Bump Steer - Adamson Method

[Andy8421]

I haven't found a solid answer from a traceable source on how much, if any, bump toe Lotus designed in. I CAD modelled up the front suspension a while back using the information I could find. The result was some toe, but a couple dimensions used were measured from my Elan and not from period component drawings. IIRC the ball joint height in the steering link and the trunnion to axle height were the 'as measured' dimensions.

Scott,
I have (somewhere) a book about suspension design for racing cars. It suggested that with suspension design similar to the Elan (rack ahead of the suspension) that there is a tendency for the suspension to go toe-in under heavy braking as the front suspension loads up. It therefore might make sense to design some toe-out under bump conditions to compensate. How this trades off against bump effects during cornering is beyond me I am afraid.

Andy.

[661]

As you have it, the pen is following the arc of the wishbone travel AND factoring bump (because the hub is free to move toe in or out on suspension movement). Which is why I thought it interesting that the two lines look so similar. Overlaying them in different colours would be fun...........if you have too much time on your hands.
Having the pen pointing out from the car ( perpendicular, as you say) eliminates the arc interference. A laser would be better as with a pen your board would have to move towards the car and away from the car as the arc is followed.
Frankly it's all a bit of a compromise:

[snowyelan]

Hi Andy,

In general, front vs rear rack position doesn't dictate bump steer.

Front racks typically create a problem with steering ackerman in that they cause increasing toe in conditions when steering angles are introduced. I'm wondering if your book suggests bump toe out as a way to compensate for this effect?

[nmauduit]

As you have it, the pen is following the arc of the wishbone travel AND factoring bump

though this factoring in would only translate (the desired toe in/out variations) in the pressure of the pen onto the writing surface, hence mostly hidden at first glance...

In my modest experience I found that getting both front wheels doing the same thing under heavy breaking and if at all minimizing said bump steer is the dominant effect for an average driver, and certainly what counts for road use. That mostly translates in breaking straight, or not jerking around too much on bumpy roads.
Secondary effects such as displacing the neutral bump steer point above or below, or adopting a non minimal bump steer setting (toe in/out variations not minimal i.e. bump steer curve not closest to vertical at the desired operating points) require quite advance driving skills to be put in action effectively.
As for measuring and setting up bump steer on an elan I find attaching a laser on the rotor (so beaming perpendicular to the spindle) and reading out a few meters back while jacking the front wheel up and down easy enough with front springs not attached. In my experience one can see differences in shimming of 1/10mm to 1/20mm with this method.

[Tmac897]

But as I mentioned above, I think my setup is wrong, or at least suboptimal.

Tony, I am afraid your setup isn't suboptimal, its definitely wrong. As pointed out above, bumpsteer is the change in toe in / toe out as the suspension goes up and down, you are measuring the arc of the wishbones.

Thanks for the feedback, Andy. I went back and read a few more posts and looked at some YouTube videos. As a result, here is my new setup and results:

I cut a couple of blocks of wood from 1x3 stock, about 5" long. In one, I drilled two holes to match the spacing of 2 wheel studs, each being the outside diameter of a stud. In the other, I cut a dado the width of my ruler, perpendicular to the long side of the block, and a little shallower than the thickness of the rule. I put the rule in the dado and then screwed the two blocks together, trapping the rule at the proper angle. The pen was then taped to the end of ruler, and parallel to the long edge of it. This afforded the ability to move the apparatus from side to side without the hassle and inaccuracy of taping it to the hub. It also gave me a flat surface to check for level, which ensured that the ruler was at TDC.

I set up the erasable board on the recommended “two-boards connected with a hinge” to allow the erasable board to ride on the tip of the pen when the suspension was moved up or down. I also drew a vertical line down the center of the board for reference, and for each measurement, started with the pen tip touching the bottom of this line. I made sure that the erasable board and the disc were as close to parallel as possible.

I also made sure that the front suspension was at normal ride height, and that the front-end alignment was zero toe-in/toe-out. I expect to set it up at about 1/16” toe-in for regular operation, but the setup only needs to be close to the final – not exact.

I found from some preliminary tests that accuracy in the setup made a big difference in the measured results. Maybe not an earthshattering revelation, but the little things seemed to matter quite a bit.

With that as the backdrop, I was able to get the whole operation done in about an hour. As it turns out, the passenger side didn’t require any shims. I didn’t conclude that until I had run tests on both sides, though.

Here is a summary of the testing:
Passenger Side (U.S.):
1. No Shims – No bump steer measured
2. Lots of shims – Heavy bump steer > 2” away from the centerline at full compression
Driver Side:
3. No Shims – Moderate bump steer
4. Lots of Shims – Moderate bump steer, but other side of centerline, about the same distance from the centerline
5. Half the Shims – Pen tracked the centerline, i.e. no bump steer
Passenger:
6. No Shims: Pen tracked centerline, i.e. no bump steer

I did this test to make sure adding shims to one side didn’t affect the other side. In Paul Adamson’s original post, he mentioned that changing the number of shims on one side changed the bump steer on the other. I did not see that, but from the time of his original post, I suspect that he was using the original rubber mounting bushings. Those would have been more “deformable” than the aluminum mounting brackets I’m using. While this might change if one side required a very large number of shims, with the four I was using, it had no effect.
For me, this was the “last thing” before I would be ready put the body back on the chassis. Thanks everyone for helping me to reach that huge milestone in my restoration.

I'll post a pic of the setup in another entry.

[Tmac897]

Measuring Bump Steer

Here’s a picture of my setup.

[snowyelan]

Hi Tony,

As I mentioned earlier in the thread, a vertical reference line will give you bump toe in. When viewed from the side the suspension travels up and to the rear at 2.5 - 7.5 degrees from vertical, depending on your particular chassis. For setting to bump steer you need to start with a line that follows the suspension travel. Earlier I suggested disconnecting the steering tie rod to make this line using the same ruler and marker setup you have. A second, likely easier, option is to clamp a similar setup to the upper suspension arms to make the line.

I also had a different amount of shims on either side. Changes on one side did impact the other but I only had to switch between sides a couple times to dial it in.

[Tmac897]

Not sure I understand your explanation. If we are defining bump steer as toe change on suspension movement, then absence of bump steer would essentially mean that the disc/wheel would remain in the same plane throughout suspension travel from droop to compression. If the disc remains in the same plane, then an arm attached at a 90 degree angle to the disc would also remain in the same plane throughout its cycle, I.e., a straight vertical line.

I’m not sure how you get to the conclusion that that equates to toe in.

[alan.barker]

Could it be possible with a Laser light and Graph paper on the Board.
Alan

[Andy8421]

Not sure I understand your explanation. If we are defining bump steer as toe change on suspension movement, then absence of bump steer would essentially mean that the disc/wheel would remain in the same plane throughout suspension travel from droop to compression. If the disc remains in the same plane, then an arm attached at a 90 degree angle to the disc would also remain in the same plane throughout its cycle, I.e., a straight vertical line.

I’m not sure how you get to the conclusion that that equates to toe in.

Tony,

Scott has raised a good point that I had overlooked in the past, that because of the geometry of the wishbones and their attachment points to the chassis, the hub doesn't go up and down vertically, but at an angle to the vertical. If you adjust your bump to a vertical line, then because the hub itself doesn't track up and down vertically, you will be introducing bump steer.

Scott's suggestion of disconnecting the steering arm and moving the suspension up and down to draw a reference line, then reconnecting the steering arm and adjusting the shims to match the reference line sounds like a very good suggestion as it compensates for any other vagaries in the frame and suspension components.

Andy.

[Andy8421]

Hi Andy,

In general, front vs rear rack position doesn't dictate bump steer.

Front racks typically create a problem with steering ackerman in that they cause increasing toe in conditions when steering angles are introduced. I'm wondering if your book suggests bump toe out as a way to compensate for this effect?

Scott,

The effect was due to the compliance in the mounting of the wishbones. Under heavy braking, the hubs are 'pushed' toward the back of the car, while the rack remains in place. This has the effect of causing the wheels to toe in.

Andy.

[ericbushby]

Good morning all,
Hi, Alan, if you look back to 5th November I described using a Laser pen fixed on the hub and a piece of cardboard for my set up. Unfortunately I called it an LED pen. I have now corrected that.
I now realise that to be more accurate my vertical line on the cardboard should have been at the same angle as the
suspension travels, that is sloping to the rear by a few degrees as it rises, if I have understood it correctly.
Should I ever be there again I now know how to improve it.
Thanks for the explanation.
Eric in Burnley
S3SE DHC

[Tmac897]

Good morning,

So what we are talking about is “camber?” I did a rough measurement with a “spirit level” app on my iPhone, and the angle of the hub goes from about 1 degree at droop to about -3.7 degrees at compression.

I did not realize that camber could cause a toe change effect. Thought it was mostly about tire wear.

Once more into the breach.

[nmauduit]

IMG_2032.jpeg

Here’s a picture of my setup.

I will still prefer using the method I used in the past as I find it to provide more sensitivity (larger readable signal for a given variation, here the extent to which the line departs from a vertical straight segment), since the leverage comes from the distance from hub to pen (less than a meter here) while with a laser pointer is can easily achieve several meters, hence helping to discern better lines away from a vertical straight.

In my experience with more or less shims one could get the curve becoming almost flat and vertical around the desired operating point (suspension harms horizontal) while the line would incline one way or the other when altering the shim stack thickness (as well as become more curvy).

Could it be that with the decreased sensivity of a relatively small leverage (plus spurious drag motion due to stick flex or tape play?) you may not discern the fine adjustment between some shims thickness? Maybe it's overkill on my part, but as most of the work is to get the car in position, I'd rather have an adjustment as good as I can get, esp. since I vividly remember the first time I experimented the result on a (bumpy old) track (Monthléry); absolutely day and night!

[Tmac897]

IMG_2032.jpeg

Here’s a picture of my setup

I will still prefer using the method I used in the past as I find it to provide more sensitivity (larger readable signal for a given variation, here the extent to which the line departs from a vertical straight segment), since the leverage comes from the distance from hub to pen (less than a meter here) while with a laser pointer is can easily achieve several meters, hence helping to discern better lines away from a vertical straight.

I would agree that the longer the “arm” used to measure, the more accurate the result. I might try the laser pointer method if I can find one easily.