Trials Australia Forums

With the tendency to modify classic/twin-shock trials bikes towards more modern geometry I’ve often wondered about many such modifications on these old bikes. I don’t doubt modern geometry on modern bikes, but they are so different in so many ways, how much ‘piecemeal’ geometry can we transfer to twin-shocks with benefit?

When there is so much interdependence/overlap between different parameters, if you want to change one, what other changes does it bring with it? For better or worse?

So I did some research… & found out how much I didn’t know! (As we do) And this is the result. It’s not meant to be the last word, so feel free to join in. You may find some surprises in store. I did!

Unfortunately most of the info out there is for bikes at speed – roadies or racing bikes – which may or may not be relevant to the slow motion of trials. But far & away the best material is by Tony Foale; author, racer, bike-builder & professional engineer:

http://www.tonyfoale.com/" onclick="window.open(this.href);return false; or http://www.tonyfoale.com/book/" onclick="window.open(this.href);return false;

He is lucid, readable, practical & informative – the doyen of this field with a wealth of experience in motorcycle design & construction. And trials bikes at least get several mentions in his books.

Foale also has a trail/rake/offset program that is downloadable free from his site, which is extremely useful for those considering steering mods done properly.

John Robinson (‘Motorcycle Handling: Chassis’) is also readable, but is more a general overview that I found of limited use. Beyond that there are several more academic works (G Cocco, R Sharp, V Cossalter) with hi-brow maths on almost every page, but of little/no use to the average trials rider.

But first a qualifier or two: i) different steering effects may be more/less relevant to some degree depending on your riding style: non-stop or hop-stop, ii) I’m ignoring gyroscopic effects which are minimal at slow motion, & iii) I’m not infallible.

In other words, the aim is to gain a clearer understanding of factors affecting steering on twin-shocks in trials, ... as best we can.

To be continued (TBC)

Steering is absolutely fundamental to a bike’s behaviour: “The way in which a bike steers is the focal point of its handling & stability” (John Robinson, ‘Motorcycle Handling: Chassis’)

Rake, trail & offset are simple concepts which are reasonably well understood, although their full effects & interdependence are often not. Weight distribution, including centre(s) of gravity (C of G), & wheelbase also play significant parts.

Let’s start with some basics. A picture is worth a thousand words:
All diagrams courtesy Tony Foale; used with permission (http://www.tonyfoale.com/book/" onclick="window.open(this.href);return false;)

Trail is the distance that the centre of the tyre’s contact patch trails behind the steering axis. It (not rake) is the real factor which causes the self-centring (self-correcting, stabilizing) effect when in motion.

It works because anything that causes the wheel to steer left or right also moves the contact patch to the side of the steering axis so the friction force (from tyre on road) acting on the contact patch produces a torque about the steering axis forcing the contact point back behind the steering axis. ie the self-centring effect

The longer the trail, the longer the ‘lever arm’ & the greater the self-centring torque produced about the steering axis, ie the greater the self-centring effect.

It is perhaps the most critical parameter in steering: “To tune the handling, trail & weight distribution are the important things; everything else is secondary” (Robinson).

Although trail is related to rake, offset & tyre diameter, it is quite possible to have trail with zero rake & some offset (ie rear-set), or with zero offset & some rake.

It is also possible to have Negative Trail - ie. if that contact patch were in front of the steering axis - which is quite undesirable to put it mildly! But it does happen:

TBC

JC I can't wait till you get to the bit about why some bikes have angular offset, and some don't

David, none of the books mentioned angular offset in the triples, unfortunately. What I've done is just comment on where they're of benefit as different issues come up throughout the article - much of it as we have discussed before.


Offset: The distance the axle is offset from the steering axis gives us the desired trail. Typically trail is inversely proportional to axle offset; ie more offset gives less trail. But perhaps just as important is the offset of the centre of gravity of the steering mass, for that is the point through which gravity/weight & centrifugal force act to affect steering.

These offsets may or may not be the same. Yokes that ‘kick-out’ the forks a couple of degrees & sliders that have the axle offset in front of them reduce the offset of the C of G of the steering mass compared to parallel yokes & sliders with in-line axle, which is a good thing (see later).

Trail = wheel radius x Tan rake angle – offset/Cos rake angle

Rake - ie the angle of the steering axis measured from the vertical. Yokes (triples) that ‘kick-out’ the forks a couple more degrees than the headstock have many benefits but they do not change rake because they do not change the angle of the steering axis.

It is generally thought that the more rake you have, the more straight-line stability & the ‘heavier’ feel to the steering. That’s only partly true. Read carefully.

Foale says, “The basic reason for rake is less easy to explain than for trail because it is just possible that we don’t need it”! Sounds outrageous, but he has both the physics & experiments to justify his assertion.

Effects of rake:
Perhaps the best known effect of rake in trials is that steeper rake gives quicker or ‘sharper’ steering. This is because with less rake, the angle you steer the bars corresponds more closely to the angle the front wheel steers. Consider zero rake (vertical axis): the angle you steer the bars is exactly the same as the wheel. But for 90deg rake (horizontal axis), it doesn’t matter how much you ‘steer’ the bars left or right, there is zero degree of steer at the wheel; the front wheel just leans instead.
All diagrams courtesy Tony Foale; used with permission (http://www.tonyfoale.com/book/)

It’s a simple relationship of trigonometry: the steering angle of the wheel = the steering angle of the bars x the cosine of the rake angle. (Cosine 0deg = 1; Cosine 90deg = 0)

That is why the steering is more precise or ‘sharper’ with steeper rake – it is more direct, ie “quicker”. eg At 0deg rake the wheel steers 100% that of the bars; at 26deg rake it is 90%; & at about 36deg it is 80% & at 90deg it is 0%.

Or to put it the other way round; the more rake, the more vague the steering. Although this is the best known reason for minimizing rake on trials bikes it is not the only one.

Wheelbase plays its part here too. Obviously a shorter wheelbase allows tighter turns, which in a sense means quicker steering. “The path’s radius (ie of a bike when cornering) is directly proportional to wheelbase” (V Cossalter, Motorcycle Dynamics)

You can see in that diagram that a longer wheelbase requires a greater angle of steer for the same centre of curvature & the front wheel follows a larger radius of curvature (wider path).

Now to delve a little deeper (still considering effects of rake).

Apart from the self-centring effect usually associated with rake, it also has an opposing potentially destabilizing effect which can be considerable at rest &/or slow motion & therefore is more significant in trials.

We all know how, when you lean the bike to put it on the side-stand, the steering flops to a certain point of equilibrium (ie a certain angle of steer; see later why it stops there). This is a consequence of rake.

To visualize the connection to rake, consider an upright bike with zero rake (ie vertical steering axis). It does not matter how far you steer it, or how much trail you have, the headstock does not rise or fall. Now consider it with 90deg rake (horizontal steering axis). If you ‘steer’ the bars left or right the headstock falls considerably - by as much as the radius of the tyre/wheel at 90deg of ‘steer’. Hence it is rightly called head-drop.
All diagrams courtesy Tony Foale; used with permission (http://www.tonyfoale.com/book/)

The flip-side of this is that, at slow/no motion on a bike with typical rake & trail, the weight on the front of the bike wants to force the headstock down, steering the wheel in the process, especially when you start to lean. This is another undesirable state for trials as it can be quite noticeable at low speed & can upset balance or catch the rider unaware, even though friction slows it down somewhat at/near standstill. It is another reason for minimizing rake in trials.

It is added to by what we might call flopping-in. At low speed & with lean, the gravitation force of the weight of the steering mass acting down/inward through its C of G exceeds the minimal centrifugal force acting outwards & so produces a torque about the steering axis turning inwards.


So several things make this head-drop effect feel stronger - more weight on the front of the bike (laden & unladen), greater steering mass & more offset of its C of G. They don’t affect how far the headstock drops & the wheel steers (which are directly related to each other as we shall see), but do affect how strongly it is felt & the force required to raise the headstock again, ie to return steering to straight ahead. They are not the primary cause of head-drop; rake is, but they do exacerbate it.

Thus kicked out yokes (angular triple clamps) & forks with offset axle reduce how strongly this is felt to some degree because they reduce the offset of the steering mass C of G. But they don't address the fundamental cause - ie too much rake. Some rake is a benefit because, up to a point, rake gives you trail with little/no offset of the axle, so the offset of the steering mass C of G & the flop-in effect are minimized in such cases. So is steering inertia.

This head-drop effect is not necessarily all bad because if the geometry is right, it can help steer the bike into the corner ‘automatically’ – the self-steering effect - so that you can "steer with the footpegs". It's not the main self-steering effect though. That comes from trail as we’ll see later.

TBC

Effect of Rake continued:

There is another problem with rake; perhaps the worst of all. As we have seen, when you turn the bars, rake transitions into camber so that at 90deg steer, a bike with 26deg rake (eg) would effectively now have 26deg camber & zero degree rake (in the direction it is steered). Worse still, whatever axle offset you started with now puts the tyre’s point of contact with the ground that far ahead of the steering axis.
All diagrams courtesy Tony Foale; used with permission (http://www.tonyfoale.com/book/" onclick="window.open(this.href);return false;)

So positive trail has reduced to zero then turned into negative trail as you steered. Hardly ideal! You can feel this transition when riding as the front end wants to turn-in further or tuck-under when the trail goes negative.

If the bike had more trail the transition point (to negative trail) would be closer to 90deg steer where it’s less noticed. But if it had zero rake, the trail would remain constant no matter how far you steer the bars right or left.

So the steeper the rake the less this reduction in trail as you steer & the closer you get to 90deg steer before trail goes negative. Since trials bikes so often operate at extreme angles of steer, this is perhaps one of the foremost reasons for steeper rake on modern bikes. But it must be remembered that in designing for steeper rake, offset must also be adjusted so that you still have an appropriate amount of trail for the steeper rake. If you don't, tucking under will likely get worse because of insufficient trail.

It is possible at the design stage, by minimizing rake & optimizing trail, to ensure that trail does not go negative at any angle of steer. See Fig 3.9 below for a road bike (89mm trail, 61cm dia wheel). At a typical 27deg rake, trail reaches zero at 67deg of steer & beyond that, goes increasingly negative. But at 20deg rake it’s delayed till almost 80deg of steer. And by the look of it (if you extrapolate the graph) at 15deg, trail wouldn't go negative at all on this bike.

But please note: They all start with 89mm trail at zero steer angle. ie the offset has been adjusted to maintain the same trail at the different rakes



However with most motorcycle set-ups, including twin-shock trials bikes, the self-centring effect disappears completely at some angle of steer & the tuck-under effect gets worse the further you steer beyond that transition point (ie the closer you get to 90deg of steer).

Ironically, the ‘quick fix’ for this is widely regarded as to “give it more rake”, perhaps because it’s easy to do; eg fit shorter shocks. That works to a degree, because of the extra trail you get with more rake. If you start with more trail the point of transition moves closer to 90deg steer where it’s encountered less often, but it is really a case of one step forward (more trail), half a step back (more rake, which is counterproductive). Ideally what’s required is more trail with the same rake or (better still) less rake with the same trail.

When modifying older bikes, especially those in the lower margins of trail, reducing rake while keeping the same yokes is often fraught with this problem. The reduction in rake would improve it, but the associated loss of trail seems to have more noticeable effect & the overall effect is that tucking-under becomes worse (ie occurs at lesser angle of steer).

This tucking-under effect is probably felt more by non-stop riders than by hop-stop riders

TBC

Keep up the good work John.
Just thinking if they mention anything about what effect cornering lean has on the theories, also the effect on handling with step down hills and drop off's.

Off topic slightly, I did inadvertently try a stepper rake angle on my 158 bultaco one time after hitting a vertical bank at full throttle in 1st gear. The end result had me in half a meter of water with the bike on top of me. End result was about 60mm shorter wheel base and 4 degrees less rake. It looked a lot like a modern bike, it was still rideable but the handling was terrible. My recomendation after that episode is leave it stock unless of course it's a one of those make believe Jap trails bikes.

Thanks George. Yes we can certainly look at effect of lean. I'll just get to the end of this segment on Effect of rake.

With regard to effect on handling, did you mean "step downhills" or "steep downhills"?

JC1 wrote:Thanks George. Yes we can certainly look at effect of lean.

Did you also mean "step downhills" or "steep downhills"? (for effect on handling)

It was steep downhills. But a step down could leave the front wheel on flat ground with the rear well above the front. That would put the contact point a long way forward giving quite a lot of negative trail and rake.
Doing a full lock turn on flat ground I would guess a lean of around 25 degrees would be normal.