Uhh, bump steer correction may have a side effect of reducing kickback, but that's not the main purpose of the correction (i.e. excess toe change during suspension compression).

Sherwood

Try to get rid of bump/roll steer or make them equal
 

In Carroll Smith's First book, he addresses bump/roll steer, (ppg 141-149). He basically concludes that anything that detracts from the linear controll of the driver is a disadvantage because it is anther variable the driver must try to anticipate.

On a street car driven to the market, this is a non issue. On the other extreme of the controll issue, a car trying to balance oversteer/understeer & maximum speed, while trying to deal with competitors entering/during or exiting a corner---Linear response to steering input is almost critical.

So, If you lower the car and you are looking at more rapidly increasing toe with bump this is not a good thing. For a moment, disregarding the rear suspension steer characteristics, I am thinking you want the toe steer curve to be inthe middle of the curve, the most gradual and predictable section of the suspension movement.

My guess is that you want to try to match front roll steer to rear roll steer, so the car goes where you point it rather than wondering---how in the hell did i get here??? Lord knows there are enough other variables to worry about/try to plan for.

Having finished my theoretical diatribe, i will tell you that this is a relatively subtle change in behavior, at least with my hot rod 914.

YMMV,
chris :-)

Give me the distance from the floorpan to the ground (or the outside surface of the A-arm to ground) and the diameter of your tire. I will show you your bump steer and how it changes with x amount of tie-rod spacers.

Hi Max,

What do you think is happening at the back with roll steer? I have no idea.


Curious,
chris

I am not sure yet. I am working on the rear simulator this weekend.

I have seen this graph posted though. I think vorspur is toe.

http://forums.pelicanparts.com/uploa...1331356381.jpg

Vorspurverlauf, Hinterachse is rear toe.

The front toe does not look correct to me, though.

Unless you want a psychotic car, roll steer is always set up as understeer (toe out in front, toe in in rear, for the outside tire). The plot Max showed is qualitatively similar to what I've measured on my car - the bump steer curve does not change all that much when the car is lowered.

Perhaps the reason the toe graph looks different is the low ride height my sime runs at. I generate a curve symmetric about the x axis (or a shifted axis parallel to the x). I don't see why the Porsche graph shows an odd function, looks like tangent, when my graph shows an even function. I am graphing bump steer as the angle of the spindle assembly as it turns on the strut. Even if they measure projected angle it seems that there would not be a reason for such a disagreement.

Could it just be the stock ride height they are using?

I see what you mean, Flieger - the "S" shape of the ride steer curve. I was just looking at the jounce portion, which has a bit more curvature than what I measured, but otherwise is not too far off.

When I give my cheesy suspension model a ridiculous amount of rebound travel (like -150mm), I get an S-shaped curve also, but not when I use what seems like reasonable values. That said, the plots go past 100mm, which is probably a tad optimistic.