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).
Tolerances and interference fits of the Phantom 111 gearbox first motion shaft bearings.
| Identifier | ExFiles\Box 94\2\ scan0133 | |
| Date | 23th December 1936 | |
| Rm.{William Robotham - Chief Engineer} RHC.{R. H. Coverley - Production Engineer} C.BY/TWD.{T. W. D???Aeith} 328 BY.3/G.23.12.36. PHANTOM 111. GEARBOX - RE: FIRST MOTION SHAFT BEARINGS. Referring to Rm{William Robotham - Chief Engineer}/Gry.{Shadwell Grylls}2/AP.21.12.36. the tolerances are the best compromise we know how to make, and cannot be modified without running into other trouble. It is necessary, of course, to have interference fits between the shaft and the bore of the inner race, without such fits the ball bearing rotates and wears the shaft. Again, it is necessary when the bearing is cold to have interference fits between the outer race and its housing, particularly when the ultimate reference for the housing is aluminium. One other point, which has been completely ignored in your memo, is that the outer interference fits completely disappear when the temperature of the oil in the box rises by 25°C, and it is in this condition that the boxes are tested. The tolerances as they stand are the result of discussions held between the Works, the D.O. and the ball bearing makers, and with the latter we have insisted that they cut their tolerances in half, although for this privilege we have to pay an increased price for the bearings. Obviously the tolerances on ball bearings in the gearboxes are a matter of compromise from beginning to end, and the idea that by modifying the tolerances we can obtain satisfactory results is entirely wrong. In the case of the front bearing, it is plain that by the handicaps imposed by the Design Department in fitting the bearing to a sleeve, which itself is fitted on a shaft and then again housing it in a steel housing to fit in the aluminium box that conditions have been imposed which are impossible to meet without some possibility of criticism, but here again I am satisfied that the tolerances provided represent the best compromise we know of. In the case of the front bearing, whilst a figure of 1.35 thousandths interference is quoted in your memo, the actual interference when the oil is warm is .65 thousandths, which gives an interference between the balls and their bearing surfaces of one and a half tenths of a thousandth, an amount of interference which certainly is not responsible for the noise complained of. In the rear bearing, when the interference in the external housing is deleted there is a working clearance of half a tenth of a thousandth. | ||