13 replies to this topic

[mariner]

I had the pleasure today of being in the company of John Miles the Lotus ex F1 driver and senior vehicle dynamics expert. He came to our local 750 club centre meeting as the guest speaker.

 

As well as recounting his racing and journalistic career ( he was technical editor of Autocar for 5 years) he also shared some of his vehicle dynamics knowledge from 15 years at Lotus as a  dynamics engineer on road car development for many clients.

 

John made point which I had never even thought of - whilst setting a racing car up for zero bump steer may be good thing its  no-no for  a road car which needs bump steer to work effectively. John's point was that the tyres don't know where the forces come from or how it is connected to the car so they can react as much to road inputs as driver inputs.

 

For average driving those road inputs ( bumps, grooves, sidewinds etc.) can be close to the low angular inputs by the driver for 90% of the time. If the car didn't use bump steer effects to cancel out road induced slip angles the driver would spend all the time making minute corrections. Hence most road cars are designed to give significant steer angle change on bump and rebound.

 

So, if your car is for road use and a dial guage check shows bump steer it's probably designed in and doesn't need fixing.

 

John also showed a neat trick for instant torsional stiffnes testing. Stretch two pieces of masking tape across the cockpit or door aperture and jack up one corner. If one of the masking tapes sags or breaks its not stiff enough!

[NotAPineapple]

Bumpsteer has been used to induce understeer for a while now. Bump toe in at the rear, bump toe out in the front. It means that when the car rolls, the front wheels steer out of the turn and the rear wheels steer into the turn.

 

I suppose this works like John mentioned (though I need to think about it a bit more before I'm convinced) because steering a car out of a turn is a stable response.

 

It all a compromise anyway. E.g. on the rear axle, bump toe in means you have rebound toe out which is exactly what you don't want under brakes. Of course you can add braking toe in compliance but then that has its own knock on effects.

 

My opinion is that bumpsteer and compliance steers are delibrately put into road cars to enhance the driving "feeling" by influencing the sideslip angle of the body. Since the driver feels the slip angle of the body, he feels these steering effects. I also see them as effects which are lying (for lack of a better word) about what the tyres are really doing. Eg lots or rear axle toe in (either through compliance steer or bumpsteer) will give a very low sideslip angle in a corner which feels nice to drive but its lying because the tyres on the rear are actually at a higher slip angle (maybe near their limit) but the driver doesn't sense it.

 

I'd guess this is why its avoided in race-cars

Edited by NotAPineapple, 05 September 2013 - 12:03.

[Bloggsworth]

John also showed a neat trick for instant torsional stiffnes testing. Stretch two pieces of masking tape across the cockpit or door aperture and jack up one corner. If one of the masking tapes sags or breaks its not stiff enough!

 

You forgot to say "diagonally"

 

Lovely bloke, does he still walk as if his legs and feet aren't quite connected?

[Lee Nicolle]

Bump steer is undesireable on any car. Though all cars have some. Limiting it to acceptable amounts in the normal suspension travel range is normally desireable.
Some manufacturers try to make the car work around bad bump steer. But are bad cars!

[Joe Bosworth]

Lee, I wonder if we need to separate roll steer from bunp steer in this discussiuon.  I suspect that they are closely inertwined.

 

For the everyday driver oversteer is highly frowned upon.  Just go back the the "unsafe at any speed Chevvy Corvair.  The fact that the sporting driver believed it to be about the best handling American car of the day is the best example.

 

The best way to avoid oversteer, or induce understeer, is to have the loaded rear wheel go to increased toe in.

 

Likewise, in a race car we are mostly very comfortable to have nuetral steering in the front and let the car find its own way going down a bumpy straight with a bit of castor to heavy up the feel.  But for the everyday run of the mill driver a steady as she goes is the desired feel.  This requires generous castor and again the loaded wheel tending to pointing in on bounce upwards.

 

Any real arguement on these principles??

 

Regards

Edited by Joe Bosworth, 06 September 2013 - 05:05.

[Kelpiecross]

No arguments from me.   In a similar fashion -  I also have read that car makers sometimes use an offset from the steering contact patch to give an added "feel" to the steering.   (I don't think I expressed that very well). 

[Lee Nicolle]

Lee, I wonder if we need to separate roll steer from bunp steer in this discussiuon.  I suspect that they are closely inertwined.
 
For the everyday driver oversteer is highly frowned upon.  Just go back the the "unsafe at any speed Chevvy Corvair.  The fact that the sporting driver believed it to be about the best handling American car of the day is the best example.
 
The best way to avoid oversteer, or induce understeer, is to have the loaded rear wheel go to increased toe in.
 
Likewise, in a race car we are mostly very comfortable to have nuetral steering in the front and let the car find its own way going down a bumpy straight with a bit of castor to heavy up the feel.  But for the everyday run of the mill driver a steady as she goes is the desired feel.  This requires generous castor and again the loaded wheel tending to pointing in on bounce upwards.
 
Any real arguement on these principles??
 
Regards

Roll steer is often bump steer anyway. A live axle car will roll steer and too a degree that also translates to bump. A swing axle car has sh*t loads of both roll steer and bump steer. Roll steer is when one side is loaded, bump is to a degree when one wheel hits a bump and the wheel either cambers or toes separate to the other. While I am no expert I have spent hours farting around simulating these scenarios. On both road race cars and speedway cars. front and rear. Remove the springs, but keep everything else intact. Jack the vehicle centrally and then one corner at a time and measure the changes at the tyre diameter. A piece of wood or steel bolted to the hub makes it far easier. Then allow a bit more for everything flexing! It can be scarey. And then you have to work out what is good and what could be helpful. And what is bad. And what you can actually do anything about. Which is even scarier.And with production cars how much extra the rubber bushes ultimately effect everything. Some flex 10mm!In all directions! Though a heim joint to drive down the road on is pretty dreadfull.

As for the sportiness of a Corvair, I have driven a later one ,left hook, nicely restored but it really did not do much for me. And earlier 911 Porkers which to me are as bad. Both roll steer and bump steer in the rear. That on brief suburban drives. A Sports Car muttering rotter sometimes gets confused about the engine position making the car good and in those days rear engine was ooh aah! Just like a Porsche!

As for American cars. Most have built in pig understeer, as do most 60s and 70s Aussie ones too. But a big pendulum in the back is hardly sporty handling! Or pig understeer in and oh so loose out! A Mustang with a wheel alignment is a far better handling car! And can be made to handle very well with little work. My 2 tons of Galaxie handles very well these days.Decent shocks, a lot more caster than original and decent wheels and tyres. As a road car though.Not a Sports Car. Though I can drive it for a 300k trip and not get out exhausted unlike most sportys.
A Corvair with far more work I am sure can be a decent handler too for a race track and does have better rear weight, though really way too much. As does all cars with the engine at the back!

Edited by Lee Nicolle, 07 September 2013 - 00:08.

[Greg Locock]

A typical beam axle has no bump steer unless you have some weird arm geometry, or a badly designed panhard. It can have roll steer.

 

A twistbeam suspension has no bump steer, but with a bit of finessing it can have roll steer.

 

IRS tend to have zero bump steer, obviously.

 

I'm not a huge fan of big amounts of rear roll steer as it makes the car jump around on irregular road surfaces, but a little bit of toe in in bump is better than absolutely neutral.

 

For the front you can be running anything up to 10 deg/m  roll steer (toe out in bump), the compromise is between linear range understeer and sensitivity to vertical disturbances.

 

Bear in mind that the typical spec for roll or bump steer at the front relies on you getting the relativity between the hardpoints right, in the relevant direction, to within 1mm or so across the car, no mean feat in production.

[Ray Bell]

Peugeot 404s seem to drive all right on the road without bump steer...

 

In fact, I think they are very good to drive.

[Lee Nicolle]

A typical beam axle has no bump steer unless you have some weird arm geometry, or a badly designed panhard. It can have roll steer.
 
A twistbeam suspension has no bump steer, but with a bit of finessing it can have roll steer.
 
IRS tend to have zero bump steer, obviously.
 
I'm not a huge fan of big amounts of rear roll steer as it makes the car jump around on irregular road surfaces, but a little bit of toe in in bump is better than absolutely neutral.
 
For the front you can be running anything up to 10 deg/m  roll steer (toe out in bump), the compromise is between linear range understeer and sensitivity to vertical disturbances.
 
Bear in mind that the typical spec for roll or bump steer at the front relies on you getting the relativity between the hardpoints right, in the relevant direction, to within 1mm or so across the car, no mean feat in production.

Greg, ALL beam axle cars have roll steer. Even with a parallel link system. Which no road car has. A leaf spring rear too rear steers [watch all that spring wind up!] though not as much as a long arm short arm live axle rear. Some are better than others. A true IRS has very little though a swing axle style like a Commodore or older Benz, Datsuns etc has heaps. The toe changes through the arc, yet alone camber changes steer the things everywhere. Which is why they destroy tyres so quickly.
Front drivers usually has less. But the rear suspension on those cars keeps the back of the car off the ground! Not take acceleration, and carry a load. Twist beams to me are an accident waiting too happen. Flexing an axle like that seems so wrong. Though I have ever only seen one break, the car then proceeded to hit a tree! Ouch. luckily at low speed. I have however seen a couple with the starting of cracks around the welds for the brackets.

[gruntguru]

Greg, ALL beam axle cars have roll steer.

 Lee, with your speedway experience you would have seen many beam axles with zero roll steer. Anything with a longitudinal Watts on each side - sprintcars, "5th coilover" sedans.

[Lee Nicolle]

Lee, with your speedway experience you would have seen many beam axles with zero roll steer. Anything with a longitudinal Watts on each side - sprintcars, "5th coilover" sedans.

Sprinters and Midgets really are all bound up. The torque tube really does that well. A few years back a Nuz Zealander came over and won our title with an open drive, a tailshaft and a 3 link [like my 1968 Supermodified has]. More traction and more travel. It was banned very quickly after.
Jacobs ladders, a rattly loosely connected series of flat metal straps can be a black art, get it right and they are good until you bang wheels or the fence, and then you have something evil! And they make the things by the dozen! Interestingly a few years ago many sprints had different length radius arms left to right. One upper, one lower on the right and one [shorter] from the centre of the front axle. Now that really made bump steer! What were they thinking!

Super Sedans create rear steer with the things falling over on the right rear and the extended shocks to allow it. You can watch the left rear walking back and forth in the wheel arch. V8 Dirt modified were as bad too.Though that 'phase' seems to have calmed down these days. Fifth arm rear creates traction, but you have to keep your foot in the bucket to manage it, probably why they crash so often! My one and only drive in one did not create confidence, though when you got in the throttle it really hooked up. Then you just had to be brave to keep your foot down through the corners! I was not very brave! So quite untidy. Especially on dry skatey track at a practice. Thankfully by myself!

[dynatune]

In general most car manufacturers try to minimize bumpsteer and camber gain and tire contactpatch lateral movement in order to have the best vertical behaviour creating the least input into the steering system. As we all know that what is good for ride is not necessarily good for handling brings directly a compromise which is called roll steer and compliance steer. So in order to have some rollsteer on IS a certain need of bumpsteer is needed. Typical values are -3° to -5° deg/m on the front toe and +1° to +3° deg/m on the rear (toe-out negative).
Now here comes a tricky part to understand about roll steer and compliance steer: Their main objective is to help the tires to create a car with "sufficient" understeer in the LINEAR range in order to guarantee its high speed stability. You have certainly seen motorcycles starting to pendulum movement at very high speed straight line running. This comes due to the fact that yaw damping decreases with speed and the bike does not have anymore sufficient understeer. In order to counteract this natural tendency a car needs to have sufficient understeer - created either by roll understeer or lateral compliance understeer - this is not a point of view but a basic law of physics (a "famous" car that was known for being inherently instable was the first series Audi TT). This "understeer" is typically known in the Vehicle Dynamics Community as LINEAR UNDERSTEER GRADIENT and is usually for a sound car around 2.5° /g and important CANNOT be sensed directly by 99% of the driver-population. Only very experienced test drivers are able to identify it.
The understeer most drivers/racers refer to is the NON-LINEAR understeer that comes during cornering with high g causing high vertical load transfer, suspension progression and tires saturation is largely overshadowing the effects of kinematics. If anyone wants to learn more about this have a look at my site www.dynatune-xl.com, one can also download for free a demoversion in which you can try out all the effects of bump/rollsteer on handling and straight line stability.

Cheers,

dynatune

[RDV]

^ ...this...