Some state highway agencies have implemented high-performance concrete (HPC) for bridge decks for durability rather than strength. While the low permeability of HPC is used to protect reinforcing steel and prevent corrosion, any concrete cracks can diminish the intended protection. One unintended consequence of HPC can be early-age cracking. This paper explores two ways to prevent bridge deck cracking - internal curing, and paste reduction by using an aggregate blend with a larger maximum size of aggregate. Results also apply to other classes of concrete used in transportation and other infrastructure. Mixtures were tested with large and small coarse aggregate. Fine lightweight aggregate (LWA) was added to some of the mixtures to reduce cracking tendency. The ring test, modifi ed from ASTM C1581, was used to determine the time of cracking of a concrete specimen due to drying and autogenous shrinkage against the restraint of the steel ring. Tests were carried out until cracking or for a maximum of 90 days. The strongest effect on cracking was due to the replacement of a small maximum size coarse aggregate (No. 8) with an aggregate blend of No. 8 and No. 57. Increasing the coarse aggregate absorption level from low to medium had a less dramatic effect, as did the introduction of LWA for internal curing to the low absorption coarse aggregate.

The partial substitution of natural sand by lightweight sand has been used to reduce autogenous shrinkage in concretes with a low water/binder ratio, but when this substitution is combined with quasi-adiabatic curing conditions during the first 24 hours, it has been found that autogenous shrinkage can be mitigated and controlled. During an experiment done at Sherbrooke University on large concrete blocks measuring 0.6 × 0.6 × 0.6 m (2 × 2 × 2 ft) where 28% by volume of the natural sand in the concrete was replaced by the same volume of saturated lightweight sand, with absorption of about 20%, it was found that autogenous shrinkage was mitigated within the concrete blocks. Moreover, it has been found that the compressive strength and the elastic modulus of the substituted concrete were not affected by this substitution. For the first time in large concrete specimens, it can be reported that autogenous shrinkage can be mitigated and controlled without the help of any chemical product added to the concrete to induce an initial expansion to neutralize autogenous shrinkage. It seems that quasi-adiabatic conditions favor the development of large crystals that result in swelling of the apparent volume of the concrete block, and that the temperature increase also contributes to reduce chemical shrinkage. This could explain why Le Chatelier found more than 100 years ago that when a paste was cured under water, after a certain time, it swells enough to break the vase in which it had been placed.

High-performance concrete (i.e. water-cementitious ratio below 0.40) for bridge-deck applications has been shown to develop shrinkage-related cracking. This study explores the concept of internal curing using pre-soaked lightweight fine aggregate (LWFA) as partial replacement of sand for mitigating autogenous shrinkage and moisture warping. Concretes with water-cementitious ratios (w/c) of 0.35 and 0.45 containing LWFA to sand ratios of 20% and 40% by volume were investigated. Results show that pre-soaked LWFA is effective in mitigating autogenous shrinkage but also reduces slab uplift from moisture warping due to combined drying shrinkage at the top surface and wetting at the bottom surface.

While typically used to reduce early-age autogenous shrinkage and cracking, internal curing will also strongly influence the microstructure that is produced in cement-based materials. In this paper, the microstructure of a set of three different blended cement high-performance mortars produced with and without internal curing will be compared. For these mortars with a watercementitious material ratio of 0.3 by mass, internal curing has been provided by the addition of pre-wetted lightweight fine aggregates. Their microstructures have been examined after 120 days of sealed curing using scanning electron microscopy of polished surfaces in the back-scattered electron imaging mode. Clear distinctions between the microstructures produced with and without internal curing are noted, including differences in the unreacted cementitious content, the porosity, and the microstructure of the interfacial transition zones between sand grains (normal and lightweight) and the hydrated cement paste. These microstructural observations will be related to previously measured performance attributes such as autogenous deformation and compressive strength development.

This research explores the influence of internal curing on time dependent strains under no water exchange with the environment, i.e., basic creep and autogenous shrinkage. The behavior of high-performance concrete mixtures (HPC) containing pre-wetted lightweight aggregate (HPLC-1) for internal curing was compared to companion mixtures containing air-dried lightweight aggregate (HPLC-2). It was found that internally stored water reduced basic creep in direct relation to the amount of water held in the aggregate. Reductions in basic creep, up to 49%, were found between dry and prewetted lightweight aggregate mixtures. Further, internally stored water reduced autogenous shrinkage to the extent that actual expansion occurred proportional to the amount of internal water.