Taking Advantage of Bump Steer?

[Gyro]

Hey all,

I had an idea the other day that I wanted to ask you all about. I was somewhat concerned about the angle of the outer front wheel in a hard turn, since I plan on using a (mostly) Ackermann geometry with some static toe-out. I was picturing the outer wheel understeering in the harder turns, and not getting the right slip angle to generate the necessary cornering forces, especially with the weight transfer involved.

So I had an idea, namely that bump steer could be harnessed to toe the outer front wheel in when the car takes its tightest turns. Everyone always seems to be terrified of bump steer, and I understand why, but couldn't it be potentially useful? Or would the magic Ackermann geometry take care of the problem? Or is there no problem at all, and I've just thought of this all wrong?

I'd appreciate any input. Thanks!

[Crispy]

Hi Matt,

At the simplest level, if the outside wheel toes in during bump, won’t the inside wheel toe out in droop? Assuming the travels are similar when rolled, wouldn't you just be adding steering to what was originally requested by the driver (not significantly changing the relative inside and outside slip angles)?

To add a little more complication. If your roll centers are above ground, when the chassis rolls it will jack up as well and the inside tire should move (and change toe) more than the outside tire. In this case I would say that bump-steer has a larger effect on the inside wheel's steer angle.

More directly to your comment, what makes you say the outside tire under-steering? In a hard corner, shouldn't the outside tire be creating a majority of the lateral force and more or less setting the slip angle at that end of the car? My though on steering geometry (at least concerning steady state) is that the outside tire is doing most of the work and the inside tire is pretty much along for the ride (and hopefully does not get in the way).

I hope that this made sense and was helpful.

[Dr Tron]

outside wheel in a turn is at a slight less angle then the inside wheel and while the most load is on the outside tire that is not to say the inside is insignificant, furthermore using bump/roll steer adjustments such as shimming the tierod ends or rack you can easily change you understeer or oversteer

[Crispy]

Originally posted by Kyle "steering and suspension":
outside wheel in a turn is at a slight less angle then the inside wheel

Doesn't that depend on the steering kinematics (& compliance)?


Originally posted by Kyle "steering and suspension":
and while the most load is on the outside tire that is not to say the inside is insignificant,

I agree that it can be significant, but it depends a lot on the car. We have a very narrow car with a relatively high CG and a stiff front suspension, so mid corner, the front inside really does very little and sometimes nothing (it does a bit more now that we have aero). During corner entry I could imagine it being more significant especially as toe changes under braking.


Originally posted by Kyle "steering and suspension":
furthermore using bump/roll steer adjustments such as shimming the tierod ends or rack you can easily change you understeer or oversteer

I don't doubt that you can play with bump steer to tune the car and make it quicker. My personal feeling is that it is something that should be left alone (minimized) just for the sake of simplicity until the rest of the car is well sorted out. I think there are many other pieces of the puzzle that are more valuable than bump steer.

[PatClarke]

Isn't the outside front wheel steer angle adjustable in real time by the driver?

Pat

[Sormaz]

Originally posted by PatClarke:
Isn't the outside front wheel steer angle adjustable in real time by the driver?

Pat
Valid, but I believe what he is getting at is harnessing bump steer such that in 'low speed' cornering (when steering is primarily geometric) you have both wheel normal vectors pointed at your steering center (ackermann geometry, as I understand it) but in high speed cornering with weight transfer (more importantly, body roll) introduced, the bump steer would allow you to develop a higher slip angle on the outside tire than the inside.

@Crispy: I see no reason why it should not be possible to design a linkage that would toe in with both bump and droop

Not saying I am convinced of this idea, though it is curious.

[exFSAE]

Bump steer is certainly a big tuning tool for handling. Think you'll find that the effects on transient handling blow away anything you'd notice in some steady-state thing.

Besides... (a) didn't you see the other thread? If you have 100% load transfer (or near it) then who cares about the inside tire? Gaining or losing a pound of grip there.. there are bigger fish to fry.

(b) With how outrageously (borderline silly) stiff so many teams run their cars, if you have no suspension travel then bumpsteer isn't going to do shit anyway.

[Dr Tron]

true, most suspension setups in these comps are soo stiff that you don't have enough suspension travel to make bump steer a contributing factor. Aim for neutral on bump steer and if you total suspension travel 1 in then your looking at .033in of tie rod end displacement, if you have two inches of suspension travel its .133in of tie rod end movement (looking only on one plane considering suspension travel) thats between one and six percent of your total steering input considering 2 in of rack travel (even less at the rod ends) which might as well be your system slop

[Crispy]

Originally posted by Sormaz:
@Crispy: I see no reason why it should not be possible to design a linkage that would toe in with both bump and droop

It certainly is possible, but I don't think it would be practical. Using a conventional linkage, there would be zero bump steer in its neutral position and it would need to gain bump steer with travel. Getting a significant angle change would require either large travels or a steering link that has a very different length relative to your control arms (or both).

This is fundamentally how our steering works (zero bump steer at neutral position), but we have relatively small travels and we choose the steering link length to minimize bump steer.


Originally posted by exFSAE:
Besides... (a) didn't you see the other thread? If you have 100% load transfer (or near it) then who cares about the inside tire? Gaining or losing a pound of grip there.. there are bigger fish to fry.

(b) With how outrageously (borderline silly) stiff so many teams run their cars, if you have no suspension travel then bumpsteer isn't going to do shit anyway.

I agree

[flavorPacket]

Originally posted by PatClarke:
Isn't the outside front wheel steer angle adjustable in real time by the driver?

Pat

Isn't the driver a fool who should rarely be trusted to do anything?

Put another way: if the vehicle dynamics engineer can dampen (or sharpen) the driver's input to improve the steering/lat accel/yaw rate behavior of the car, why shouldn't it be done?

For example, for the past 35 years, Porsche has built hilarious amounts of roll understeer into the 911's front axle in order to keep doctors and lawyers from killing themselves. Effects of similar magnitude (but different direction) can be achieved on many vehicles by playing with bumpsteer. Of course, first you'll need to let the suspension move, as exFSAE said...

[Gyro]

@Crispy: Yes, the outside tire should be generating most of the cornering forces in a hard turn. That's why I was concerned with making sure that it could maintain a proper slip angle with its path of travel; I thought that maybe with static toe-out and Ackermann geometry (which is essentially dynamic toe-out), the outside wheel may stray too far from the ideal angle.

@Sormaz: I think you've got my meaning. The idea is to have the steering dynamically adjust in the harder corners to produce more lateral acceleration. And I'm not convinced of it either, it's just a thought.

[ZAMR]

So let me see if I understand this... You'd use bump steer to reduce the front outside tire's slip angle when the driver turns the wheel hard and oversaturates the front tires, but then you would recover extra effective steering angle as the car rolls?

Sounds logical to me for a corner entry without a braking zone, but not when you take into account that most drivers in FSAE will brake heavily then initiate their turn. Coming off of heavy braking and entering a tight corner leaves both front tires in bump, so your front outside outboard will not be moving significantly wrt the chassis, even if you are sprung lightly.

Your yaw acceleration into a tight corner is caused by a momentary imbalance of lateral force at the front and rear tracks, so one of the best ways to create this imbalance (in my view) is to use overdamped compression damping coefficients, with the rear being noticeably higher than the rear (or using high rear roll centers, or a combination of the two) to make decent use of both front tires as you begin the maneuver, as they are both momentarily highly loaded and can produce a lot of lateral force. Keeping the front inside highly loaded for as long as possible is worth more to me than perfecting your slip angles using geometry.

[swong46]

What about dynamic toe out of the outside rear when jacking the rear from a spool setup? http://fsae.com/groupee_common/emoticons/icon_redface.gif

In my mind I don't see why not. It reduces the slip angle and you don't need to worry about what happens under droop of the inner wheel (to certain extent when designed correctly)

[Jersey Tom]

Going to chip in 2 cents here before disappearing off this forum again for a while.

Don't overthink this. Or perhaps more correctly - understand what's going on at a fundamental and practical level.

I've seen some talk here of "maintaining proper slip angle" etc. In some (read: my) ways of thinking, slip angles are a system output - not an input.

Think of it this way. A driver going into a corner, at some point decides that they are going to take corner X, of radius Y, at Z speed - and hope they can pull it off. Driver has no idea or care about what slip angle that takes. They're going to demand whatever force it takes to accomplish that maneuver - regardless of if it takes 2 deg slip angle or 4 or 6.

Perhaps a better example is at the rear axle. The rear tires do not care at all about what your kinematics and bump steer are doing. Whatever unbalanced moment the front tires are creating, the rears will answer the demand of FORCE to balance it. Assuming you have symmetric bump-steer, in a given maneuver the rear tires are going to go to the same slip angles no matter what you do with bump-steer.

In a pure cornering event, bump-steer will affect HOW you get to those ultimate slip angles - the steering and body slip reactions.

Does it affect what your driver feels? Sure. Can it affect your end lap time? Definitely. But I would say you have to really understand the cause-and-effect relationships here.

Good thing to be able to design into your car as an adjustment parameter. Get your car pretty well set up steady-state, then you can proceed to the dick-a-thon of dampers, bump-steer, etc etc. Might learn some interesting things about vehicle dynamics, though it also may not be the best use of your time as you're dealing with lower control arms which have yielded, and a FARB which has seized in the bronze bushings you've chosen. Or front hubs which shear on first application of the brakes.

I remember those days... (CU '05 car - sorry Graham!)

[ZAMR]

Originally posted by Jersey Tom:

Think of it this way. A driver going into a corner, at some point decides that they are going to take corner X, of radius Y, at Z speed - and hope they can pull it off. Driver has no idea or care about what slip angle that takes. They're going to demand whatever force it takes to accomplish that maneuver - regardless of if it takes 2 deg slip angle or 4 or 6.



I agree with what was already said on the forum, if you as an engineer can make the car perform better based on the driver's inputs, do it.


Originally posted by Jersey Tom:


In a pure cornering event, bump-steer will affect HOW you get to those ultimate slip angles - the steering and body slip reactions.

Does it affect what your driver feels? Sure. Can it affect your end lap time? Definitely. But I would say you have to really understand the cause-and-effect relationships here.

Good thing to be able to design into your car as an adjustment parameter. Get your car pretty well set up steady-state, then you can proceed to the dick-a-thon of dampers, bump-steer, etc etc. Might learn some interesting things about vehicle dynamics, though it also may not be the best use of your time as you're dealing with lower control arms which have yielded, and a FARB which has seized in the bronze bushings you've chosen. Or front hubs which shear on first application of the brakes.



I don't see how this contributes to the discussion. To many times people will say "just get it set up pretty well." The problem is judges don't settle for "pretty well," they want to see your, as you say, dickathon of dampers, bump steer etc. This forum is about trying to understand what is really going on. If you are dealing with yielding lower control arms and you use that as an excuse not to explore these different vehicle control options, the judges won't buy it. If you are competing in the middle of the pack, sure don't worry about this just try to finish. But if you want to win, practice as many dickathons as you can to prepare for the real dickathon that is the design event.

[Luniz]

Just another thought: As you're saying, you are picturing the outside front tire understeering a lot, and you want to correct that by bump/roll steer. Suspension movement or roll in a corner is a result of lateral force. When the car is understeering heavily, there is not much lateral force, so there's little suspension movement, and hence little bump steer to correct the slip angle.
Back in 2010 we kind of made a similar mistake: according to our tire data, we decided that it would be beneficial to run 60% ackermann, which was the right choice for a steady state corner. The basic idea behind that was that the laden wheel required a smaller slip angle than the unladen one to achieve maximum lateral force. What we didn't think of was that on steer in there is no load transfer yet, so we never achieved the situation we wanted, the car would just rund straight on ;-)

[Jersey Tom]

Originally posted by ZAMR:
I don't see how this contributes to the discussion. To many times people will say "just get it set up pretty well." The problem is judges don't settle for "pretty well," they want to see your, as you say, dickathon of dampers, bump steer etc. This forum is about trying to understand what is really going on. If you are dealing with yielding lower control arms and you use that as an excuse not to explore these different vehicle control options, the judges won't buy it. If you are competing in the middle of the pack, sure don't worry about this just try to finish. But if you want to win, practice as many dickathons as you can to prepare for the real dickathon that is the design event.

I am merely suggesting that before a team goes down the road of trying to "take advantage" of bump-steer, they should have a better fundamental understanding of what it really does. Otherwise you're chasing your tail.

If for argument's sake you have a linear bump-steer curve at a given axle, in pure cornering/roll as one quarter suspension goes into jounce and one into rebound, the effect is going to be some axle steer. As an aside, I think linear bump-steer would be a good target if anything. Otherwise you will have dynamic toe-in/out in roll which I'd rather decouple and do with Ackermann.

Depending on which axle(s) you're talking about, axle-steer will change your steering and/or body slip angles for a given speed and curvature but it will not change the tire slip angles. Will not change whether or not your car is limit U/S or O/S or by how much.

It will however change the dynamics and feel of how the car responds. Perhaps your driver(s) and engineer(s) are truly talented enough to separate the two out and tune them independently. If so I would be very impressed as in my previous experience this was typically not the case on a large majority of teams. In any event, this is why I suggest you (a) build in adjustable bump-steer so you can get rid of it if it doesn't do what you expected, (b) save it for some separate testing or at the end of your test & tune cycle - if you have everything else sorted out [which again, I feel as if few teams do a great job of].

It's involved enough in a pure cornering or slalom type maneuver with the roll-steer aspect. Dynamic toe-in or toe-out with pitch adds an added level of complexity.

Then we circle back to level of significance. As has been said, if you aren't expecting much suspension travel then this all becomes a moot point and things like diff setup probably completely drown out any effect bump-steer will have.

[Gyro]

Thanks for the opinions, guys. I'm glad to see that my little idea generated a real discussion.

At this point, I'm sort of agreeing with Tom & others that it's probably not something really worth pursuing at this point. Better to design for zero bump steer and be able to tinker with it later to examine the effects.

But I'm still somewhat curious as to how an outside tire with static toe-out can generate adequate cornering forces. Maybe I just can't quite get the right mental picture, but it seems like the slip angle will be in the wrong direction (ie, in a right-hand turn the front-left tire will be pointing to the left of its actual heading, and thus pulling in the opposite direction of the turn). Can anybody set me straight on this?

[ZAMR]

Why would it matter? Any car with positive ackermann will be in the situation you describe. The inside and outside tire will each arrive at a final slip angle that balances the yaw moment/latG FBD. In the special case when the inside front lifts, the driver would only have to turn the wheel a little more than he would have to without static toe-out in order to reach the exact same operating condition.

[Jersey Tom]

Originally posted by Gyro:
But I'm still somewhat curious as to how an outside tire with static toe-out can generate adequate cornering forces. Maybe I just can't quite get the right mental picture, but it seems like the slip angle will be in the wrong direction (ie, in a right-hand turn the front-left tire will be pointing to the left of its actual heading, and thus pulling in the opposite direction of the turn). Can anybody set me straight on this?

What is the definition of slip angle? Can say it is the difference between the direction a wheel/tire is pointed, and the direction it's heading.

In your example of a right hand turn, even with some static toe out... once you get into the corner, the tire slip angles will all be of the same sign (by SAE convention, negative slip angles). That means that ALL the tires, including the left front, will be pointing RIGHT of their "actual" heading.

Not sure if that helps to clarify things, but I think it's an important point to establish.

[Gyro]

Tom: Okay, that makes sense. I guess I was thinking only about the geometry of the steering, and not about the dynamic motion of the car. Thanks for pointing that out.

[swong46]

Originally posted by Luniz:
Back in 2010 we kind of made a similar mistake: according to our tire data, we decided that it would be beneficial to run 60% ackermann, which was the right choice for a steady state corner. The basic idea behind that was that the laden wheel required a smaller slip angle than the unladen one to achieve maximum lateral force. What we didn't think of was that on steer in there is no load transfer yet, so we never achieved the situation we wanted, the car would just rund straight on ;-)

With this, I am thinking that ackermann steering with the necessary slip angle change under WT will take away from Turn In. So it is better to design ackermann for cornering radius purposes only and let the bump steer take care of the required SA changes under WT.

[flavorPacket]

Originally posted by swong46:
With this, I am thinking that ackermann steering with the necessary slip angle change under WT will take away from Turn In. So it is better to design ackermann for cornering radius purposes only and let the bump steer take care of the required SA changes under WT.

That's the conventional approach that most vehicle designers choose for cars needing high steering angles. For road racing, NASCAR, or other series with low steering angles, designers have more freedom to mix the two. In either case, the thing to remember is that the yaw frequency and the roll frequency do not change in the same manner with speed. Thus the directional stability of a car can change considerably from one speed regime to another if roll steer effects are not properly accounted for.