Push rod on upright suspension - an examination

[strad]

Bump-steer means any time the wheel encounters a bump the steering geometry changes. Toe in or toe out, camber.??. Most all cars have some bump-steer some times and is usually undesirable.
When you lower a cars ride height you usually create bump-steer and have to adjust the steering rack and tie rod height.
This thing of purposely inducing bump-steer has me quite confused.
I'm not a front suspension expert but when I lowered my Mustang Cobra I had to learn and fix the problem.
It's un-nerving to have your steering change on it's own mid corner.

[Greg Locock]

I agree, but if we start from the premise that when reverse engineering the original designers weren't stupid, there must be some effect they were aiming at.

[Maritimer]

What's the advantage of rolling the car when turning? Why is this better than CGZ/RCH/sta bars?

Well, further thought says that this will be speed dependent, SWA and hence wheel steer is greater at low speeds. But I'd have thought rolling the car at low speeds is not especially beneficial.

Believe it helps load the inside wheel during low speed corners, as well as drops the nose/FW which increases downforce marginally. It raises the inside tire and lowers the outside to generate negative roll in a sense, although I'm not sure how much mechanical grip this contributes on these cars.

[Zynerji]

I would expect the jacking of the inside wheel would effectively allow the car to "lean" into corners, and load the ends of the front wing (ground effect) instead of just the middle, thus inducing uneven front wing loading to help mid-corner bite.

[Greg Locock]

So in plan view is the pushrod to spindle joint in front of, or behind, the kingpin axis?

[PhillipM]

Believe it helps load the inside wheel during low speed corners, as well as drops the nose/FW which increases downforce marginally. It raises the inside tire and lowers the outside to generate negative roll in a sense, although I'm not sure how much mechanical grip this contributes on these cars.

In these implementations with extreme offsets inboard of the upright, it actually lowers the entire front end of the car to the track. Tim is on the right idea here, and yes, it does make the car want to wander because the forces from the springs/dampers then try to overcome the self aligning torque from the tyres - the forces *want* the wheel to be turning, instead of tracking straight like usual, so it's a bit of a balancing act on your other geometery to get it to work *and* still feel natural to the driver.

All these people talking about bump steer are completely missing the point - this is not the steering arm attachment like your road car, it's the pushrod for the springs and dampers.

[Tim.Wright]

Yup I'd say so. The system creates a "negative" steering torque which effectively means the driver doesn't feel a centering force at the wheel like normal cars but an anti-centering force which tries to increase the steering angle by itself. The driver has to fight against this to keep the car going straight.

So how does the setup know whether it's turning left or right?

Sort of a trick question...

It doesn't. It's done to lower the car in both directions, not to jack one side up and the other down.

[godlameroso]

Bump-steer means any time the wheel encounters a bump the steering geometry changes. Toe in or toe out, camber.??. Most all cars have some bump-steer some times and is usually undesirable.
When you lower a cars ride height you usually create bump-steer and have to adjust the steering rack and tie rod height.
This thing of purposely inducing bump-steer has me quite confused.
I'm not a front suspension expert but when I lowered my Mustang Cobra I had to learn and fix the problem.
It's un-nerving to have your steering change on it's own mid corner.

All cars will undergo suspension geometry changes as the suspension is loaded and unloaded. It all depends on the initial geometry, the forces, the compliance of the suspension and other mysterious factors that create imaginary lines, such as COG and roll centers.

If you change the arm slant but not the suspension pick up points you change a lot of variables. Geometry, KPI, roll center, forces on the control arms which affects bushing compliance and more.

The suspension on F1 cars is hard to change in layout. On the one hand raising the suspension pick up points as high as possible is good for aero, but potentially bad for kinematics if you can't create the proper geometry. Furthermore aero is king so suspension has to be designed around not just being out of the way, but also to control the aero platform of the car.

As you can imagine these are contradictory goals. I feel the extreme angles and geometry we see in F1 goes unappreciated.

[Zynerji]

As you can imagine these are contradictory goals. I feel the extreme angles and geometry we see in F1 goes unappreciated.

This is why I don't want to see active suspension return to F1.

It would be 99% aero platform driven, and 1% wheel control.

[godlameroso]

Believe it helps load the inside wheel during low speed corners, as well as drops the nose/FW which increases downforce marginally. It raises the inside tire and lowers the outside to generate negative roll in a sense, although I'm not sure how much mechanical grip this contributes on these cars.

In these implementations with extreme offsets inboard of the upright, it actually lowers the entire front end of the car to the track. Tim is on the right idea here, and yes, it does make the car want to wander because the forces from the springs/dampers then try to overcome the self aligning torque from the tyres - the forces *want* the wheel to be turning, instead of tracking straight like usual, so it's a bit of a balancing act on your other geometery to get it to work *and* still feel natural to the driver.

All these people talking about bump steer are completely missing the point - this is not the steering arm attachment like your road car, it's the pushrod for the springs and dampers.

In other words the steering tie rods are angled relative to the push rod and control arm geometry so that when the wheel toes out, it tries to unload the opposite side of the suspension and the ARB jacks down on the whole front axle? I assume the wheel toeing in would have the opposite effect in this setup, so the combined forces would produce such an effect. In any case isn't this now restricted by the regulations?

[Tim.Wright]

In other words the steering tie rods are angled relative to the push rod and control arm geometry so that when the wheel toes out, it tries to unload the opposite side of the suspension and the ARB jacks down on the whole front axle? I assume the wheel toeing in would have the opposite effect in this setup, so the combined forces would produce such an effect. In any case isn't this now restricted by the regulations?

In the effect we are describing, the ARB doesn't play a role.

If the pushrod mount has a pure inboard lateral offset from the kingpin axis in the static condition it will cause the suspension to lower in both toe in and out movements. If you imagine the trajectory of this point with steer, it will always move in a way which tries to extend the pushrod which will reduce spring load and lower the car.

If the pushrod mount has a pure longitudinal offset from the steering axis, it will cause a roll movement with one wheel lowering and the other raising. In this case the ARB will be actuated.

The choice of the pushrod mounting point will then be driven by what mix of the two above effects are desired. The first is mainly a mechanical grip consideration, the second is mainly aerodynamic.

[strad]

sorry if I mis-understood but a lot of what y'all are saying sounds like bump-steer to me.

[PlatinumZealot]

Yup I'd say so. The system creates a "negative" steering torque which effectively means the driver doesn't feel a centering force at the wheel like normal cars but an anti-centering force which tries to increase the steering angle by itself. The driver has to fight against this to keep the car going straight.

So how does the setup know whether it's turning left or right?

Sort of a trick question...

It doesn't. It's done to lower the car in both directions, not to jack one side up and the other down.

Sounds like the effects of caster with the centering torque you are talking about. Ok gonna look into this. Caster angle is another
" complication " in this topic but going by photos of mercedes W10 car it is postive caster angle so there is a self centering tendency. So this is the perfect example.
OK so now you say the Mercedes setup of the pushrod is to lighten the steering a bit by counteracting the effect of the postive caster. Fine let's explore.



You mentioned "negative" steering torque so automatically one will seek to clear up how the geomtery is designed to work whether car is going left or right?

I have an idea how such geometry can still can "know" (and work) whether the car is steering left or right.. But i wanted to see somebody else's take.

I haven't gone in depth into the kinetics of it yet tho.. Just working peice by peice.

[Zynerji]

FRIC suspension is banned (front/rear interconnect), but I'd imagine FFIC could still be used.

Maybe that's how it can "know" which way it's turning, but Im just speculating.

PS: Concerning caster. In sim racing about 2007ish, my friend and I did a ton of setup mapping in rFactor. Now, obviously, it is a game, but we found tons of time with the toe/caster settings. Oddly, whenever I've spoken about going slight negative on caster with high camber and slight positive rear toe, people don't believe it is a good thing. I watched him win the 2007 and 2008 league championships with that setup philosophy. My friend called it the Snap and Hold setup (smooth), as it was a bunch of small, quick steering maneuvers. He disliked it at first, as he was a very "stair step" corner entry driver (choppy), and it took hours for him to adapt, but it was worth 1.3s in the end.

With modern machine learning, we will probably see lots of "rules of thumb" be overturned with "wacky" setups.

[godlameroso]

FRIC suspension is banned (front/rear interconnect), but I'd imagine FFIC could still be used.

Maybe that's how it can "know" which way it's turning, but Im just speculating.

PS: Concerning caster. In sim racing about 2007ish, my friend and I did a ton of setup mapping in rFactor. Now, obviously, it is a game, but we found tons of time with the toe/caster settings. Oddly, whenever I've spoken about going slight negative on caster with high camber and slight positive rear toe, people don't believe it is a good thing. I watched him win the 2007 and 2008 league championships with that setup philosophy. My friend called it the Snap and Hold setup (smooth), as it was a bunch of small, quick steering maneuvers. He disliked it at first, as he was a very "stair step" corner entry driver (choppy), and it took hours for him to adapt, but it was worth 1.3s in the end.

With modern machine learning, we will probably see lots of "rules of thumb" be overturned with "wacky" setups.

F1 cars already run a fair bit of toe out on the front, a hint of toe in at the rear, and as much camber as Pirelli lets them get away with. What is not known to me is caster angles, although I'm guessing looking at some of these pictures that 6-10 degrees wouldn't be out there.

2018 tires had wear marks on the inside of the front tires which is common signs of having tons of toe out. Probably like half a degree.