Fig. 2612.
Fig. 2613.
If, for example, a sleeve be split at four equidistant parts of its circumference, and from each end nearly to the middle of its length, as in [Fig. 2611], any pressure that may be applied to its circumference to cause it to grip the shaft it envelops will cause it to grip the shaft with greater force at one part than at another, according to the diameter of the shaft and the location of the external pressure. Thus, if the pressure be applied equally along the length a b, the weaker end b will close most readily, while at a the support afforded by the unsplit section offers a resistance to closure at the ends a of the split, hence the shaft, even though a working fit to the sleeve bore, will be gripped with least force at the end a. If the shaft were simply a close fit, as, say, just movable by hand on the sleeve bore, the form of the coupling bore would, when compressed upon the shaft, be as shown in [Fig. 2612], the bend on the necks a, b, c, d, being magnified for clearness of illustration. If the compressing piece covered the compressed sleeve for a lesser distance, the grip would be more uniform, because there would be a greater length of the sleeve to afford the curves a, b, c, d, as shown in [Fig. 2613]. The grip may be more equalised by boring the sleeve of slightly smaller diameter than the shaft.
Fig. 2614.
[Fig. 2614] represents a sleeve carrying out this principle. It is composed of two halves, as shown, bored slightly smaller than the shaft diameter, and is to be compressed on the shaft, which, acting as a wedge, would spring open the sides of the bore until the crown of the bore bedded against the shaft. This, in the case of parallel shaft ends of equal diameter, would hold them with great force axially true, and with equal force and bearing, thus meeting all the requirements. If, however, the end of one shaft were of larger diameter than the end of the other (as has hitherto been supposed to be the case), the end accomplished by boring the sleeve of smaller diameter than the shaft is, that the end of the sleeve is afforded the extra elasticity due to the transverse spring of the sleeve, which permits the edges of each half of the sleeve to bear along a greater length of the shaft end than would otherwise be the case; but the bearing is in this case mainly at and near the edges of the split.
It will be perceived, then, that under this principle of construction, when applied to shaft ends of varying diameters, the metal is left to spring and conform itself to the shape of the parts to be connected, and that there is nothing outside of the condition of relative diameter of shaft to sleeve bore to determine what the direction of the spring or closure of sleeve shall be; but, on the other hand, the principle possesses excellence in that the sleeve being cylindrical and its closure taking place equally at similar points of contact the shafts will be held axially true, one with the other; or in other words, the movements of the metal while sleeve closure is progressing are equally radial to the axis of the sleeve, and there is no element tending to throw the shaft axis out of line one with the other.