Estimating pneumatic trail & true scrub radius

[murpia]

This is not for FSAE, but I thought I'd post here as I'm confident of a useful discussion...

I need to estimate pneumatic trail and the true 'pneumatic' scrub radius in the absence of tyre data. Both should be related to the shape of the contact patch, so I'm planning to try to model that based on tyre diameter, width, camber and a range of inflation pressures. I haven't yet decided how much sidewall stiffness to include...

Therefore I'm going to assume:
1) that the true 'pneumatic' scrub radius is the distance between the steer axis and the contact patch area centroid.
2) that the pneumatic trail is ~50% of the distance from the tyre centre to the edge of the contact patch, at the centroid location (Milliken Fig 2.5).

How does this approach sound?

Regards, Ian

[flavorPacket]

I think it's dangerous to assume that the force distribution at the CP is centered at the centroid. This does not take into account thermal conditions at the CP.

Do you have the ability to put the tire on a car and drive it? I think that would be the only way to back out dynamic trail information.

(edited for spelling)

[Chris Aho]

If I was understanding Paul Haney's book correctly, you can neglect the sidewall stiffness as a major contributing factor to the shape of the contact patch. I suppose that would depend, however, on how stiff your tire carcass is and what pressures you are running.

ie. A soft carcass tire with higher pressure is going to have a CP that is influenced much less by the sidewall than a stiff carcass tire with low pressure.

Again, this is just my thoughts and I am not a tire expert so take it an grain of salt.


Also, the shape of the contact patch, the shape of the normal force distribution, and the resulting frictional force distribution is going to be dependent on speed as well as other things. Speed dependent because the leading edge of the contact patch has some inertia due to the fact that the tire is rotating and the rubber has some initial velocity in the down direction. Again, I would think this would be pressure dependent because if you are running a low pressure and the longitudinal length of the contact patch is long with respect to the size of the tire then the leading edge of the CP will have greater component of a downward speed because the tire is doing less of the slowing it down before it hits the pavement.

Again, these are thoughts of my own and make no claim that they are true or significantly affect the data you are trying to get.

[murpia]

Originally posted by caho:
If I was understanding Paul Haney's book correctly, you can neglect the sidewall stiffness as a major contributing factor to the shape of the contact patch. I suppose that would depend, however, on how stiff your tire carcass is and what pressures you are running.

Also, the shape of the contact patch, the shape of the normal force distribution, and the resulting frictional force distribution is going to be dependent on speed as well as other things.
I have a copy of Haney, I'm afraid I don't agree that sidewall stiffness can be neglected... At least not on the tyres I'm concerned with which are very wide and very low profile. Hence my modelling idea to include sidewall stiffness.

As far as driving the car on these tyres, the driver comments are that the steering is too heavy. Looking at the steering & suspension design the only realistic option we have is to reduce steering trail, but my concern is that the reduction we can package is insignificant compared to the pneumatic trail... Therefore we might have to add some power-assist.

Regards, Ian

[Chris Aho]

Is it heavy as the car gets faster or heavy all the time?

[murpia]

Originally posted by caho:
Is it heavy as the car gets faster or heavy all the time?
Yes, because the car has downforce...

Regards, Ian

[Buckingham]

Is it possible to measure steering effort and subtract out the mechanical trail effects to give you your pneumatic trail estimate?

[flavorPacket]

Originally posted by dhaidinger:
Is it possible to measure steering effort and subtract out the mechanical trail effects to give you your pneumatic trail estimate?

That's also what I was wondering above. If you have a car, why spend so much time on the model when you can get the data from physical testing?

[murpia]

Originally posted by flavorPacket:
That's also what I was wondering above. If you have a car, why spend so much time on the model when you can get the data from physical testing?
Let's put it this way, car mk1 has too much steering effort, but doesn't run any more and we have no data on column torque or from strain gauged trackrods.

Car mk2 gives us the opportunity of some redesign (subject to budget) so we need to model the effect of design changes as best we can. That's a pretty standard real-world situation if you ask me, we need to get the design right and verify it with testing, not just say 'lets try this' when we get to the track.

Regards, Ian

[flavorPacket]

That does clarify the situation. So you're not able to make development uprights with adjustable geometry?

I see the difficulty now: you don't have a quantified target, all you know is that it's too heavy.

Can you put the tire on a similar car, do the manuevers, and get the ptrail data from that? Then you know the tire, you know a little bit of the mk1 behavior, and you can choose your new geometry.

[Buckingham]

Is it possible to extract the data from your driver's head by putting him through a simple test to correlate "heavy" to N*m, and use his estimate of where mk1 was, and your calculated values of mechanical trail, to come up with your estimate?

I would suggest doing whatever you can to approach the problem from three ways (one of them being the initial estimates you are making regarding contact patch) and hope that two of the estimates agree with each other.