Fig. 884.
Another plan would be to hold the work by dogs applied on the outside, and turn the bore and both of the faces. To these fasten four plates on the chuck plate, and turn their ends to the size of the bore and place the work on them, as in [Fig. 883], in which a, b, c, d are the four plates, and are clamping plates. This plan is often employed, but it is not a desirable one in heavy work, because the weight of the work is quite apt to move the plates during its setting. A better plan than either of these is to first turn off one face and then turn the work around in the lathe and hold it as in [Fig. 884]. The bore may then be turned, and all that part of the face not covered by the plates. Four holding plates must then be applied with the bolts within the bore, and when screwed firmly down the outside plates may be removed, leaving the work free to have the remainder of its face and its circumference turned up. In this way the work may be turned more true than by either of the two previously described methods, because it has no opportunity to move or become out of true.
Fig. 885.
Cylindrical work to be chucked with its axis parallel to the face plate is chucked by wood workers as shown in [Fig. 885], in which b, b are two blocks screwed to the chuck c, and having Vs in to receive the work as shown; the work is held to the blocks b, by means of the straps s, s, which are held to b, b by screws. An example of a different class of chucking by bolts and clamps may be given in the engine crank. A common method of chucking such a crank is to level the surface of the crank in a planing machine, and to hold that surface to the chuck-plate by bolts and plates, while boring both the holes, merely reversing the crank end for end for the second chucking.
This method has several inherent defects, especially in the case of large cranks. First, it is a difficult matter to maintain large chuck plates quite true, and as a result by this method of chucking any want of truth in the surface of the chuck will be doubled in the want of parallelism in the bores of the crank.
Suppose, for example, that the chuck surface is either slightly hollow or rounding as tested with a straight-edge placed across its face, then the axial line of the hole bored in the crank will not be at a true right angle with the planed surface of the crank. When the crank is turned end for end on the chuck-plate and again bolted with its plain surface against the surface of the chuck, the second hole bored will again not stand at a true right angle to the planed surface, and furthermore the error in one hole will be in a directly opposite direction to that of the other hole, so that the error in the crank will be double the amount that it is on the chuck surface. To this it may be answered that if such an error is known to exist it may be corrected by placing a piece of paper of the requisite thickness at the necessary end of the crank for both chuckings. But this necessitates testing the chuck on each occasion of using it, and the selection of a sheet of paper of the exact proper thickness, which is labor thrown away so long as an equally easy and more true way of chucking can be found. Furthermore there is a second and more important element than want of truth in the chuck to be found, which is that of the alteration of form which occurs in the crank (as each part of its surface is cut away) as explained in the remarks with which the subject of chucking is prefaced.