The purpose of these installations was to determine the durability of concrete containing aggregate larger than 1½ in. in size. First Installation In November 1962, 10 mass concrete cubes (8 cu ft) were installed at half-tide elevations. After 2042 cycles of freezing and thawing, all of these cubes had failed. The cubes were made of air-entrained concrete (4.8 to 5.3 percent air) using Type II portland cement, a 6-in. maximum size coarse aggregate, and a manufactured limestone sand as the fine aggregate. The test variables were kind of coarse aggregate and w/c; eight coarse aggregates were used and the water-cement ratio was either 0.5 or 0.8 (by weight). The coarse aggregates were three different limestones, a natural gravel, and a graywacke.Table 1-QA of TR 6-553 lists these specimens and gives their exposure record along with other pertinent information.

The principal conclusions drawn are the following:
a. The 0.5 w/c cubes are more durable than the companion 0.8 water-cement ratio cubes.
b. The most durable aggregate mixtures (all 0.5 w/c) were Limestone A, Limestone B, and Natural Gravel A. Second Installation In December 1963, six concretes cubes (8 cu ft) were installed at half-tide elevation. All of these cubes had failed after 2490 cycles of freezing and thawing (20 winters). The concrete in these cubes was made to the specifications of the concrete in the first installation except that three different coarse aggregates were used. These aggregates were a dolomite, another natural gravel, and a gneiss. Table 2-QA of TR 6-553 lists these specimens and gives their exposure record along with other pertinent information. The principal conclusions to date are as follows:
a. The 0.5 w/c cubes are more durable than are the 0.8 w/c cubes.
b. The most durable mixtures (0.5 w/c) are those containing Natural Gravel B and gneiss.