From the Rolls-Royce experimental archive: a quarter of a million communications from Rolls-Royce, 1906 to 1960's. Documents from the Sir Henry Royce Memorial Foundation (SHRMF).
Page from 'The Autocar' magazine discussing the performance and design of different types of ball bearings.
| Identifier | ExFiles\Box 20\7\ Scan016 | |
| Date | 6th April 1912 | |
| THE AUTOCAR, April 6th, 1912. 635 away the size of the journal bearing for a given diameter of stub end. The form of thrust bearing employed is open to two objections, urged by the author, against the older cup and cone bearing, one that since it is adjustable it can be wrongly adjusted and ruined, and the other, the degree of slip that takes place. Figs. 3, 4, 5, and 6. I enclose a series of sketches (figs. 3 to 6) based on the ordinary journal bearing 72 mm. outside, and 20 mm. inside, showing the effect of the axes of the ball remaining parallel to the axes of the journal, and at 30°, 60°, and 90° to it, the last being the form of thrust bearing which the author advocates. In each case I have assumed that the radius of the race is about 10 per cent. greater than the radius of the ball, and that in consequence the virtual contact between the ball and the race is about 1/8th of the diameter of the ball. I have had calculated out the slip that takes place at each of the four points b, b₁, c, and c₁, (see table and figs. 7 and 8). I have assumed that the axis of the ball does not change; this is probably not correct, because the ball, I think, in all bearings, constantly changes its axis to a slight degree. This constant change is beneficial, as it distributes the wear, and it does not affect the point that I wish to call attention to, namely, that in the thrust bearing the slip is far greater than in the cone bearing which the author condemns on this ground. He also shows another form of thrust bearing (fig. 32), of which he speaks highly, but this also is at all times subject to the same disqualification as the cone bearing, namely, the slip of the balls on their races. It will be observed that the total slip in the journal bearing is, as would be expected, less than under other conditions, but the very fact that this slip is always present, even in journal bearings, brings out strongly why it is necessary that a ball bearing should have good lubrication, which can only be secured by forcing lubricant in at one end, and preventing dirt and water getting in at the other end. There is one further point. The arrangement recommended by the author relies for the security of the wheel on the axie on the dust excluding ring which is screwed into the hub. This ring is locked by three screws, the shanks of which are embedded in the ring, while parts of the heads penetrate into the hub itself. I do not think this plan is sufficiently positive and fool proof; besides which it does not allow for any very fine adjustment. If fresh holes have to be drilled they are liable to be drilled incorrectly, and the engagement of the hub with the head is insufficient for security. I have known of a case where an error of this kind led to an accident. It was not a case of the screw coming out, but of the ring actually turning round in the hub and carrying the screw with it. The following list gives other makes of cars of which the hubs correspond more closely to fig. 2 than to fig. 1; some, however, are rather different, but in all cases there is no provision other than journal bearings to deal with thrust: La{L. A. Archer} Buire, Chalmers, Detroit, Charron, Clement, Bayard, De Dietrich, Delaunay Belleville, Gobron Brillié, Gr{George Ratcliffe}égoire, Hotchkiss, Leon Bollée, Martini, Métallurgique, Minerva, Mors, Opel, Panhard, Peugeot, Schneider, Unic, Armstrong Whitworth, Bergmann Electric, Dayton, Maudslay, Russell, Clément Talbot, Bianchi, Chenard Walcker, Daimler (English), De Dion Bouton, Fiat, Itala, Lancia, and Rolls-Royce. SLIP OF BALL BEARINGS. Type, Angle (°) of Axis of Ball, Slip per Revolution of Ball at b, c, b₁, c₁, Total Slip. Journal, 0°, 0.00676, 0.00676, 0.01285, 0.01285, 0.03922 Cone, 15°, 0.01774, -0.00410, 0.03292, -0.00765, 0.06241 Cone, 30°, 0.02862, -0.01463, 0.04962, -0.02548, 0.11835 Cone, 45°, 0.03824, -0.02456, 0.06107, -0.03839, 0.16326 Cone, 60°, 0.04831, -0.03431, 0.06409, -0.04409, 0.19490 Cone, 75°, 0.05726, -0.04086, 0.06726, -0.04802, 0.21340 Thrust, 90°, 0.06373, -0.04602, 0.06373, -0.04602, 0.21950 The Author's Reply. Mr. G.{Mr Griffiths - Chief Accountant / Mr Gnapp} F.{Mr Friese} Barrett replies to Mr. Pugh's contribution as follows: I have read Mr. J.{Mr Johnson W.M.} V.{VIENNA} Pugh's letter with interest, and note his criticisms on my paper. The old and simple form of employing journal bearings in front wheels, as illustrated in his fig. 1, has, I am aware, been used in the past by a very large number of firms, as indicated in his list; and, as far as my knowledge goes, some of these firms still continue to use this arrangement on all their cars, while others are discontinuing its use for the larger cars, and are only using it for the smaller ones. I also know that some of the firms given in this list have had very considerable trouble owing to this method of mounting due to permanent side thrust being put upon the two rows of balls. Considerable difficulty is still being experienced with this type owing to dirt and moisture getting into the bearings. Referring now to Mr. Pugh's remarks on journal bearings subjected to side thrust: it is, of course, an advantage to spread the two races as far apart as possible, and therefore resolve the horizontal forces acting on the tyre into nearly vertical components, but this is not found in practice to relieve the journal bearings of side thrust sufficiently to prevent undue wear, except perhaps in the lighter form of touring cars. The journal bearing in the line of high speed drills referred to, which has to withstand the pure end thrust and no journal load from the drill, does so by virtue of the fact that the end thrust is evenly distributed over all the balls, and therefore is very much less than when subjected to a combined journal and thrust load; in the latter case the journal load is carried on two or three balls, and these are held by this load in the bottom of the tracks. A side thrust coming on to this bearing at the same time would be carried by these same two or three balls which are not under journal load, and which are therefore slack in their races, will not be carrying any of the side thrust. In regard to the clamping of the revolving elements, it is true that the 'creep' of bearings is a pure rolling motion, but the action of the outer ball race rolling round in the hub shell produces a knurling action, which is found in practice sufficient to cause the seat in the housing to wear larger in diameter. With regard to protection of the bearings from dirt and moisture, this is nothing like sufficient, and endless trouble | ||