Taper keys are tapered on their surfaces a and b in [Fig. 466], and are usually given 1⁄8-inch taper per foot of length. There is a tendency either in a key or a set screw to force the hub out of true in the direction of the arrow. It therefore causes the hub bore to grip the shaft, and this gives a driving duty more efficient than the friction of the key itself. But the sides also of the key being a sliding fit they perform driving duty in the same manner as a feather which fits on the sides a, d in [Fig. 467], but are clear either top or bottom. In the figure the feather is supposed to be fast in the hub and therefore free at c, but were it fast in the shaft it would be free on the top face.
Fig. 468.
[Fig. 468] represents a shaft held by a single set screw, the strain being in the direction of the arrow, hence the driving duty is performed by the end of the set screw and the opposite half circumference of the bore and shaft. On account, however, of the small area of surface of the set screw point the metal of the shaft is apt, under heavy duty and when the direction of shaft rotation is periodically reversed, to compress (as will also the set screw point unless it is of steel and hardened), permitting the grip to become partly released no matter how tightly the set screw be screwed home. On this account a taper key will under a given amount of strain upon the hub perform more driving duty, because the increased area of contact prevents compression. Furthermore, the taper key will not become loose even though it suffer an equal amount of compression. Suppose, for example, that a key be driven lightly to a fair seating, then all the rest of the distance to which the key is driven home causes the hub to stretch as it were, and even though the metal of the key were to compress, the elasticity thus induced would take up the compression, preventing the key from coming loose. It is obvious, then, that set screws are suitable for light duty only, and keys for either heavy or light duty. It is advanced by some authorities that keys are more apt to cause a wheel or pulley to run out of true than a set screw, but such is not the case, because, as shown in [Figs. 466] and [468], both of them tend to throw the wheel out of true in one direction; but a key may be made with proper fitting to cause a wheel to run true that would not run true if held by a set screw, as is explained in the directions for fitting keys given in [examples in vice work].
Fig. 469.
If two set screws be used they should both be in the same line (parallel to the shaft axis) or else at a right angle one to the other as in [Fig. 469], so that the shaft and bore may drive by frictional contact on the side opposite to the screws. Theoretically the contact of their surface will be at a point only, but on account of the elasticity of the metal the contact will spread around the bore in the arc of a circle, the length of the arc depending upon the closeness of fit between the pulley bore and the shaft. If the bore is a close fit to the shaft it is by reason of the elasticity of the metal relieved of contact pressure on the side on which the set screw or key is to an amount depending upon the closeness of the bore fit, but this will not in a bore or driving fit to the shaft be sufficient to set the wheel out of true.
If two set screws are placed diametrally opposite they will drive by the contact of their ends only, and not by reason of their inducing frictional contact between the bore and the shaft.