FF front suspension

Front Suspension for open cockpit, FF Single Track vehicles

Definition; “FF”

A single track vehicle with a seat base less than 20” above ground level at ride height, fitted with a seat back capable of supporting the rider. The front suspension should not be steered.

Application.

This applies to all FFs but is specifically relevant to 'agile' or 'performance' FFs.

Sources.

The history of vehicle suspension was enthusiastically plundered. “The Motor Vehicle” (Newton, Steeds and Garrett, ISBN 0-408-01082-7) is a good source for multi-track derived systems. Contact with early seventies Formula One cars was also instructive. Of the various systems bolted to motorcycles only Normans Hossack's system, recently appropriated by BMW and the system pioneered by Tony Foale, subsequently used by Yamaha on the GTS, bear consideration for road use..

Many of the vehicles featured on this site have contributed development knowledge.

Introduction.

At first glance it might seem that the front suspension system universal in motorcycles and 'performance' scooters, Telescopic Forks, would be entirely adequate for a simple FF. This is entirely true, but chiefly to demonstrate beyond doubt that this 'legacy' system is not adequate for even an FF directly converted from a motorcycle.

This doesn't mean that such systems become more unsafe when applied to the FF layout, just that their performance limits that of the vehicle.

There are three reasons for this, all inherent in the FF layout, and this subject is also considered in the paper on 'basic dynamics'.

Braking performance.

The secure riding position and low CG of the FF layout means that braking performance is limited only by available adhesion, rather than the rider losing control or rear wheel lifting. This means more than 1G deceleration in reasonable conditions and telescopic forks are quite incapable of dealing with these loads - while also providing suspension. Only the dive under braking and lack of rider security, inherent in high CG motorcycles, make telescopic forks tolerable under braking.

A system is needed that, at least, can withstand full braking forces without deflection. Any vehicle engineer will assume that the suspension and steering will also continue to work.

'Violent Correction'.

The low CG and secure riding position also allow very fast direction changes, including those needed in correcting some loss of adhesion. Performance use naturally exploits this ability and reasonably lightweight FFs like 001approach kart racing levels of agility. Even FJ can be slalomed violently enough at low speed to lift the front wheel.

Only systems in which the mass of the suspension is not steered can be used this violently without feedback wobbles. Applying the 'opposite lock flick' common in a momentarily oversteering vehicle to a motorcycle normally produces a 'tank slapper' as the considerable mass of the steered suspension resonates with the geometry.

Torsional stiffness.

Any single track vehicle is steered by moving the front contact patch from side to side, relative to the CG, in order to control the lean angle of the vehicle. Much of the 'steady state' is dealt with by the geometry and inherent features like tyre coning but rider inputs in an FF can be rapid and violent.

If this is to result in a precise steering reaction the movement of the contact patch must result in a prompt reaction from the vehicle. Control precision in an FF is defined by the torsional stiffness of the whole vehicle. A flexible single track vehicle will behave like a flexible multi-track; prolonged understeer, followed by an unpleasantly sudden transition to oversteer as everything is finally wound up and the message gets through to the rear wheel.

First of all the steering system used to direct the contact patch must be torsionally stiff in order to transmit steering inputs and detect feedback. Secondly, the structure of the front suspension must precisely transfer the torsional load imposed by a suddenly displaced contact patch into the structure of the vehicle.

Summary.

These three requirements can only be ideally met by systems that are stiff enough to cope with the braking and torsional loads and do not steer the suspension.

However the Comfortmax points to a possible lightweight FF 'window' where the loads involved may be low enough to permit adequate performance with Telescopics. A small, lightweight, wheel may avoid steering wobbles in violent inputs, especially if combined with BMW's telever system, which does not steer the suspension. This should also improve the torsional stiffness of the slider assembly. The actual forks can be better supported with the bottom yoke much closer to the ground than possible on a motorcycle.

The existing Comfortmax example does not meet these criteria but makes noticeable movement towards them. It is still limited by it's front forks,.particularly under braking, but less so than motorcycle derived cut'n shuts.

Telescopic forks are basically an unfortunate accident of history. If the people who developed 'girder forks' especially those used by Vincent on their post-WW11 models, had possessed the understanding and innovative ability of Norman Hossack much pain could have been avoided.

Suitable systems.

Hossack.

It is appropriate to start with Norman Hossack's system. This is, geometrically, “double wishbone” but with the 'upright' extended downwards to form a pair of forks with a wheel axle across their lower ends. It is virtually identical to the “Girder Forks” almost universal on motorcycles until BMW introduced telescopics before WW11. The crucial difference is that the pivoting links and springs of the girder system are not steered on the Hossack. The steering head is moved out to the apex of what are now two wishbones, firmly based on the chassis and the suspension picks up on the lower wishbone in normal automotive practice.

As the fork unit is a single component it can be made light and stiff. The lower wishbone can just clear the tyre, minimising the fork extension and bending loads and even torsional stiffness can be reasonably good. Resistance to lozenging in the wheel axle/fork leg/bottom yoke assembly can be obtained by adding thickness to the fork extensions and the FF layout allows this.

This system has performed reliably on motorcycles and providing the higher loads imposed by FF use are accounted for, it should be worth consideration on any cut'n shut. Comfortmax-style scooter conversions would seem to be an appropriate application for this system where no lightweight full-HCS system is available. Fabrication is easy, no special engineering is required and there is full control of geometry.

“Full HCS”

Single-arm

The very simplest system, Single-Arm, is used on the Reliant three wheeler, which has sacrificed so many engines to the FF cause. Although good in terms of steered weight and braking performance it lacks a top wishbone and any claim to torsional stiffness. The geometry also varies with suspension movement and although five or six degrees can be tolerated over full suspension travel, it seems unlikely that Single-Arm would be a stable arrangement in a full-size FF. The Italjet scooter uses this system and there are some reports of reasonable handling, although this particular example appears to suffer rapid 'king-pin' wear

One might consider it for an ultra-lightweight, a fuel efficiency challenger for instance.

Double-arm

The obvious development, this system has a long history in cars, with the pivots in line with the direction of travel instead of across the front as in a single-track chassis. Both the Citroen DS series and the pre WW11 Auto-union racers used this system. Turned through 90 degrees to become a single track system, the French-based ELF racers of the mid-eighties are probably the best known examples. As is usual in automotive practice suspension loads are taken on the lower arm and the upper arm is used to control the geometry, brake torque and provide torsional stiffness.

This system, in common with other single-sided systems, needs a wheel with considerable offset. The steering axis must be on the wheel centre and the brake rotor necessarily outboard of that, with the wheel disc furthest outboard. Making such a wheel is the major constructional task.

It is also impossible to reconcile the need for the top arm to be reasonable stiff with anything like adequate steering lock. This arm has to tuck in under the wheel rim and inevitably restricts lock to that side unless it sweeps through a wide curve that is clearly incompatible with stiffness.

The Elf racers, like other race bikes, were untroubled by the lock restriction and it seems that racing is the only real application for this simple system. However it is also worth noting the single brake disc, another disadvantage when considered for a heavyweight road going FF.

Arm-and-Wishbone

This is a logical development of double arm, where the upright is continued up, around the tyre and over the top to the wheel centreline. Here a simple top wishbone can be fitted giving excellent torsional stiffness, full braking and geometrical control and no restriction of steering lock.

Normally attributed to the English motorcycle engineer Tony Foale, this system is probably more familiar globally due its use on the Yamaha GTS cruiser.

Significantly this design uses a fairly exotic ventilated disc to overcome the single disc problem and this does cost more than two conventional discs. Also the thickness of the single lower arm needed to give stiffness under braking loads intrudes on the lock to the right hand side, opposite the arm. It would also seriously reduce footbox space, or increase overall width in an FF design with a 'forward seated' rider.

For the fabricator the design of the single lower arm presents a real challenge if all the stresses are to be satisfactorily contained. Again a deeply offset wheel disc is required. However, once these problems are solved this system offers everything that is needed in an FF front suspension system. Rider reports of both the GTS and other privately made systems of this type report no problems with asymmetrical deflection under braking, the most apparent weakness.

Double wishbone.

This is evolved from the oldest and most successful FF yet produced, the Sheffield Simplex Neracar. This used what may be called a double-sided single arm, where the arm, instead of ending at a king-pin pivot to provide steering at the wheel centre, continued through the wheel centre and back to the fork pivot across the chassis.

A component I call the “Barrel” in my designs, mounts on the king pin, to provide the steering movement, and the wheel rotates on bearings mounted on the outer diameter of the barrel.

This would have balanced the braking loads if front brakes had been fitted but confers no other advantages over single arm except for a symmetrical wheel. It appears to have worked reasonably well on the Neracar.

Many years later Jack Difazio allowed the king-pin at the wheel centre to rotate on the axle, and then controlled the geometry and brake torques by carrying an upright, bolted to the outer faces of the barrel, over the wheel and steering it from the upper cross piece using two track rods connecting to a transverse steering arm on the chassis.

This eliminated geometry change and permitted twin brakes. The double-sided lower arms are ideal for braking loads and apart from the upright, only the wheel assembly is steered.

Torsional stiffness still relied on the king-pin at the centre of the wheel and, as this rotated on the axle necessitating free play, stiffness on the original systems was poor. The use of one of these systems on the Banana highlighted this weakness and a top wishbone capable of taking torsional loads was fitted to the top of the upright, exactly as is done in Arm-and-Wishbone.

This system allows the Banana to be driven with all the enthusiasm of later Double Wishbone systems

The Difazio systems were fairly lightweight, with limited lock, each one turned from solid billet and nearly all a slight development on the last. Later versions of the 152 systems made had ball bearing king-pins rather than plain bearings and some had rubber bushes on the axle pivot instead of greased bushes.

Both Malcolm Newell and myself needed a simpler, more rugged system with adequate lock and lower costs for our respective FF projects. Bob Tait, Malcolm's designer and I both used spherical joints instead of a king-pin at the wheel centre, relying entirely on the top wishbone to keep the wheel vertical.

The Mk11 Voyager system shown on my site and used on FJ, is the most developed of these systems with ample lock, minimal component count and reasonable weight. It uses a common automotive ball joint at the wheel centre, from a similar application in a small car.

This system copes fine with heavyweight open-cockpits like FJ and might be considered the simplest in design terms, with symmetrical loads and ample stiffness. In common with the other systems major parts have to be designed and made and in this case the most difficult is the design of the axle, ball joint and barrel set at the centre of the wheel.

Application.

These appear to be the only systems to meet the requirements for an FF. For a cut'n shut, where the priority is simplicity, ease of conversion, quite possibly to study some aspect of FF performance apart from outright dynamic performance - aerodynamics, ergonomics, etc. - the Hossack system appears ideal. One might even find one of Norman's conversion on a suitable donor motorcycle. In any case fabricating such a system to use an existing frame and wheel is not particularly difficult.

Only Arm-and-Wishbone and Double-Wishbone offer practical solutions to full scale FFs and the decision as to which to use might be based chiefly on the braking requirements. Ultimately the double-sided system can carry more brakes.

Footbox clearance is also a factor. If the FF is intended to fit through a European doorway (a useful feature) and the rider is to sit forwards, as in FJ and the Comfortmax while enjoying normal steering lock, the room for the lower fork is severely restricted. FJ, with an overall width of 660mm, 38 degrees of steering lock each side and a 16” front wheel, has 35mm left on each side for the lower fork, 5mm of which is clearance to the footboxes. It is difficult to see how this could be achieved with the much greater thickness of a single lower arm.

This problem can be avoided by moving the foot position and GTS FF conversions are in progress.

Steering.

It is inherent in all these systems that the steering is entirely separate from the suspension.

On a motorcycle a lever is simply attached to the top of the suspension and steering passes through it. With either Hossack or HCS some part of the unsuspended assembly must be firmly attached to a steering system. This system, as noted above, must be as torsionally stiff as the suspension while maintaining a low weight and good packaging. Several options are available.

Single link.

The simplest way to transfer a steering load from the unsuspended wheel assembly to the suspended chassis, is a single link (“Drag link”) parallel to the vehicle centre line and exactly aligned with with the pivot points of the top wishbone of any of the suitable systems. At the outer end this is attached to the upright or Hossack fork unit and at the inner end to a lever pivoting on the vehicle centre line at the same angle as the steering axis itself.

This provides bump-steer free steering. This system was used on Malcolm Newell's Phasars.

Anyone familiar with the geometry of these parallelogram suspension linkages will be aware that the steering link can be designed to fit anywhere on the the upright, providing that the geometric requirements in terms of angles, lengths and pivot points, are met. This may be useful to improve packaging but a link below the top of the wheel needs to be wide based to clear the wheel on lock.

Double link.

The Voyagers use this geometric method but 'couple' the steering links so there is one on either side of the top wishbone. This may be seen as extravagant but is an effective way of resolving all the resultant loads in the steering, providing exceptional torsional stiffness with minimal weight increase. Steering input torques are high enough to make flex in single link pivots detectable from the cockpit. In theory a coupled system can use much lighter individual components, while maintaining a degree of fail-safe. In practice component availability tends to define actual size and weight.

Connecting to the rider.

Once a chassis-mounted pivot has been established, the actual steering control can be connected directly to it or a further linkage can be used.

A direct connection is obviously ideal and the Comfortmax shows that, if the rider sits well forward it is possible to connect directly to the fork tops without linkages of any sort. This certainly allows direct use of the chassis-mounted steering pivot of a Hosaack or HCS system.

The most obvious further linkages is to repeat the single, or coupled, drag link system further to the rear where a second pivot may be established to support the actual hand control. Such a link was used on the Quasar to connect to the rider to the motorcycle-style 'Earles Forks' and they were normally used on the Phasar series. The Banana, with coupled forward links, used coupled rear links in a similar way, with Jack Difazio exploiting the limited lock to use very light aircraft-style rod ends to minimise size and weight.

This approach has two major disadvantages, one is the further multiplicity of joints on the steering, tending to increase stiction and free play, the other the packaging problems imposed by the clearance needed for link rods and levers in an area already filled with components and constrained in size.

001 introduced a universal joint on the top of the chassis mounted steering pivot with a steering column taking steering inputs upwards and rearwards to the hand control. This feature continued through all the Voyagers up to FJ. This is very efficient in packaging terms and allows a steering control at an ergonomically, rather than geometrically, set angle.

However the universal joint must be converted from needle rollers to close tolerance bushes if irritating free-play in the steering is to be eliminated.

In any case development has shown that a forward riding position is desirable and that this allows a direct connection to the chassis mounted pivot in most cases. The hand control can be some distance behind the steering axis to bridge moderate gaps to the riders hands, and it seems that even FJ might achieve direct connection with more sweep back on the hand control. Excessive tiller effect can cause ergonomic and packaging problems but the exact limits of this compromise have not been identified. These details can be studied by looking at the Voyager engineering drawing, side view, on my site www.hightech.clara.net

From a design point of view it might be better to move the rider further forwards, rather than replicate the UJ/steering column. The Reliant engine prevents this in the Voyager series.

Hand control support.

The final point worth looking at in the steering system is the hand control support. This structural feature will have to withstand the braking loads imposed by the riders upper body at around 1.2G, plus safety factor. It will also be used to lift the vehicle back to upright when it falls over and this usage seems to have broken more support structures than anything else.
Finally, under the most extreme circumstances, it should collapse non-injuriously if the riders upper body crashes into it. This is considered in the paper on crash performance.

Summary

Now that BMW have adopted a front suspension system that meets some useful requirements apart from traditional appearance, it is possible that similar systems may appear from other manufacturers. The Mk11 Voyager system is rather cheaper to rebuild than the GTS arm-and-wishbone but there is no doubt that it could be further improved by a full production design process. At present it represents what can be done relatively cheaply, fairly simply.

Yamaha could certainly improve it - but probably not as much as Ford!

Whatever system a designer chooses for an FF design, the front suspension is one of the two key components. The other is the seat back, a part that can be lifted straight out of a car and this is considered in the paper on ergonomics. The front suspension cannot be lifted from any existing vehicle except the GTS and unless this, or a Hossack system is used, must be created from scratch or bought new or second hand from very limited prototype supplies. None of these routes are particularly cheap and anyone seeking to become an FF builder must first define and acquire a suitable front suspension.

Royce Creasey
Jan. 2005
Copy free for credit.