Fig. 683.
Supposing now that we have a roll r ([Fig. 683]), placed in position and slowly revolved, and that the carriage c is fed along by feed screw e, then the line of motion of the emery wheels will be parallel to the axis of the roll, provided, of course, that the bearings b ([Figs. 681] and [687]) are set parallel to the V-guideways in the bed, and that these guideways are straight and parallel. But the line of travel of the emery wheels is not guided by the Vs except in so far as concerns their height from those Vs, because the swing-frame is quite free to swing either to the right or to the left, as the case may be. Its natural tendency is, from its weight, to swing into its lowest position, and this it will obviously do unless some pressure is put on it in a direction tending to swing it. Suppose, then, that instead of the roll running true, it runs eccentrically, or out of true, as it is termed, as shown in [Fig. 683], when the high side meets the left-hand wheel it will push against it, causing the carriage c to swing to the left and to slightly raise. The pressure thus induced between the emery wheel and the roll causes the roll surface to be ground, and the grinding will continue until the roll has permitted the swing-frame to swing back to its lowest and normal position. When the high side of the roll meets the right-hand emery wheel it will bear against it, causing the swing-frame to move to the right, and the pressure between the wheel and the roll will again cause the high side of the latter to be reduced by grinding. This action will continue so long as the roll runs out of true, but when it runs true both emery wheels will operate, grinding it to a diameter equal to the distance between the emery-wheel perimeters, which are, of course, adjusted by the slide rests a a. If the roll is out of true in the same direction and to the same amount throughout its length, the emery wheel will act on an equal area (for equal lengths of roll) throughout the roll length; but the roll may be out in one direction at one part and in another at some other part of the length; still the emery wheel will only act on the high side, no matter where that high side may be or how often it may change in location as the carriage and wheels traverse along the roll. Now, the roll does not run true until its circumference is equidistant at every point of its surface from the axis on which the roll revolves, and obviously when it does run true its circumference is parallel to the axis of revolution of the roll, because this axis is the line which determines whether the roll runs true or not, and therefore the swing-frame is actually guided by the axis of revolution of the roll, and will therefore move parallel to it.
It is obvious that if by any means the swinging of frame f is slightly resisted, as by a plate between it and c, with a spring to set up the plate against f, then the emery wheels will be capacitated to take a deeper cut than if the frame swing freely, this plan being adopted until such time as the roll is ground true, when both wheels will act continuously and simultaneously, and f may swing freely.
A screw may be used to set up the spring and plate when they are required to act.
Suppose now that the roll was not set exactly level with the V-guideways of the bed, there being a slight error in the adjustment of the roll journals in the bearings on b, and the emery-wheels would vary in height with relation to the height of the roll axis, and theoretically they would grind the roll of larger diameter at one end than at the other.
Fig. 684.
This, however, is a theoretical, rather than a practical point, as may be perceived from [Fig. 684], in which r is a part of a section of a roll, and w a part of a section of a wheel. Now, assuming that the V-ways were as much as even a sixteenth out of true, so far as height is concerned, all the influence of the variation in height is shown by the second line of emery-wheel perimeter, shown in the figure, the two arcs being drawn from centres, one of which is 1⁄16th inch higher than the other. It is plain, then, that with the ordinary errors found in such V-guideways, which will not be found to exceed 1⁄30th of an inch, no practical effect will be produced upon the roll. Again, if one V is not in line with the other, no practical effect is produced, because if the carriage c were inclined at an angle, though the plane of rotation of the emery-wheel would be varied, its face would yet be parallel to the roll axis. If the Vs were to vary in their widths apart (the angles of the Vs being 45° apart), all the effect it would have would be to raise or lower the carriage c to one-half the amount the Vs were in error. It will be thus perceived that correctness of the roll both for parallelism and cylindricity is obtained independent of absolute truth in the V-guides.