Bump Steer What is it and How to Measure It!

Bump Steer What is it and How to Measure It!
A Series of Email postings April 2000

******* FIRST MESSAGE **********

From: "Ian Phillips" <ian@ianphillips.demon.co.uk>
Subject: Measuring Bump steer

When I posted my message re bump steer I forgot to ask the main question!

An earlier message I read mentioned the method of measuring bump steer and I wonder if someone could point me to it?

Ian '68 S4

----- Original Message -----

From: Ian Phillips <ian@ianphillips.demon.co.uk>
Subject: [LotusElan.net] Measuring Bump steer

I have recently joined this Elan list and am following the Bump Steer topic with interest as after '(yet another) full rebuild, my 1968 S4 just does not drive like it did when I first bought it 32 years ago. Its out of warranty now!

The steering generally feels 'dead' and the front of the car is higher off the ground than it used to be. The coil spring damper units are standard and I have tightened all the suspension nuts with two two persons in the car and two more on the front wings to try and lower the front.

Does anyone have any ideas?

***************************************************************

From: Keith Franck <kdfranck@pacbell.net>

Ian,
The archives of this list on Onelist should have a posting called 'Autocollimate the Bump Steer' which I posted about six months ago. Try it, you'll like it. Particularly if you've experienced the dial indicator method before. It takes about one thousandth the time to do a measurement (two seconds vs 30 minutes). I'm guessing it's also at least one hundred times more accurate.

***************************************************************

From: Keith Franck <kdfranck@pacbell.net>

Peter,
Okay, try holding the steering wheel still while tracking straight ahead over moderate bumps with someone following behind about 50 meters. They should be able to tell if and how much the car is darting from side to side. It is hard to do while driving because unconsciously you are reacting continuously.

There is a correlation with increasing speed, the inertial effects are more pronounced and easier to observe and sense. Therefore, a race car at speed can have really wicked bumpsteer and it's easy to observe. A few I've seen would've spun on the straightaway if it weren't for the driver fighting it.

As far as the most correct position to place the rack to minimize the bumpsteer I think you'll find every car needs it's own solution derived. If one spatial solution worked I think Colin would have gotten it spot on. If one were to measure many cars I'll bet the stock position is dead center of the envelope though.

The biggest benefit is by minimizing the bumpsteer it helps to reduce the driver workload.

If the setting is way off and you have adjusted the toe with the car unladen then the toe is wrong when you get in it. This also slightly effects the Ackerman alignment for those of you who enjoy the slip angle limit.

A U-shape curve is what I referred to as a reversal.

Could I ask you to post your opinion of the autocollimation technique? A few words of caution are appropriate here concerning this measurement technique. The precision of this process can be ultra sensitive so don't get caught up with trying to make everything perfect.

 

***********************************************************

From: "Peter Boegli" <peter@rolotec.ch>

Keith
Thank you for your elaborate eMail. As soon as I shall have the car back on the road (second half of 2000) I shall test the result of our "bump steer exercise", and I'll keep you informed.

As to your autocollimation technique - I find it super-easy to apply. No fiddling with cumbersome gauges trying to make measurements repeatable in sub-millimeter precision. And its inherent precision seems to be much better than actually needed. So even a sloppy setup gives a correct first impression. In fact the biggest problem I had was the laser pointer automatically switching itself off after about 1 minute. I then had to

restart it manually and its original position was lost. But even so I managed beautifully. Thank you for posting a really helpful method.

Peter

 

***********************************************************

From: Keith Franck <KDFranck@lbl.gov>

Peter,
The laser pointers can be a pain. Would you like me to describe a simple white light focusing source you could cob together for about $25 and about an hours work? My dad built his in the sixties.

--Keith

 

***********************************************************

From: Rob LaMoreaux <RLaMoreaux@BTE-US.com>
Subject: RE: Bump Steer - Low Tech 29 cent Pencils [SSHG Reference#: ]

All the wear points in the front have been replaced, and the rack is the height specified by the manual. I was trying to undertand the causes, and how much of the bump steer is designed in. If there was a published distance from the lower control arm pivot to the steering arm ball joint we would be able to determine how much bump steer was designed in.

I haven't measured bump steer on my car yet, but I suppose i should dig out one of my Helium Neon lasers and see what I get. This week though I think I will work on the engine parts since it got cold in Michigan again.

***********************************************************

From: Keith Franck <kdfranck@pacbell.net>
Subject: Re: Bump steer

Peter,
May I add one more major point of adjustment which can have a huge effect? The steering arm to tie-rod rotation point can be moved up or down to help minimize the slope and even reshape the profile.

For listers with limited technical experience, please don't heat up your steering arm with a torch and bash it with a hammer.

If your car was really out of sorts then the bump steer tweaking will be apparent.

My pet peeve with the Elan is I've got to constantly compensate for the car wanting to veer off the crown of the road. I'm running the stock suspension still.

--Keith

***********************************************************

From: Rob LaMoreaux <Rob_LaMoreaux@compuserve.com>

I was thinking in the shower this morning about bump steer and how I could be sure my Elan won't have a problem after I get it back together and I came up with a couple of ideas why some cars would be different than others. I would like to bounce them off everyone to make sure I'm thinking correctly. Of course these are probably written somewhere, but I learn better figuring things out. So here goes:

Given:
Bump steer is caused by the steering arm not being parallel with the lower control arm.

I am not sure it is quite as simple as that. As far as I know bump steer is defined as unwanted changes of wheel direction with suspension travel.

The Elan front suspension is basically a 'Watts' linkage composed of unequal length members. Apart from the top and bottom radius arms not being parallel to each other their pivot points are at the corners of a trapezoid.

My understanding of the requirements to minimise bump steer is that in 'straight ahead' position,

1, The pivot point of ball joint on the end of the rack should be on a line drawn between the inboard top and bottom suspension arm inner pivots.

2, The outer track rod ball joint should be on a line drawn between the upper and lower pivot points of the front hub assembly.

3, The actual height of the rack unit is not what matters (at the design stage) but it should match the height of the outer ball joints. so that as the top and bottom links travel through their (different) arcs, the arc described by the end of the track rod stays in 'contact' with the line in (2) above.

I hope that makes some sort of sense, when built up the steering on my new chassis I intended to leave off the coil spring damper units and manually move the front suspension up and down with a straight edge fastened to the face of a brake disk to act as a pointer. That way I could have seen the effect of shimming the rack. In the end I just used the packing sizes marked on at the factory and have the nagging feeling that it is not quite right.

Ian

***********************************************************

From: brassrngfm@aol.com

All - I just recently checked bump steer on my Elan S2 by fabricating a piece of 12 ga. steel - bolted to two wheel studs on the front hubs - with an opening and clamp that would hold a #2 pencil firmly, without wobble. I traced the original line in black onto a chart taped to a piece of plywood I had attached to the floor so it would pivot, but wouldn't move (laterally), backed it off and then assembled the steering to the wheel and repeated with a red pencil.

Unless I'm missing something (which is definitely possible) it seemed to work like a charm.

Paul
New Richmond, Ohio

***********************************************************

From: abstamaria@syciplaw.com

Rob,
I spent a lot of time when assembling my new Tony Thompson chassis on bump steer, using a Longacre bump-steer tool. I kept ending up with a fair amount of shims (about 8mm per side), so this woried me a bit and I kept measuring, convinced I was doing something wrong. But the results were the same.

One of the Elan Listers, Randall Fehr I recall, then confirmed that in his experience a fair amount of shimming was always required. That was very helpful, so I used the shims indicated by my measurements. I would suggest you use new wear items (bushes, ball joints, etc.), and just measure carefully.

My car is almost finished, but no track time yet.

Regards,
Andres
Manila

***********************************************************

From: "Paul Lane, Jr." <planejr@worldnet.att.net>

I have been trying to think of how to resolve the bump steer thread amicably. Racer Parts Wholesale (1-800-397-7815) has a bump steer gage for $119.99. It can be used with removing spring/damper or, for those with bolt-ons, with the suspension assembled.

***********************************************************

brassrngfm@aol.com wrote:

All - I just recently checked bump steer on my Elan S2 by fabricating a piece of 12 ga. steel - bolted to two wheel studs on the front hubs - with an opening and clamp that would hold a #2 pencil firmly, without wobble. I traced the original line in black onto a chart taped to a piece of plywood I had attached to the floor so it would pivot, but wouldn't move (laterally), backed it off and then assembled the steering to the wheel and repeated with a red pencil.

Unless I'm missing something (which is definitely possible) it seemed to work like a charm.

Paul

***********************************************************

Subject: Re: Bump Steer - Low Tech 29 cent Pencils

Paul,
You are the second person to use a similar method. The methodology is sound but suffers from low resolution because a pencil is only about half the length of the wheel diameter. Plus the amount time you invested building that setup you could have assembled an optical metrology solution that is at least 100 times more accurate. At least the data acquisition time is similar, that being seconds, you have a graph and saved about $50. The bottom line it's probably okay for highway speeds, so give it a try.

--Keith

***********************************************************

From: "Paul Adamson" <paul@adamson43.freeserve.co.uk>

Sounds good to me.

Hold on - Thanks the method I used! Major advantage it that you get a nice permanent graph to impress your friends!

Paul.

***********************************************************

From: "Peter Boegli" <peter@rolotec.ch>

Rob
Just my 2p:

There was a posting on this list from Jacques Mazel, dated 13.11.1999 containing

In 1997 I wrote to the Dereham Lotus club, about rack shims, and Graham Arnold answered to me:

"The distance from the centre line of the lower wishbone to the top of the rack mounting should be 2.19 inches plus or minus .005 inches."

IMHO your theory about bump steer is close, but not quite on the spot. FWIW let me try my version:

When a (front) suspension assembly moves up and down, the the ball joint at the end of the steering arm (outer BJ) describes a curve in space. What do we know about this curve? Because the two axis' of the upper and the lower wishbones are parallel to each other, the curve lies in a plane perpendicular to either of the wishbone axis. If the two wishbones were of equal length and parallel to each other, the curve would be an arc, i.e. circular. If we place the inner ball joint of the tie rod at the center of this arc, we have zero bump steer. This also implies that the length of the tie rod should be equal to the length of either wishbone. Lets hope that the sum of the lengths of both tie rods plus the steering rack is equal to the distance of the outer BJs. Otherwise bump steer cannot be set to zero.

In real (Lotus) life the wishbones are not of equal length and therefore not always parallel, and the outer BJ does _not_ move on a circular arc. There is not _one_ definite center of the curve the outer BJ describes, but for each position of the suspension there is a momentaneous center, which are still in the plane passing through the outer BJ and perpendicular to the wishbones' axis. The art of cancelling bump steer is to find a point in the middle of all these momentaneous centers, valid for the mostly used range of suspension positions. The are three dimensions to play with:
1) up/down: i.e. shimming the steering rack


2) fore/aft: i.e. variation in rack mounting block shape (minor influence as long as the inner BJ is not too far out of the plane of movement of the outer BJ)


3) left/right: given the length of the steering rack, we can shim the steering arms to shorten the tie rods. If we shim each steering arm bolt by an unequal amount we also alter the operating point in the Ackerman function. Maybe we want that.

It is easiest to measure bump steer after having removed the spring/damper unit and with all rubber bush bolts/nuts slightly loosened. It is easy then to detect any worn trunnions, bushes, ball joints or mounting holes in the wishbones. There is no point in trying to eliminate bump steer if there are any worn parts in the suspension. If anything is bent, it is either visible or highly probable that left and right shim thickness results to be very different. If this is the case I would put every effort in finding the cause.

So why would there be differences in rack shim thickness at all? I can think of a number of causes:
* Wear of trunnion mounting holes in the lower wishbones


* Measuring errors in determining bump steer.


* Chassis assembly (welding) tolerance i.e. steering rack platform height vs. position of wishbone fulcrums


* Lotus designed in some bump steer, but never published it. Some purists (such as us) try to cancel bump steer and so arrive at different shimming. Some (lesser) purists only partially cancel bump steer and arrive at some other shimming.

As I wrote somewhere else I decided for myself to cancel bump steer as well as possible and then assess the result both on the road and on the track. Hopefully I can gain some insight in the intents of Lotus concerning bump steer on the Elan: Is the car less nervous on straight but bumpy roads at the price of losing its responsiveness in the twisty parts? Give me a few month's time to finish my restoration work and I shall share my "wisdom" with all interested.

Sorry for this lengthy posting, I hope it is of any value to you.

Peter