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).
Critiquing a servo design, proposing modifications for simplification and improved performance.
| Identifier | ExFiles\Box 73\1\ scan0141 | |
| Date | 1st April 1924 | |
| To E.{Mr Elliott - Chief Engineer} & ST.{Capt. P. R. Strong} from R.{Sir Henry Royce} c. to CJ. BJ. RG.{Mr Rowledge} DA.{Bernard Day - Chassis Design} HS.{Lord Ernest Hives - Chair} SECRET 49910 R1/M1.4.24. EAC.2 & 3. X.9940 SERVO - LEC. 1861 Very disappointed with this design. It is extremely involved: wanted someone to clean up. Too great a variety of parts, and costly parts. For instance big central spring and floating plate could be replaced by providing main plate with 3 pins thus - SEE ORIGINAL FOR SKETCH. Do not reverse bolts because waste of dia. Use thin nuts or clearance can be less as ferodo is shewn fully thick, and centre plate could be .2" or .15" and ferodo 3/16", clearance of faces .02" instead of .06". Ratchet nut has steps too high - suggest .020 as in last paragraph. Here is where pedal is losing stroke through servo clearance being too great - (see Hs.{Lord Ernest Hives - Chair}report). I am strongly in favour of flat plate servo because stress should be self contained and friction faces with such type must control end play. The 4 hardened ball thrust washers should be alike if possible, end of lever just sliding fit over, both providing support. Thrust thread may want to be 45º and might be better scintillaed or case hardened to ensure release, having coarse thread, and ample clearance. Nut might be bronze. There must be no doubt about a free release, but this scheme with double ball thrust is different to the agreed scheme and may have no risk of sticking with even 30º thread. The agreed scheme was said by Hs.{Lord Ernest Hives - Chair} to be slow because of flat plate clutch, but that could not be the reason. I believe the servo was very slow speed with fairly large angle of action. This one is 7 to 1 instead of 28 to 1. Mr. Rowledge and I sent a good cross section which we believed workable. I have not it by me, but DA.{Bernard Day - Chassis Design} sent something about the same; (N.Sch.1782) here are some more - SEE ORIGINAL FOR SKETCH. Since connecting rods do not need adjustment (tension only) they could be flat pieces of steel or milled thin in centre as you shew. There is no wear to worry about, the stresses cannot be much beyond those known, so that the enormous pins and areas provided seem to be quite unnecessary. contd:- | ||