Steering wheel twitchy on bumps / racecar 2

[Ibrahim M]

Hi

What geometry problem causes the steering wheel to twich when passing small bumps / road defects like on this video :

https://youtu.be/X1i90PivtaU

Around 0:20 you can see it going in the turn


Excessive scurb radius ?

 

[BrianPetersen]

I see nothing out of the ordinary.

The scrub radius can be a factor but there's a whole lot more going on.
Vertical load on the tire is varying. (The suspension is moving up and down, the spring is compressing and extending, the damper is absorbing compression and rebound.)
Available lateral grip is a function of vertical load on the tire, which is varying.
The moment around the steering axis is a function of the lateral force, which is varying.
Bump-steer geometry won't be perfect; it never is.
The bumps are never perfectly horizontal and parallel to the track surface. The tire is practically never hitting bumps square-on, so now there's more moment around the steering axis because of not hitting bumps absolutely square-on with no moment-arm to the steering axis.
The car has a responsive steering rack, probably not "dead" to road feel like many power steering systems are. Forces/motions at the tire contact patch are transmitted back to the driver to some extent ... the "feel" lets the driver understand what the tires are doing.
The driver is driving the car harder than one would normally see on a trip to the grocery store.

What's on the video is quite mild.

 

[cibachrome]

Steering torques originate from tirerod loads. It's "twitching" because there is a large momentary imbalance between the tierod loads. Tierod loads come from tire Fy, Mx, and Mz forces & moments. Caster is also a factor in this effects because momentary loss of Fy means a large imbalance in net rack force.

That being said, Here's where I would go:

Set back the caster to figure the effect of caster on the problem, like an optometrist: "Better or worse". Caster helps you acquire better steering feel of max Fy, but also reduces max Fy.
Reduce your scrub radius. Even though it's set mechanically, tire Mx, both static and no slip at wheel loads can be very large AND can be of the opposite sign of your wheel aligner's reading. Thurn the tires around on the rim to see if it changes, however only if tire conicity is the amplifier.

By any means, reduce the wheel offset. This acts just like my farm tractor when I need the steered wheels spider facing outward to match the rears. Wheel kick is severe,

Ride steer is also a candidate, as is too much rebound damping: you are "snatching" a wheel. When lifted and reloaded, you are again facing a tire Mz initial transient.

Get a video of your steered tire / road interface. Loss of Mx ("snap thru") will do this to you, too. Watch the tread package shift laterally back to a more central position. This will also show you whether the wheel is actually turning (steer) or just the force(s) are disconnecting.

Consider a steering damper. You'll appreciate the reduced aggravation.

 

[Ibrahim M]

Ciba, sadly tjis isnt my video but i could probably film my car next time i go to the track.

Can you explain how caster reduces the Maximum Fy ???

 

[gruntguru]

Being an Audi, it is probably AWD. If so the combination of scrub radius and sudden variations in side-to-side distribution of Fx will also contribute to steering snatch.

je suis charlie

 

[cibachrome]

The car in the video is not likely FWD as I saw it. But, if it is FWD, "Torque Induced Steer" is a common problem cause by a positive NET scrub radius. Sure, the geometry indicates a negative scrub radius, but the static + dynamic pneumatic scrub added by some tire constructions (Tire Mx / Fz ) will cause the "twitch" as you describe. Basically the P-Scrub is quite a bit larger than the geometric value and is the opposite sign. A -15 mm scrub radius plus a +50 mm P-Scrub is a positive number. Turn the tires around on the rims for a Big Science undertaking. It is common for tires of the same size, rim, pressure, but different brands to have very different P-Scrub values. This is a real problem on high hp FWD and 4WD vehicles.

As for caster induced understeer, caster increases net tierod load. Tierod loads add compliance from steering gear mounts, rack/pinion separation forces, hydraulic valve reaction rotation, steering column isolator twist and steering intermediate shaft U-joint displacements.

In just about all cases, peak tire Mz occurs before peak Fy. Mz is a contributor to steering effort, and a component of "road Feel" is its derivative (gradient). Most drivers will operate up to the point where this gradient goes to zero, even though more Fy is available. BTW, the Mz induced compliance component to vehicle understeer is also going away, so the car's 'balance' will change.

So, adding caster moves the apparent peak steering moment gradient upward on the slip angle function so it coincides with peak Fy (if done properly). But, this adds understeer which lowers the vehicle's max Ay capability.

 

[BrianPetersen]

If the front wheels are driven ... Is the steering wheel "twitching" happening on the power, or on the brakes, or at neutral throttle coasting, or while cornering at neutral throttle, or throttle/braking doesn't matter and it does it all the time (on irregular surfaces)? Does the throttle or the brakes have an influence on the steering that you can feel? If so, describe?

If the front wheels are driven ... does the vehicle have a limited-slip front differential?