Bump Steer Vs Roll Steer

[samiam520]

Just got done taking some suspension measurements on a friends car this weekend and I believe I have more questions then answers.

I guess my main question would be, if you had the choice to optimize either bump steer or roll steer which one would be more important? This particular car is in a spec class so there are limited adjustments.

[Adam Brouillard]

I've always wondered what the difference is. It seems if your wheel doesn't turn through suspension movement, you would have neither.

[2BWise]

As with any change there's going to be some sort of compromise. Both bump and roll steer have andvantages and disadvantages. You could use them to tune into a car or tune them out to improve them, but like anything its going to be specific to the type of car.

Ultimately, zero of both would be the most predictable, but the use of either could also make the car quicker. Using the measurements you've got along with the subjective feel of the car you might be able to find a direction in which to change. them

[Dave Harmison]

It sounds like you've done bump steer. You measured the amount of steering angle change as you exercised the suspension up and down from the nominal ride height. You would plot this data as degrees of steer versus inches or mm of suspension travel.

Roll steer is the amount of steer motion you get from rolling the vehicle. You'd plot this as degrees of steer versus degrees of roll.

Roll steer is really only of interest on solid axles. Picture a solid axle from the top with trailing links where one side of the axle moves rearward as the wheel goes into jounce and the other tire moves forward as it goes into droop. If you did this by rolling the car you'd quickly realize that this is roll oversteer. Likewise, if you picture the same solid axle suspension stroking up and down on both sides at the same time, you'd get no steering, just sysmetrical motion of the axle forward and backward.

Sometimes you'll see cars with independent suspension that transfer forces through sway bar links or the such that causes one wheel to react to the motion of the other, but that doesn't happen on a formula car with rod-ends.

On a formula car just a bit of bump understeer on both ends is desired, so the front should toe-out a hair in bump and the rear toe-in a hair. Roll steer dxoesn't really mena anything.

On a solid axle I'd look for a hair of roll understeer on each end. Assuming the front is independent (you're not working on a sprint car are you?) it will be bump understeer just like a formula car and in the rear you'll actually need to roll the axle to measure it. Keep in mind that roll steer on a solid axle changes as a function of ride height so jacking forces and all that come into play.

Hope this is helpful.

[samiam520]

Dave,

Thanks for your thoughtful response. I am a little confused by your use of the terms understeer and oversteer when it comes to toe changes. You say that bump understeer is preferable at both ends but doesn't toe out in the front usually cause oversteer? BTW, the car I'm working has independent rear suspension.

[Dave Harmison]

The terms oversteer and understeer are meant to describe the behavior in a turn.

Picture the car from above going around a left hand turn. As the car goes around the left hand curve you want the front tires to turn towards the outside of the curve to decrease the amount of steer you get, understeer. So the left suspension will be in droop and you want the left front tire to toe to the right, so that would be toe in with droop. Conversely, the right side suspension will be in jounce and you want the right front to toe out. Toe-in in droop and toe-out in jounce are the same thing, as the suspension moves from the bottom of it's range of travel upward, the wheel should progressively toe-out.

The exact opposite is true in the rear, both tires should steer towards the inside of the curve as the car's body rolls, again, understeer. So toe in with jounce.

This is pretty hard to describe over the internet. At this point, everybody I've ever talked to about this has held their hands out infront of themselves like two tires and moved them up and down simulating toe in and rolling their shoulders to simulate body roll. If someone sees you doing it they'll think you're a little nuts, but if they are vehicle dynamics people they will instantly recognize what you're doing. Lots of books have good diagrams of this.

Now having said all this, the most important part is to keep the steering motions SMALL. Like spices in cooking, some is good, too much is horrible. If you make the bump steer too big the car will hunt around in corners and be twitchy over bumps. And don't worry that your putting understeer into the car. This isn't the grinding type understeer that your Dad's Grand Marquis had when you were a kid, this is transient understeer which happens as the car makes transitions from straight to corner.

Cars have been made to work with roll oversteer at one or both ends but they are exceptions and it is done to offset some other defect in the car. So it's more like two wrongs make a very poor right so I strongly suggest against it.

Good luck.

[Daryl DeArman]

if you had the choice to optimize either bump steer or roll steer which one would be more important? This particular car is in a spec class so there are limited adjustments.

I prefer to keep it simple. What does the rest of the suspension/steering do during bump/droop and roll? Do the camber curves stay nice with bump/droop versus roll? (like a FV front beam) Or does the camber stay relatively mild with roll but go whacky with travel (like a FV zero roll swing axle)?

In other words, no sense concentrating on roll steer if the car doesn't roll much but has lots of travel, and no sense concentrating on bump steer if the car has more roll and very little travel.

Lastly, the bump steer and roll steer curves you plotted--are they both "steering" in the desired direction at all times? Perhaps more importantly than how much they change is the fact that they are changing in the direction you wish for them to change.

[samiam520]

Lastly, the bump steer and roll steer curves you plotted--are they both "steering" in the desired direction at all times? Perhaps more importantly than how much they change is the fact that they are changing in the direction you wish for them to change.

Daryl,

"The desired direction" Yes, this is the fundamental question that I am seeking an answer to. Like many theories on suspension set up, sounds like there could be a few different opinions as to what is optimal. I guess I need to get my suspension books out of the storage unit again.

BTW, the car that I am working on is not a FV

[samiam520]

Lastly, the bump steer and roll steer curves you plotted--are they both "steering" in the desired direction at all times? Perhaps more importantly than how much they change is the fact that they are changing in the direction you wish for them to change.

Daryl,

"The desired direction" Yes, this is the fundamental question that I am seeking an answer to. Like many theories on suspension set up, sounds like there could be a few different opinions as to what is optimal.

BTW, this car that I am working on is not a FV

[Daryl DeArman]

the car that I am working on is not a FV

I was just using a FV as an example easy for me to explain and you to visualize.

As I am sure you've figured out there isn't one answer for all cars at all tracks.

FWD / RWD
High HP, Low HP
lots of downforce, no downforce
bumpy track, smooth track
long straights, extremely technical.

Lastly, you mentioned it is a spec class. Does the spec class allow relocation of the rack and/or alternate suspension geometry?

[Daryl DeArman]

...

[samiam520]

Daryl,

The car is a SRF.

Yes, the rack can be adjusted, mainly in a downward direction. During dis-assembly we did find several worn out bushings and rod ends. Additionally the left and right front tie rod ends were of a different manufacture and dimensionally quite different. The car has always been very dynamically unstable, thus, difficult to drive quickly. I suspected that there was some geometry problems which were confirmed when we measured everything. During reassembly we will set the bump steer in the front and back and try to make everything proper.

Thanks,