Two causes operate to produce a settlement in masonry:
The contraction of volume due to the setting of the mortar or concrete.
The weight of the masonry results in actual compression. The first cause operates while the cement is setting, the second operates for years.
Fig. 21. Use of salt in mortar during freezing weather.
N. B. - A very fine and noticeable example of the settlement of masonry is seen in the Washington Monument, where the settlement is still going on. It is not necessary to know the original thickness of the mortar joints. One will notice around the base minute fragments which have been spalled off from the corners of the stones. The process may be analyzed as follows: Universally the joint on such stone work is thinner at and near the face than it is generally throughout the interior of the joint. The relatively thick mortar in the center permits a greater settlement than does the relatively thin joint at the face; consequently there is an unequal distribution of the load, more being carried at and near the edge, which tends to spall off the unsupported corners. It is probable that this spalling off is somewhat furthered by a slight creeping of the mortar toward the face. This process will continue until the corners of the stone have spalled off to what might be termed the angle of repose.
Except at or near the surface of the masonry the effect of the first cause is distorted or concealed by the accompanying second cause. In other words, while the mortar or concrete is setting and while it is still in a condition to be easily compressible, a slight superimposed weight will squeeze out the mortar in opposition to any horizontal contraction. The net effect is more settlement than that due to either cause alone.
The objection (see page 63) to overhanging large stone, and subsequently building up with spalls and mortar under the overhang, is that both contraction and settlement operate to pull the mortar away from the stone and to produce a joint less liable to become tight from subsequent operations. In other words, on account of the bonding of the masonry, the load over such a joint would rarely act on the masonry in the joint enough to produce any lateral expansion counteracting its tendency to pull away from the overhanging stone.
As to the effect of the settlement on the face work, it is of course only at the faces that the effect of the long-continued settlement may be observed and measured. However carefully the face courses may be set to line and grade, subsequent testing will show that on the plumb upstream face they are full and low, i.e., have been pushed slightly outward and settled. On the downstream face they have settled and probably drawn in somewhat according to the batter. It may be claimed that the possible effect of such settlement upon the pointing vitiates the argument in favor of thick face joints rather than thin and points to the thin joint as desirable, but an analysis of the entire case justifies the thick joint. Thus we may admit that the less settlement the better, but it must be remembered that this should apply to the entire structure as a homogenity. A 2-in. face joint is much nearer comparability with the thickness of the joints that it is feasible or economical to obtain in the interior work, and it conduces to a condition such that face work and interior work will settle together. Notwithstanding the intimate adhesion of the face work to the interior work and proper bonding by headers in the former, it is entirely conceivable that serious stresses between the two classes might exist in a plane below which were say fifty 1/2-in. joints in one and fifty 2-in. joints in the other. If the bond between the two classes of work is not effective there will be some separation; if it is effective the result will be that the face work is carrying more than its share of the load, somewhat as in the Washington Monument.
Without attempting to assign a figure to the total amount of settlement in any given case or to the rate of settlement at any time during the total period, it may be safely assumed that the rate is constantly diminishing. If the time element is plotted as abscissa and the settlement as ordinates, the result will be a curve which continually approaches some horizontal line as the limit of settlement. A tangent to this curve at any point indicates the rate of settlement at that time. In the earlier and steeper portion of the curve two rates, say six months apart, will be quite different. It follows then that if a column of masonry, say 50 ft. high, be built and allowed to stand for several months and if a second column be then built alongside of and in contact with the first, the subsequent settlement of the two would be at different rates. The result would be a broken and imperfect contact between the columns, with the settlement of any point in the second column, below that of the corresponding point in the first column. Whether the total settlement at all points in the column is in exact proportion to the height of the point above the base depends on the effect of age on the compressibility of the mortar, the effect of the added weight and the uniformity in rate of building. However, the answer is immaterial to the argument. Now let us see what appreciable effect varying rates of settlement may have in building a high masonry dam.
The exigencies of actual construction often require that progress be made more or less in vertical steps instead of continuous horizontal layers. Then any and all portions are alternately higher and lower than adjacent portions. The derricks at any given time are actually building on one-half of the area while they occupy the other half. A derrick will build up say 10 ft. or 12 ft. above the level upon which it sets, next it will move up onto the higher portion and fill in the depression it came from and then in turn it will carry that portion higher than the first. It is not to be doubted that this procedure results in unequal rates of settlement and more or less distortion in the vicinity of the contact of the various parcels. However, as the differences in height and age are not very large and as the distortion occurs when all the masonry may be called fresh and better able to accommodate itself to distortions, the effect is generally so small as to be entirely negligible or even inappreciable.