A 1922 40/50 "Silver Ghost"

In this article A. C. Molyneux critically examines the design and construction of one of the most famous and most typical of Rolls-Royce cars. We think this is appropriate, for last Christmas we published a description of a Rolls-Royce amongst bicycles, and we now enable those who wonder just how the “Silent Ghost” was laid to digest this great design and form their own opinions of this engineering masterpiece. – Ed.

I would recommend to everybody who is genuinely interested in real engineering the excellent book of Nockolds, “The Magic of a Name,” in which is described the hard life of a hard man, and everybody knows that this hard man is Royce. As a Manchester man it is good to know that ” the world’s best motor” was born and reared in Manchester. Having just completed the reading of this book, it was most appropriate that a client should telephone and ask us to collect a “Silver Ghost” from a breaking-up yard in the Wilmslow direction. This was indeed a brilliant discovery on his part, for the price of this thoroughbred was a mere £26 10s. The owner’s intention was to convert it into an ambulance, so the first operation was the removal of the 7-seater limousine body by Barker’s. This body was a beautiful piece of work, and was panelled throughout in 1/8-in. aluminium sheeting, and the panels, to prevent drumming, were lined with 3/8-in. 3-ply wood, and even the floor boards were made of cypress. At this point I think a remark made by an old engineer is well worth mentioning. After the gigantic body had been lumbered off and the complete chassis exposed this engineer, on viewing the transmission, gearbox, etc. said, “Even if I had not seen the radiator I could tell that this was the work of Royce, as I was familiar with his electrical cranes, and his name is written in every detail.”

The chief measurements of the frame are: track 4 ft. 9 in., and wheelbase 12 ft. 6 1/2 in. The depth of the sidemembers at the deepest point is 6 in., with 2 1/4 in. flanges, tapering to a depth of 2 3/4 in. at the rear, and 1 3/4 in. at the front, with a constant width of flange. At the rear there is a slight upswept portion to clear the rear axle. The cross membering of the frame is most thorough. Starting from the rear there is the usual light channel-section member situated just to the rear of the axle casing. The next member is a 2 in. tubular one situated precisely where it should be, namely, where the cantilever springs are anchored to the chassis. As may be expected, the next member appears in the precise position for it, where the shackled ends of the cantilever springs lie. Some 8 in. further forward there is a larger-section tubular member, 2 3/4 in. in diameter. This member also supports the two rear anchorages of the gearbox and ball-ended torque tube. From the outer ends of the first-mentioned tubular cross-member two members, tubular in section, extend forward towards the member supporting the gearbox and are substantially clamped close up to the torque-tube socket. Carrying, on forward again the next member lies just ahead of the gearbox, and is of the same section as the last; to this member is attached the forward gearbox anchorage. Approximately at this position the chassis sweeps slightly inwards until it reaches the rear anchorage of the front springs, whence it again becomes parallel. The only other cross-member is a 2 1/2 in. diameter one slightly aft of the front axle, and incorporated with this member are the two diagonal arms which strengthen the dumb irons. A noteworthy feature of the chassis is the absence of rivets. The suspension at the front is by half elliptics, 44 in. long; the complete spring comprises 14 leaves. The rear cantilever springs are 4 ft. 7 in. long, with ten leaves, The anchorage of the centre part of the spring is a delightful piece of work and is arranged so as to relieve the side-members of twist. This is achieved by extending the fulcrum to the inside of the chassis and dropping a forged bracket from the cross-member. The side-members are also braced by the dropping of two king-posts approximately 8 in. in length, separated by some 4 ft., and united by round forged members. An interesting point here is that where the forged members are attached to the king-posts a hardened steel bush is used and the whole clamped together by a half-inch nut and bolt. At the rear, single Hartford shock-absorbers are used, with shock-absorbers of Rolls-Royce design at the front, which work on the cone-clutch principle and appear on the small side for their job. The front axle is the usual I-section forging, and there is nothing outstanding in the general lay-out of this axle. The rear axle is of the fully floating type, and the most astounding point of interest here is that the axle shaft is splined on to the hub, instead of using the usual taper and key. The main brake drums are some 17 1/2 in. in diameter, and the inner brake drum is 6 in. smaller. The normal two-shoe braking system is used, but there is an adjustment for resetting the brake lever on the camshaft. Both sets are compensated, and there is a master adjuster for the foot-brake. I can give little information as to the crown wheel and pinion assembly, but it is well worth while mentioning the astonishing number of studs which are used in the rear-axle casing. For instance, the tail pinion housing is held in place by no fewer than 23 3/8-in. bolts, and there are some 20 for the axle tubes. The torque tube is split in the middle and there must obviously be a large-size ball race in this position. A ring of 19 bolts make good this joint. Over the complete rear-axle assembly there was not a trace of escaping oil. (Compared with this attention to detail I remember some few weeks ago examining an E-type “30/98” Vauxhall and noted with astonishment that the tail pinion housing relied on four 3/8-in. bolts to keep it in position.) Rough calculation shows that the rear axle ratio is 3.714 to 1 and the tyre size is 895 by 135.

As previously mentioned, the gearbox is suspended at three points, and there is nothing unusual about the design except the attention to detail. The constant mesh uses helical wheels and the first motion shaft is supported on both sides of its gear wheel. The main shaft is supported at its mid-point, and at the rear there is a substantially-sized ball race. The selectors are of particularly heavy section. The gate change is spring loaded, so that the gear lever always flies into the neutral position, and there is a very definite reverse stop. The gear lever itself follows the design of the normal hand-brake lever. The constant mesh wheels are in a separate compartment, and the main and layshaft wheels are in additions of 70; starting from the spigot end of the crankshaft they are as follows: 37/33, 40/30, 47/23, giving ratios of 3.7, 5.5, 8.27 and 12.65 to 1.

A cone clutch is used and it is of very generous dimensions. Situated on an extension of the crankshaft which appears to form a spigot bearing, there is a valve which permits a leakage of oil on to the clutch. A clutch stop is, of course, provided, and is fully adjustable. Oil-drain plugs are provided in case the clutch should be over-oiled when using an oil gun.

Turning to the 7,670-c.c. [4.25 in. by 5.5 in.] 6-cylinder engine, there is nothing really outstanding about the design, and it is again that attention to detail which is most obvious. The most unusual appearance is created by leaving the valve springs exposed, and as far as I can see there is no reason why they should not be provided with covers, even in the year 1922. The lubrication system is perfectly straightforward. The oil is sucked through a large filter at the rear of the sump by a gear-type oil pump. From the pump it is distributed to the timing gear wheels and main bearings, and then through the drilled crankshaft to the bigend bearings and, finally, via an external pipe on each connecting rod, to the gudgeon pins. All the piping is exposed and there appears to be a complete absence of internal ducts. The pressure of this system is 20 lb. maximum, and can be regulated by means of a relief valve. The drive to the pump is via a vertical squared shaft, and a squared tube which is intended to open up should the pump become jammed by foreign matter; this appears to be going just a little bit too far! There is also an auxiliary oil supply to the base of the cylinders on the off side. This supply comes into operation at about 7/8-throttle opening. Rolls-Royce appear to have forgotten themselves when they mention that as soon as the engine is running this auxiliary oil valve should be depressed by means of a screwdriver or other suitable tool, in order to add a little additional oil to the cylinder walls. Surely the typical thing to do would be to have another lever on the dashboard, connected to this valve by divers cunning levers, bellcranks and rods? Provision is made for priming the pump in the event of its having been dismantled. A reserve oil tank is fitted on the near side and the filler is provided with a tyre-valve adaptor so that air pressure can urge the oil along its lengthy passage in the 3/8-in. pipe. The crankcase breather is the main oil filler and is about 1 in. in diameter.

The crankshaft runs in seven main bearings, the size of which I cannot quote, as the engine is in such perfect running order that there is no excuse to drop the sump. The pistons appear to be just normal pot-type, with three rings above the gudgeon. The cylinders are cast in threes and possess a separate exhaust manifold, but a common induction pipe. This induction pipe is an amazing and somewhat stupendous piece of work, as it comprises a copper tube running the length of the engine, and in the centre a branch is formed which proceeds over the top of the engine and then inclines down to the carburetter on the off side. It is a little difficult to gauge the internal diameter, but 1 7/8 in. would be a fair estimate. There is no hot spotting or heating arrangement, and it is difficult to conceive why these cars average 15 m.p.g. The carburetter is of Rolls manufacture, and is a most complicated but, nevertheless, as the consumption bears out, a most eflicient piece of work. Very briefly it comprises two adjustable jets with individual air passages, one jet functioning as a slow-speed jet and the other as an automatic power jet. In addition, there is an automatic extra air valve. There are no throttle stops and slow-running adjusting screws, as the governor takes complete control of the throttle. The slow running is adjusted by the adjustable slow-speed jet to the front of the carburetter. Both these jet needles are coupled, and the mixture as a whole can be varied from the steering column. Before starting a form of Ki-Gass is used. (Who were the first to use this method of starting, Ki-Gass or Rolls?) This is cleverly arranged by having a two-way tap in the air hand-pump delivery line, which enables petrol to be drawn from the base of the float-chamber and blown into the manifold via two jets situated at the branch on the near side. One shudders to think what would happen to the engine if this priming tap were not turned to the running position as soon as the engine started, for neat petrol would be drawn into the cylinders. That this method of starting is as near perfect as possible was proved by the fact that this car, although it had been standing outside for two years, started up on the first press of the button after its carburetter and ignition had been overhauled. As though to confirm the desire for no heating of the induction system, there are round the jet assemblies two very small water passages fed by 5/16-in. piping. The circulation of water must be exceedingly sluggish, and one is tempted to say that the scheme seems inefficient. The throttle is governor-controlled and this scheme was, to my knowledge, used up to the “Phantom II” series.

The device was very straightforward in its application and comprised a tension spring inserted between the hand throttle lever and the carburetter throttle. This spring applied a varying tension tending to open the throttle to a degree dependent upon the height to which the throttle lever was raised by the driver. With no other counteracting force this spring would be capable of opening the throttle fully when the lever was at the top of the quadrant. The restoring and counteracting force was supplied by a centrifugal governor driven from the engine timing gears at half engine speed, the force of which was directed by control rods to the throttle in such a way that increase of engine r.p.m. would tend to close the throttle completely. The throttle was compelled to take up a position decided by the point of balance between the manually-operated spring tending to open and the governor force developed by engine speed trying to close the throttle. For example, with the engine running, the driver sets the hand control roughly one-third open, the spring, without any opposing force, would considerably over-open the throttle, the engine speed would increase and continue until the governors developed a power which would close the throttle and reduce the engine speed to such a point where the governors could only balance the initial tension given to the spring. The engine speed would remain constant at that figure until some disturbing factor upset the balance. This may be due to the position of lever being changed, or to load being applied to the engine through engagement of the clutch to get the car under way. In the latter case, under any sort of load, providing the clutch was subjected to a reasonably smooth engagement, the car would get away with the governor control in the slow running position only, because the application of load would reduce the engine speed and weaken the closing force of the governor and the tension spring would open the throttle to provide the power necessary for the particular load applied. In actual practice to avoid a too critical clutch engagement, the governor lever would be set in a position to give an r.p.m. slightly faster than slow running.

Owing to the tremendous inertia of the clutch and gearshaft assemblies a quiet gear change on the “Silver Ghost” was not easy to obtain, except with the most skilful of operators double de-clutching with a high degree of precision, and it was here that the governor proved most useful, as the engine could be set to give a change into a lower gear at certain specific readings of the speedometer. Alternatively, another method was to take the control momentarily from the accelerator pedal to the hand lever, when a change could be made from top to third by dividing the road speed by five, which gave the number of notches to advance the hand lever to obtain perfect synchronisation of engine speeds. Lastly, the accelerator pedal was quite independent of the governor control, and through a pick-up operated the carburetter throttle directly, although the governor, under the influence of engine speed, would be attempting to close the throttle completely. This effect could never be felt, as the leverage obtained by the foot on the pedal was too great to be influenced. At the same time, on “luffing,” immediately the foot was removed from the accelerator pedal, the governor would come in and completely close the throttle, in an endeavour to bring the engine speed down to the hand-lever setting, which would probably be the idling position.

Dual ignition is provided. A Watford magneto lies on the near side and is driven by an extended shaft with a vernier at each end; a self-extracting nut is fitted on to the appropriate timing wheel, so therefore the timing of the magneto is a very simple operation. The coil-distributor is on the off side and is driven from the same vertical shaft as the oil pump. The contact breaker assembly is, I am sure, the largest and most substantial one ever made. Likewise the rotor and distributor are a sensible size and the leads are most carefully marked. The magneto and coil are coupled together and the manual advance is situated on the steering wheel. Mounted just below the distributor is the automatic air pump for providing pressure in the petrol tank. This comprises a very neat cylinder and piston driven by an enclosed eccentric on the vertical drive. The compression ratio of this pump is arranged so as not to permit a greater pressure than 3 lb. per square in. in the petrol tank. Needless to say, a hand pump is provided to produce the initial pressure, and at the near side just below the bulkhead there is a tap which gives the following variations: hand pump only; power pump only; hand and power pump; release. On the off side, in approximately the same position, there is the petrol tap. Consider the necessary actions for starting this car. They comprise turning the priming tap, priming the engine, and then turning on the main petrol supply by walking round to the opposite side of the car and setting the tap to the desired position. Then comes the actual starting of the engine, and finally a nimble leap from the driving compartment to depress the valve controlling the extra oil to the base of the cylinders – quite a formidable programme. There are one or two more things about the engine which are worthy of mention. One is the fitting of a thermostat, in a very neat manner in the cylinder head, which is still functioning to this day. Circular nuts are used on manifolds, water connections, and the exhaust system; these nuts are a particular curse when used on exhaust manifolds, as they corrode, and a ring of small-section bolts is a much more practicable proposition. There is a very crude brake situated on the magneto, which is most obviously an afterthought, and for this afterthought I can offer no explanation whatsoever. Lastly, there is the absolute wizardry of the coppersmith’s art in pipe bending.

The outstanding feature of the electrical system is the operation of the self-starter. This is slung between the two crossmembers supporting the gearbox, and drives by chain to a shaft running inside a hollow layshaft. At the rear of the layshaft there is a dog clutch. Pressing the starter button energises a solenoid which, by means of a lever and rods, draws this dog clutch into engagement, and when the other solenoid closes the main contacts, the starter motor delivers its power through the constant-mesh gears to the crankshaft. The only noise heard is the slight threshing of the chain, and as the solenoids are of Rolls-Royce manufacture, this system must be as efficient as it is silent. A very interesting point is the use of carbon contacts to make the initial contact and final break, while copper contacts are arranged to take the weight of current. A dynamo is slung in a similar manner on the off side of the gearbox and is belt-driven. Two-pole wiring is used throughout, black and red wires being utilised, and all the switching is done in the negative leads. The dynamo is of the plain, shunt-wound type, and is of a size more than sufficient for requirements. A substantial Lucas switch box is carried on the instrument panel, and contains three switches for head and side lamps and dynamo respectively, and the dynamo field fuse and the cut-out. A Weston moving-coil ammeter is used, and conduits and junction boxes grace the engine side of the bulkhead. Strangely enough, the magneto-and-coil ignition switch is on top of the steering column.

The foregoing is an unprepared series of notes which, it is hoped, will be of interest, but which could be enlarged upon many times, as every detail on this real engineering job provides in itself a lesson in the art of engineering. There is one point which I should have mentioned, and it is a fitting which has annoyed, I think, nearly every person who has called in to view this vintage motor, and that is the diminutive steering wheel, which is 16 in. in diameter and has a horribly thick rim, some 1 1/2 in. in diameter. Such details as this, and the magneto brake, “do cause one furiously to think.”