All long masonry structures which are exposed to appreciable changes of temperature will contain temperature cracks. In a masonry dam they manifest themselves principally at or near the top where the section is comparatively thin; if they extend down the faces to where the dam has a thickness of 30 ft. or more it seems to have been demonstrated that they seldom extend entirely through.
Cracks may show on the surface down where the dam is much thicker but they seem to be analogous to cracks produced in mud from drying out in the sun, namely, the cause operates with diminishing force as the mass is penetrated and the resulting crack diminishes in thickness until it dies out entirely. Along the top of a dam of uniform cross-section the cracks may be expected to appear at fairly uniform intervals of 30 ft. to 40 ft. Even a slight change of section such as produced by a buttress or pilaster or an abrupt change in elevation of the foundation, will fix the location of the crack at that point, and probably tend to localize the effect, i.e., make that crack wider than adjacent ones. A crack may also be started as a settlement crack and continue as a temperature crack.
Temperature cracks in a dam need never be regarded as a serious matter. As above stated they exist chiefly in the upper and thinner part. In a dam without overflow that part of the crack below water level is thinnest and is small in extent. The water against the upstream face of the dam limits to a very large extent the range of temperatures acting on the masonry under water and consequently limits the thickness of the cracks. Exposed to the sun and atmosphere the temperature of the face masonry might range during the year 130 deg. Fahr. or 140 deg. Fahr., while water in the reservoir might range from freezing to say 70 deg. Fahr., or far less than half as much.
Temperature fluctuations diminish rapidly in extent as the mass is penetrated for the thicker the mass the nearer the temperature of the center comes to being constant at the annual mean; even so the effect of the water is most beneficial. An overflow dam is usually thinner at the top, and in that respect is more susceptible to cracks, but the effect of the water is exerted at the top and (if the water is flowing over) also over the top and down the other face. A temperature crack is in a vertical plane at right angles to the line of the dam. Hence even if water did penetrate some distance into the dam it could exert pressure only in such a direction that no harm could result. The worst that can be said of a temperature crack is that it allows the water to penetrate the upstream face masonry which it is desired to keep impervious; but at any such depths as are necessary for any considerable pressure temperature cracks are insignificant if indeed they exist at all. In a curved masonry dam the pressure of the water tends to keep the dam in compression and consequently to close the cracks. Indeed the temperature variations in length of the dam are largely taken up by deformation of the arch, the radius of the curve being longer in cold weather and shorter in hot.