Bump Steer

[JGC403]

On my '76 Duster, the car has a little bit of bump steer and I would like to fix this. I have a old book "Chrysler Performance Upgrades", there is a section showing a procedure to fix bump steer in these cars. Basically describes a trial and error process, that involves either adding spacers to raise the steering box or ovaling the lower bolt hole for idler arm to re-position it. Anyone familiar with this process? Is there a better way?

 

[69MOPE]

I am interested in this also. Is there a safe way to "bend" the pitman arm and the idler arm?
In theory, to have the least amount of bumpsteer, the horizontal angle of the center link has to be the same as the tie rods, at ride height. That means that the angle where the inner tie rod meets the center link has to be zero at ride height. Even at factory specs, the tie rods hang down. Raising the car, at all, makes it worse.

 

[JGC403]

I am interested in this also. Is there a safe way to "bend" the pitman arm and the idler arm?
In theory, to have the least amount of bumpsteer, the horizontal angle of the center link has to be the same as the tie rods, at ride height. That means that the angle where the inner tie rod meets the center link has to be zero at ride height. Even at factory specs, the tie rods hang down. Raising the car, at all, makes it worse.

To minimize or eliminate bumpsteer, the center link should be parallel with the ground. Because the wheel travels in an arc, the tie rods that go to the wheel need to travel in the same arc ideally.

 

[69MOPE]

To minimize or eliminate bumpsteer, the center link should be parallel with the ground. Because the wheel travels in an arc, the tie rods that go to the wheel need to travel in the same arc ideally.

Correct, the centerlink has to be parallel with the ground, so, that means the tie rods need to be parallel with the ground at ride height.

 

[69MOPE]

To minimize or eliminate bumpsteer, the center link should be parallel with the ground. Because the wheel travels in an arc, the tie rods that go to the wheel need to travel in the same arc ideally.

Also, the arc of the tie rods during jounce must equal the arc of the tie rods during extension.

 

[Dubob]

My chassis book has the process. If I get time over the weekend I can try and post it if there's interest in it. It doesn't seem like it's a trial and error process.

 

[375inStroke]

Having the tie rods level sounds very simplistic. A line drawn through the upper control arm, lower control arm, and tie rod end pivot points must intersect at the same point, and to do so, the tie rod can't be parallel to the lower control arm. Impossible to actually do on a vehicle, but the effect can be measured. One needs to measure toe change on a wheel as it is moved up and down. This is something I'm working on now getting our team's car ready for the next race. It's a Mustang. People buy bump steer kits for them, which adjust the tie rod height with shims. Following are instructions for a Mustang, but measuring on a Mopar would be the same, so reading that will give you a good understanding of what to do. I already have a dial indicator, so I'm just making my own fixture tool. I'm sure you could also tape a plum bob to the front and back of the tire, mark on the ground, raise the wheel an inch at a time, and check the difference. You'll have to measure the change on the ground in each plumb bob, and subtract those values from each other to see the change in toe since the wheel moves in and out as it moves up and down. It's the left and right movement that you need to measure. Make a reference of where the wheel is at ride height, take the tension off the torsions, and move the wheel through it's range of motion in one inch increments. There's always some bump steer, but you want as little as possible +1" and -1" from ride height. You want less than .020" toe change per side for every inch of wheel travel.

[ame]http://www.maximummotorsports.com/Assets/install/pdf/steering/MMTR-1-3-5-6r3.pdf[/ame]

[ame="http://s183.photobucket.com/user/375instroke/media/bumpsteer_zpsed10d4ac.png.html"]bumpsteer_zpsed10d4ac.png Photo by 375instroke | Photobucket@@AMEPARAM@@http://i183.photobucket.com/albums/x265/375instroke/bumpsteer_zpsed10d4ac.p@@AMEPARAM@@183.photobucket.com/albums/x265/375instroke/bumpsteer_zpsed10d4ac.p[/ame]

 

[player1up]

I just went through this last weekend with my 73 Duster. Aligned correctly but would veer slightly when driven over lines in the road or uneven pavement. Not real bad but enough for me not to be happy with it.
My center link was about 1/2" off from one end to the other, but the ears only had to be tweaked ever so slightly.

If your center link isn't too off level, remove the idler arm and bend the ears for the idler arm mount. I was able to bend the ears enough to fix my issue, just tweak it in little increments and make sure the bolt will still go back through. Remember that small movement at the mount translates into a greater distance at the end of the idler arm.
The holes in the ears on my car were big enough to where the change in angle didn't effect bolt fitment.
The ears are easy to bend with a 12" crescent wrench. I may end up adding some bracing to keep it from happening again. Oh, just make sure that the steering box bolts are tight and the box mount doesn't look damaged or have any cracked welds or you'll be chasing a moving target.. literally.

 

[autoxcuda]

On my '76 Duster, the car has a little bit of bump steer and I would like to fix this. I have a old book "Chrysler Performance Upgrades", there is a section showing a procedure to fix bump steer in these cars. Basically describes a trial and error process, that involves either adding spacers to raise the steering box or ovaling the lower bolt hole for idler arm to re-position it. Anyone familiar with this process? Is there a better way?

What specifically is the steering condition you are feeling while driving the car?

Systematically and throughly check every suspension component. This often takes two people. One to move the wheels or steering wheel, while another feels and looks at (12" or closer) every joint for movement.

To feel, put your hand around both ends of the joint and feel for movement between them while the other person moves the component. Move the rim back and forth either side to side to check: tie rod, pitman, and idler. Move the rim up and down to check: upper/lower ball joints and whl brg. Move the centerlink ends up and down to feel: idler/centerlink joint, idler/k-frame joint, pitman/centerlink joint, and steering box output bearing. Move the steering wheel back and forth to check: steering coupler joint, steering box lash: input vs output, steering output bearing, and pitman to centerlink joint.

99% of the time the problem is not a bump steer adjustment. More likely it's a bent, worn, broken part.

.
.
.
.
I have a Longacre bump steer gauge. My car measured out with very good bumpsteer. No reason to adjust further. But when you drive, many would think it had bumpsteer. That is because the steering box is worn out, idler has a little up and down movement (less than 20K miles :sad10: ), and slightly worn tie rods.

So you can have a car great bumpsteer that still follows the cracks in the pavement.

 

[player1up]

Good point Steve. The term "bump steer" seems to get thrown around a lot in regards to our cars and most assume it's a design problem. Your post proves that's not really the case.
Most people define it as this regardless of cause:
Bump steer -- when I hit a bump, my car steers itself. and while that may be a correct observation, the cause could be incorrect geometry, could be bent / worn / loose anything in the system.

Seems aftermarket parts don't hold up either and cause problems way sooner than one would expect. In addition to my idler arm mount being bent, the idler arm itself had some play in it as well. Less than 20k miles here too.

 

[omahamoparguy]

I think in the Herb Adams book, he shows how to check bump steer with a home-made tool. I think he used plywood. Yeah I have the old direct connection chassis book with the procedure. it can be quite involved if you want every thing dead nuts on. Also remember that the steering box itself has a tendency to move around because of flex in the K frame. this can be removed with strategic strengthening of the K frame. The power steering made this issue worse because of the forces imparted on the box during "fast cornering" the idler arm was not the greatest design on the older a-bodies but they were improved a bit in the later ones.

 

[autoxcuda]

I think in the Herb Adams book, he shows how to check bump steer with a home-made tool. I think he used plywood. Yeah I have the old direct connection chassis book with the procedure. it can be quite involved if you want every thing dead nuts on. Also remember that the steering box itself has a tendency to move around because of flex in the K frame. this can be removed with strategic strengthening of the K frame. The power steering made this issue worse because of the forces imparted on the box during "fast cornering" the idler arm was not the greatest design on the older a-bodies but they were improved a bit in the later ones.

Yes, in the Herb Adams book Chassis Engineering he shows a flat object held against the tire at ride height. Then moves the tire up and down (springs off/disconnected) with a jack. Then you see the gap at the front or rear top of the tire against the board.

Richard Ehernberg made a similar tool to the Longacre gauge I have in a Mopar Action magazine with plywood, piano hinge, and dial indicator.

I got my Longacre bump steer gauge at a swap meet for $25 from a circle track racer that was selling off stuff. Came with original digital cheapy dial indicator too (lens cracked on edge). I don't think I could build a plywood gauge with a nice piano hinge and indicators for $25.

 

[HemiDenny]

Having the tie rods level sounds very simplistic. A line drawn through the upper control arm, lower control arm, and tie rod end pivot points must intersect at the same point, and to do so, the tie rod can't be parallel to the lower control arm. Impossible to actually do on a vehicle, but the effect can be measured. One needs to measure toe change on a wheel as it is moved up and down. This is something I'm working on now getting our team's car ready for the next race. It's a Mustang. People buy bump steer kits for them, which adjust the tie rod height with shims. Following are instructions for a Mustang, but measuring on a Mopar would be the same, so reading that will give you a good understanding of what to do. I already have a dial indicator, so I'm just making my own fixture tool. I'm sure you could also tape a plum bob to the front and back of the tire, mark on the ground, raise the wheel an inch at a time, and check the difference. You'll have to measure the change on the ground in each plumb bob, and subtract those values from each other to see the change in toe since the wheel moves in and out as it moves up and down. It's the left and right movement that you need to measure. Make a reference of where the wheel is at ride height, take the tension off the torsions, and move the wheel through it's range of motion in one inch increments. There's always some bump steer, but you want as little as possible +1" and -1" from ride height. You want less than .020" toe change per side for every inch of wheel travel.

http://www.maximummotorsports.com/Assets/install/pdf/steering/MMTR-1-3-5-6r3.pdf

Regarding bumpsteer, on our factory Mopar front suspensions I think it is important that the tie rods run parallel to the lower control arms. It is difficult because of the rear steer design to use the alternate location shown in the chart.

 

[Bozwel]

a search about bump steer came up with this. those readings pretty are much what I measured on my '66 Barracuda. those readings work out to about half an inch of tracking change over the full range of movement (1/4 inch each side obviously).

I made a bump steer gauge from scrap I had kicking around

i'm afraid I would beg to differ about the readings being very good. all zero's would be the best you can get.

anyway the reason i'm looking into this is because i'm changing my Barracuda to RHD and fitting a power rack from a popular old UK car. problem is that the rack is three inches longer than the drag link. so as an experiment I shortened the track rod by one and a half inches to see what that did to the bump steer figuring it'll make things worse.

to my surprise it virtually eliminated bump steer. One inch either side from ride height is was down to 0.005 thou and Two inches from ride height it was only 0.015 thou. (the last inch of extension it went to 0.060) in both directions it went toe out which shows that the track rod is a fraction too short but i'd be surprised at that tiny amount of movement it'll cause a problem.

just thought you may find this interesting?

 

[Bozwel]

so i did some more measurements. Caster, camber and bump steer on lock.
I can see why the standard suspension and K frame gets ditched, the readings are a joke.
nine degrees of Caster change over the full range. four degrees in just the one inch either side of ride height change. Camber changes by a degree over the same two inch range :roll:
also anything other than straight ahead the bump steer goes all to pot. reason for that is the drag link is held in place by the steering box and idler arm which are angled. basically the track rod changes angle by five degrees. when just a few 10's of thousandths can affect bump steer that much angle movement made too much difference. from 5 thou over one inch straight ahead to 50 thou over one inch on 15 degrees of lock.

at least a steering rack moves in a linear motion so that wont be an issue to change bump steer on lock and for the rest i'll just have to do what I can and put up with it.

 

[autoxcuda]

so i did some more measurements. Caster, camber and bump steer on lock.
I can see why the standard suspension and K frame gets ditched, the readings are a joke.
nine degrees of Caster change over the full range. four degrees in just the one inch either side of ride height change. Camber changes by a degree over the same two inch range :roll:
also anything other than straight ahead the bump steer goes all to pot. reason for that is the drag link is held in place by the steering box and idler arm which are angled. basically the track rod changes angle by five degrees. when just a few 10's of thousandths can affect bump steer that much angle movement made too much difference. from 5 thou over one inch straight ahead to 50 thou over one inch on 15 degrees of lock.

at least a steering rack moves in a linear motion so that wont be an issue to change bump steer on lock and for the rest i'll just have to do what I can and put up with it.

Suspension geometry is full of compromises. The rack will have bump steer in turns too.

When your turning at speed that tire gets a slip angle. And the Ackerman geometry for slow speed turning is less needed. An open wheel rear engine race car rarely uses Ackerman.

Yes you want negative (top in) camber gain in roll. Thats LCA and UCA geometry, doesn't have to do with rack or no rack. Most stock cars go positive camber in significant roll so the front tire looses traction before the front (under steers) because it is safer for the average driver than oversteer.

You can install taller upper ball joints to gain neg camber, but you have to check roll center height and movement.

Go install an aftermarket k-member and rack and compare those same measurements. Track width, motor width, turning radius, etc all put constraints on the design.

 

[Bozwel]

I understand that the suspension was a product of the age. had to be cheap to produce and tyre technology was not as it is now.
also don't forget i'm in the UK where it's pot hole heaven and a little bit of wandering means changing lanes

the reason for measuring was to make sure I didn't make things worse by fitting a rack. seems it would be quite easy to make some improvements

 

[Bozwel]

iv'e got more things to try to see what happens to the geometry.

i'll need to move the engine to the left to either centralise it or bias it slightly right, and maybe raise it slightly to clear the rack. but reading up lowering the K frame is another option (and also reported to reduce bump steer?). so i'll try half inch spacers to see what that does and report back soon..

 

[Mopar92]

I have a bump steer gauge that I use for
My late model. I had a weird pitman arm that caused BS. I got the correct Pitman and honestly, I can't see how it would be worth it seeing that I had like .030" of bumpsteer full travel difference. It's actually quite impressive. Maybe look at that first.

 

[AJ/FormS]

Bozwel, post #15
The same issue that gave you an outrageous 9 degrees of caster change, is probably also responsible for the other creepy changes: None of that is normal on any A-body I have ever put on an alignment machine. I think you may need to revisit that, and figure out what is going on. Forget locking the steering wheel. Keep the roadwheel on the side that you are testing,dead ahead. The locked steering wheel will jack the toe around as the suspension cycles up and down, and that will jack the caster readings.

 

[Bozwel]

the bump steer measured is pretty much the same as Autoxcuda's measurements so probably pretty normal. the Caster change is dictated by the angle of the antidive top arm so I would have said pretty normal?

don't forget i'm fitting a steering rack. the reason for measuring everything to see whats happening and what issues may be caused by fitting a steering rack 3 inches longer than the centre link. off the bat it seems that the extra length reduces bump steer.

I can improve the bump steer but the Caster change is fixed unless I relocate the top arm. what I need to do for caster though is set caster at ride height at 4ish degrees instead of the 0.5 it is now. a thicker torsion bar will reduce the movement it has over the soggy suspension.

 

[Bozwel]

so i put the steering back to stock and dropped the K frame by 0.575 inch.

and the result is YES it does drop bump steer to just a few thou. virtually eliminated completely at one inch either side from ride height. Camber and Caster gain are unaffected.

with the shorter steering link it is still excellent apart from the last inch of bump and droop where it jumps a fair bit but still within the original measurements. so it has four inches of travel with very little bump steer.

 

[Bozwel]

this should show the differences easier than just numbers.

View attachment 1715003525

Green is the standard factory set up
Pink represents just fitting the rack
Orange line, which is the best at present, is just a K frame drop on the standard steering.
Blue is the rack and a K frame drop

so looking at the graph my best option is a very slight K frame drop for the rack conversion but there is no more point experimenting until I have a rack in place to fine tune as the suspension arms will need setting into the best position for Caster etc before hand.

 

[72bluNblu]

There's an article on swapping spindles that Mopar Muscle did awhile ago that analyzed the stock suspension pretty well, detailing the caster, camber and toe changes for a '73+ that's been lowered an inch. Keep in mind too that just because the number is lower doesn't necessarily mean it's better. Toe out with compression is an easier thing to deal with from a driver's standpoint than toe in under compression.

The other thing to consider is that ride height by itself makes a big difference. Lowering a car from stock improves the camber gain curves, and reduces bump steer (travel is reduced). A car with the stock K frame and steering that has larger torsion bars and has been lowered an inch or more will have much better geometry than a car at stock ride height. The factory numbers for these cars are pretty good, lowering them improves those. In most cases not very much bump correction is needed at all once the car has been lowered. On most of the mopar platforms the best suspension geometry for handling occurs when the UCA is roughly parallel to the ground. Camber gain is all negative at that point, and the roll centers are much improved compared to stock.

Here's the article, Swapping Disc-Brake Spindles - Mopar Muscle Magazine

 

[Bozwel]

the only problem with that article is that it doesn't show bump steer on partial lock. when I measured that it threw the readings way off.

also it shows that with the B body spindle it increased the bump steer.

but that doesn't matter to me as I have an early Kelsey Hayes disc set up to go on. and the rack being more linear on operation will not have the same affect as to bump steer more on lock. all the readings iv'e taken show toe out on bump and with the track rod being shorter it also goes toe out on droop. so in theory on cornering and leaning the total toe will go 'out' too much but that could be cancelled out by reducing ackerman.

plan is to also run stiffer suspension anyway.


my day job means I got to drive some nice old cars over the years as I work on classics for a living.

MK3 I restored in 2011