6 replies to this topic

[Paolo]

At a Claude Rouelle seminar, a couple years ago, we had a discussion about the usefulness of "soft" springs on a race car.
Soft springs upset a car's behaviour; this is even worse if ground effect comes into play.
When a street car is tuned for circuit racing, one of the first steps is to increase the spring stiffness.
Now one wonders why there is an upper limit to this stiffening.
Some reasons are obvious : fatigue (of both driver and chassis), limited stiffness of chassis and tyres, that would flex in place of the suspension over a certain level.
Anyway Rouelle added another reason: the tyre friction coefficient decrease with spring stiffness.
He quoted a 30% decrease as possible, surprising even a momentarily attending Gianpaolo Dallara.
Now, a tyre has a nonlinear load sensitivity; this would explain why the lower the load variation, the higher the friction coefficient.
I' ve found references to a Dynamic Load Coefficient ( DLC ), and this is estimated, on streets,. to be between 0.1 and 0.3 ; it is the standard deviation of dynamic tyre force divided by the static tyre force.
Yet, I'd like to find a way to calculate the loss for a given stiffness and surface roughness.
It is very easy to do this if I assume the vehichle is divided in indipendent "front" and rear" part, both of which travel at constant height from the ground; then, knowing the spring coefficient I know the force in every moment.
Yet, I'd like to go beyond these simplifying assumptions, but still have an "easy" mathematical model.
Ideas, info ?

(Knowing the surface roughness of circuit asphalt would help)

[Greg Locock]

Ideas: use ftire (www.ftire.com) to model an acceleration event over a 3d road surface.

Surface profile of roads is well represented in the Noise and Vibration literature by a straight line of amplitude vs frequency, probably on log-log axes, with a negative gradient. I cannot rememebr much more than that. It would be relatively easy to use that to generate a road profile.

Although we measure 6 dof tire forces we don't do so during limit handling events. When braking the ABS cuts in. So I don't have any useful data to derive any hints from.

In the real world you could just try plotting your real world max longitudinal acceleration vs spring rates.

You could try putting very lumpy sandpaper on a Flattrac machine!

[Ben]

Originally posted by Paolo

Anyway Rouelle added another reason: the tyre friction coefficient decrease with spring stiffness.
He quoted a 30% decrease as possible

Is that tyre friction decrease with increase in wheel rate?

Originally posted by Paolo
Now, a tyre has a nonlinear load sensitivity; this would explain why the lower the load variation, the higher the friction coefficient.

I would look at this differently. What causes tyre load sensitivity? I've never read a good explanation of why the grip should be proportional to load. Tread models such as those of Trevorrow and Hallum predict that load sensitivy shouldn't happen, but also postulate two alternatives as to what does cause it. Neither of these postulates has been tested. Trevorrow thinks that Hallum's is unlikely but hasn't tested his own fully.

It is possible that nonlinear rubber draping over road asperities influences the grip generation with a fluctuating load, but again no one has ever objectively modelled or measured this phenomenon.

Higher load fluctuations using stiffer springs can actually get you more grip quickly from a hard compound if Qualifiers are banned. Greater load fluctuation meaning less grip is only true for identical thermal conditions of the tread rubber.

Ben

[Greg Locock]

/an aside - what is tire chatter ? - Team Banal drivelled on about it all the time during the Shanghai race/

So far as I know the load sensitivity effect is due to the interaction between the different mechanisms by which grip is developed. With zero normal load some grip is developed by mechanical cogging, as more vertical load is added (a) work is done dragging the rubber past the road and (b) the engagement of the rubber into the road is deeper.

Having said that I haven't read anything more advanced than Dixon.

[Ben]

Originally posted by Greg Locock
/an aside - what is tire chatter ? - Team Banal drivelled on about it all the time during the Shanghai race/

So far as I know the load sensitivity effect is due to the interaction between the different mechanisms by which grip is developed. With zero normal load some grip is developed by mechanical cogging, as more vertical load is added (a) work is done dragging the rubber past the road and (b) the engagement of the rubber into the road is deeper.

Having said that I haven't read anything more advanced than Dixon.

No such thing as "tyre chatter". Chatter is just a resonant vibration of the unsprung mass. Given that frequency is proportional to stiffness and inversely proportional to mass one needs to change one or both of these parameters in either the tyres or suspension. Suspension includes the lateral bending of the frame, this is why Mr Rossi has a new chassis this weekend. There aren't many classically trained engineers in bike racing though so the voodoo persists.

Ben

[GregorV]

There is one effect that I haven't seen mentioned in the thread, namely the relaxation length of the tyre. To generate grip, a tyre needs to deform slightly, and for the deformation to occur it must travel a certain distance, namely the relaxation length. If the tyre for some reason looses the normal force acting upon it, it would quickly revert to a lower grip and therefore a lower deformation. If the normal force is the reapplied, it takes a certain distance travelled for the tyre to again the initial grip. If the hopping frequency is comparable to this relaxation length or smaller, it can be seen that the tyre grip will be reduced a lot. Softer springs will mean that the variations in normal force will be less and therefore the overall grip will be larger.

[Goran Malmberg]

1
Soft springs upset a car's behaviour; this is even worse if ground effect comes into play.
2
When a street car is tuned for circuit racing, one of the first steps is to increase the spring stiffness.
3
Now one wonders why there is an upper limit to this stiffening.
4
Anyway Rouelle added another reason: the tyre friction coefficient decrease with spring stiffness.
He quoted a 30% decrease as possible, surprising even a momentarily attending Gianpaolo Dallara.

This is a hard question to make an easy answer. Especially since I am Swedish and don’t
speak English to good. But Ill give it a try.

Tire coefficient of friction (µ) is decreasing with load in a falling rate. But a load increase does always result in a bigger friction number. Let us leave the rubber temperature affection of µ values to concentrate on load effects.
This is the main reason why larger tires produce better cornering power as they result in lower loads per square area of tire contact patch to ground thereby raising µ number. Friction is a product of µ*load. This is why a stiffer bar creates less grip for the axle in question as it produce a higher load per tire contact pavement area on the outside tire.

A softer spring setting tend to even out contact patch loads thereby rising the average friction over a distance. However, an initial high contact patch load will raise the grip momentary, for example, in a turn in situation. While sustained high load outer front wheel will make the car to loose grip and become understeered.

2
We need a suitable spring rate for the car (and track) in question. Tire is the item that create grip, not spring stiffness. So, (2) it is the forces from tire grip that governs the spring setting of the car. Low cgh is essential, but it will limit wheeltravel wich is also a factor. On a bumpy track we may need to raise the car in order to get more wheeltravel and use softer springs. If the track is flat there is mainly camber change (depending on A-arm geometry) that limits wheeltravel in roll. And we may use stiffer springs in order to reduce load variation within each contact patch.

1 and 3
The best µ number is had when all of the tire contact patch area is having equal load per area.
Both in between the tires and within each tire contact patch. So, we must figure out what will upset this. Softer springs will smoothen out roads and dynamic forces but it makes the car to roll and squat-dive thereby alter (1) the camber situation. So of course, (3) there is an upper limit of spring stiffness as well as there is even a lower limit for best grip. But not independent from a lot of influence factors.

4
I cant see other than that this must be a dynamic situation or different road levels for each tires. A specific rubber area at a certain (temperature) and load against a road surface does not change becourse the load is created by a spring or a dead weight.

Also, µ number for tires is almost impossible to get from a tire manufacturer, so there is no easy calculation to do.

Don’t know if this clears anything up….
Regards
Goran Malmberg