I must now leave these remarks to persons having more means and leisure than myself, to pursue the subject; wishing only, that useful truth may result from them: and that this unbelief of mine “in several special kinds of friction,” may at least be found to have some reasonable ground to rest upon.
But I may be opposed in some of my statements by the fact, that friction rollers, with centres, have been used with little advantage; and often laid aside. This I acknowledge; and go a step further. Friction is by no means of so much consequence as it was once thought to be: and is not the source of the greatest defalcations that occur in the use of power. Yet, to get rid of it, in some cases, would be of considerable importance; and the subject deserves at least the attention of every intelligent mechanician.
Those who have used friction rollers, know that it is a thing of great difficulty, to place their axes exactly parallel to that which they are intended to support: and even, if rightly placed at first, that a small degree of abrasion, greater on one pivot than another, will soon destroy that parallelism; and thus introduce a growing friction, capable, at length, of rendering the whole completely useless: for although the original friction is lessened by being transferred to a slower-moving axis, yet the latter still resists in some degree, say 1⁄4 of the whole; (its pivots being 1⁄4 of its whole diameter) so that the cohesion, or something else, between the main shaft and the friction roller, (thus resisted) must be sufficient to drag round the latter, against about 1⁄4 of the original friction; which in a word it cannot do without some relative motion between those surfaces, the friction roller lagging behind the main shaft, until its own friction is overcome by another. And thus it is, that a friction roller of this kind, does not make so many revolutions on its pivots, as its diameter compared with that of the main shaft, would imply; for example, if the shaft were 4 inches in diameter and the friction roller 8 inches, the latter would not complete one revolution against two of the former. There would thus remain a difference spent in real friction, in addition to that on the axis of the friction roller. Besides this, we have the want of parallelism above mentioned; which occasions a rubbing, in the direction of the shafts, small indeed in quantity, but for that reason very powerful in bringing on a change of form, and thereby hastening the common destruction. Both these accidents, therefore, make friction rollers, in general, an unsatisfactory and perishable expedient: and it is to make them less so, if not entirely to cure these evils, that the two following articles are designed.
In [fig. 6] of [Plate 17], A B is an axis which it is desirable to divest of its friction. To do this, as nearly as may be, I connect with it two rings of hard metal C D, formed as truncated cones; and under the shaft, in the same vertical plane, I place two smaller shafts E F, carrying on their tops, other two cones, similar to the former. The summits of each pair of cones meet of course in the points a b of the main shaft; and, on the principle of bevel geer, every contiguous part of the touching cones moves with the same velocity: so that there is no sensible rubbing between them—for, 1st. the pivots c d, are hard and pointed, and run on the hardest steps that can be obtained; and, 2ndly. the tendency of the cones u toward each other, is repelled without friction by the cylinders e f, attached to them, and which lean right and left against each other, turning with the same velocity, without causing any friction, or any creeping, between the two pairs of cones e C, and f D. All the weight therefore, of the shaft A B, (which of course is kept in place in the other direction by proper side cheeks, &c.) rests on the points of the vertical shafts E F, accompanied by no sensible tendency of these points to quit the places assigned to them.