Caster/Camber/Steering shims

[rjwoods77]

I know this has been talked about in the past a little but I would like to start a thread that kinda gets the questions answered once and for all. Its all a comprimise as we all know but I would like to try to find the overall solution the best I can.

Camber Adjustment:

1: shim between the ball joint and the upright?
2: shim at the a-arm and the frame?
3: shim/lengthen individual aarm pieces(Auburn)?

If using shims for Number 1 does it matter if they are tapered or flat?

Castor Adjustment:

1: shim at the a-arm and the frame?
(upper aarm rotation)
2: shim at the a-arm and the upright?
(longitudal slots on a-arms at balljoint)
3: longitudal translation of upper a-arm?


Steering Adjustment:

This is a little more fuzzy due to the many different ways to have a steering arm. My interest is how to maintain the same steering geometry when you change all of these. Since the steering attachment is tied into the upright and/or lower balljoint holder(denny style), when you change castor you are rotating the upright which rotates the steering pick-up point of the upright. This surely will affect any designed in (non-bump steerness)? and any ackerman.

Steering will get screwed up from camber adjustment unless you do Camber style one with a lower ball joint / steering holder or if you have camber shims for the steering node as well.


My choices:

I am personally using camber #1 and castor #2. Camber #1 because i think it is the correct way to go and castor #2 because I wanted to change castor by rotating the rim/tire by its axis and not by just moving the upper or lower point because it allowed me to better compensate for the steering change. I didnt want to use castor #1 because it made for a variable castor change which is too many variable for me personally to keep tabs on. It also screws up your steering because of the arc that it swings on. Basically my intent was to keep the wheel moving in consistant arcs with suspension movement. I have gotten the changes real small(under .010) and in some cases perfect retention of specs. How much I really have to worry about this, well..., i always knit pick so....... I was just wondering what all of you have found about this. Oh yeah. One last thing. With all castor adjustments that dont allow for a "resizing" of the uprights balljoint locations(which nobody does) will cause the a-arms to come together. This will comprimise geometry based on how much the difference of your castor change is relative to how close it was to zero to begin with. My current method can solve this by having variable thickness castor shims. I know this is a ton of shit but I would like everyone to pitch in to find out what the best comprimise is to get consistant geometry with adjustments. We spend enough time working ont he geometry just to see it change, in some cases a bunch, with adjustments.

[rjwoods77]

I know this has been talked about in the past a little but I would like to start a thread that kinda gets the questions answered once and for all. Its all a comprimise as we all know but I would like to try to find the overall solution the best I can.

Camber Adjustment:

1: shim between the ball joint and the upright?
2: shim at the a-arm and the frame?
3: shim/lengthen individual aarm pieces(Auburn)?

If using shims for Number 1 does it matter if they are tapered or flat?

Castor Adjustment:

1: shim at the a-arm and the frame?
(upper aarm rotation)
2: shim at the a-arm and the upright?
(longitudal slots on a-arms at balljoint)
3: longitudal translation of upper a-arm?


Steering Adjustment:

This is a little more fuzzy due to the many different ways to have a steering arm. My interest is how to maintain the same steering geometry when you change all of these. Since the steering attachment is tied into the upright and/or lower balljoint holder(denny style), when you change castor you are rotating the upright which rotates the steering pick-up point of the upright. This surely will affect any designed in (non-bump steerness)? and any ackerman.

Steering will get screwed up from camber adjustment unless you do Camber style one with a lower ball joint / steering holder or if you have camber shims for the steering node as well.


My choices:

I am personally using camber #1 and castor #2. Camber #1 because i think it is the correct way to go and castor #2 because I wanted to change castor by rotating the rim/tire by its axis and not by just moving the upper or lower point because it allowed me to better compensate for the steering change. I didnt want to use castor #1 because it made for a variable castor change which is too many variable for me personally to keep tabs on. It also screws up your steering because of the arc that it swings on. Basically my intent was to keep the wheel moving in consistant arcs with suspension movement. I have gotten the changes real small(under .010) and in some cases perfect retention of specs. How much I really have to worry about this, well..., i always knit pick so....... I was just wondering what all of you have found about this. Oh yeah. One last thing. With all castor adjustments that dont allow for a "resizing" of the uprights balljoint locations(which nobody does) will cause the a-arms to come together. This will comprimise geometry based on how much the difference of your castor change is relative to how close it was to zero to begin with. My current method can solve this by having variable thickness castor shims. I know this is a ton of shit but I would like everyone to pitch in to find out what the best comprimise is to get consistant geometry with adjustments. We spend enough time working ont he geometry just to see it change, in some cases a bunch, with adjustments.

[Charlie]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Rob Woods:
Camber Adjustment:

1: shim between the ball joint and the upright?
2: shim at the a-arm and the frame?
3: shim/lengthen individual aarm pieces(Auburn)?

</div></BLOCKQUOTE>

Not sure what you are thinking of but Auburn's last 4 cars have all had a system that was effectively the #1, although it hasn't always used shims, the a-arm length stayed constant.

[Kirk Feldkamp]

I know all the current Forumla Atlantic cars (Swift 014a) have the shim style camber adjustment. It sure made things easy to change! What do most IRL cars have, Charlie? http://fsae.com/groupee_common/emoticons/icon_wink.gif

-Kirk

[rjwoods77]

Charlie,

Last pic I saw was the upper ball joint connected to one clevis with the other inside it. Almost looked like the a arm was two "tie rods". Maybe I was looking at something else.

[Charlie]

Yes the upper a-arm was two members, this kept them 2 force members so the threads weren't in bending (or at least, neglible bending). I think the idea was also intended that strain gauges could be mounted to measure forces, though it was never done.

Regardless of the a-arm comstruction, the camber adjustment was on the upright....

All the modern formula cars I've seen (IRL, Champ Car, F1) have used shims... I seem to recall learning from a reputable source (unfortunately, can't remember when or who) that shims are not only quick adjustments but they are also more accurate and repeatable than threaded adjustments, which is why they have become so common.

[clausen]

Camber number one is great because it means you don't have to re-toe it after doing a camber change. I don't like the idea of doing big changes this way though, becuase it also changes the KPI and scrub radius. I prefer to do fine tuning this way, and if you need to do drastic changes because you haven't done your tyre homework, or you change tyres, do it somewhere else.

[Z]

Rob, Ok, its the weekend so here is an extra long-winded answer.

First of all, why make any adjustments at all?

1) To fix chassis build errors. This only requires one post-build adjustment.

2) To "tune" handling. The ideal here is to do all your testing with complicated, heavy, adjustable parts, then, when you know what you want, you race with lighter, stiffer, stronger, non-adjustable parts. This is how many big $ teams do it these days. The point is that "adjustability" adds weight, reduces strength, etc..

So, which adjustments really have to be made, how often, and how accurately? And which parts can you leave as non-adjustable? (By accuracy I mean that if the "optimum" is half way between the adjustment "steps", then you won't lose much performance by being at the step either side.)

Taking it from the least important:

CASTOR
======
Range = 0 - 10 degrees (or maybe up to 20). Accuracy = 1 degree (or 3deg at large angles).
(& TRAIL: Range = 0 - 50mm. Accuracy = 5mm)
Castor, in itself, is for beneficial camber change with steering. This is interlinked with static camber and the type of tyres you have. Castor change usually changes Trail and this changes steering feel. It also changes bump steer behaviour, which then has to be reset. In my opinion Castor and Trail aren't hugely important. You have to be very fussy to spend a lot of time chasing an optimum.
LEAST OFTEN CHANGED. (No change necessary on rear wheels)

CAMBER (Static)
===============
Range = 0 to -5 degrees. Accuracy = 1 degree (or maybe 0.5 deg)
Static negative camber is there to compensate for wheel camber and tyre distortion during cornering. So the amount you need depends on suspension roll stiffness and camber gain geometry, and on your tyres. So once you have a good setup in testing you shouldn't have to change it much. But different tyres (rain) or softer springs (track bumpier than expected) may demand changes.
OCCASIONALLY CHANGED.

STEERING
=========
ACKERMANN: Range = from parallel steer to max. dynamic toe-out.
==========
Adjustable ackermann is a good education in testing. I think you will find that the max. dynamic toe-out setting you can get works best in most conditions. So make non-adjustable steer-arm and LEAVE IT THERE.

STATIC TOE ANGLE: Range = -1 to +1 degree. Accuracy = 0.1 degrees.
================
This is by far the most important setting to get right. If camber is wrong you can usually see it either by just looking at the car, or by the tyre wear. But you can't see wrong toe angles by eye. It is very helpful to make a good "toe-alignment" rig - eg. pipes clamped to front and rear of chassis, with taught string lines down each side - and check toe angles as often as possible (certainly after any other chassis change). Bump steer should also be checked often. Note that if the steer-arm is 120mm long (BJ to steer-axis) then you need 0.2mm (0.008") adjustment on toe-link for 0.1 degrees - so fine L&R threads, or differential threads, or fine shims (ok, you can get away with 0.3 degrees error, but you will lose some performance).
MOST OFTEN CHECKED/ADJUSTED!


BOTTOM LINE
===========
Many race teams love lots of adjustments - 'cos lots of "toys" to play with. But this often leads to the teams getting hopelessly lost, at least at the more amateur end of the sport. For example, some go-karts have adjustable chassis torsional stiffness because it effects handling and feel. Do you need this in FSAE? I would say not. Just build a reasonably stiff chassis and leave it like that. Also adjustment adds weight, and extra weight at the wheels is about the worst place to have it (increased yaw inertia and unsprung mass).

So my suggestion would be to make a rod-end-less suspension (eg. all sphericals), non-adjustable uprights (except for steer-arm BJ height, to fix bump steer), non-adjustable lower arms, do camber and castor adjustment on the non-adjustable upper-a-arm chassis mounts via shims, make new a-arms for big changes, and keep checking your toe-angles.

Of course, I wouldn't bother with double wishbones in the first place, but I would provide some camber and toe-angle adjustment. http://fsae.com/groupee_common/emoticons/icon_smile.gif

Z

[Charlie]

Z-

Sorry I don't mean to follow your posts and knock them but once again my FSAE experience disagrees with your design suggestions. I guess I'll start with what I agree with. You did give good ranges and accuracy levels I'm sure teams can start from.

I also agree that castor is difficult to change and once you find an acceptable value, I'd say stick with it. Doesn't mean you shouldn't play with it in testing though, because FSAE might be one of the series where castor matters most. (high steer angles).

In 2003 we built a car purposely with an incredible amount of adjustability. I don't think our added adjustability added more than a few ounces. Some of it was just building easily fabricated replaceable parts such as steering arms, ARBs, etc. Some of it was designed for adjustability, with only tiny losses to mass and no durability issues. That's what engineering is about, finding a solution with the least negative impact and I think we did that. A barely longer steering arm with an extra hole is not adding significant mass or sacrificing durability.

Ackerman is very important in FSAE. And we've been on both sides of our eventual best. Depending on your steer arrangement you could easily get 300% ackerman, so why tell people to get the most when you have no idea what thier limit is? Ackerman is easy to test and you can do it with static toe out if you want, so I wouldnt build an incredible amount of adjustability in it but don't make your car so it can't be changed easily, very bad idea.

Basically here was our philosophy in 2003. We had no idea how to set up a car, and all our testing sessions ended up with more questions than answers since the 2002 car was tough to make adjustments on. So we built a ton of easy adjustability in the 2003 car, and we were able to zone in on a much better setup quite quickly. We had a better driving car than ever, one the drivers got very comfortable with, and we had a great finish.

Saying don't have adjustability because you might get hopelessly lost is totally against what I beleive in. It's possible, yes. Would I rather have a sub-par setup I can't adjust but feel good because at least I won't let my stupid self mess it up more? Definitely not. If you want to learn car setup, add as much adjustability as possible. If you aren't sure why you are changing things, step back and find out, and you won't get lost.

In 2004 we had a much better idea of the ranges we wanted, so we built less adjustability in some areas. So yes, having less or no adjustability will give you a marginal benefit in weight and strength, I acknowledge that. But we also added more adjustability in areas we weren't sure about, like roll center and anti-squat. So I say, if you haven't ever adjusted parameters, and aren't totally sure what they should be, build in quick and easy adjustability, and your performance increases will FAR outweigh any weight penalty (provided you have a good design for adjustability).

[rjwoods77]

Charlie and Z,

Guys thank you very much for your comments. I appreciate both of your responses on this. This cars purpose is to get a practical feel for suspension design. I am the most knowlegable person on the team for suspension and that is scary at best. So I decided to put the car in a postion of extreme adjustability much like Charlie was saying. Good thing is I am very good at designing things and I have gotten the system down to just using spherical joints and system of swappable, pre machined shims. Overall weight added to the car is probably .25 lbs so I am not sweating it too much but I understand your point very well Z. I however dont like toys. I hate extra crap but I feel that this is necessary in order to dial in a range of useful variables for the next car to work from along with any increased knowledge I pick up on the way. Always thinking ahead. Also so of intersted I came across in doing geometry analysis. I found that there are greater changes in suspension geometry if you adjust at the frame and not after/at the balljoints. Thats why I am doing everything at/after the upright. Makes everything a design challenge but I already signed up for that by going with 10" rims anyway. Again both of you, thank you for your comments. Much appreciated.

[Z]

Charlie and Rob,

Despite my extra long-winded post above, I guess I didn't manage to fully and clearly explain my point of view (hey, its a big subject, and I think in pictures).

After reading and re-reading your post, Charlie, I think we are in full agreement! My point is that adjustability is certainly very important in testing, and for learning what is important. But come race day, most of the adjustable bits just don't get adjusted.

I was going to add the comment "Don't paint anything until just before competition.". I agree that well designed adjustable components don't add much weight, but they do take longer to design and build. I think that the "replaceable component" type of adjustability is a good idea (especially for large adjustments) - eg. different a-arms, steer-arms, etc. But since a lot of test pieces never get used more than once (they didn't work) they can be made a bit rougher than normal. Making multiple components is also good practice, and sometimes inevitable when learning fabrication skills. In fact, you can keep "cutting and butting" a simple a-arm until it works, then make a neater "real one" the same size. (Many engineers grew up with a "Meccano" set so they like to do their adjustments with threaded fittings, shims etc. I grew up with an angle grinder and stick welder, so I reckon everything is adjustable! http://fsae.com/groupee_common/emoticons/icon_smile.gif Well, to an accuracy of a mm or so, and it buggers the paintwork.)

Regarding race teams that "get lost". Not so long ago racecar suspensions typically had at least three (sometimes five) chassis mounting points for each end of the a-arm (hence adjustable roll/pitch centres), every dimension of the a-arm was adjustable (3 x rod-ends, giving adjustable camber gain), multiple steer-arm BJ holes, shims by the bucket load, different spring rates, different rocker ratios and rising/falling rates, ARB rates, N x damper adjustments, and I'm sure a lot more that I've forgotten... So after a respray and rebuild - "What f@&*^# hole does this piece go in???". Ok, http://fsae.com/groupee_common/emoticons/icon_smile.gif so that was amateur motorsport, but...

On race day the two main adjustments you want to think about (providing the car is well tested, and not including springs, tyre pressures, etc., etc.) are camber and toe. That was the point I was trying to make.

Z