MG-Cars.net

Welcome to our resource for MG Car Information.

Recommendations

Parts

MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.

MG Midget and Sprite Technical - How to measure bump steer?

As an experiment I tried to measure the toe of my midget at both "neutral spring height" (7") and at full bump/jounce (tire touch the inner fender). To do this, I removed the springs and sway bar, and supported the tires at those two heights (this is with a fully rebuilt front end, and with the jack stands shimmed to hold the frame level, side x side).

Measuring this way, I got a 1" increase in toe, which is way more than I expected to see.

This makes me wonder if maybe this method I am using is proper, so I thought that I would ask the suspension experts here. Has anyone else (racers maybe) measured this before, and what have you found?

I researched what information is in the archives here, and read for hours about suspension, but found little more than a mention of this topic, not any data posted from anyone about having measured it.

I understand that there isn't much/anything I can do about the amount of toe change during suspension stroke, on a road car with no intention of any major surgery, I was just curious about it and quite surprised by the amount that I found. Perhaps the same limitation is why others have not measured it before?

Maybe I need to buy a new copy of Daniel's book? The copy I've got is about 10 or 15 years old and I understand it has been updated a lot since then.


Norm "stubbed a toe" Kerr
Norm Kerr

Bump Steer (as you know) is a Toe Change relative to Bump/Rebound that is the result of Camber change as the suspension travels. It occurs both on the Front AND Rear axles.

The reason for it is as follows:

1. No suspension system gives a perfectly flat Camber curve as the wheels travel from full bump to full Rebound.

The worst case scenario is the 'Swing Axle' - imagine your arm is the axle and swing it up and down; your fingertips (the wheel) will describe a circular arc tangential to the length of your arm - this arc is the Camber change curve.

If one was to decouple the axle (as per any front set up, or any good independent rear set up), then we can reduce this Camber change, although we can not eliminate it. Even the best double-unequal length wishbone set ups merely minimise Camber change (and the A35 'Peanut' derived Spridget front suspension is light years away from one of these); in fact they give a flattened 'S' shaped curve.

2. The requirement to provide steering at the front means that the front axles aren't really fully decoupled (because they are connected through the steering rack). By its very nature, the steering axis MUST be ahead or aft of the axle line in order to provide the steering leverage. Thus, as the Camber change occurs on one or both wheels, the effective steering axis will either shift left or right, or it will reduce in length or increase. Thus changing the Toe of the front wheels, thus causing the steering effect known as Bump Steer.

Other factors involved include:

1. If the rack is above or below the axle line then the Steering axis effect can be increased (or in some cases, reduced). The big factor here is steering arm inclination which should be minimised and equalised from side to side in order to reduce the effect.

2. Inclination of the Steering axis (ie the rack is not parallel to the axle line) exacerbates the problem. British Sports cars such as the Spridget are NOTORIOUS for this.

3. Castor angle, while it WILL change the Bump Steer, MAY mask some of its effects. Ie the driver feels it less.

4. Static Laden Camber will not alter the overall Camber change but WILL alter the slope of the Camber curve and thus alter the degree of Toe change for a given amount of Bump/Rebound. Ie you can dial in some Static (negative) Camber to alter when the Bump Steer occurs.

5. Ride height. The Static Loaded ride height can be altered (and controlled) such that the vehicle remains on the flattest part of the Camber Curve for most of its operation. This is done on Race Cars but does pre-dispose robustly controlling the ride height (ie stiffer springs).

Given a vehicle where you are not going to alter the suspension pick up points, swing arm lengths, etc, then the only real way to dial out Bump Steer is to alter the rack position.



In your case Norm your Toe Change seems somewhat excessive. This may be a vagary of your particular car, or it may be how you are measuring.

The way I do Bump Steer measurements on Race Cars is as follows:

1. Ensure the area you are working in is both FLAT and LEVEL. This is CRITICAL for accurate measurement.

2. Set the car up on ride height blocks having removed the wheels and springs, and having disconnected the ARB (Sway Bar). You remove the springs so that you can jack the axles easily, the ARB so that it doesn't interfere with the results (it links the axles and thus can and will produce anomalies), and the wheels because these are the last thing you want to measure upon to get an accurate Bump Steer curve. I Measure using the brake disc as my point of reference (it will have less anomalies than a wheel/tyre combination), or for Race cars I use a 'Bump Steer Plate' which can be bolted accurately to the hub.

3. Prior to measuring the Bump Steer curve I use a Camber Gauge to measure the Camber Curve at each axle and record to ascertain whether this is equal to both sides - if it isn't you will NEVER deal with controlling the Bump Steer and will end up confusing yourself.

4. Measure the Toe change through Bump/Rebound at each hub (brake disc) INDEPENDENTLY side to side and record. That is to say DO NOT measure overall total Toe Change wheel to wheel but Toe Change each side relative to the vehicle's longitudinal axis. You can use a chassis rail for this (assuming it isn't bent) or (better still) a known reference line marked on the ground.

4. Measure your Steering arms in length and equalise.

5. Take a spirit level and record each side's Steering arm inclination from full bump to full rebound.

Using these measurements you can work out what you have to alter and where in order to minimise the Bump Steer.

For the calculations involved see: 'Prepare to Win' by Carrol Smith.



The trick now is to equalise everything from side to side, Steering arm length, steering arm inclination and how the steering arms pick up at the vertical link.

Essentially this means shimming the rack and going through all the measurements again and again until you reach the minimum amount of Bump Steer. Initially you will have to shim diagonally to get the Toe Change on each axle equal. Then you shim equally to minimise the Bump Steer. On the Race cars we have it a little easier because we use Rose Jointed (Heim Joints) steering arms which allows us to alter the steering arm inclination independent of steering rack inclination.



Once you've done all that then you can get 'trick' by dialing in the amount of Toe In/Toe Out on Bump/Rebound to alter the vehicle's dynamics and handling characteristics. Eg a tad of Toe Out in bump at the front reduces the slip angle of the outside front tyre at small steering angles during corner entry which will reduce Corner Entry Understeer. However go too far and you will make the car unstable under braking or on bumpy surfaces.

But that's a story for another day.......
Deborah Evans

Deb,a comment from someone whose understanding of suspension physics pretty much tops out at understanding oversteer and understeer! You say to remove the ARB so it doesn't produce anomilies that will interfere with the results. But, if you are measuring the bump steer on a car you will be driving in the real world (with the ARB intact), wouldn't you want to measure it so you will get real world results, which would include the anomilies produced by the connected ARB? Or, is the idea to get the pure measurement, knowing that there will be some differences attributed to those anomilies?
Jack Orkin

I checked mine very simply using a cheap magnetic laser tape measure, but a laser pointer will do. Springs out as you said, wheels off. Piece of white card on garage floor, laser fixed to disc (rotor) so that beam is parallel to it. As you rotate the disc the laser will track across the card. Mark the tracks at different deflections and see if they are parallel.......hey presto! But remember, before you get too stressed, that the only significant tracks are those made at normal loaded and heavy compression positions.
Allan Reeling

Jack,

The ARB will introduce anomalies owing to the fact that it is:

1. Connected to both axles.

2. Owing to 'stiction' in the system (ie the Drop Links) the effect of the ARB will vary (in the Static Case) dependent upon the amount of 'swing' of the bar (ie how far you jack the suspension). This does not accurately represent what happens during cornering because in the real case the bar only comes into play when there is lateral load transfer. If you think about it, the ARB tries to resist suspension deflection in the Dynamic Case, so it resists Bump on one wheel and Rebound on the other. Thus in the Dynamic Case (aside from the Diagonal Load Transfer effects of the ARB) it has little overall (well some) bearing on the overall Bump Steer Curve.

3. Additionally, given Point 2 above, if you leave the ARB connected and jack each wheel in turn, then it will 'think' the car is cornering and add its input (which will also make jacking the suspension more difficult because it will resist this). Thus you don't get the true Bump Steer Curve of BOTH sides in Bump and BOTH sides in Rebound, which is what you are looking for. Remember, we are not just concerned with cornering but with how the suspension reacts to bumps when going in a straight line (where there is no Lateral Load Transfer).

The trick is, as you say, to get the 'Pure Curve' and minimise this as I explained and accept the ARB will have an affect in the real world.

This can then be tuned out by playing with the ARB settings - yet another reason why Racers like to control the suspension with springing and fine tune with ARBs.
Deborah Evans

Thank you Deb for taking the (considerable) time to explain all this. I've been through the exercise a few times with the race car and it's a PITA, but it did provide improved results.
And thank you Norm for giving me your findings on standard Spridget suspension. My race Midget had fabricated top arms and I was wondering if the bump steer was inherent in the original design or was introduced by the changed length of the top arm.
My method was less sophisticated than Deb's. I had the front wheels on dollies that were free to rotate and show toe changes, and jacked the car up and down from static height, measuring toe change from wheel to wheel. Extremes of bump/rebound produced extremes of toe change. I graphed the results and the shape of the curve on the graph indicates where the problem lies and what change needs to be made to fix it.
Used "Race Car Vehicle Dynamics" by Milliken and Milliken as my instruction since I don't have Carroll Smith's "Prepare to Win".
Because the race car is so low, the rack needed to be lifted substantially, well beyond the realm of shims. There is limited scope inside 13" wheels for lowering the outer ball joint to achieve a similar outcome. The rack still hasn't been lifted. I was able to minimise the bump steer by lifting the ride height a little and using different springs to keep the car away from that extreme zone in bump. A shortened rack with longer tie rods also helped. However, the aerodynamic lift at the front of the car at top speed will still increase toe out. Since Conrod Straight is the only place where this can occur, it's not a problem until the car returns there.
I've since sold that car but still do the work on it, so I'm pleased to see that others too are taking notice of the bump steer in these cars. Thanks again Norm and Deb.
Mike Allen

Interesting findings and information. The extra toe in under suspension compression might explain why Spridgets tend to respond to larger front sway bars which usually induce a lot of understeer, by helping to control suspension compression under cornering they reduce the extra toe in's tendency to indue oversteer. Thanks all for helping me understand a bit more about our little cars traits.
B Young

Close but not exactly!


In fact Toe IN in Bump (Compression) will INCREASE Corner Entry Understeer because it increases the slip angle of the outside front tyre as the Load Transfer is occurring and the Slip angles are building (thus increasing its grip) during the Corner Entry phase. Ie the car doesn't want to respond to the upsetting (cornering) force.

Remember Toe OUT is unstable, Toe IN is stable. Having a tad of Toe OUT in Bump will get the car turning in better, but go too far and the car will be twitchy under braking.


The reason Midgets like large front ARBs (as opposed to, say, Spitfires) actually has more to do with the Front Roll Centre and Roll Axis Inclination rather than Bump Steer, coupled with the short lower A-arm that produces little leverage upon the spring.

Normally it would be true to say that increasing the ARB will increase the Corner Entry Understeer (owing to the increased rate of Lateral Load Transfer) and this is ALSO true in Spidgets, especially if you go psychotic with the front ARB (anything over 3/4" on softer springs, anything over 11/16" on race springs).

HOWEVER

In a Spridget, the inclination of the Roll Axis is such that, during the Corner Entry phase, the car wants to roll more about its front axle than its rear. This causes Camber change especially at the outside loaded wheel (tends towards positive reducing grip), but lifts the inside tyre which reduces that tyre's grip more and induces Corner Entry Understeer.

Remember all this is going on in the transient phase before Load Transfer is complete - once THAT has happened then the ARB no longer has any effect.

In other words, what you get with a Spridet is Corner Entry Understeer induced by ROLL STEER.

In order to stop that roll (and hence the Roll induced Understeer) you need to control the front suspension.

You do this EITHER with Incresed Spring Rate OR a larger ARB. The former on a race set up, the latter being more suited to a road set up where high spring rates are not suited to the driving surface, the fact that the surface is changeable (eg rain), and the fact that Diagonal Load Transfer is less of an issue because you aren't (or shouldn't be!) achieving the levels of Corner Speed that you would in a Race car.

The reason the Spring and ARB rates are so much higher is to do with the reduced leverage provided by the lower A-Arm.

A race Midget, for eg, will run 600lb front springs and an 11/16 ARB. A race Spitfire (similar weight and weight distribution) owing to its better front suspension design will run 450 - 480lb springs and a 1/2 - 5/8" ARB.

Once the vehicle is in the Steady State then these self steering effects cease to apply (until you hit a bump, or accelerate or decelerate).

Having a car that is set up well to give a good response during the transient phase(s) (which is where race cars spend 99% of their lives) makes the car a lot easier (and nicer) to drive.

The steady state in a Spridget (ie mid corner when the car has settled and taken its 'set') tends towards Understeer owing to the Roll Steer effect caused by rear axle skew due to the fact that the cart springs cause 'Tramp' when unevenly loaded.

But that's another matter...
Deborah Evans

Norm,

How did you actually measure the toe change as there is no way any standard suspension will give 1 inch / 25mm of bump steer?

A common mistake is to just measure say the position of the leading edge of the disc / or wheel assembly, this way you have track change as well as toe change?

Cheers

Spencer

S Deakin

oh, that is true, isn't it? I really couldn't believe that 1" result, and you are right Spencer, I should have measured the difference in front and rear toe, at each position. Now I feel a little bit silly for not having considered the track change, at the time.

it will be a lot easier for me to do this again, properly next week, I have ordered some tools, rather than the jury rigged carpenter's square, caliper and tape measure (and trigonometry) that I've been struggling with!

Also, I am planning to re-do the camber change measurements that I did last week, as the gear I've got coming will be able to measure toe, caster and camber (yay!).

Norm
Norm Kerr

Deborah
have you ever mapped the various data points on a Spridget and input them to a suspension modelling programme?
I installed SusProg3D with great intent but haven't yet found the time to do all the required measurements. I've always hoped I could find someone else who has already done it.
http://www.bevenyoung.com.au/suswin.htm
Mike Allen

This thread was discussed between 25/04/2011 and 28/04/2011

MG Midget and Sprite Technical index

This thread is from the archive. The Live MG Midget and Sprite Technical BBS is active now.