Some necessary differences in treatment, both in cutting and setting the stone, between the upstream and downstream faces of the dam deserve mention. The upstream face of a masonry dam is usually plumb for the greater part of its height. Where it is battered below a certain elevation the latter is usually about one in twenty, or too slight to affect the methods of cutting and setting from those applicable to a plumb wall. Thus for a course of 30-in. rise with the beds kept horizontal, the top pitched line is 1 1/2 in. back from that at the bottom. This distance is not enough to have any practical effect upon the cutting or the size of block required to furnish a given stone, and to have no effect whatever upon the setting. However, when we turn to the downstream face, with a slope varying between say 45 deg. and practically or quite vertical, we are faced by a situation that calls for a choice between several difficulties. At once the question arises whether the bed joints shall be horizontal or normal to the face of the dam. Let us see what each scheme involves.

Showing different shapes and methods of setting downstream face stone or blocks.

Fig. 20. Showing different shapes and methods of setting downstream face stone or blocks.

If the beds are to be kept horizontal it does not mean that the stones can be cut as in Fig. 20, A, because such an edge is not allowable in stone work; they must be cut as shown in Fig. 20, B, with the bed joints normal to the face for say 4 in. before turning to the horizontal. For a stone of say 12-in. horizontal top bed and 24-in. rise on face of 45-deg. slope, the original block of stone must be of the dimensions shown on two sides by the dotted line, or say about 1 ft. 8 in. X 2 ft. 8 in. In this case about 30 per cent, of the volume of the stone must be wasted, and the expense of cutting is very much increased. For a header of 3 ft. horizontal top bed 23 per cent, of the original block must be wasted. The template for cutting the stones must of course be changed as the work progresses in order to follow the slope of the face and the varying angle between it and the horizontal bed.

If the entire bed is kept normal to the face as was done on the New Croton dam, as illustrated in Fig. 20, C, the result is a regular shaped stone with minimum waste and labor of cutting, but troubles are immediately encountered when it comes to setting the stone. As the backing must be left low enough and far enough back so that the stone may certainly be set without having to cut away anything from either the stone or the backing, one must leave room in excess of the actual requirement of the face stone. The Lewis hole may of course be such that the stone hangs as it is to be set, but as it is being set the rubble backing must be built to such an extent and in such manner as to hold the face stone securely in place against its tendency to slide on its bed.

This process, however conducted, is slow and expensive, and can result only in considerable delay to the work on the adjacent backing for it cannot proceed until the course under construction has hardened sufficiently to hold the face stone up to line. Fig. 20, C, shows how the surface of the masonry must be at certain stages. Thus when starting to set course A the surface of the masonry must exist as from C at the face, to E, thence to F and G; when starting to set a header course B the surface of the masonry must be D-F, G-H.

This necessary procedure in setting such face work merges, without very much awkwardness, with the procedure of building a backing of rubble masonry in mortar, as was the case on the New Croton dam. Both processes are slow and the necessary steps and recesses in the rubble backing are perfectly feasible. When, however, we come to contemplate such face work in connection with a backing of cyclopean concrete, and analyze the various steps in the process, we see that expensive expedients and much delay inevitably result. Soft concrete would not hold a face stone from sliding; even after hardening in the horizontal strata natural to its deposition, some rubble backing would have to be built in connection with setting the face stone. Such rubble backing would not be as effective when laid on the horizontal concrete as it would if laid between the face stone and line E-F, Fig. 20, C. The only feasible method would be first to build the concrete by means of forms to the lines E-F, G-H, and then set the face stone as in the case of rubble backing. This procedure would result in a much less effective bond between face work and backing, and if this effect were to be minimized by keeping the two along at the same elevation, serious delay would result to the progress on cyclopean concrete.

A compromise between the two foregoing methods of radial and horizontal joints, which avoided many of their difficulties, was employed on the downstream face work of the Wachusett dam, as illustrated in Fig. 20, D. Necessarily some stone was thrown away in cutting the headers, but there was only a small waste in the stretchers, which were light stone cut with their natural beds parallel to the face of the dam. Although some rubble backing had to be built in connection with setting the stone, it was only enough to hold the stretchers from rotating, as there was no tendency to slide. No awkward steps or recesses were necessary; and the backing could at all times be kept level at about 3 in. below the face work. The change in pattern for varying face slope applies only to the headers, the stretchers being practically uniform throughout.

The downstream face of the Roosevelt dam (except the pilasters near the top) was formed entirely of selected stone as they came from the quarry, set with horizontal beds. (See Plate VIII, Fig. B.) No work was done upon them except a little with a sledge hammer to square them roughly, or occasionally (after setting) to trim roughly the outer top edge to line. The steps protruded beyond the face, i.e., the re-entrant angles were kept at the neat line. Naturally the joints were thicker, probably up to 5 in., and lacked the regularity of the other face. The general effect, however, is entirely satisfactory.

Concrete blocks are made very nearly as shown in Fig. 20, A except that they have a smooth face. Of course, the considerations of labor and wasted material to produce such shapes in concrete do not apply as in the case of cut stone. The stability while setting does not require any attendant backing up, hence the backing when built may be concrete as well as rubble.