Saturday, August 09, 2008

Castor, Camber, Toe, Bump Steer.

Well, after 4hrs every evening this week my front suspension is setup.

Firstly Camber, then Castor, recheck camber, recheck castor. Obviously Anti-Roll Bar is not connected at either end.

I do not know the exact castor figure, as my patience with maths is pretty low. However, I can measure two things with a digital inclinometer and use logic. Firstly I can attach it to the hub on the wheel studs horizontally, zero it. I can then turn the steering 1 turn on the steering wheel. I made a pointer and on the dash and a mark on the wheel to ensure reliable repeatable results.

On both hubs the change on the meter is the same for 1 turn either way of centre. This tells me that both sides are doing the same thing :) They are the same to within 0.1-0.2deg! Thats cause I planished all the lower wishbone shims down to get it EXACTLY equal, this took AGES. I found my measurements entirely repeatable.

At the same time I also ran tests using meter running vertically on the hub, measuring camber change over the steering wheel movements from centre, side to side. Again I got this within 0.2deg each side and keeping 0.1-0.2deg on the other measurement...

My rough calc is 4.5 of castor, which is book figure. It was not a quick job to get this equal...

Satisfied with my efforts I moved on the toe and bump steer measurements.

Now...Something has changed massively here, I never used any great accurate ways of checking it before, however I had shimmed the rack to get ride of any bump in, I had shimmed it give a little bump out under compression. I had a quick fiddle the other day and something in the swap from GT6 uprights to Caterham Spitfire ones, the lack of Trunnions and Mikes wishbones have really changed the desired location of the rack. Actually if anything the system is more suited to the normal rack height than it was. I had to shim to the rack before to get rid of bump in under compression.

Now with my previous settings I was gettings more bump out than before, indicating the rack needed lowering or the trackrods raising to get the suspension back where it was. So I removed the 8mm of shims from rack. This inclines the steering arm more, so as the suspension bumps the arm gets shorter in relation to the kingpin pivot points, so arches faster and shortens quicker compared the lower wishbone. The top wishbone is more steeply inclined to begin with so shortens alot giving camber gain under bump...So the steering arm needs to match the mean radius of both kingpin pivots etc...+/- a little if you want a bit of toe out or toe in.

Anyway removed the shims from the rack, rechecked with the method below using the dial gauge against the hub. I was still getting a little toe-out under compression. I wanted to get the point of having some toe-in, not sure what has changed in my suspension, it had more shims on before, more camber, different parts.... So time came for another mod, as I cannot really lower the rack enough, well maybe I could JUST in hindsight. However never one to turn down the chance to change another part of the car!

Below are some tubes made from good qrade alloy. I knocked these up in a few minutes, its the thinking and planning and why and why not that take the time! The big ones were threaded right through.

I happened to have some TOP quality 1/2 x 1/2" male rod ends, jam, nuts and sidespacers, also a top quality bolt. So this thing was easy to make. Lucky I had just enough thickness on the steering arm to cut the thread further up the shank, I had to shorten the ends of the steering arms by about 3/4", the two ends inside the alloy tube virtually meet in the middle, only 2mm spare. The alloy tube is now a perminant fixture on the end of the rack arm, it has a jamnut obviously, however for some extra safety, it's been throughly loctited in place. It will not be coming off again even if you want, no biggy if the rack needs changing I can make some more, the bar just needed cutting, cleaning up in the lathe and drilling a 11.5 hole through to take the 1/2" threading.

I see some guys in the USA using alloy radius arms and rods to replace the rotoflex wishbone, with rod-ends in each end! So a bit of alloy on a pretty low load joint doesn't bother me.

Any adjustments to toe you just use a 1/2" spanner on the rod end case and undo the jam nut, the large O/D of the alloy tube means steering rack arm twisting is now a hand affair, not a pair of grips.

To test the bumpsteer firstly I need to set the toe parallel, this was done with some string and couple of poles, the steering was then locked.

Obviously also the ride height has been dummied up on the nice flat floor, the suspension is lifted with a range of bricks and shims to the correct place to correspond to static ride height I want and the chassis is all leveled etc.

I then attached the dialgauge to a piece of 8mm angle iron attached to a pair of small steps. I made two very faint lines on the hub either side of centre, to keep where the dial gauge will run the same each time to allow me to get repeatibility. I had to establish a reliable measuring technique, that would give readings within a close range when the steps and dial gauge are taken away and replaced. I got this down to about 0.001"-0.002" or complete repearability within 0.0005. Using the same line point each time I was able to get really good resolution on my adjustments. I initially shimmed up the old track-rods with two 2mm washers before making anything, this gave me still a trace of bump out. So I made the smaller alloy bush slightly oversize.

So then I had my first toe-in under bump, something I do not want bar to see where the cross-over point is from in to out bump steer. I planished down the smaller alloy bush about 0.010" at a time, it was fascinating to see the distance travelled on the dial gauge changing with each adjustment, almost in a linear curve, as you'd predict! I had the front of the hub deflecting 0.088" at the start and the rear about 0.058"...

This tells me the front of the hub is moving further away from the dialgauge under bump. So its toeing in, as predicted and designed in to the spacer thickness for the start point.

I then removed 0.010" and it was 0.075" front and 0.065" rear, then another 0.010" off and it was 0.066 rear and 0.067 front. That final figure is 0.001 of toe-in (in) 1" of unloaded supension deflection (bump).

(above figures are just an example)

I want a little toe-out under bump, this time I want it accurately set! Hense all this work.

I ended up with passenger side setup to give 3 thou bump out over the hub during 1" of suspension travel (REAR 0.075", FRONT 0.072") the driverside 3 thou also( REAR 0.0685", FRONT 0.0655".

I have no idea if these are good settings :) However having no bump steer is good, I can shim up the trackrods again or test different settings from a perfect base now. Just shim the rose-joints evenly.

Having eliminated the bump steer completely I have attempted to give a little toe-out in bump to resist any backwards movement made by the upright, under braking the upright and suspension will deflect backwards a bit as the car tries to push through the suspension! Also the car will dive a bit, so if you setup with a little toe-out under bump, when the front is loaded and the bushes are deflected back a bit, the toe-out you designed in may go back to even, or may need more, or fine tuning for driver style and whatever.

I am happy enough with all that, I am just doing the best I can with the limited equipment.

The setup has been to extremes of movement also, and there is basically no bump steer :) Seemed to gain the initial 3 thou over 1" of bump then gain another 2thou over the next inch. If the suspension is drooped from static placement there is loss of 1 thou of the original toe setting, a very small turn towards toe-in. Everything from static upwards will give either even readings, of as I have tuned in, some bump out in a linear amount.

My shocks and springs were designed to give 18mm of droop past static before the wheel is lifted by the shocker becoming fully extended, this is also adjustable! SO from two extremes, all bump steer has gone bar a TINY amount in the 18mm droop zone before wheel lift, gone, obviously bar the compression bump out that has been added by me.

Each wheel is setup for bumpsteer on its own, obviously. One side needed roughly 2.5mm raising over the standard axis of the track-rod and the other side (passenger) needed 4mm. Obviously the rack was centralised first before any adjustments so each steering arm is the same length from its inner axis.

Shimming up does increase the torsional load on the steering arm thats attached to the upright. However its only 2.5mm and 4mm over normal. So doesn't concern.

The steering arms needed the hole opening out a little to make way for the 1/2" pinch bolt. All paint and coatings removed from mating surfaces under washers and stuff...To get a reliable clamp.
Thank f*** that's over. It may need some final adjustments when its on its wheel's, but just the toe-in I think! Which shouldn't change much anyway, cause there is bugger all bump-steer! That was the bloody idea!

I hope it makes a difference, cause via the textbook it's doing exactly the same thing on both sides!

There was no play or untoward movement in any part, as the hub was locked also... the canley bearing carrier that's currently fitted, that gave me 0.001" of my 0.002" of unexplained movement. You can hear it clonking if you move the suspension and this relates into a change of toe, inorder of 0.001" and a change in camber of around 0.05 degrees. Its fascinating to see the tolerance of a bearing in its housing effecting your measurements, just cause it means your measurements are not 0.5mm or about right, they are microscopic! down to engineering tolerances and lower! My suspension is as accurate as a blueprinted engine!

Shame its attached to a rubber chassis :) oh well.

1 comment:

Anonymous said...

excellent work Dave. Time consuming but worth it as your measurements bear out.