The Ferguson R5 Prototype
Photograph, Ferguson Research Archives.
The anti-spin and anti-lock mechanisms described by Charles Bulmer B.Sc., A.F.R.Ae.S.
Dual-purpose – a car intended to set new standards in road-holding and road safety and yet to have traction. ruggedness and ground clearance making it equally at home off the road.
We described the Ferguson single-seater racing car in July, 1961, and the estate car prototype, forerunner of the models featured in our road test, later in the same month. Since that time the special Ferguson features and principles have remained basically unaltered but there has been a great deal of development; these changes, which are found mainly in the transmission and brakes, we shall describe below to bring the story up to date, but before doing so it would be as well to outline very briefly the reasons why the car takes the particular shape and form it does.
It will be remembered that four-wheel drive is an essential part of the design and this demands a longitudinal propeller shaft extending from front to back. Clearly. the engine is then most easily accommodated if it goes right at the front, ahead of the front final drive unit or right at the back behind the rear drive; it could. of course, go anywhere between if it were mounted above the shaft but this would be less economical of space and probably raise the centre of gravity.
Now with the Ferguson transmission it makes no difference to traction or braking power which end you put it-most of the old arguments in favour of one arrangement or the other become invalid. There remains, however. the effect of weight distribution on road holding and cornering; with a forward weight bias stability on the straight, in cross-winds and on corners is easier to achieve without resort to extremes in suspension design which usually carry undesirable penalties and side-effects.
So the engine was put at the front, leaving the maximum accommodation for luggage and also making it possible to use a body of estate car pattern with a low rear floor level; a type which many designers now regard as the normal family body of the future. With this layout an engine of minimum length is desirable to reduce front overhang and of minimum overall height for forward visibility and to harmonize with modern low-fronted styling. Both considerations suggested a flat four or six and for an engine capacity around 2.2litres, the former was thought entirely adequate.
Regarding the Ferguson as a car designed for the near future. it would have been an anachronism to have anything but independent suspension all round, or a near-equivalent like de Dion. The front suspension is by the usual unequal length transverse wishbones with coil spring/damper units mounted very high and bearing on the outer part of the upper wishbone, very much like the Triumph 1300. Rear suspension is shown in Figure 1I. Geometrically it is closely related to the almost standardized type used by modern Grand Prix cars: mechanically it is engineered in a sturdier way more appropriate to a touring vehicle.
fig.1 The Independent rear suspension is strong but simple.
Ordinary cross-ply tyres were used at first but when it became clear that radial-ply covers offer higher cornering power, higher cruising speeds at standard pressures without overheating and better grip on slippery surfaces. a car dedicated to safety could hardly ignore them. Adapting the suspension to accept their different characteristics without excessive harshness and road noise has been a development headache.
fig. 2 Engine and transmission layout.
A steel body/chassis unit of conventional construction completes the design, the general layout being shown in the plan-view drawing (Figure 2). We have made no attempt to describe the car in great detail because its main purpose is to provide a vehicle of sufficiently modern design and performance to demonstrate the virtues inherent in the transmission and braking systems.
fig. 3 The actual interrelationship 01 the various transmission components is illustrated in this diagram and described in the text. Note the Maxaret unit on the extreme right.
Transmission
This is best described in two separate parts-first the semi-automatic gearbox and then the four wheel drive and special centre-differential assembly. The whole assembly is shown diagrammatically in Figure 3.
From the engine flywheel, the drive goes directly to the Ferguson Terramala hydraulic torque converter which has an unusually wide conversion range–it can give torque multiplication ratios as high as 2.7-3 when starting from rest. ,In previous designs a two speed epicyclic gearbox (like an overdrive) was interposed between the engine and the converter, an unusual arrangement with interesting characteristics but rather expensive.
Ferguson estimate that to add a torque converter to a conventional synchromesh gearbox involves only half the extra cost of replacing it with a fully automatic transmission, particularly as the gap-bridging effect of a converter makes it necessary to have only three speeds instead of four.
This, as the drawing shows, is what they have done. A conventional foot-operated friction clutch is retained between the two to separate them for gear changing-otherwise gear changing on the move would be impossible or very destructive–but, of course, this clutch need not be used for take-off from rest. The car can therefore be driven as a two pedal vehicle (in one gear) in towns or open country or a three pedal machine for maximum performance but, in any case, the wide-ratio converter reduces to a minimum the necessity for gear changing.
The Centre Differential
The offset between the gearbox output shaft and the drive shaft line is bridged by a twin duplex Reynolds roller chain drive running
at very high efficiency in; an oil bath. This chain drives the planet cage of an ordinary differential and the two sun wheels are connected to the front and rear final drives respectively. Up to this point, therefore, we have an ordinary four-wheel drive system delivering equal torques to front and rear but with nothing to stop the wheels at one end spinning.
The special feature of the Ferguson drive is best illustrated by a simplified diagram (Figure 4).
Fig. 4 (above) This shows the basic principle (not the actual layout) of the Ferguson differential. The drive is transmitted through the central gear set. To the differential cage: the other two gears on the two output shafts are driven at a higher speed and idle on their free·wheels. But if either shaft accelerates to the same speed as the gear revolving on it. The free wheel locks and prevents any further increase.
This shows the main drive taken from the input shaft to the differential cage, as before, but in addition two more gears of different ratio are also connected from the input to the output ·shafts. This can only work because the gears on the output shafts are mounted on free wheels (one-way roller clutches) which allow them to rotate at higher speeds. If, however, either of these output shafts speeds up to the same r.p.m. as the gear running on it, the free wheel will lock up solid and prevent any further increase. So the centre differential will allow the front wheels to rotate faster than the back (or vice versa) but only between certain limits of rotational speed dictated by the ratios of the two additional gear sets.
In the Ferguson car this principle is retained but the layout is different (Figure 3) in that the control gears and clutches are physically sited all on one side of the centre differential; this is more convenient because of the large centre distance between input and output shafts. In this application the gear A is driven directly from the differential cage, the front drive propeller shaft passing freely through the middle of it. It follows then that A and B and the whole “duolok” layshaft are in effect coupled directly to the input from the gearbox; by using the two free-wheeling gears on this layshaft to control the speed of the front propeller shaft within limits (which will be a little more or a little less than that of the differential cage) exactly the same effect is achieved as with the simple system of Figure 4. In this case the free wheels will operate in opposite rotational directions; for reversing they have to be put out of action altogether otherwise they lock up solid.
So for steering purposes or to accommodate variations in tyre diameter one pair of road wheels can rotate slightly faster than the other but as soon as one or more wheels try to spin, the control action comes into effect by prohibiting the large relative speeds involved. It is, of course, still possible to spin all four wheels or, in special circumstances, to spin one front and one back wheel simultaneously.
On slippery corners splitting the driving load between all four wheels enables a skilled driver to use a great deal of power without any danger of producing a sudden breakaway at one end of the car-perhaps more to the point it enables an unskilled driver to be as ham-footed as he likes in these circumstances without dire results. Although the arrangement we have described gives a 50/50 torque split between front and rear wheels, this is not an essential feature of the design and by replacing the ordinary centre differential with one of different pattern it is possible to divide the torque in other proportions. The Jensen FF, for example, has an epicyclic differential passing 63% of the drive torque to the rear and 37% to the front; in this way normal handling characteristics can be modified without prejudicing the speed limiting effect which will override other considerations when abnormal (incipient wheelspin) conditions are reached.
Now the intercoupling of front and rear operates, of course, in deceleration as well as in acceleration. It is not effectively possible for a single whee to lock unless all four do so or, at the very least, unless another wheel locks at the other end of the car. But either of these occurrences can be prevented entirely in normal circumstances by the Dunlop Maxaret control unit-Figure 3 shows its location in the Ferguson layout.
Braking system
We have described the Maxaret unit before on several occasions. It contains a small flywheel which is driven (in this application) at input shaft speed (or at some fixed fraction of this speed) by a spring drive of limited torque capacity. A sudden angular deceleration, of the kind which accompanies wheel locking, collapses the spring drive and the relative (angular) movement between the flywheel and its drive shaft is used to operate valves which automatically unload the brake hydraulic line pressure before locking actually happens. In this way the braking can be controlled to oscillate (or “cycle”) around the region of maximum braking without allowing rotation to stop.
In the form used by aircraft for many years, each landing wheel has its own Maxaret operating independently on the very high pressure hydraulic supply to that wheel. The centre differential of the Ferguson allows one Maxaret to operate all four brakes since, as we have already pointed out, single wheel locking is prevented. This means a considerable saving in cost. But in its original form it still needed the high pressure hydraulic supply which cars in general do not possess (Citroen and Rolls-Royce are exceptions) and which would be very expensive to fit.
The development of the system to ‘use a vacuum servo and ordinary direct hydraulic brake operation is the significant accomplishment of the last few years in evolving towards a design which is economically as well as technically possible; the current layout is shown in the simplified diagram of Figure 5.
fig. 5 The latest, much less expensive braking system has been developed to work with an ordinary vacuum servo instead of a high pressure powered hydraulic system.
The brake pedal acts directly on a tandem master cylinder feeding separate front and rear brake hydraulic circuits. A large direct acting servo – a standard Kelsey Hall unit made under licence by Dunlop – is coupled to the pedal operating rod. All this is standard practice.
In the “off” position vacuum is present on both sides of the Servo diaphragm because the two chambers A and B are in direct communication through the drilled passages and plate valves in the centre of the servo unit. First operation of the pedal seals off this connection and further pressure then pushes a plate valve off its seat to open a connection between C and B. Now C. at this stage, is at atmospheric pressure because it is. in communication with the right hand chamber of the control unit and this in turn is connected to atmosphere via the pipe running below it in the diagram.
So air is admitted to chamber B and the servo diaphragm IS pushed to the left, assisting the driver’s own efforts on the pedal; the valve between Band C is so balanced that it closes agam when the servo force reaches a certain value and in this way the assisting force is kept strictly in proportion to the driver’s own efforts. This, of course. is ordinary servo operation because at this stage, as indeed for all normal braking, the Maxaret control is inactive.
As soon as a wheel starts to lock the Maxaret contacts close and it sends an electrical signal to energize the solenoid in the control unit. This moves the double shuttle valve from the position shown (at the right-hand end of its travel) hard over to the left which cuts off the connection between the vacuum reservoir and chamber A and transfers it instead (via the control valve) to C and thence to B: at the same time A is connected to atmosphere so that the net result of the Maxaret intervention is to reverse the pressure and vacuum conditions on the servo diaphragm. It then pushes against the driver’s foot, forcing the brake off again until such time as the “locking” signal ceases and the solenoid allows the control valve to return to the right hand position under spring loading.
. In practice these alternations of pressure and suction across the two sides of the diaphragm will occur very rapidly-several times a second-to keep the brake hydraulic pressure fluctuating around the value which just corresponds to the locking peak. If the driver pushes harder still on the pedal, the servo will oppose his effort more strongly to keep the net force on the master cylinder the same.
There are a number of refinements and fail safe devices in the system. For example, a connection is shown to the lower part of chamber B through a non-return valve; this is not essential but it by-passes the valves and small passages between Band C and greatly speeds up the rate of pressure changes. Then there is a threshold pressure switch in the front brake hydraulic line which isolates the whole system electrically until the pedal is pressed and a little hydraulic pressure is developed. This prevents the Maxaret from energizing the solenoid in response to jerks in the transmission which can arise from gear-changing or even from exceptionally bad bumps or potholes. The latest developments are beginning to make this pressure switch redundant.
If a fault or short circuit developed either in the Maxaret or in the external wiring: in such a way that the solenoid was energized inadvertently; it would be dangerous because the brakes would then be held off \\’ith full servo force. At the right-hand end of the control unit is the fail-safe valve which in these circumstances would experience a vacuum to the right of the small plate valve and atmospheric pressure to the left of its diaphragm; the areas of the valve and diaphragm are so calculated that the differential load across the valve would then be just sufficient to move it to the right against the seating provided, cutting off the vacuum supply to the servo diaphragm. Since similar differential pressure conditions can exist for short periods in ordinary Maxaret use, a deliberate delay IS built into the operation, the air to the left hand side of the fail-safe diaphragm being supplied through a very small restrictor so that it needs about half a second for the pressure to rise high enough to push the valve across.
Facts and fallacies
Perhaps the impression has got around that the Ferguson Maxaret combination is infallible-the complete answer to all skidding problems and a device which makes brakes super-normally effective. This is only true up to a point and .it is certainly a wider claim than its sponsors would make for It. It might be as well to look more closely at its limitations as well as its virtues.
Its greatest virtue is safety because on all normal. wet or dry road conditions, both on the straight and on corners, It w1l1 allow even a novice driver to react to an emergency in the most panic-stricken way without losing control of the car. We needn’t stress this point any more because it emerges clearly enough from. the road test. As regards actual stopping distances, it makes little difference from very low speeds or even from high speeds on dry roads. On wet roads it is a different matter; on May I, 1965 we published some figures and graphs showing the tremendous difference in the available coefficient of friction between tyre and road just before and just after the wheel locking point which may differ by a factor of 2 to I; the Maxaret can keep the wheels oscillating around peak grip but no driver, however skilled, can do so. So the higher the speed and (up to a point) the more slippery the road the more dramatic the gains it can show.
It is not, however, really effective in such conditions as ice and snow. There are various reasons for this; the Maxaret unit has to allow quite rapid wheel deceleration on dry roads without coming into operation-the sort of deceleration that corresponds to a 1g stop with rotating wheels – only at appreciably higher angular decelerations should it send a distress signal to thee solenoid. On ice. wheel accelerations and decelerations tend to be low because of the small amount of braking used and the low grip. A Maxaret can be designed for these conditions but only by prejudicing its behaviour on surfaces with coefficients in the region of 0.25 to 1, on which the average motorist drives for nearly the whole time. At present an attempt is being made to extend this range by the use of the all-weather valve; we shall not describe this in detail but it acts as a variable restrictor in the line to chamber A, altering its own restriction in accordance with the vacuum in the line in order to slow down the re-application of the brakes. It is able to discriminate between road conditions because in dry weather you need little vacuum to take the brakes off sufficiently-in wet weather you need a lot more.
There is one other limitation in exceptionally slippery conditions and that is the possibility of “pushing through” against maximum servo resistance. For example, taking· the servo diaphragm area as 48 sq. in. and the maximum available depression as 10 Ib/sq. in., then the greatest force which the servo can exert against the driver is 480 lb. or, allowing for a mechanical brake pedal leverage of about 3 to I, say 130lb. at the pedal. Since a strong driver may be able to exert 200 lb. in a panic, he could in fact overcome the servo and still apply a respectable braking effort. There are obvious ways of changing the parameters to reduce this possibility but with production servos and master cylinder sizes giving adequate fluid volume it is not as easy as it would seem.
But even here, as in all slippery conditions expected or unexpected the brake system operates as an early warning device; a driver of any sensitivity will be warned by an early “kick-back” on the pedal that things are not what they seem to be.
First published in Motor magazine. Published in Journal 32, Autumn 1999. Charles Bulmer B.Sc., A.F.R.Ae.S.
