3rd. The tool gets duller as the cut proceeds, causing more strain from the cut, and, therefore, placing more strain on the follower rest; and,
4th. It would be necessary, on account of the length of the cut, to resharpen the tool before the cut was carried from end to end of the spindle, and it would be almost impracticable to set the reground tool to cut to the exact diameter.
The second, third, and fourth of the above reasons operate together in causing increased work spring as the tool approaches the middle of the work length; thus the deflection of the follower rest, the increased weakness of the work, and the comparative dullness of the tool would all operate to cause the work to gradually increase in diameter as the cut proceeded towards the work centre (of length).
Suppose, for example, a cut to have been carried from the dead centre, say, five feet along the work; at the end of this five feet the tool will be at its dullest, the shaft at its weakest, and supported the least from the dead centre and follower rest.
Suppose, then, that the reground tool be placed in the rest again and set to just meet the turned surface without cutting it, then when it meets the cut to carry it farther along the work the cut will produce (on account of the tool being sharper) less strain on the work, which will therefore spring or deflect less. Precisely what effect this may have upon the diameter to which the tool will turn the work will depend upon various conditions: thus, if the top face of the tool be sufficiently keen to cause the strain due to bending the shaving cut or chip to pull the work forward, the tool would turn to a smaller diameter. If the depth of the cut be sufficient to cause the work to endeavor to lift, and the tool edge be above the centre of the work, it would be cut to smaller diameter. If the tool cutting edge were below the centre, or if its top face be at an angle tending to force the work away from the tool point, the diameter of the work would be increased.
From these considerations it is obvious that the finishing cut should be started at the centre of the work length, and carried towards the lathe centres, because in this case the tool will be sharpest, and therefore will produce less tensional strain on the work at the point where the latter is the weakest, while the resisting strength of the work would increase as the cut proceeded, and the tool became dull from use. Furthermore, if it were necessary to regrind the tool, it would be reset nearer to the lathe centres, where the work would be more rigidly held; hence the tool could be more accurately set to the diameter of the finishing cut.
By following this plan, however, it becomes necessary to have the shaft as near true and parallel as possible before taking the finishing cut, for the following reasons:—
Let the diameter of the spindle before the finishing cut be 11⁄32 inches, leaving 1⁄32 inch to be taken off at the finishing cut, then the ring in the follower rest must be at starting that cut 11⁄32 inch bore, and if the rest is to follow the cut the bush must be changed (so soon as it meets the finishing cut) to one of an inch bore. But if the spindle be turned as true and parallel as possible before the finishing cut the rest may lead the tool, in which case the bush need not be changed. There are differences of opinion as to the desirability of either changing the bush or letting the tool follow the rest, but there can be no dispute that (from the considerations already given) a spindle turned as true and parallel as may be with the tool started from the dead centre and carried forward can be improved by carrying yet another cut from the middle towards the dead centre. In any event, however, work liable to spring or too long to be finished at one cut without removing the tool to grind it, can be more accurately finished by grinding in a lathe, such as was shown in [Figs. 676] or [679], than by steel-cutting tools, and for the following reasons:—
If it be attempted with steel tools to take a very fine cut, as, say, one of sufficient depth to reduce a diameter, say 1⁄500 inch, the tool is apt to turn an uneven surface. There appears, indeed, to be a necessity to have the cut produce sufficient strain to bring the bearing surfaces of the rest into close contact and to place a slight strain on the tool, because under very light cuts, such as named above, the tool will generally momentarily leave the cut or take a reduced cut, and subsequently an increased one.
It may be accepted that from these causes a finishing cut taken with a steel tool should not be less than that sufficient to reduce the diameter of the work 1⁄64 inch. Now an emery-wheel will take a cut whose fineness is simply limited by the wear of the wheel in the length of the cut. Some experiments made by Messrs. J. Morton Poole and Sons, of Wilmington, Delaware, upon this subject led to the conclusion that with corundum wheels of the best quality the cut could be made so fine that a 12-inch wheel used upon a piece of work (a calender roll) 16 inches in diameter and 6 feet long, would require about forty thousand traverses to reduce the diameter of the work an inch, leading to the conclusion that the wear of the wheel diameter was less than one eighty-thousandth part of an inch per traverse.