CSR bump steer

[Shortshift]

No sooner than one issue is resolved than the next one rears its head.

First decent drive in the CSR two days ago, following a lengthy tidy-up project that started as soon as I bought the car last year.  Would you believe it - the car is quite lively as soon as the road gets bumpy.  It hasn't been flat floored but it is pretty well set up with static toes (both front and rear) being spot on, camber settings (front and rear) as per spec, ride heights and rake sensible.  So it was a tad disappointing.

Gave me something to do yesterday afternoon, though, and what I found by measuring the o/s wheel was that going from full droop/rebound to full bump/compression (vertical travel of around 150mm at the wheel) gave a total change of toe (toe-out, actually) of 12.3mm , or something close to 1.8 degrees, over that range of travel.  I've plotted it out and it's fairly linear throughout the travel range so approximately 3.6mm of toe out (more than 0.5 degree) sweeping 20mm either side of nominal ride height.  Just to confirm: from nominal ride height, toe-out going into bump and toe-in when falling into rebound.  Should add that the rack is already sitting on what looks like 6mm of spacers.

Now, I'm used to seeing recommendations (mainly on S3 cars?) to add shims under the rack to "improve bump steer" but I seldom read whether the problem that's being fixed is toe-in on bump or, as I have here, toe-out on bump. My own analysis suggests that I need to reduce the height of the rack to reduce toe-out on bump...

Need a buddy check on this, please.  Lower the rack to reduce toe-out on bump?

And any CSR-specific experience would be welcome, too.

James

 

[SM25T]

If you do adjust the height of the rack, do check that you still have clearance around the gaiters where they pass through the sideskin. I had to enlarge the tops of the apertures on my SV after the rack was raised.

[Shortshift]

I know, Ian, and I really don't want to get into that unless there's no other viable option.  But a first step would be to change the height just to see how much alteration is needed (and in which direction - up or down!) and I can then decide on whether the aperture is an issue or not.

I've also been looking at variable height track rod ends (as per Jack Webb Motorsport) but (a) I don't want to have to drill the existing control arm castings to take these and (b) I am also concerned that they may not work in the direction that I need them to (equivalent to reducing rack height).

James

[7 wonders of the world]

James,

the common thought on here for an S3 is add a 9mm spacer.....

However having checked a few now I find them all to be different and some need lowering, sometimes you get in and out at the same time on opposite wheels, possible due to the variation in machining on the older spitfire uprights and steering arms.

Has your castor been checked as a variation side to side does effect bump steer.

Everything must be correct prior to checking, ride height, travel, toe, camber, caster etc all affect bump steer.

You must also rest the static toe after making adjustments, prior to rechecking otherwise you will chase your tail since resetting toe alters the length of the track rod and therefore its arc.

Assuming your checking the front but have you also checked the rear on the CSR too as being IRS it can also suffer.

I used rack shims and I now run spherical joints instead of normal TRE's  which can be shimmed too, since altering the different ends affords different results depending on your requirements side to side.

BM if you need more.....

[7 wonders of the world]

I did'nt use JWM for that reason ! mine are tapered..

[Shortshift]

Thanks Neil

I haven't checked the n/s wheel yet (nor the rears), just the o/s, but this wheel alone is exhibiting more than enough toe change for it to be a cause of bumpy road wandering! 

Your taper-fit spherical joint TRE sounds interesting.  But I still have the concern that if I need to do the equivalent of removing shims at the TRE (to bring the ball joint lower, or closer, to the control arm) then this isn't possible, irrespective of the type of TRE I might move to.  Easy to add shims with an adjustable TRE, of course, to increase the distance between control arm and ball joint centre, but going the other way...?  Which is why I fear that lowering the rack might be the only option, assuming I am right about the need to lower the rack to reduce toe-out with bump/compression.

James

[7 wonders of the world]

the rose joints sit about 6mm lower than the OE TRE's, so it puts you at a different starting datum point.

I checked both wheels together using reward lasers on mm caliabted toe scales on the rear wheels, just by moving through your full travel range you can see slow movement and qucikly highlight and rapid 'kicks' in change, usally where the track rod goes through its longest point.

I checked mine on max and min travel and 2 points in between based on where the kick was, I set the front on plastic slips, jack under the chassis and measured to the floor from the lower nose mounting tab as it was a clear easy point.

I would suggest you increase your checking range as you might as it nears full bump it does something odd.

I ended up with lots of test data notes, a ball aching job and I tried lots of options untill I was happen with what I was seeing.

[Tom_Arundel]

Have you fitted new tyres recently?   Have you checked the tyre pressures?

[Shortshift]

Thanks Neil.  If the Rose joints are 6mm lower (closer to the control arm) than the standard ball joint, then that sounds veruy interesting.  I've sent a BM but woudl really like to more details!

No worries on the measurement aspect; I used a laser device and measured around nine points through the full wheel travel range (as I said, around 150mm of travel).  It proved to be repeatable and consistent, and the plot is reassuring as to accuracy and correctness.

Looking forward to learning more about your spherical joint TREs!

James

[Shortshift]

Tom - thanks, and yes I have fitted new ZZS's (a very tight fit under the CSR arches) and they are correctly pressured but that's not really a consideration as it's the measured wheel travel and associated toe change that is confirming the excessive (bump steer) movement.

James

[Paul Deslandes]

James

if I read your measurements correctly, the direction of travel is the same as I had on my S3, the car going toe out on bump (compression) and toe in on droop.  A large part of my problem was that the car had been incorrectly repaired after a shunt and the rack lower clamp halves were missing   Replacing these and adding another, I think, 3mm cured the BS completely.

My Low Flying article dealt mostly with the laborious measurement process but the measurements might be of interest.

http://www.fastgrandad.co.uk/documents/Measuring%20and%20Correcting%20Bump%20Steer%20on%20a%20Caterham%207.pdf

Paul

 

[Shortshift]

Thanks Paul.  I had, of course, searched and read your article before starting.  I am using a different method (with a laser fixed to the wheel hub and projecting onto a board some 3m distant) but the results are the same.

I have just completed measuring the n/s wheel as well and (somewhat pleasingly) this exhibits the same characteristics as the o/s wheel: increasing toe-out (in a fairtly linear manner) as you go from full droop/rebound to full bump.  It's just that there's too much of it!

I'm pretty sure this trait can be tackled by lowering the rack (remember - it is already sitting on 6mmm spacers) but this will bring its own challenges (impact on column, possible enlargement of body apertures).  So my preference is to tackle it at the track rod end, if I can. 

Someone tell me - what is the right way to adjust the TRE in this case?  If the right course of action is to lower the rack then is it correct to raise or lower the TRE joint centre, relative to the control arm, to achieve the same effect as dropping the rack?  I appreciate I should know this, or be able to work it out (and I'm confident I could, given time) but maybe someone here just happens to know the answer...

James

 

 

[paul jacobs]

Oops, double posted, see below.

[paul jacobs]

In the days that I was looking to give a little more clearance under the sump on my 260, I played around with raising the front ride height to see what would happen - bump steer is what happened, this disappeared as soon as I lowered it.  In the end, as detailed elsewhere, I just raised the engine a bit, but that's another story.  So, just try dropping the ride height a bit James, and see what happens.

[7 wonders of the world]

No altering the TRE does not have the same effect.

Lowering (or raising) the rack alters the relationship of the pivots points to one another, upper & lower wishbones and track rod.

Altering the TRE height maintains the pivot relationship but alters the datum point for the TRE in it's prescribed arc, by looking at what your trying to dial out you can gauge which to alter by looking where the problem is occuring, ie if your seeing a sudden kick around the point the track rod is at its longest in relation to the other wishbones

This is where a laser is ideal as its a very visual thing and easily spotted as you can take the suspension smoothly thro full travel with a jack.

[Shortshift]

Thanks Neil.

Hmmm...  The laser curves are very smooth; there's no sudden point of inflection or sudden change of curve throughout the full sweep of wheel travel.

I have just removed the spacer (turns out to be 5mm, not 6mm as previously estimated) from under the n/s rack mount and that side is now pretty much bang-on.  Around a 90% reduction of toe travel as I go from full rebound/droop to compression.  Result! 

But there are reasons why I don't want to drop the rack (need to enlarge the body holes around the rubber bellows; all sorts of restrictions to do with the steering column on the o/s).  So, can I get the same effect as a 5mm lowering of the rack by TRE adjustments?  I was thinking that a spherical/Rose joint arrangement with height adjustment (as JWM or by other means as suggested by Neil) would do the job.  But now I'm not sure!

James

[7 wonders of the world]

It wont be like for like in dimension, but you should be able to achieve the same result maybe wit a 3mm spacer and shims on the TRE.

A good selection of penny washers is handy for testing rack heights and then have shims cut if required.

I did exactly what you are doing, 

I dropped my rack back so my gaiters now clear the side skin and with a little playing with both ends improved the results with a combination of the two.

Don't forget those results will alter slightly once you drop the other side of the rack as that will see-saw this end back up a little.

[7 wonders of the world]

Forgot to mention earlier James, you need to lock the steering dead straight otherwise small movements in steering angle will allow caster and camber to influence your results.

I used a small ratchet strap thro the steering wheel and secured in the holes in the lowered floor.

[Shortshift]

More good advice, Neil - thanks.  I locked the steering column with a wedge under the column sliding adjuster piece - it wedges between the bolt heads (which are inverted) on the adjuster and the horizontal plenum/bulkhead shelf.

More progress.  A bit more fiddling around showed that going from 5mm to no spacer under the rack did reduce the overall toe movement dramatically (as #16 above) but it also introduced a touch of inflection in the curve - going from initial toe-out to toe-in around and above the nominal ride height position in the sweep from full rebound/droop to full bump, which is not good.

So I have made and tried a 1.5mm spacer and that seems to be pretty good on all counts.  Toe change through the sweep is always in the same (toe-out) direction when going from full rebound/droop to full bump, and an 80% reduction from where I started.  I'm still concerned about having to alter the body hole size, though, which is why the TRE solution might be the best bet in the end.

I'll try the o/s now with a similar 1.5mm shim, but I think I'll end up with steering column/body foul problems around the CSR pedal box area.  But I want to confirm the effect on bump steer on the o/s with this thin shim, check back on the n/s (the see-saw effect Neil mentioned), and also check for clearance around the body aperture before committing to the adjustable TRE solution.

Interesting day!

James

[Shortshift]

A further post to round this off and to explain the final outcome.

The more I thought about it, the more I realised that things just weren't right.  One of the plus points of the CSR is that it is meant to have optimised suspension and steering geometry without needing things like rack spacers/shims, yet here was my car with the rack sitting up on 5mm spacers which I had already established were having a serious, adverse effect on bump steer.

So I decided to get to the bottom of the issue and address the root cause of the problem.  I had already found that lowering the rack caused the upper steering column to foul and then bind solid in the pedal box/bulkhead area (remember that the CSR is quite different from a S3 here).  So I removed the cheese wedge and - wow - signs of fettling in and around this area (presumably from when the car was built) just to get the column though to the rack, even with the rack raised.   Hmmm...  Maybe the case some 15 years ago of: "we need to do something to get the car out of the door - and to hell with the geo.."  Piccy of the condition ex-build below:

It took me the best part of yesterday to cut, file, fettle, refinish and generally make the mods that were needed to give more clearance around the column upper - and I still need to get or make some small ali infill panels to replace the ones that were (ahem) sacrificed in the process.  But I finally got to the point where I could remove the 5mm spacers under the rack, drop the rack on to the bridge, optimise the column/rack alignment and still have at least 10 thou (!) clearance all around the column.  Phew.

But the results make it worthwhile.  I'm including a couple of images of some very rough plots below for the O/S wheel (the N/S shows an even bigger improvement) but you can see the dramatic improvement that has been achieved just by reducing the rack height and without the need for adjustable TRE's to raise the outer joint height. 

In summary, over the full range of travel (approx 200mm of wheel travel) total toe-out has reduced from 13mm down to 3.6mm but, perhaps more importantly, over the 20mm of travel either side of nominal ride height (ie: 40mm of travel in total) toe-out has reduced from 2.7mm down to just 0.3mm - so a 90% improvement.  I still need to try it on the road but I think the outcome will be clear.

I also had a further concern that, in dropping the rack height, the rack gaiters/bellows might end up a tad close to the lower edge of the body side panel holes but, with the car sitting on its wheels again, it looks as though this will actually be OK as well.

So - faith in the claims made for CSR geometry has been restored.  Not sure if this was a one-off or not.  So whilst I'm quite prepared to believe that the build/column alignment issues on my car are not representative of all CSRs, maybe this thread will be of help to others who come across similar issues.

Many thanks to everyone who chipped-in with help and advice on here and elsewhere - particularly Neil (7WOTW) and Jack Webb of JWM. 

(By the way - whereas I had to lower the rack on my CSR to get good bump steer performance, I understand that S3 cars often need the rack to be raised to address the same issue - so please be aware that what I've described here may not apply directly to the majority of cars).

James

 

[Shortshift]

And maybe Jonathan could do the honours and correct the orientation of the two graphs for me??

James

[Jonathan Kay]

 

 

[Shortshift]

Thank you Jonathan!

And, just for completeness here's the near-identical improvement for the n/s wheel - toe deviation reduced from 13.0mm down to 2.6mm over the complete stroke, and from 2.6mm down to just 0.3mm over the +20mm/-20mm range of wheel travel either side of normal ride height.

James

Edit: ignore the red legend saying "O/S" - both curves are for N/S.

[Jonathan Kay]

[Shortshift]

Thank you again, Jonathan!

And just to clarify that both curves in the image above are for the N/S wheel - ignore the erroneous reference (in red) to "O/S".

James