The Ultimate Guide to Suspension and Handling Part X: Bump Steer/Toe Steer

by Mike Kojima

It's time to think about suspension.  Some of the most common handling problems that you are the most likely to see on modified street cars and race cars based on production chassis, especially in racing classes that do not allow relocation of suspension pick up points are bump steer and toe steer related handling issues.  Bump and toe steer can have significant negative effects on a cars handling which is manifested by steering pull, twitchiness, instability, steering inaccuracy and unpredictability.  All of these things are scary for drivers to deal with.

For most production based modified cars, the most common reason why bump and toe steer becomes a problem is due to over lowering the car.  Most modern cars have bump or toe steer by design and over lowering the car puts the suspension geometry in a position that increases bump and or toe steer to the point where it becomes a serious handling problem.  This is one of the many reasons why we always suggest lowering a car only 1-2 inches at the most.  

So now that we've established that bump and toe steer are common and serious issues, how do you avoid and fix them?  Read on!

Read the whole series and learn how to make your car handle!

The Ultimate Guide to Suspension and Handling

Bump Steer

Steering precision and stability are both affected by bump steer. Bump steer is steering input independent of what the driver is doing with the steering wheel created by the suspension moving through its stroke in response to bumps and cornering induced body roll. It is caused by the suspension links moving in different arcs than the steering linkage as the suspension follows its stroke.


A common cause of bump steer on the popular McPherson strut front suspension car is that the lower control arm and the steering tie rod are out of plane and pivoting on a different axis.  The difference in the arcs is the amount of bump steer that results.  Usually, the further from the design ride height a car is, the worse the bump steer will be as the difference in angles becomes more extreme.
If a McPherson strut suspension has the inner tie rod in line with the lower control arm pivot, bump steer is greatly reduced.
On this race car, a spacer was made to fit in between the steering rack and the inner tie rod to move it outwards which placed it in line with the lower control arm pivot point.  This alone got rid of 80% of the bump steer.  With the stock steering, this car had a severe issue with bump steer.
Page 1 of 6 Next Page
Bookmark and Share
Wednesday, September 03, 2014 8:35 AM
Finally an easy to understand explanation of this. I already knew about but failed to explain to people ...

One important thing is that it is not hard nor (much) expensive to measure bump and toe steer. It can be hard to fix though. The easiest fix being ... raise that car!

All that is needed is a bump steer gauge and some time.

step 1: car on the ground, measure the height of the center of each wheel vs something fixed on the chassis (fender height is fine)

step 2: car SECURED in the air but not too high, remove wheels, disconnect ARBs, on coilovers set springs free so they will not prevent from moving the suspension. On multi link suspension you could remove the coilover. Cannot do that on mcPherson, so i hope your springs are not too hard.

step 3: install a lift under the knuckle and lift the suspension so it is in the same position as it was with wheels on the ground (remember step 1 ? That is what it is for). Install the gauge, and set its zero point.

step 4: lift/lower suspension and graph how much it toes in and out.

Now to adjust ... I only worked on S chassis, so this is what i learned.

To adjust it in the front, the easiest is to use an adjustable rod end and change the spacers between the knucke mounting point and the heim joint. The hardest way to adjust is to move the steering rack forward. The most expensive, easiest and maybe the most effective is to buy a set of good lowering knuckles and suspension arms. Kits exist for front and rear, and i do believe they are worth it. I have a cheap keisler, it just works (although bump steer is not adjustable on it, i just love how fast the steering is. Half a turn in each direction helps a lot, no more passing the wheel on tight corners). TDP / gktech are expensive ones but i guess worth every penny.

On the rear, this is corrected (mostly) by lengthening the traction arm, but do not overdo it as a small change affects toe curve a great way.

4 things to remember:
1- do not eyeball. the correct length window is less than 1/2" .
2- The rear spindle rotates when moving up and down: dont forget to adjust gauge level before reading the toe change.
3- do NOT bind your suspension to achieve the desired toe curve. At one point, you may need to chose between your desired static camber and correct toe curve. Chances are you can't have both. IMHO, a bit more static negative camber than needed is well worth a nice toe curve, but as said in the article, a lot more static toe in prevents the suspension from opening too much. It depends on application and driver. I am not a very fast driver, but i found the car way easier to control with very few toe change and too much static camber than quite some toe change and correct static camber. Toe change wore me off. Maybe a better driver won't mind though.

4- no PU bushes on an S suspension, except the RLCA arms. Every other bush works in more than 1 direction, using PU there will create bind, which you don't want because it will wear something fast and will make the suspension unpredictable. If the bushings are worn and you dont want to go heim/rose joint, use new OEM bushings. Went there, bought heim/rose joints quite fast after.

If you dont want to pay for the bumpsteer gauge, you can also use a magnetic laser level, millimetric paper ducted in the wheel well, and weight suspended on on wheel stud to keep it straight when rotating in the rear. It will be harder, longer and may lack a bit of precision, but it will also cost less.
Wednesday, September 03, 2014 10:40 AM
I love this series, glad it's back!
Mike Kojima
Mike Kojimalink
Wednesday, September 03, 2014 11:21 AM
One thing I have noticed in drift cars, in the beginning we ran a lot of toe steer and as the cars evolved to be grippier and faster, we reduced it greatly.
Wednesday, September 03, 2014 12:35 PM
Easiest way to measure bump steer?

Tape a laser pointer to the hub and point it at the wall. Draw where the dot goes.

Adding rear toe-in during roll is a bad idea. Sure you'll be going faster but let off the throttle and now you have weight transfer induced oversteer AND toe induced oversteer. Best to just keep as little movement as possible.

You should really add something about compliance-toe changes. Maybe you can fix it on Dai's car and claim it as another one of your 'special sauces' that work so well.
J Finken
J Finkenlink
Wednesday, September 03, 2014 1:07 PM
Great article! Very interesting as always Mike.
I have semi trailing arms on my car and will definately try with stiffer springs and more static toe-in.
Concerning the E30 M3. The pictures in this link shows how BMWs motorsport engineers went to great lengths (at leeast within the limited regulations) to get the old fashioned suspension setup to work as well as it did.
It is after all the most successful touring car of all time.
Mike Kojima
Mike Kojimalink
Wednesday, September 03, 2014 1:31 PM
@doozer, in no way does rear toe in under roll equate to more oversteer. I think you are thinking backwards. Dai's S13 and BR-Z have very little compliance toe change as both those cars have bearings in all of the pivots. The only compliance is in the deflection of the links and chassis.

The amount of toe steer under roll is a tuning variable which changes mostly due to the size tire, the skill of the driver and the power of the car.
Dan DeRosia
Dan DeRosialink
Wednesday, September 03, 2014 2:00 PM
Love this series. Bump steer is so overlooked it's not funny.

Thinking of steering stuff, any opinions on ackerman? One ex-GM suspension engineer who did a lot of racing work on the 1st gen RX-7 did some interesting wrinkles there.
Mike Kojima
Mike Kojimalink
Wednesday, September 03, 2014 2:07 PM
J Finken, the best way to approach a E30 is a combination of springs, swaybar and roll center location to reduce body movement. The idea is to limit roll yet keep the softest possible spring. Too much roll center height and jacking is an issue, too much sway bar and inside wheel lift gets to be a problem. You have to find the happy medium. I read your write up on the DTM E30 and enjoyed it. One note, the probable reason why the diff was raised was to save the cv joints on a real low car.
Wednesday, September 03, 2014 7:01 PM
I road race a 93 4th gen Camaro in NASA's CMC and it's a perfect example of an overlowered car not allowed to correct geometry. I currently suspect that the ~1/4" of scrub that I have is insufficient, so I'm looking to add more per your previous articles although they don't really mention any consequences of having less than 3/4" of scrub. The mention of mid-corner understeer due to bump steer grabbed my attention. My car has a bit of that now that I'm trying to quell. My first move was to drod front spring rates and noticed immediately that the car tolerated more input before the fronts broke loose, but the car really isn't any faster this way (been to the track 3 times like this now trying to relearn driving it). I feel like I took a step in the right direction, but that I'm still missing something. Wider spacers transformed the car, but made it illegally wide. Maxxing the track width via LCA adjustment got me some of the befits realized with illegally wide spacers, but not all of it. That and there's probably more to be had on top of the spacers anyway.

I've been wanting to correct the steering geometry because every time I hit a bump, the car rips the wheel out of my hand. I did raise the car a bit and make it less bad, but at the same time increased geometric anti-roll so probably lost some of the effects of lighter springs.

Instability entering corners is another one that drives me nuts (I'd like to help it without taking out more bias if possible). I can keep the car on track, but I lose a lot of speed when the car breaks loose on entry and if I do it leading onto a straight no defensive line in the world can save me. It's happened and it sucks no matter how badass I look drifting on someone else's Gopro. I didn't realize it could make a car suck to this extent, hopefully this makes the car less bad because the driver is running out of creativity. An iron will and lack of fear can only get me so far. . . Are there any glaring deficiencies not already addressed in this article I might rectify in my car?
Mike Kojima
Mike Kojimalink
Wednesday, September 03, 2014 7:15 PM
Are you not allowed any changes to pick up location at all? Even simple stuff with bushings and spacers?
Wednesday, September 03, 2014 7:51 PM
I can correct bump steer with aftermarket outer tie rod ends and I can use sphericals and drill the knuckles for a non-tapered bolt. I'm allowed to hog out the subframe to allow for greater range of motion on the camber/caster adjustment bolts on the LCAs (both front and rear bolts are adjustable from the factory on the inboard side of the LCA). I'm also allowed to hog out the upper control arm inboard mounting holes. Strano used to sell offset bushings for this, but no more. The intent is to allow for camber/caster adjustment, but I could easily use this to also increase scrub if I narrow up the track width by sucking both upper and lower arms inward. With no spacers, I have -1/4" of scrub. Bushing material is unrestricted as well. I happen to have sphericals on my LCAs. No mention of longer ball joints being allowed, and there is a catch all at the beginning that basically states if it doesn't say you can, then you can't. I do plan to beg for a rule change here though as it's getting close to that time of year. So at this moment, bump steer can only be corrected to the roll center (per your pic), and the only way to alter the roll center is to change the ride height of the vehicle.
Mike Kojima
Mike Kojimalink
Wednesday, September 03, 2014 9:12 PM
Don't go overboard with scrub. I generally like to run between 0.5 and 1 inch if possible. You can probably go a long ways to reducing bump steer via the rules.
Wednesday, September 03, 2014 10:27 PM
Realistically I could probably only get about .5" total scrub by getting all creative with the upper control arm slots and wider spacers. No way I could eek out a full inch of scrub within the constraints of the rules due to track width limitations. Someone said that on these overlowered pony cars the added weight jacking induced by scrub was helpful with rotation. I'll have to read more to understand why that is.

Just measuring scrub was very empowering, merely knowing something about my car instead of making assumptions was very eye opening. Prior I thought I had too much scrub. Now I'm pumped to see where I am by the numbers for bump steer as well. Knowledge is very powerful indeed. Thanks for making these articles!
Mike Kojima
Mike Kojimalink
Wednesday, September 03, 2014 10:55 PM
The super small amount of weight jacking changes induced by minimal changes in scrub are probably not a factor at all. Caster would make a bigger difference. The weight shift might help in corner exit traction but I don't think the car would rotate better, might rotate worse. I am a fan in low scrub geometry as it turns in better to mid turn, has less wheel fight, feels better in split mu conditions and is less sensitive to tram lining, bump steer and other wheel shaking things. It does reduce steering feel and feedback, hence why I like a little.
Thursday, September 04, 2014 9:16 AM
Great article Mike. I think everyone can agree on that :).
Thursday, September 04, 2014 10:01 AM
I've heard of some FF2000 cars running a bit of toe-in bump steer (toe steer?) in the rear to help put power down coming out of corners...any credence to this?
J Finken
J Finkenlink
Thursday, September 04, 2014 12:30 PM
Mike, yeah it was great pictures in that E30 M3write up. I cant take the credit for that though, just found the link. They have basically done the same as the modification you described for the 510 in your article, except they have all the adjustment in just one joint.

I dont have an E30 actually, but a modified 1987 Nissan Laurel. It still has a similar semi trailing arm rear suspension, so i would think the same principles should apply. This type of car is probably not known in the US, but the suspension is pretty similar to Z31 300ZX and the S12 Silvia.
Mike Kojima
Mike Kojimalink
Thursday, September 04, 2014 12:58 PM
Nbruno, If you mean S2000, the AP1 has a lot of toe steer toward toe out. These cars need a lot of static toe in or they have toe links that relocate the pivots to correct this that help tremendously. The toe changes were corrected in the AP2.
Attridge 300ZX
Attridge 300ZXlink
Thursday, September 04, 2014 7:42 PM
Looks like I'll be getting some SPL tie rod ends for my Z32!

Mike do you still have your Z32, are there going to be any updates on that build? It would be awesome to see how you'd set that car up.
Friday, September 05, 2014 10:05 PM
I can confirm he still has it... I sat in it a few months ago. But I don't know if Mike really has any plans to modify/update the car further.
Mike Kojima
Mike Kojimalink
Friday, September 05, 2014 11:35 PM
I would like to get rid of it.
Saturday, September 06, 2014 1:47 PM
Thanks for putting out tech articles, I always enjoy reading yours.

The limitations of the semi-trailing arm suspension is a frustrating aspect of the E30 type cars. Between the toe change and then the seperate spring/shock setup, it's nothing but comprimises. Lock it down with stiff springs and now you have about 1" of droop travel before the tire is in the air.

Thinking coilovers with helper springs might be the only real answer there?
Monday, September 08, 2014 10:28 AM
Nice article. I recently dialed in the front end of my autocross FC (corrected roll canter and bump steer)

The RX-7 rear trailing arm toe solution has nothing to do with the rear steer (They're completely separate systems). It has 2 additional links, one of which is between the control arm and the subframe.

I measured the rear toe steer on my rear suspension and it's essentially 0 (max of -0.006")
Mike Hawken
Mike Hawkenlink
Monday, September 29, 2014 11:21 AM
Hey Mike, what happens to the tie rod axis when you turn the steering wheel? I'm still learning, thats why I'm asking, but isn't the axis going to move when you turn the wheel? If so is the change not that noticeable, or is there always going to be inevitable bump steer while you're turning?
Post Comment Login or register to post a comment.

MotoIQ Proudly Presents Our Partners:

© 2018 MotoIQ.com