International Concrete Abstracts Portal

Previous work demonstrated that carefully monitored length change measurements during the first freeze period of a concrete specimen containing the aggregate being evaluated show a "fingerprint" that can be successfully correlated with the durability factor that is obtained after many later cycles of freezing and thawing. Six different coarse aggregates were used in this study to further substantiate the conclusions of the previous work and to attempt to shorten the test evaluation procedure from 8 to 3 days. The slope of the cumulative length change versus temperature and the length change versus time curve of the first freeze cycle near the freezing point of water was used as the fingerprint. Although attempts to shorten the procedure by using a boiling water accelerated curing procedure were considered successful, it was recommended that other methods of accelerating early strengths be attempted. The tests indicated that the procedure was satisfactory for screening aggregates having a durability factor of less than 30 and greater than 50 percent.

The properties of concrete--stress-strain behavior and permeability, for example--differ significantly from those of hydrated cement paste having the same water-cement ratio. The explanation for the difference lies mainly in the existence of a transition zone between the hydrated cement paste and the coarse aggregate particles. Results from a previous study at the University of California show that improvements in the mechanical and permeability characteristics of concrete brought about by the addition of pozzolans can be attributed largely to their effect on the transition zone. Results of a study on the transition zone as affected by the addition of varying amounts of granulated blast furnace slag to portland cement are reported. It is found that with a cement containing 10 percent slag, the thickness of the transition zone, as determined by the thickness of the region in which calcium hydroxide crystals showed a preferred orientation parallel to the aggregate surface, was not significantly different from that of a portland cement paste of the same water-cement ratio and curing conditions. For the cement with 30 percent slag, however, no preferred orientation was observed at or near the paste-aggregate interface. Since 30 percent slag is not sufficient to combine with all of the calcium hydroxide produced by portland cement hydration, the lack of preferred orientation can be attributed largely to the nucleation of numerous randomly oriented calcium hydroxide crystals on the fine particles of slag, with the pozzolanic reaction having a secondary effect.

Laboratory and theoretical studies of the individual factors that influence the durability and long-term behavior of concrete have the virtues that they provide evidence of performance that can be analyzed and understood in physical and chemical terms and also that many experiments covering a wide range of variables can be undertaken at modest cost. Concrete, in reality, does not necessarily conform with the laboratory idealizations; the scale of operation is much greater, the temperatures and environmental histories are less finely controlled, and the composite actions between one part of a structure and another will add a further complication. There is thus the need to use the knowledge gained from the fundamental studies to help to explain and to predict the performance of real structures. The way in which this problem is being tackled at the Cement and Concrete Association is discussed. Predictive models based on basic understanding have been developed and site observations from structures in the U.K. are used to validate or to challenge these predictions, and so lead the way to improvements both in realistic understanding and prediction.

Concrete mixtures were designed to nominal 28 day compressive strengths of 6000, 8000, and 10,000 psi (41, 55, and 69 MPa) using mix designs typical of commercial production of high-strength concretes. To produce the higher strength concretes, additions of fly ash (Class C), water reducers, and high-range water reducers were utilized. Concretes were subjected to both moist and air cures. Durability test procedures included rapid freezing and thawing in water, and application of deicing agents. All moist-cured, non-air-entrained concretes performed poorly, exhibiting rapid deterioration, irrespective of strength level. Entrained air contents, measured in the fresh concrete, of 3 to 4 percent were found to be necessary to assure adequate durability when concretes were subjected to freezing and thawing in water. However, moist-cured, air-entrained, high-strength concretes, prepared at 8000 and 10,000 psi (55 and 69 MPa), while performing satisfactorily with respect to freezing and thawing in water, were less resistant to applications of deicing agents than were air-entrained concretes prepared at the lower strength level. This was true even with air contents between 7 and 8 percent in the fresh concrete. Air curing had generally beneficial effects on resistance to freezing and thawing and application of deicing agents to normal strength air-entrained concretes, but had little positive influence on durability of high-strength mixtures. Performance of non-air-entrained concretes during freezing and thawing in water was somewhat improved when given a period of air drying; however, all non-air-entrained concretes performed poorly when exposed to deicing agents.

A full-scale prospective durability experiment was established in the spring of 1970 at Covenham, Lincolnshire, England, when five different concrete mixes were used to construct portions of the wave wall of a 88 Ha inland reservoir. The reservoir is approximately 1.0 x 0.9 km in plan, and the maximum water depth is 14 m. The concave face unreinforced wave wall at the top of the embankment faces southwest into the prevailing wind and is subject to wave action in winter. Variables tested were a, increased sand proportion, b, air entrainment, c, increased cement content, and d, use of a lignosulfonate-based water-reducing admixture. The standard concrete mix used for the rest of the reservoir was used as a control. The alternate bay method of construction used for the wave wall insured adequate replication for both test and control concretes. To date, all mixes have performed well, although weathering differences began to show at 4 to 6 years when the alkalinity of the concrete surfaces had been reduced sufficiently by carbonation to allow growth of microorganisms, particularly lichens. Yellow lichen species were most prominent at first, but subsequently were overtaken by grey/green lichens. After 10 years of exposure, all the modified concrete mixes showed less weathering effects than the control mix with least improvement given by the air-entrained concrete a and the mix containing a higher sand percentage b. Increasing the cement content c gave a significant improvement, but the best performance has been obtained with the concrete d batched with a lignofulfonate-based water-reducing admixture.