OILING THE FUTURE.
(c) Tony Foale 1985 - 1997.
With the recent unveiling, at the Tokyo show, of the Suzuki Falcorustyco, and the Mark.3 Bimota Tesi at the Milan show, it seemed a good time to have a look at hydraulics. The Suzuki has hydraulic drive to the rear wheel, and both feature hydraulic steering. Do either have any real merit or are they just marketing men's gimics to satisfy the insatiable appetites of technology junkies? To find out, let's look at the basic principles involved and how these are applied by the two makers.
Consider the elementary system shown diagrammatically in Fig.1.
As the handle-bars are moved to turn right the piston in the control cylinder displaces hydraulic fluid, which is fed into the slave cylinder. This causes the piston in that cylinder to retract, pulling, as it does so, on the steering arm and hence turning the wheel to the right. Well, the basics are that simple but why go to that trouble anyway? On a conventional telescopic forked bike, there would be little point as it would be hard to devise a more reliable, simpler, cheaper or lighter system than just bolting the bars directly onto the top of the forks. But with alterative front suspension designs, loosely termed hub-centre, this is not possible. With these systems it is usual to provide the steering through a series of links and levers. Depending on the individual layout this may be done simply with one drag link, directly connecting the bars to the upright, or it may be necessary to have a more complex arrangement if the route between the handle-bars and wheel is more torturous. Simple or complex, there is one common requirement that does not occur with telescopic ( or any head-stock mounted ) forks, viz; the need to avoid what is called "bump steer". This is a new term to most motorcyclists and is often not properly understood. So I shall digress for a moment to explain. Fig.2 illustrates an Elf. type of double link front end.
As the wheel rises under bump the end of the steering arm describes an arc, defined by the length and angles of the two supporting links. Now, the steering arm will be forced to follow the path of the end of steering drag link, and unless these two arcs are coincident the steering will be deflected. This can be prevented by the correct length and location of the drag link. I have deliberately set a bike up with some bump-steer to check the practical effects, and I can categorically advise against it, the feeling is hard to put into words other than to say that it is definately 'spooky'.
With some hub-centre designs, particularly those with the links low down near the hub, like the Tesi for example, it may be difficult to physically locate a steering column and/or drag link in a position to avoid this bump-steer. Engine cases, exhaust pipes, oil coolers and radiators all compete for space in the vital areas. This is where hydraulic steering may be of use, there is much more scope for varying the location of the cylinders as they can be connected via flexible hoses. However, as I have mentioned in these pages before, we seldom get anything for nothing. The fly in the ointment, is that the basic system in Fig.1 is not suited to practical use. Expansion and contraction from temperature changes and seal leakage are two bugbears.
Expansion causes increased pressure in the cylinders, which at best, means that seal friction is increased resulting in sticky and stiffer steering. At worst the seals or cylinders may burst, resulting in total steering failure. Contraction may lead to a partial vacumn which would cause a dead spot or backlash in the steering. i.e. The bars could move a small amount before the wheel responded. If the seals were less than perfect, then this vacumn would be filled by air seeping past them. This would eliminate a true dead spot but introduce a compliance to the steering, ------ hardly any more desirable. I believe that the Tesi.1 used this simple system and had problems when the temperature changed. But the Tesi.3 ( and I think also on the Tesi.2 ) has a modification to prevent a reoccurrence. Fig.3 shows the idea,
The hydraulic system is connected to a pressurized gas cylinder, similar to those used with some suspension units. The gas and oil are kept apart by a flexible diaphram. The gas pressure must be selected carefully, too high and there would be problems of seal friction, too low and undue compliance would be introduced. This compliance can be likened to preload on a suspension unit, until the preload is exceeded the unit acts as if it were solid. So too with the Bimota steering, unless the force required to make the wheel follow the handle-bar movement exceeds the force possible on the pistons from the static gas pressure, then no additional compliance is introduced. (I should have clarified this point a bit more in my article of May 85. on alternative steering systems) I have not had the opportunity to ride the Tesi but at Milan I did operate the steering (with the front wheel off the ground) and it felt very smooth with no sign of sticking.
Full technical details of the Suzuki 'Rusty Falcon' are a bit thin on the ground, and so, many of the following comments are based on a close scutiny of photos. It is also not clear just how close the show machine is to a properly engineered and working prototype. From various sources I have been told that it is on the one hand a ridable machine, and on the other that many components are nothing more than wooden mock-ups. So if some features appear a bit odd it may just mean that these have not been fully engineered or are included just to evaluate public opinion. Unlike the Bimota, the Suzuki has four pipes leading to the slave cylinder, there could be several explanations for this. e.g. Some crafty way of avoiding the expansion problem described above or perhaps it is just a fail-safe feature. Without being able to see through the bodywork I shall stick my neck out and suggest that two control cylinders are used as in Fig.4., one to each hand-grip. Thus, if only one grip was operated the other would be forced to follow, because of the flow back from the slave cylinder. So even if the two grips are not mechanically connected the action would still be fairly normal.
Provision for expansion is still needed and one can only surmise on how this is done, if at all. Perhaps the Tesi system is followed, in which case two pressurized gas cylinders are necessary. Other aspects of the steering system appear to be copied from (or should I say "similar to") the Bimota. Apart from the obvious styling differences, the main variation seems to be that instead of using two torque stays (one on each side of the wheel) above the main swing arm, as on the Tesi, the Suzy makes do with just one torque arm, on the left hand side, below the swing arm. Two arms have the advantage that one could fail without dire con- sequences. But the disadvantage is that the dimensions and location of both arms must be accurately controlled, because any variation between the two may lead to undue stresses being placed on some hub components.
There are two possible types, hyrostatic and the hydrodynamic system which was used on the 400cc. automatic Honda and the Moto- Guzzi "Convert". Both having fluid clutches or torque convertors connecting their engines to automatic gearboxes. In its simplest form (the fluid clutch), one impeller is connected to the engine and a second to the gearbox input shaft. The assembly being sub- merged in oil. As the engine speed rises the oil starts to circulate, gradually transferring power to the gearbox. The principle of operation is the same as a paddle wheel in a jet of water, the kinetic energy of the jet is transformed into power at the shaft of the paddle. There is no fixed speed relationship between the driven and driving components. A certain degree of slip is inevitable and results in power loss. The faster the engine spins the lower the degree of slip, for a given torque. Early car automatic transmissions were of this type and were quite inefficient. Current practice is to introduce a third fixed element between the two rotors which redirects the internal flow, and uses the slip at low engine speeds to multipy the torque transmitted from the engine to the gearbox, thus increasing efficiency. Hence the term torque convertor.
No doubt there are history buffs out there who can correct me, but as far as I know the 'rusty falcon' is the first bike actually constructed with hydrostatic transmission. Although others have suggested its use in the past. I know that at least as far back as 1936 it was proposed for car use, and in the mid 60s. I even designed a future bike using it to provide two wheel drive. But I could not afford the high cost of the pump and motors to construct a prototype. But I did hear a rumor a year or two back that Honda had it in mind if not in metal. Hydrostatic transmission is accomplished by the interconnection of a positive displacement pump on the engine and a similar positive displacement motor driving the wheel(s). This is similar to the steering systems above, i.e. the driving pump displaces a given amount of fluid and this in turn forces the driven motor to rotate by a fixed amount, with negligable slip. While the hydrostatic system may be a Suzuki first, for motorcycles, it has long been used on other vechicles. Many of which we like to regard as low tech. compared to some of the current crop of new bikes. Several types of earth moving and construction vehicles have one pump on the engine connected to a motor mounted at each wheel. Infinitely variable automatic gear- ing can be incorporated and installation problems of different- ials and drive shafts are avoided. To all intents and purposes the design of the pumps and motors are identical and fall into two main classes. i.e. the radial piston type and the axial piston swash-plate variety. Both can easily be produced with the facility for changing the pumping capacity per revolution. Fig.5 shows the basic principle of the radial type, and how the capacity can be changed by simply moving the outer cam ring.
This ability makes possible a transmission with infinitely variable gear ratios. Consider for example, a variable capacity pump mounted on the engine with a maximum capacity of 100cc. and a minimum of 25cc., driving a motor with a fixed capacity of 100cc.. For the same pump speed the motor output speed will vary over a 4 to 1 range. The mere mention of automatic transmission is almost guaranteed to make any red-blooded motorcyclist gasp with horror. Quite understandable if we were talking about the mushy car type systems, but the feel of a good, well controlled hydrostatic device would be totally different. Performance would be enhanced not reduced and so too would economy. Too good to be true? No, just a matter of combining a bit of the 80s. electronic technology with the much older technology of hydraulics. Let's see how. Consider first the basic requirement for maximum performance, i.e. a gear ratio that allows the engine to spin at the RPM for peak power, at any road speed. The trouble with a normal manual gearbox, with perhaps 5 or 6 ratios, is that the engine can only be kept at peak power for a short time, each time a change is made the revs drop below the optimum. Road conditions dictate that we cannot ride with full power all the time, regardless of the fun that would be. As we have to pay dearly for our fun, would not it make sense, that if during those periods that we do not require max. power, that we have the bike set up for maximum economy. The need here, is to have a gear ratio and throttle opening that allow the engine to produce the power, necessary for the desired road speed, most efficiently. Modern electronics allow us to do this, with a magic black-box. As inputs to this box let's use the RPM., a fuel flow transducer mounted in the fuel line and a position transducer on the twist- grip. The actual throttles on the carbs., or better still, on the fuel injectors, would be controlled not directly from the twist-grip but by a small servo-motor controlled from the micro- processor in the black-box which also would be used to set the effective gear ratio between the pump on the engine and the motor at the wheel, by changing the displacement of the hydraulic pump or motor. Rather than setting the throttle position the twist-grip would basically just send to the control box, a signal indicating the rider's desired speed and acceleration. The first half twist, or so, would indicate definite speed requirements and the control box would set the gearing and throttle position to give that speed with the minimum fuel flow. A full handful, on the other hand, would indicate that maximum power is wanted, and so the gearing would be set to give optimum RPM. and the throttles would be fully opened, regardless of fuel flow. The last half twist, for example, could indicate that acceleration greater than that for maximum economy is required and the throttle and gearing would be adjusted accordingly. The closer the twist-grip to fully open, the more the power being made available, up to the maximum, of course.
Criticisms sometimes levelled at hydraulic drive include --- Cost, weight and poor efficiency. Let's look at these in order;--
Cost, this is hard to evaluate, because at the moment no suitable units are in mass production suitable for bike use. The production levels are relatively small for the present uses of such equipment and consequently prices are high. But assuming future wide use in the bike/car field I am sure that the industry production engineers would soon work out methods of low cost mass production. The cost of a gearbox and transmission is saved anyway.
Weight:- Similar components have long been used in weight critical installations such as aeronautics, the use of light alloys and high hydraulic pressures can easily keep this aspect under control. Do not forget that we can throw away the weight of a normal gearbox and rear drive components.
Efficiency:- This depends on detail design features and can be over 90%. Sometimes higher figures are quoted for a conventional transmission, but I feel that it is unlikely that this is achieved in practice. If an hydraulic drive is electronically controlled as described above then it is likely that, even if it had a lower mechanical transmission efficiency, the overall efficiency of the bike would still be higher, except perhaps at top speed. But with today's megabikes it is seldom possible to travel at max. speed anyway. ---------- Once we have this type of drive all sorts of other possibilities occur. e.g. As the connection between the engine and wheel is by hoses, which may easily take any route between the two, it is no longer necessary to consider drive train alignment when deciding on engine location and orientation, however careful thought must be given to hose movement. Hoses that seem quite flexible when assembled become very stiff when pumped up with 1000s of psi. pressure.
Another aspect is that the drive motor can be used as a pump when slowing down. If a restriction is placed in the return hose then the back pressure will act on the wheel motor and provide a braking effort. If the degree of restriction is controlled by the brake pedal then there is clearly no need for a normal brake, further saving cost and unsprung weight. Also, if we install a pressure regulator (limiter) in the return line, then the maximum braking effort will be limited and it will not be possible to actually lock the wheel, although under slippery conditions the wheel could be made to turn slower than the equivalent road speed. This would act as a crude anti-lock device, far from perfect but much better than nothing. But for a small amount of additional electronics, the regulator could be controlled to give a very sophisicated anti- lock system. The opposite could also be done to provide an anti- wheelspin device. Heat would obviously be produced if braking was provided like this, but that could easily be handled by plumbing in an oil cooler. Even with present regulations, manufacturers are troubled with noise reduction, and the traditional chain drive is threatened, even with shaft drive we still have whine from the gears and bevel box to contend with. Hydraulics would be much quieter.
If we have a source of high pressure oil from the pump it seems reasonable to consider other uses for it. e.g. The front brake, the control lever would only have to operate a pressure controlling valve, with the braking effort coming from the pump. This simplifies any anti-lock system that may be fitted, it only being necessary to bleed off surplus fluid. To still give some braking capacity with the engine stopped an accumulator pressurized by compressed gas would be necessary. But none of this is new, Citroen cars have had this system on their fully pump powered brakes for years. Incorporating these features gives us a bike with no gear or clutch lever, and if we have anti-lock brakes on both ends we only need one brake control and this can be of light action if powered by the pump. So only two controls are necessary, one to go and the other to stop. Why not combine them into one? The conventional twist grip works quite well, so let's use it in the normal way for acceleration but rotate it backwards to control the braking. Many of you will have Pavlovian reponses against such an approach to riding, uttering comments about removing the skill and enjoyment from it. But what real pleasure exists in our ritual toe and finger exercises that we are presently forced to endure, every time we wish to change speed? One control operation would free us from this burden to really get on with the job of enjoying ourselves, and allow us to concentrate on the riding. Another possible use for the supply of pressurized oil would be for a self levelling suspension system, again Citroen have beaten us to it, but their hydraulic pump is of small capacity, being only for the brakes and suspension, not for the transmission. Hence the response rate of the suspension is quite slow, but satisfactory to cope with different load conditions. With the much greater flow capacity of a pump designed for transmission use, it would be possible to have suspension with a much quicker reaction time. So under cornering loads on a bike, when normally the suspension compresses under the action of so-called G force, the suspension could pump up to maintain the static ride height. Thus reducing cornering clearance problems, hence we can build a lower machine to start with, as well as enabling the suspension to cope with a full range of bumps, whilst leant over.
How many of the above possibilities are included in the Suzuki is a matter for some guess work. There are no foot controls and no clutch lever, and a pressure sensitive pad on the right hand grip controls the braking. From this I think that we can safely assume that the gear ratio is automatically controlled, although the actual selection criteria may differ from that which I have suggested. The method of brake operation points to a pump powered anti-lock system, but there is no indication of whether the front gets its pressure from the transmission system or a separate pump, Citroen style. I have seen no reference to any pump powered, self levelling suspension so it is probably not used. If they show that on a Mark 2. version remember where they got the idea. Perhaps they are waiting for Lotus to finish developement of their 'Active Suspension' system, if that ever gets on a bike our standards for assessing ride comfort and handling will need to be changed upward. Is the Rusty Falcon the bike of the future? Well it's the closest I've seen for some time and I've been looking and waiting for over 20 years.