Conserving Petrol Mechanically

The Editor Contemplates the Free-wheel

With fuel conservation uppermost in the minds of many car-owners, it seems opportune to look back on a mechanical contrivance which was aimed, among other things, at petrol economy, namely, the free-wheel. When I pleaded for a revival of this device a few months ago many readers were in agreement, for coasting along with the engine consuming fuel only through the idling jets must clearly save the expensive spirit. You can toast simply by putting out the clutch or sticking the gear lever into neutral, of course. This, however, is said to be an illegal practice, although my friend Anthony Bird rightly asks, how so, seeing that we all coast for greater or smaller distances every time we withdraw the clutch at a traffic stop, or when we sit in a towed car to steer it ?

The fact is that a car fitted with a mechanical means of free-wheeling saves the tedious process of shifting the gear lever every time a tempting down gradient gives rise to an itch to save fuel, apart from other advantages.

I have driven various cars with in-built free-wheels, notably Rovers (whose makers at one time also used this as a means of obtaining 2WD on the Land-Rover) and in later times Subs, the latter make having inherited this piece of equipment from the triple-cylinder two-stroke DKW of which the original Subs were a crib. In the days before political and scientific bogey-men had made our flesh creep with tales of a drying-up of the World’s petrol resources, with the resultant savage priceincreases we have since experienced, the Sporting driver was apt to sneer at the freewheel, locking it out of action, because he could change gear without its assistance and wasn’t troubled by the cost of petrol. Today we live in a different motoring age, when to improve fuel consumption is an aim few can disregard. This being the case, I think the free-wheel might well be speedily re-introduced, as a means of enabling the engine to run at idling speed whenever the accelerator is released, without any other action on the driver’s part, the drive becoming “solid” again as soon as the throttle is opened. Modern engines idle smoothly, or should. Nor do they easily stall, so one disadvantage of freewheeling, namely that if the engine stops, Power-assisted brakes may be affected, is immediately removed. In fact, modern servobraking systems usually incorporate a vacuum reservoir so that several brake applications can be made before the servo energiser is exhausted. (Conversely, I well remember driving a Big Six Bentley before the war and allowing it to coast in neutral to off-set its great thirst for Petrol. At the very first red traffic light I tried to stop, but found the brake pedal useless, because the vacuum was a vacuum no longer— only hauling with both hands on the brake lever saved an alarming situation.) Today, too, so many cars have alternators instead of mere dynamos that, even if the engine is idling for some 25% of a journey under free-wheel action, the battery should remain effectively charged.

Being unable to think of any snags to freewheeling, let us look back to the days when it was quite often encountered. It was at the very end of what we now call the vintage-era of motoring that the free-wheel came to the notice of the Industry and the keener purchasers of its products. Quite why this was so is difficult to determine. Unless you allow that the financial slump of 1929/30 made fuel saving of interest, even at the then very low cost of petrol, it is at first difficult to see why a free-wheel was thought necessary, for its main attribute then was foolproof gearshifting and synchro-mesh gearboxes were already well-known, having, I believe, been pioneered by Cadillac and La Salle in 1928. The answer is probably that this American innovation had not yet spread universally to British cars and thus the free-wheel was welcomed by an expanding wave of new drivers, of whom a growing proportion were women, who were fearful of a “crash” gearchange, or so the inventors tried to convince themselves.

The first free-wheel of any note to gain publicity here seems to have been the French De Lavaud, which The Motor tested on a 2-litre Rover for 3,000 miles in 1928. This was a rather elaborate device, which replaced the normal back-axle differential and, apart from automatically disconnecting the drive when the accelerator was lifted, gave something of the effect of a limited-slip differential unit. It was claimed that as only the axle had to be dismantled for fitting, this free-wheel was easy for a garage-hand to install, at a cost of about £12. Its reputation was enhanced by the statement that it had been standardised for the six-cylinder Riley and was offered as an option by other British manufacturers.

This De Lavaud differential was being made at Letchworth as the Delavaud and was of the normal cam-and-rollers type, with a patented elastic connection between the roller cages, achieved by using spring-loaded balls engaging “dimples” in the opposite cage. A spring-loaded brake band was used to locate the rollers in the free-running position, but this caused them to move into the locked position when drag was withdrawn from the band, as when the engine was pulling or the car being reversed. Each half-shaft carried a centrepiece for the roller races, so that differential action as well as free-wheeling was provided for. The usual locking was accomplished by pulling up a floor control but there was a safeguard against it operating when the engine was running fast, the driver in this case letting-up the throttle after the control had been operated, whereupon the free-wheel would lock before the throttle was next opened. When reversing the free-wheel was locked automatically.

Apart from declaring that the Delavaud mechanism worked well and reliably, The Motor was interesting on the subject of freewheeling in general. It decided that coasting constituted up to 30% of the journey-distance in hilly country, about 20% in towns, but might sink to 10% or less in fast main road motoring. Fuel saving was reckoned to range from about 15% to 20% over normal consumption and the silent slowing down was said to reduce driving fatigue, although the noise of the engine when it was revved-up was now more noticeable. However, with the drive locked, the noise on the over-run and in the gears was found unpleasant by a driver accustomed to free-wheeling; as the cars of the 1970s are relatively quiet, here is another point in favour of reviving the device.

Stability was much improved, at all events with the Delavaud gear, even though the Rover to which it was fitted was on tyres “practically devoid of tread”, such was the freedom of the 1920s. Gear changing was rendered instantaneous without so much as using the clutch, with the proviso that when changing-up a very slight pause was desirable, but before the lever was moved out of a lowergear position. Very rapid gear changing was further facilitated on the Rover because it had a clutch-stop. The brakes were adequate when coasting. It was felt that a driver might not at first take to free-wheeling but that after a week he “will begin to like it” and after a month’s trial “would not willingly be without it”.

Nothing much more seems to have been heard of this French free-wheel but soon afterwards the Millam free-wheel and sprag was in the news. It was sponsored by Frictionless Mastagear Ltd. and derived prestige by being made by Sir W. G. Armstrong Whitworth & Co., Ltd. of Newcastle-on-Tyne, the aeroplane builders, at their Scotswood factory. It was fitted behind the gearbox and operated on the customary cam-and-roller principle but had, in addition, a separate sprag which prevented a car from running backwards in neutral. A further rather-endearing feature was an oilsump below the mechanism, with its own little dip-stick, from which lubricant was flung into the cam-and-roller gear by means of a fibre wheel pressed by spring action against a rotating drum, the idea being that above a certain speed the fibre wheel would commence to slip and so reduce the oil splashage! However, unlike the Delavaud, the Millam required the driver to rev-up the engine when he wished to lock the drive, which was done with a lever beside the normal gear lever, and the lock could only be freed if the engine was pulling; also for reversing it was first necessary to drive a few inches forward so as to release the locking device from the free-wheel position, or else to bang in the clutch with the hand brake on and a forward gear engaged. This method was introduced as a modification, it was said to enable the driver “to avoid subjecting free-wheel and transmission parts to abuse”. But perhaps a less complex-tooperate lock had proved unreliable ?

The Press tried the Millam on a flatradiator Morris-Cowley saloon early in 1929 and found it to act smoothly and quietly, with the expected easy gear-changing. The Autocar reporter admitted to fumbling urgently with the door handle in his eagerness to vacate the car when the sprag was demonstrated to him, in neutral, hands off the steering wheel, up a 1 in 3.1/2 hill! Extra durability was gained from the Millam by substituting a series of cages on the gearbox tail-shaft, for the more normal separate bearings for the rollers, the latter resulting in very heavy loading between driving and driven shafts should one roller lock “early” and try to crowd one shaft away from the common axis.

It was said that 10,000 Millam units were to be made in the first batch and that the device would fit some 20 of the most popular British cars, at prices of approx. £12 for small cars, £16 for cars of 12 to 22 h.p. (Such prices are now of only academic interest, apart for giving some idea of the manufacturing cost of such equipment.) It was claimed by the Millam’s makers that it gave a 25% saving of petrol, 25% to 60% economy of engine oil, eased the wear on engine and transmission, as the freewheel was likely to be in operation for 25% of a journey, and that it made for greater safety, the sprag apart, because of the ease with which a novice driver could get into a low gear and lock the drive by revving-up (provided the engine had not stalled, of course, which is really why the Lucas “Startix” control was introduced, which caused some interest when I referred to it in the February MOTOR SPORT). Testimonials from Austin car, Austin van, Standard, Morris-Cowley and Wolseley owners were published.

An even better-known device was the Humfrey-Sandberg uni-directional clutch. It was a product of Humfrey-Sandberg Ltd. of Transport House, Victoria, who had registered offices at Imperial Chemical House, Millbank. This was again a free/”jammed” roller-race mechanism, claimed to obviate local loading of its circumference and not to rely on special friction surfaces or powerful springs for its engagement. Engagement was effected by a ratchet-reducing-gear operable through the very smallest of practicable radii or angular movement. Used as a free-wheel behind a car’s gearbox it was claimed to save from 15% to 30% on petrol bills, conserve engine oil, reduce the frequency of decarbonising (which dates it!), and reduce wear on engine, transmission and tyres. A stubby auxiliary “gear” lever actuated the free-wheel lock.

The Humfrey-Sandberg was intended as a ratchet-reducing-gear applicable to all kinds of feed controls, for mechanical paralleling or series-ing of prime movers, and as the main clutch for low-starting-torque electric motors, etc., by manual or centrifugal operation. As a free-wheel it was made for the Austin 7, the Austin 12, the six-cylinder Morris cars, Morris-Oxfords up to 1929, and Morris vans, and was an option on AC, Alvis, Armstrong Siddeley Twelve, Jowett and MG cars.

The Toothill free-wheel used a hexagonal inner member and locked its roller cages with keys and keyways. The Maurice Bone unit appears to have been a cruder device, using pawls and ratchets held out of engagement by bob-weights acted on by centrifugal force to free the drive, and engaged by cams to lock the drive, a sliding sleeve holding them in engageI ment to give a “solid” transmission.

As there was no need to be unduly parsimonious about petrol consumption in the days I am recalling, and as easy-change gears were soon to be introduced, the free-wheel was neglected by the majority of car manufacturers. With, however, some significant exceptions, for in their wisdom Hudson-Essex, Nash, Studebaker, Reo, Plymouth, Pierce-Arrow, Franklin, Volvo, Standard, Rover, DKW, and Saab adopted it. Today the position is different and as the only anxiety about free-wheeling in the 1930s was how the brakes would stand up (although Rover and others had no qualms on that score 40 years ago) and as brakes have since been so much improved, and the thing can be locked out for dicing down an Alp or when driving on sheet-ice, there seems absolutely no objection to re-introducing a simple piece of mechanism which materially reduces petrol consumption.

Especially as our rulers are determined, by every means in their power, to stop us from driving at more than 50 m.p.h. on roads where free-wheeling will save fuel. So what about it ?—W.B.