratio of 0.42, an air content of 8.0 ± 1.0 percent, and a slump of 1.5 to 3 in. (38 to 75 mm). Cement is the only cementitious material permitted. The temperature of the concrete at point of placement must be between 55 and 70°F (13 and 21°C) to control the temperature differential between the concrete, as placed, and the supporting beams. Even on a hot day in June in Kansas, concrete was placed within these specification limits using ice in the concrete and casting at night. The lower temperature also slows the setting time and allows for easier finishing of the deck. EvapoRATE(7) software is available to evaluate and document evaporation rates expected at a site.
A key aspect in obtaining workable concretes with low cement contents is the use of optimized aggregate gradations (using a proven method such as the Shilstone Method,(8) or KU Mix(9)). Workability is enhanced using water-reducing and high-range water-reducing admixtures. A high-quality aggregate with a maximum absorption of 0.7 percent is specified. Bridges in Kansas have used granite from Arkansas and Oklahoma. The low absorption helps improve freeze-thaw resistance, but also helps maintain a constant slump through the pump. Concrete with a slump as low as 1.5 in. (38 mm) has been successfully pumped during this program.
To limit problems on the job, the construction specifications require that the concrete must be placed using buckets or conveyors, unless the contractor can demonstrate that low-slump concrete batched to satisfy the specifications can be pumped. Four out of the five LC-HPC bridge decks completed in Kansas have been placed using pumps, and the fifth was placed using a conveyor belt system only because the coarse aggregate had very elongated particles.
Plastic shrinkage cracking is minimized by controlling the rate of evaporation from the concrete surface. Windbreaks may be required on windy days. Fogging is required using devices mounted on the finishing equipment supplemented with handheld fogging equipment, from time of concrete strikeoff until the concrete is covered. Fogging water, however, cannot accumulate on the concrete surface or be used as a finishing aid. Finishing is accomplished using a single-drum roller screed or a double-drum roller screed with one roller removed (to minimize the cement paste
that is worked to the concrete surface), with a supplementary pan/burlap drag and bullfloating, as required. In addition to fogging, contractors are required to place the first of two layers of pre-soaked burlap on the newly finished concrete within 10 minutes of strike-off and finishing. The second layer must be placed within another 5 minutes. Once the concrete has set enough to support foot traffic, soaker hoses are placed under white polyethylene sheeting for 14 days of wet curing.
One of the most significant modifications to the construction specifications has been the requirement for a qualification slab. The slab is 33 ft (10 m) long with the same design cross section as the actual deck, including the reinforcement. It is cast using the qualified concrete mix 15-45 days prior to placing concrete in the bridge deck. The motivation for requiring the qualification slab is to prevent experimentation on the bridge deck and to identify any problem areas. Meetings are held with the contractor, materials supplier, and state DOT representatives before and after placement of the qualification slab and after the placement of the bridge deck. Problems identified during qualification slab placements have included meeting material specifications for slump or temperature, accumulation of fogging water on the deck surface and use of this water as a finishing aid, handling and placement of the wet burlap, and general timing of the construction process. Lessons learned during these placements and follow-up discussions with DOT representatives and construction personnel have significantly improved the enthusiasm and participation of all parties to produce the best quality low-cracking bridge deck.
Results
Of 14 decks let in Kansas to date, construction costs for all but the first two have been about the same as those of the control decks. Crack surveys have been completed on the first three LC-HPC bridge decks in Kansas and on four control decks. The figure shows crack density, expressed in linear meters of cracking per square meter of bridge deck (m/m2) versus the age of the deck at the time of the survey for three previous studies of monolithic bridge decks in Kansas (diamonds), new control decks (circles), and new LC-HPC decks (triangles). Symbols that are connected by lines indicate decks that have been surveyed multiple times. The amount of cracking on the LC-HPC decks is lower than that for any of the other decks and shows promise to continue the trend of low cracking for years to come.
To date, the study has been successful in identifying low-cracking portland cement concrete mixes. Several additional approaches, however, have been identified that have the potential to increase the benefits of the project, including using supplementary cementitious materials, new sources of aggregate, and new approaches to finishing. These approaches will continue to be evaluated during Phase II of this project.
References
- “Durability of Concrete Bridge Decks,” Final Report, A Cooperative Study of ten states, the U.S. Bureau of Public Roads, and the Portland Cement Association, 1970, 35 pp.
- Krauss, P. D. and Rogalla, E. A., “Transverse Cracking in Newly Constructed Bridge Decks,” Transportation Research Board, National Cooperative Highway Research Program Report 380, Washington, D. C., 1996, 126 pp.
- Eppers, L., French, C., and Hajjar, J. F., “Transverse Cracking in Bridge Decks: Field Study,” Minnesota Department of Transportation, 1998, 195 pp.
- Schmitt, T. R. and Darwin, D., “Effect of Material Properties on Cracking in Bridge Decks,” Journal of Bridge Engineering, ASCE, Vol. 4, No. 1, February 1999, pp. 8-13.
- Miller, G. G. and Darwin, D., “Performance and Constructability of Silica Fume Bridge Deck Overlays,” SM Report No. 57, University of Kansas Center for Research, Lawrence, Kansas, 2000, 423 pp.
- Lindquist, W. D., Darwin, D., and Browning, J., “Cracking and Chloride Content in Reinforced Concrete Bridge Decks,” SM Report No. 78, University of Kansas Center for Research, Lawrence, Kansas, 2005, 453 pp.
- EvapoRATE, available at www.ksdot.org/kart/.
- Shilstone, J. M., Sr., “Concrete Mixture
Optimization,” Concrete International, Vol. 12, No. 6, June 1990, pp. 33-39. - KU Mix, www.iri.ku.edu/projects/concrete/bridgecrack.htm, 2006.