Bump Steer

[69S4]

I assumed you'd do it twice - first time without the steering arm attached so you'd get a trace for the castor angle straight line, and then a second time with the arm replaced. Any change from the first line is down to bump steer. You can then start the process of trying to correct it.
Presumably if you don't get a straight line without the steering arm attached you've got bigger problems than bump steer.

Just trying to get this straight as it's going to be one of my winter jobs.

[john.p.clegg]

..and I hate to say this,but the cause of bump steer is simply the relative position of the rack to the front suspension pivots...

John

[vxah]

Yes, what 69S4 said, just to get a line on my wall to match the caster created by the turret. Then with the steering connected the line should be followed?
might be interesting to see what sort of line we get with a few degrees of steering on? Maybe there is more to the design than we give credit?

[ericbushby]

OK, I think I get that.
So in my description near the start of this topic, the vertical line on the cardboard would have been more accurate if it had been drawn at the castor angle of the front suspension.
I did not think of that !
Even as it was, it transformed the driving experience so it cannot be far out.
Thankyou
Eric in burnley

[nmauduit]

one may or may not get a straight line, depending on actual geometry I guess. That setting is done around the resting level of the car properly loaded, and the circle can be approximated as a straight line around the resting point.
Obviously if one observes a portion of circle that is tangent to the vertical at resting position (i.e. car mostly tuned already), the line will go in the same direction above and below the resting point: increasing shims will twist that circle in one direction, decreasing shims will twist it in the other (plus changing the radius of the circle a bit possibly).
As said above, this setting is quite sensitive (depending on your goal I guess), so I made a bunch of shims of various thicknesses so that I could get a combination down to 1/20mm (2 thou).

Last time I did it on my street S4, the objective was obvious enough though : when the wheels goes up and down, if the line is tilted that means there is some steering input into the wheel; when getting closer to the sweet spot, the line get closer to vertical and the radius of the circle increases (if you go too far, the line starts tiliting in the other direction).

Of course one may consider a different setting depending on his suspension (e.g. with springs on the soft side, one may opt for a tuning point a tad higher than the resting point so that optimum configuration is obtained closer to where the car loaded when under braking for instance). I'm not sure how to approach this level of fine tuning and did my car at resting level: anyway the circle diameter is large enough that the line can be considered vertical for at least an inch above and below that resting point, which would cover most of the domain of my driving.
On the following track day I did experience a tremendous improvement : when braking hard at the end of a long straight, instead of a mean shake moving the car laterally (previously considered an inevitable consequence of rubber bushings making the elan "float" as speed by track experts) and me wandering every time if I would hit someone at my side of leave the track, the car would just brake straight. My confidence in the car greatly improved from that day.

[prezoom]

The nice thing about the Elan is, with front steer, bump or droop will almost always induce toe in of the effected wheel/tire. Far better than the toe out when rear steer is employed.

[vxah]

I was thinking also that one needs to get the actual toe setting correct before setting up the bump steer as presumably the change in track rod length will ruin the previous set up? Easy in theory but maybe not so when you have the chassis on trestles! I guess raising the rack location will also change the toe setting..Maybe not enough to worry about though.

[RichardHawkins]

I used the laser attached method to project a light beam onto my garage door (which is reasonably flat and vertical). I secured the steering coupling so as not to accidentally steer while raising and lowering the the front suspension.

A couple of things I don't think have been mentioned.

1. I noticed that when tightening the clamps that hold the rack to the chassis, the rack is stiffer than the chassis mounting brackets, and the brackets deflect (I am using the original rubber bush type mounts). I fitted stiffeners underneath the chassis mounting brackets.

2. The chassis came with shim thickness engraved onto the chassis mounting brackets. The shims I ended up with are much thicker. This bothers me as I don't think I know more about the subject than Lotus, but as I am no better than an average driver I will probably not notice.

Nearly forgot, Brian Bucklands book shows a simple way to measure rack mounting bracket position. Page 424 in my copy of his book.

Richard Hawkins

[661]

To make things reproducible if you are tracing a line up and down several times, ensure you have the chassis bolted down so it doesn't shift.

I used a bump steer gauge.

[baileyman]

This issue of bump steer with a bit of lock on is interesting.

Thinking about the triangles involved, if a bit of lock, the outside front will have the rack pivot moved inboard, tending to move the virtual intersection further out the near side. (Actually the intersection no longer intersects...) The pivot being further inboard would pull the steering arm more quickly in bump or rebound. The inboard wheel with the pivot further outboard should be pulled less quickly and perhaps even be pushed. But as the racer said, the unweighted wheel is pretty irrelevant. Understeering into a corner and then hitting a bump seems like excitement!

Then if the outside corner is in bump, it seems it would be pretty easy to get beyond the roughly linear travel initially into bump and exacerbate the above effect.

Better to scoot around the corners without lock!

John

[prezoom]

My bump steer gauge is similar to Graham's, dual dial indicators, but the plate that is fixed to the hub is much wider and is marked at the outside diameter of the tire, where the rubber meets the road. The dial indicators read from these marks.

[vxah]

With those cunning bump steer gauges, how does one set the toe before starting the job? I might have to build up the hubs to put a pair of wheels on and adjust the track at assumed ride height with my Dunlop gauges?

[john.p.clegg]

Might be worthwhile waiting and doing at at ride height rather than "assumed" ride height ?

John

[john.p.clegg]

Might be worthwhile waiting and doing at at ride height rather than "assumed" ride height ?

John
Edit..Double post???

[vxah]

Perhaps I should not say presumed, I have adjustable platforms so I should say "where I want the ride height"?
Or, with the "desired" lower arm angle?