Tyre Peak MZ and Peak FY

[CWA]

Hi guys,

At the moment I'm trawling through some TTC data. In the data I spotted what I've now come to understand as normal tyre behaviour, the slip angle where peak MZ occurs is much lower than the slip angle where peak FY occurs.

I've also been reading through Peter Wright's book, F1 Technology, which suggests that the driver uses the peak in steering torque (peak MZ?) to judge how close to the front axle grip limit he is (peak FY). Now, from the diagram included by Wright, the driver finds the peak in MZ, by noticing the drop-off as he increases steer (and front tyre slip) angles, and then keeps the wheel around that position.

First of all, is this correct? Is this how driver's judge how much front axle grip there is, and how close to the maximum they are? If so, how can you design a steering system to align the total peak MZ's to occur at the same angle as the peak FY? Surely it's impossible, due to the nature of pneumatic trail in tyres? Can castor trail or ackerman angle be used to tune this?

Secondly, if the driver does not use this method of seeking the steering torque peak to find maximum grip of the front axle, how does he do it? Is there another way of sensing which steering wheel position gives the most grip? Or does it just come down to experience, and recognising how fast he had previously been going around a certain corner, and recognising when he is doing it faster the next time?

My main concern is that certain inexperienced drivers of our car will keep the steering angle at the highest torque position (peak MZ) believing it to be the slip angle with the highest FY, whereas peak FY is unintuitively a few degrees further. So they will constantly be driving around using only say 80% of maximum front cornering force.

In short; how can the steering system be modified to reduce this effect, is altering the steering system to reduce this effect common practice in steering system design, and if not, what is the alternative solution? Or am I dreaming up a problem out of nothing?

Thanks in advance!

[Owen Thomas]

From what I understand about tires and driver controls, your statements above are correct. You also answered much of your own questions it seems http://fsae.com/groupee_common/emoticons/icon_smile.gif.

There is no way to reduce the effect of peak steering torque from pneumatic trail occuring before peak lateral grip, since this has to do with the tire properties and not the steering system (aside from somehow creating a tire with different properties, of course). You can however effectively eliminate the symptom as seen by the driver by overwhelming it with aligning torque from mechanical trail and scrub radius. Basically if you increase your steering effort through geometry, the driver won't be as sensitive to the pneumatic trail effects. Typically increasing trail or scrub radius by enough to do this will warrant power steering though, which even further removes the driver from what's happening at the tires. Oh, and this is not common practice in performance vehicles (no reference, just observation).

My opinion is that you should inform your drivers of what's happening, and let them learn and adjust. This is generally good advice anyways, and gives a reason why "good" drivers most often have some knowledge of vehicle dynamics. From talking to others and from my own experience, pushing yourself past that initial feeling of tire slip and realizing that there is still grip to be had is one of the hardest parts of getting faster.

TL;DR: You can't change the tire, but you can change your steering forces through geometry, altering driver feels. Would not recommend. Inform drivers to grow a pair and test in wide open space.

[CWA]

Originally posted by Owen Thomas:
There is no way to reduce the effect of peak steering torque from pneumatic trail occuring before peak lateral grip, since this has to do with the tire properties and not the steering system (aside from somehow creating a tire with different properties, of course). You can however effectively eliminate the symptom as seen by the driver by overwhelming it with aligning torque from mechanical trail and scrub radius. Basically if you increase your steering effort through geometry, the driver won't be as sensitive to the pneumatic trail effects. Typically increasing trail or scrub radius by enough to do this will warrant power steering though, which even further removes the driver from what's happening at the tires. Oh, and this is not common practice in performance vehicles (no reference, just observation).

What you say here is interesting, and adding in lots of mechanical trail had crossed my mind (hence why I initially mentioned it). But thinking about it more, as you mention, I'm not sure it will help the situation, as mechanical trail just seems to get rid of the noticeable peak in MZ entirely.

RCVD talks about this briefly; IIRC any more than something like 50% of the overall steer torque being created by mechanical trail is not advised, as otherwise it 'dillutes' the changing MZ (from the pneumatic trail migration) felt by the driver; he finds it harder to notice the peak. And even then, it just adds torque on top of that, it doesn't make the peak MZ occur at different slip angles, right?

In my mind, ackerman might be a tuning variable here, where using the combination of the two wheels at different steer/slip angles can somehow get the peak MZ and peak FY's to align when at a certain SWA. I've been trying to do some simulations to find out whether changing ackerman could achieve this, but it's proving difficult as I haven't fully grasped tyre data, nor MATLAB, so I'm wondering if I'm not going down a pointless road..

BTW, what does TL;DR mean?

[Owen Thomas]

And even then, [mechanical trail] just adds torque on top of that, it doesn't make the peak MZ occur at different slip angles, right?
Correct. In reference to the 50% benchmark, I believe this is subjective. It is usually a better idea to quantify your total steering effort (torque) at maximum and on average and make sure your design meets driver requirements like amount of feedback and steering difficulty.

Ackermann is a very good tool for testing these sorts of things, you are right. If at all possible, it would be a good idea to have extremely adjustable ackermann settings so you can play around with the different effects in real life and get driver feedback (and skid pad/lap times). Keep in mind that your steering torque is also correlated to the distance of your tie rod pickup to the king pin axis, so changing your ackermann will very likely change the steering weight.

It is possible to make your car more "drivable" with ackermann, like you mentioned. For example, on the tire data that I have seen, it is possible to have a setting that puts your outside tire at peak My and your inside tire at peak Fy at some amount of lateral force. For all intensive purposes, this decreases your total lateral grip, but would also make it easier for the (rookie) driver to find said peaks. Of course, there are a ton of other things affected by ackermann, so it all depends on what your goals are...

TL;DR means Too Long; Didn't Read. Internet for "In conclusion..."

[BillCobb]

The occurence of peak Mz before peak Fy is true for just about all tires. Actually, a driver can not sense Mz directly because there are things like tierods and steering gear and often U-joints and a steering wheel rim in the stream of things. A driver can only sense the steering wheel rim forces resulting from the downstream stuff. Besides, there are usually two tires in the steering control system operating at different levels of slip, load, camber and sometimes even pressure. Need I mention that the use of 'pneumatic trail' is a misnomer because the tire'z Mz is not produced by sideforce and is not at all proportional to sideforce where it really counts (at a distance traveled of 0+ units after steer input. What typically matters to a 'good' or 'great' driver is the GRADIENT of NET tierod load by lateral acceleration and this can easily be manipulated into a rim force gradient. Tierod loads are usually measured because these parts are seldom changed during a practice test or evaluation, the individual wheel contributions can be extracted (sum and difference). Combined with a steering wheel torque transducer, they can isolate steering gear issues (especially true with power assisted steering) and the wires are hidden from observation by other teams (hey whatcha doin' there?).

That being the case, you can adjust the relative position of peak TRLG to net peak Ay (there are two tires, remember ?) by the manipulation of caster, caster offset and kingpin inclination. Each has a 'feature' that can augment or attenuate the steer angle or 'g'-level the driver uses to feel the gain of the vehicle. Because both rear tires also contribute a finite Mz to the rigid body degrees of freedom, the manipulation of rear load transfer via steering inputs can be brought into the act.

A good simulation can easily produce tierod loads and lateral gs. Post-processing can produce the gradients and some front end adjustments able to manipulate the crossover of the gradients (from positive to negative) are all you need to make. Without the tire data, though, you will be lost.

I believe I posted an example of how this all works (simplified of course) on the TTC forum. Its in the form of a Matlab driven primitive vehicle with a non-linear TTC tire and some influence coefficients relating gs to tierod load. Fill in the gaps and watch what happens. The tire transient responses aren't included (I don't mean "relaxation"), but that's for another day. If you had a driver model with road feel-feedback, they would be required.

[Z]

Originally posted by CWA:
... how can you design a steering system to align the total peak MZ's to occur at the same angle as the peak FY?
...
Can castor trail ... be used to tune this?
CWA,

Indeed, you have answered your own question.

Firstly, KISS! Assume Scrub-Radius/Castor/KPI/etc. all = 0, since including these now might just confuse you.

Next, do some plots of;

(Mz + Fy x Trail) vs Slip-Angle,

for several different Trail values (= Mechanical-Trail, = Castor-Offset...). Also overlay plots of Mz vs SA, and Fy vs SA.

For moderate Trails, say 1cm to 3cm, I reckon there should be enough bumps on the curves for even amateur drivers to feel "peak-Fy".

If possible, post such plots here. They should be helpful to other FSAEers.

Z

(PS. Driver's also use other means (eg. seat-of-pants) to sense when they are close to the limit. Note also that the "warning" should come BEFORE the "bad things" happen.)