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).
Component failures related to roller bearings, discussing insufficient case hardening and recommending manufacturing and design improvements.
| Identifier | ExFiles\Box 148\5\ scan0245 | |
| Date | 11th December 1937 | |
| -2- Rm{William Robotham - Chief Engineer}/FJH{Fred J. Hardy - Chief Dev. Engineer}6/R.11.12.37. (5) 3.OP.114. Some rollers crushed. Others not Paris Trials chipped. Inner and outer races in new car. good condition. Rocker shaft twisted. Failure attributed to accident. (6) 3.BT.{Capt. J. S. Burt - Engineer}179. Parts not yet received W.G. Bethel. at Works. 5,135. (7) 3.AX.149. Milne Watson. 15,543. We consider that the evidence in case No.4, 3.AX.175, is the deciding factor in attributing Nos. 2 and 3 to insufficient case hardening on the inner races. Failure No.4 is unquestionably due to casehardening. Previously it has not been obvious whether the broken rollers damaged the inner races or whether the collapsing of the inner races broke the rollers. The drawing of the inner races specifies a case 0.025-0.035 deep. The races on 32-EX were satisfactory with 0.040 case, while 3.BU.94 failed with a 0.025 case. We, therefore, recommend that all existing inner races should be scrapped and 0.040 + 5 case-hardening introduced immediately, and also used for all replacements. We have found some races which have been set up eccentrically prior to grinding so that the case is 0.010 thinner on one side than on the other. Steps should be taken to ensure that this does not recur. The races on 32-EX with 0.040 case were obtained from production stock, while the races on 3.AX.175. have only 0.020 case. We consider that this variation is too wide. We do not think that the damage caused to packing washers during assembly of the roller follower is responsible for any of the failures we have yet investigated. The roller follower which will be introduced with Wraith wheel movement is of the ball type, and has, approximately, 100% greater load capacity than the existing design. The existing cams and rollers for the whole of E.{Mr Elliott - Chief Engineer} series are already in stock so that the cost of introducing the new roller at 100 in E.{Mr Elliott - Chief Engineer} would be £400, approx. There is, so far, no evidence that the existing design will fail if the case-hardening is adequate and the use normal. Rm{William Robotham - Chief Engineer}/FJH.{Fred J. Hardy - Chief Dev. Engineer} | ||