Bump Steer And Steering Shimming
11 posts
• Page 1 of 1
Gents
Just removed my rack to regrease, change seals and fit poly bushes. In the process looked at the shims on the chassis used to adjust the bump steer (as I want to check the bump steer afterwards) and was wondering If someone could answer me the following:-
1) Why are they fitted?
2) Are the chassis not made up on a jig to a high accuracy?
3) How are the shim dimensions arrived at by Lotus? ie what datum points are used to produce the correct shim heights?
4) Any further info on this topic. on which I am not so well versed much appreciated.
thanks
Mark
Just removed my rack to regrease, change seals and fit poly bushes. In the process looked at the shims on the chassis used to adjust the bump steer (as I want to check the bump steer afterwards) and was wondering If someone could answer me the following:-
1) Why are they fitted?
2) Are the chassis not made up on a jig to a high accuracy?
3) How are the shim dimensions arrived at by Lotus? ie what datum points are used to produce the correct shim heights?
4) Any further info on this topic. on which I am not so well versed much appreciated.
thanks
Mark
-
mark030358 - Coveted Fifth Gear
- Posts: 1172
- Joined: 29 May 2004
Mark - It's been a few years since I shimmed the rack, but it was pretty straightforward. The steel supports for the rack are pretty flimsy, IMO, and with normal mfg. tolerances I feel that the use of shims was required. My car took 0.125 at each side. I made up some shims of varying thicknesses and then determined the proper height by checking bump steer so that the arc traversed by the suspension (without springs and shock attached - of course) was the same both with and without the steering attached to the suspension. I made up a simple jig to check the bump steer - following some old directons from this conference. I just used a jig that had a pencil and some graph paper, although other methods would be more accurate. I started out with thinner shims and just progressed upward in thickness and found , luckily that 1/8 "was what I needed. All you're trying to do is insure that the rack is high enough that it doesn't induce a pull up or down as the suspension moves. My 65 handles quite well over rough roads and I'm pleased with results.
Paul Zimmerman
Cincinnati, OH
026/4600
Paul Zimmerman
Cincinnati, OH
026/4600
- brassringfarm
- Second Gear
- Posts: 83
- Joined: 08 Jan 2004
Hi Mark,
To fix the stack up tolerance errors any assembly of components is subject too. No car manufacturer can afford to produce perfect parts.
Can't speak for Lotus' choice of an acceptable tolerance envelope. Seems well within reasonable engineering accuracy since an 1/8" shim stack could very well be the mean value in a bell curve distribution. The point is it's likely the chassis brackets are offset on purpose so a positive thickness shim stack can correct the rack postion. There is no such thing as a negative thickness shim at least in this dimension. :rolleyes:
Again I never worked for Lotus so no help there. Would venture to guess they measured the bumpsteer and found the stack height by trial and error methods if at all.
Start by reading this article. Their dial indicator solution is lame in that the measurement is tedious to do. It has one redeeming quality in that it mostly only measures one DOF (degree of freedom) which is the bumpsteer. It's not affected very much by the caster inclination.
What's important is the F. Diagram depiction. Line 1 would be the vertical link. Line 2 is the chassis mounting points. Keep in mind this solution only truely cancels the bumpsteer when the car is tracking straight down the road. Turn the steering wheel oneway or the other and it all goes south quickly.
<a href='http://www.racerpartswholesale.com/longtech3.htm' target='_blank'>http://www.racerpartswholesale.com/longtech3.htm</a>
-Keith
1) Why are they fitted?
To fix the stack up tolerance errors any assembly of components is subject too. No car manufacturer can afford to produce perfect parts.
2) Are the chassis not made up on a jig to a high accuracy?
Can't speak for Lotus' choice of an acceptable tolerance envelope. Seems well within reasonable engineering accuracy since an 1/8" shim stack could very well be the mean value in a bell curve distribution. The point is it's likely the chassis brackets are offset on purpose so a positive thickness shim stack can correct the rack postion. There is no such thing as a negative thickness shim at least in this dimension. :rolleyes:
3) How are the shim dimensions arrived at by Lotus? ie what datum points are used to produce the correct shim heights?
Again I never worked for Lotus so no help there. Would venture to guess they measured the bumpsteer and found the stack height by trial and error methods if at all.
4) Any further info on this topic. on which I am not so well versed much appreciated.
Start by reading this article. Their dial indicator solution is lame in that the measurement is tedious to do. It has one redeeming quality in that it mostly only measures one DOF (degree of freedom) which is the bumpsteer. It's not affected very much by the caster inclination.
What's important is the F. Diagram depiction. Line 1 would be the vertical link. Line 2 is the chassis mounting points. Keep in mind this solution only truely cancels the bumpsteer when the car is tracking straight down the road. Turn the steering wheel oneway or the other and it all goes south quickly.
<a href='http://www.racerpartswholesale.com/longtech3.htm' target='_blank'>http://www.racerpartswholesale.com/longtech3.htm</a>
-Keith
- type26owner
- Coveted Fifth Gear
- Posts: 1246
- Joined: 18 Sep 2003
Hey Paul,
Your method measures actually three DOFs (two is wrong, realized my mistake in the shower this morning) and is fraught with danger if perfomed incorrectly. It measures the suspension travel (camber is wrong too! won't reveal where I was when this popped into my head <_< ) at the caster inclination primarily. The bumpsteer is only a change of slope from the caster's. But even more importantly the bumpsteer measurement must be derived by first using the caster inclination as the baseline trace for the bumpsteer to hopefully follow and certainly be measured from it if there's any deviation. What this boils down to is you should measure the camber/caster trace without any bumpsteer by disconnecting the tierod and graphing the line. This might be nearly impossible to complete without rotating the vertical link. For instance If the suspension is the early 7 degrees of caster type then the trace will be a line which is inclined by 7 degrees off of a vertical line.....this is only true because of the trunnion btw. Then reconnect the tierod and do the bumpsteer measurements.
-Keith
Your method measures actually three DOFs (two is wrong, realized my mistake in the shower this morning) and is fraught with danger if perfomed incorrectly. It measures the suspension travel (camber is wrong too! won't reveal where I was when this popped into my head <_< ) at the caster inclination primarily. The bumpsteer is only a change of slope from the caster's. But even more importantly the bumpsteer measurement must be derived by first using the caster inclination as the baseline trace for the bumpsteer to hopefully follow and certainly be measured from it if there's any deviation. What this boils down to is you should measure the camber/caster trace without any bumpsteer by disconnecting the tierod and graphing the line. This might be nearly impossible to complete without rotating the vertical link. For instance If the suspension is the early 7 degrees of caster type then the trace will be a line which is inclined by 7 degrees off of a vertical line.....this is only true because of the trunnion btw. Then reconnect the tierod and do the bumpsteer measurements.
-Keith
- type26owner
- Coveted Fifth Gear
- Posts: 1246
- Joined: 18 Sep 2003
Mark,
I posted a way to do the bumpsteer measurement via an optical method on the Yahoo Elan site quite awhile ago. If you go there and do an archive search all you need to know is there. It measures two DOFS camber and bumpsteer but the horizontal displacement is just from bumpsteer and the caster inclination does not affect this method. Easy, fast, accurate, low cost & the metrologist's way. Can be done at the racetrack if necessary.
<a href='http://autos.groups.yahoo.com/group/lotuselan/' target='_blank'>http://autos.groups.yahoo.com/group/lotuselan/</a>
-Keith
I posted a way to do the bumpsteer measurement via an optical method on the Yahoo Elan site quite awhile ago. If you go there and do an archive search all you need to know is there. It measures two DOFS camber and bumpsteer but the horizontal displacement is just from bumpsteer and the caster inclination does not affect this method. Easy, fast, accurate, low cost & the metrologist's way. Can be done at the racetrack if necessary.
<a href='http://autos.groups.yahoo.com/group/lotuselan/' target='_blank'>http://autos.groups.yahoo.com/group/lotuselan/</a>
-Keith
- type26owner
- Coveted Fifth Gear
- Posts: 1246
- Joined: 18 Sep 2003
Keith - I believe I did it correctly. and recognize that caster and camber are both affected: Disconnect tie rod. Take a trace. Reconnect tie rod. Retrace and compare new trace vs baseline. Recycle, adding shims, n times (where n is a large number) until acceptable. OK? Is there something I'm missing here? The 'fraught with danger' comment made me raise an eyebrow!
Paul
Paul
- brassringfarm
- Second Gear
- Posts: 83
- Joined: 08 Jan 2004
Paul,
Yup, your description would be the correct application of this method. The fellow that originally posted this technique didn't have it thought through like you have and I'm fairly certain he did it wrong. Not an easy or obvious metrology process to tackle at all. Confused me too overnight! You've done well, grasshopper!
-Keith
Yup, your description would be the correct application of this method. The fellow that originally posted this technique didn't have it thought through like you have and I'm fairly certain he did it wrong. Not an easy or obvious metrology process to tackle at all. Confused me too overnight! You've done well, grasshopper!
-Keith
- type26owner
- Coveted Fifth Gear
- Posts: 1246
- Joined: 18 Sep 2003
Gents
Thanks for the help, and I will check the Yahoo group listings as recommended.
cheers
Mark
Thanks for the help, and I will check the Yahoo group listings as recommended.
cheers
Mark
-
mark030358 - Coveted Fifth Gear
- Posts: 1172
- Joined: 29 May 2004
Mention has been made in some replies on this topic to comparing pencil line traces with and WITHOUT the tie rod connected.
With the springs and dampers removed how does one move the suspension over the full travel without applying some movement to the steering axis as without the steering arm it is unrestrained?
Was the 0.125 shim measurement inches or mm? if inches it seems an awful lot.
Ian
With the springs and dampers removed how does one move the suspension over the full travel without applying some movement to the steering axis as without the steering arm it is unrestrained?
Was the 0.125 shim measurement inches or mm? if inches it seems an awful lot.
Ian
68 Elan S4 DHC. Built in a weekend from a kit (just like the advert said)
-
Elanman99 - Third Gear
- Posts: 449
- Joined: 11 Sep 2003
The optical method does not require the tie-rod be removed and a baseline trace taken. The caster contribution is canceled out by the fact the observed position of a reflected light beam is not affected by the X-Y motion of a flat mirror as it moves in the reflective plane. The reflected beam only gets displaced by an X-Y tilting (2dof) of the reflective plane. X=bumpsteer, Y=camber using the right-hand rule. Strike a vertical line and measure the X value relative to the line. Ignore the Y values. Remember the X values are doubled in precision since the light beam source is stationary.
The shim stack is the standard low-cost way to handle this type of alignment problem. It's still done this way today everywhere. Got a better solution?
-Keith
The shim stack is the standard low-cost way to handle this type of alignment problem. It's still done this way today everywhere. Got a better solution?
-Keith
- type26owner
- Coveted Fifth Gear
- Posts: 1246
- Joined: 18 Sep 2003
My 0.125 measurement was in inches. Yes it is a lot!
As for keeping the suspension from rotating - it wasn't easy so my wife helped me.
Paul
As for keeping the suspension from rotating - it wasn't easy so my wife helped me.
Paul
- brassringfarm
- Second Gear
- Posts: 83
- Joined: 08 Jan 2004
11 posts
• Page 1 of 1
Total Online:
Users browsing this forum: No registered users and 42 guests