The part a of the hollow arbor x, Fig. 172, serves for the determination of the capacity of the lubricant to decrease the friction of two surfaces sliding upon each other.

To avoid as much as possible the influence exerted by the

differences in the condition of the sliding surfaces, their width is reduced to a minimum by the interior radius of the bearing b sliding upon a being somewhat larger than the radius of the exterior surface of the arbor itself. By these means the contact of both surfaces takes place only in a line parallel to the revolving axis of the arbor, and, therefore, under always equal external conditions.

The sliding bearing b, Fig. 175, is of steel sufficiently hardened to make a change in the condition of the surfaces on the

FIG. 175,

point of contact by cutting in or working in impossible. The bearing is fitted to a strap-like piece c, Fig. 175, the mass of which is so distributed that the lower half has a greater moment of oscillation than the upper.

The test of the lubricant in the sense of question 1, page 498, is carried on as follows: The outer surface of the arbor a and the inner surface of the bearing b are sufficiently moistened with the oil to be tested, and thus kept during the experiment. The mechanical process which begins with the revolution of the

pivot is illustrated by the schematic sketches Figs. 176, 177 and 178.

The vertical line passing through the centre of gravity S of the strap (Fig. 176) appears after the commencement of the revolution of the arbor as shown in Fig. 177, and after several oscillations forms an angle constant to the vertical line. This angle depends on the intensity of the friction created on the point of contact between arbor and bearing, and is always the same for the same material at the same temperature. The

FIG. 176.

FIG. 177.

FIG. 178.

૨૦

-X

greater the total friction is, i, e., the smaller the capacity of the respective lubricant to counteract the friction between the two lines of contact of the arbor and the bearing, the larger the inclination of the strap from its vertical position will be, because the component y, which represents the intensity of the friction on the periphery of the arbor, becomes greater. The inclination of the pendulum strap d indicates also the capacity of the lubricating material to decrease friction between the two points of contact, and is directly determined and read off in units of circumference on the periphery of the lower portion of the strap.

To test the lubricant in the sense of question 2, page 498, the apparatus is provided with the following arrangements:

As explained in the beginning, the judgment in question here must proceed from the heating produced by the internal work of the lubricant in the layer between the arbor and bearingy (Figs. 173 and 174).

The construction must, therefore, be such as to offer a suf.

ficient guaranty that the heating observed originates from this source alone, i. e., from the layer itself, as only under this condition can it be observed entirely independent of foreign influences. To attain this object and to leave the working arbor, so to say, in a pendent position, the latter was given the form of a revolving body whose plane, set parallel to the axis of revolution, is curved according to the arc of a circle, the centre of which, however, lies outside the axis itself; moreover, the bearing used joints accurately to a portion of the upper surface. Upon the bearing itself rests by means of two points the lever h, Fig. 173, which is fastened in the frame, thus fixing the position of the bearing, and as this embraces the arbor, that of the latter also. Vertically downward the arbor is only sustained by the two rolls e.

In this manner the bearing, the running surface of which serves for the experimental plane, assumes at the same time the guidance of the arbor, the rolls e serving only for its support, but without the production of any material friction, there being no sliding between the circles of contact and the surface of the arbor.

This arrangement fulfills the condition of an almost entire thermic insulation of the source of the development of heat, i. e., the layer between the arbor and bearing, and of the latter themselves.

To test the lubricant about 12 drops of it are distributed upon the surface of the arbor and that of the bearing. The oil is kept uniformly distributed by the small strip of leather # moved to and fro during the revolution of the arbor by the oblique collar w.

The revolution of the arbor is effected at x by the belt r. The arbor itself is hollow and hermetically closed on one side by a spring-plate f, and is filled with vapors of sulphuric ether, the expansion of which, with the quantity used, amounts on an average to about ten times that of atmospheric air.

Now, when the arbor becomes heated by revolving and the heat yielded by the layer of lubricant, the heat developed is

communicated to the vapors in the interior of the arbor, which causes an outward expansion of the spring-plate f. The movements of the latter with the required interpretation are communicated to the pencil t, whose different positions correspond, therefore, to the various temperatures of the arbor and of the layer of oil. In front of this pencil moves eccentrically a strip of paper p. Its motion is effected from the pivot ato (Fig. 172) by wheels and screws, so that a certain number of revolutions always correspond to a determined length of paper un

rolled. The proportions are so chosen that the strip of paper only progresses 3.75 millimetres per minute, while the arbor makes 250 revolutions per minute.

In order that the pencil may mark its positions upon the paper without resistance, the arrangement is made (Figs. 179 and 180) that the strip of paper is lightly pressed against the pencil only every fifteenth second, and the latter is at once released, so that it may change its position without exerting a restraining effect upon the spring-plate f. The mechanism of the pencil is besides so arranged that its point moves in a