TUNING PERFORMANCE SUSPENSION
--A PRIMER--
By Tom O'Rourke

INTRODUCTION

Designing and tuning the automotive suspension for performance involves fact and science, as well as a good measure of experience and art.  In the following monographs I will emphasize the former, but will occasionally use a first person presentation to acknowledge the intertwining of my opinions.

Keep in mind that cats can be skinned numerous ways.  One man's stay bar adjustment is an other's sway bar change.  Often differing approaches do not mix well.  Accordingly, for the most part, my comments will be more qualitative than quantitative.

And particularly keep in mind that "performance" comprises quite a range of cats.  Completely opposite objectives and expedients are appropriate for various specialties.  My bias tends towards road and circle racing on pavement.

Organization of the various subjects under set-up or dial-in reflects the above-mentioned bias. Under set-up I address those items more readily sorted out in the shop or during testing.  Set-up items typically alter other settings, or are difficult to set and/or measure.  Dial-in items tune the suspension to a particular track or changing conditions.   Each topic is developed further in a more focused paragraph.
 

The first concern is basic housekeeping. Chassis rigidity, particularly resistance to twist in torsion, is a necessity. Chassis flex will render the other tunable suspension control mechanisms insensitive or inoperative.  The chassis should be square --check the wheelbase and wheel diagonal measurements.  And place as much weight as feasible low and toward the center of the vehicle, except high for drag racing, somewhat higher for dirt, and towards the left (driver's side) and maybe a bit more rearward for circle racing.

Ride height is the primary suspension setting. Most of the other suspension settings change when ride height is altered.  Set ride height first and check it regularly.  Static wheel weight should also be set during this process.  I prefer to treat wedge as a fixed parameter, though this is a minority view in many circles.

Locate and, if necessary, adjust the various pivot axes between the sprung weight and the unsprung weight.  These include the roll centers and dive and squat characteristics.  The front and rear roll centers define the roll axis which, in conjunction with the Cg, in turn define the roll moment, i.e. the tendency of the vehicle to lean when turned.  Dive and squat suspension geometry influence the vehicle pitch attitude, i.e., nose down under braking and tail down during acceleration.

While springs and dampers,  though often varied somewhat during dial-in,  are initially chosen to keep the vehicle from bottoming and oscillation, respectively.  Calculating the wheel-rate during set-up aids understand of the effect spring and damper changes will have on handling.

Steering geometry is also an early consideration.  The  front spindle determines the steering inclination angle to produce a reasonable scrub radius.  Caster, Ackerman geometry, and bump steer are important parameters which should at least be recorded and tracked.

With these items under control, we can proceed to dial-in.
 

My preference in "dialing-in" is to first optimize the absolute total adhesion of the tires at speed, and then work on balancing the vehicle.  Of course if the vehicle balance is vicious, it is only prudent to kill the biggest and closest snake first.

Tire temperatures are an excellent measure of tire adhesion.  Tire pressure and camber are adjusted to produce a flat temperature profile across the tire.  Roll stiffness at one or the other end of the vehicle is then adjusted to balance the handling in terms of understeer (push) or oversteer (loose).  Springs may also be varied in small increments to tune pitch stiffness as well as roll stiffness.

At this stage the tires will be flat on the ground and working.  The vehicle will have a reasonably balanced response.  If a locked rearend is involved in a circle track setting, stagger is provided at the rear tires to facilitate power-on turning.  Front to rear brake bias is adjusted to optimize braking effectiveness and stability, and /or enhance corner turn-in. Dampers can be tuned to adjust the rate of weight transfer, either fore and aft or in roll -but usually roll- as well as to control small, rapid wheel movements.

Though not one of my favorite expedients, wedge is sometimes fine tuned to balance the vehicle.  While wedge changes ride height somewhat, it is often effective as a quick pit stop adjustment to correct an unbalance roll stiffness resulting from tires fading unequally during a race.

Of course the trick is to make the best compromises and get all of the above to work together to approach (but never reach) the optimum.

My view is that we want to gain as much cornering power as possible from the inside tire.  Tire slip angles under maximum turning allow the actual radii traveled by the tires to differ from each other as well as to deviate from the direction in which each tire points!   The more heavily loaded outer tire pretty much determines the actual cornering line.  Provided the inside tire is not turned so far as to induce it to lose its slip angle and slide (wash), there is worthwhile cornering power to be gained by turning the inside tire even beyond the theoretical Ackerman angle öthough the reasons for this are beyond the scope of this discussion.

Optimum Ackerman geometry is best determined by testing.  With rear steer (the steering arms pointing rearward), start with a modest setting and move the tie rod forward and/or increase the inward steering arm angle until the driver notes a fall off in performance, and then back off a bit.  Front steer is a bit more problematic in that the desired steering arm angle often interferes with the tire.  Rearward movement of the tie rods is the easier approach.   Be sure that toe and/or bump steer don't wander and influence the results.

Brakes can also be advantageously biased to emphasize instability.  According to the traction circle theory, a tire can do only so much total work.  If asked to accelerate or brake while turning, less traction is available for turning.  Most of us are familiar with "driving with the throttle" in a turn.  Similarly, by biasing braking more strongly at the rear and utilizing trailing brake into a corner, more turning traction is available at the front and less at the rear.  Thus the vehicle rotates into a corner.  Turn-in can be greatly aided by this technique.  As can spinning!

Bias is controlled in several ways. Foot force on the brake pedal can be selectively divided between front and rear master cylinders by an adjustable bias bar.  Or an adjustable proportioning value in the hydraulic circuit will attenuate the pressure down stream of the valve. Braking can also altered by changing the diameter of the master cylinder and/or the slave cylinder ödecrease the master and increase the slave for more braking. Tire diameter and/or footprint also affect brake bias.

Many race vehicles are running with a lot of bias cranked into the front because it feels more comfortable.  Or because the factory would rather error on the side of stability at the expense of total braking.

Bump steer is identified by jacking a wheel through its normal travel and measuring the thus induced steering movement.  This is commonly accomplished by mounting a flat plate on the wheel and measuring the angular movement of the plate with one or two dial indicators as the wheel is exercised up and down.  Of course the steering mechanism must be locked.

I've had good results by hanging plum bobs  from the end of the plate and measuring the position changes at the floor by marking segments of tape with a pen and measuring with dial calipers or a good rule.  Remember that we're not interested in the lateral movement of the wheel but only in the direction it points.

Correction of bump steer is by changing the relative arcs of the suspension arms and the tie rod by moving the drag link or steering rack position.  Both pivot location and tie rod length are of concern.  If you find appreciable bump steer, investment in a more comprehensive suspension text (I haven't published one) would be a wise investment.

I find tire temperature profiles the most reliable means for adjusting camber.  Keep in mind that cornering, braking, accelerating, etc. each develop a different temperature profile.  Chose your compromise appropriately.

Still, to grasp suspension tuning Cg must be understood.  It is the "thing" to be overcome to attain performance.  It wants to keep going straight when you want to turn.  It wants to maintain its present speed when you want to accelerate or brake.  Only your tire contact patches can persuade the Cg to change velocity, i.e. speed and/or direction.

Cg pushes horizontally at maybe belt buckle height and the tires push at ground level.  This offset develops two of the more important tuning factors, weight transfer and attitude change, specifically roll and pitch.  Suspension is the monkey motion linkage we interpose between the Cg and tire patches  to optimize the job the tires are able to do.  Getting this concept in mind is useful when considering the suspension tuning factors we do control.

Dampers also function as pseudo springs during suspension movement.  They can be utilized to aid or resist a spring during transition; but do not affect steady state conditions.  Think of the role the springs  play in roll resistance  when tuning dampers as pseudo springs.

Adjustable valving is available not only for independent  compression and rebound rates, but also for fast and slow wheel movement.  Thus one rate can be selected for slow wheel movement such as during turning, while another rate is utilized for rapid movements such as hitting a bump.

By angling at least one of the A-arm's rotation axis downward towards the front of the vehicle, a reaction force tending to lift the front of the vehicle can be generated.  In moderation, this can be a good thing.  Too much and the front end gets very hard and may even lock up.  If both A-arms are not equally angled, caster can vary with suspension movement.

There are many ways to balance handling.  My least favorite correction for oversteer is the reduction of overall cornering force by compromising front cornering efficiency.  Weight jacking without a fresh dial in is the usual culprit.  It's pretty common and makes the driver happy.  A better answer is to adjust the relative roll resistance, more at the front and/or less at the rear.  Sway bars are ideal for this.

Lateral location of the vehicle body relative to the wheels is provided by linkages such as a Panhard bar, Watts linkage or angled control arms.  The Panhard bar is the simplest in that it usually comprises a link with rod ends attached at one end to the body and at the other end to an unsprung portion of the suspension.  Since body roll can cause the Panhard bar to steer the attached suspension, the Panhard rod is happiest when it is as long as feasible and mounted level.  Generally the Panhard bar determines the roll center at the associated end of the vehicle, though stiff leaf springs can compromise this.

A-arm suspension roll centers can be located by plotting the A-arms as viewed from the front, extending the A-arms on each side of the vehicle either inward or outward until they intersect and drawing a line from the intersection through the midpoint of the tire contact patch.  The intersection of  these lines defines the roll center.  Note that the roll center can move both laterally and vertically with body roll.  This tends to be clearer with reference to the drawing.  MacPherson strut suspension can be viewed as a special case of A-arm suspension with the upper A-arm extending at a ninety degree angle from the upper attachment point of the strut.  Keep in mind that the strut changes length during body roll.

A Hotchkiss rear axle has a roll center at the midpoint of the Panhard bar, or at about axle height when located only by leaf springs.  Angled control arms are a bit beyond this discussion, but are discussed in the better publications on suspension.

Locating the front roll center at about ground level and the rear roll center six or eight inches above the ground are decent starting points.  Lower decreases weight jacking and higher improves driver feel.  Minimizing movement of the roll center with body roll is desirable but often difficult with production suspension.

Body roll during turns is resisted by the springs and swaybars.  By increasing the roll resistance at a selected end of the vehicle, weight transfer to the outside tire is increased at that end and decreased at the other end.  If the greater roll resistance is generated by a swaybar, weight is also decreased at the inside tire.  The ultimate result for a vehicle with a reasonable set up is that the selected axle will run at a greater slip angle with increased roll resistance.  Thus, for instance, by stiffening the sway bar at the front of the vehicle, an oversteer imbalance can be corrected.

While for the most part we pretty much have to accept whatever scrub radius results from all the other settings we chose, it is still important to understand and locate this parameter.  Steerable wheels turn about a steering inclination axis (king pin axis) defined roughly as the line through the ball joints.  This line usually intersects the ground rather inboard of the tire contact patch.  I like to see it at the inside edge of the tire contact patch.  Scrub radius is the distance from the steering inclination axis and the center of the  tire contact patch.  Our only adjustment is the spindle configuration or wheel offset.  In any event, the scrub radius should be equal side to side unless you know why you want is otherwise .

Vehicle weight either moves directly with wheel movement -unsprung weight- or indirectly with the wheel as cushioned by the suspension spring -sprung weight.  The long and short of this is that the latter is good and the former isn't. Unsprung weight moves with the wheel.  The more unsprung weight, the stiffer the spring/damper assembly needed to absorb the energy imparted when the wheel hits a bump.  This both compromises the suspensions ability to keep the wheel on the ground and transmits the upset to the suspended weight.

Various links, control arms, dampers, springs etc. are attached at one end to a sprung member, i.e. the chassis, and at the other to an unsprung component such as a wheel carrier.  Accordingly, these are partly sprung and partly unsprung.  Put the expensive, light weight rod end on the unsprung end.  Low mass wheels have the added advantage of decreasing rotating weight and gyroscopic inputs.

Functionally, a swaybar links the wheels at an end of the vehicles such that the wheel can move synchronously up and down relative to the vehicle body without resistance from the swaybar, but resists movement of one wheel relative to the other.  Body roll constitutes such relative wheel movement.

Two important functions are provided by  swaybars.  First, they allow for increased roll resistance (roll couple)  independent of the springs.  In most instances the spring would be much too stiff if they were the sole source of roll resistance.  And second, swaybars allow for allow adjustment of the front and rear roll couple ratio.  Adjustment of this ratio modifies the tire loading at each end of the vehicle as well as the inside to outside tire loads at each end.  This is a very important aspect in dealing with understeer or oversteer.

Structurally, a swaybar is usually a transverse torsion bar with levers at each end.  One of the torsion bar or levers is attached to the chassis and the other to the suspension.  A convenient way of varying the swaybar stiffness is by use of a sliding attachment to the lever to change the lever arm length.  This can be adjusted by the driver through a cable or other mechanical linkage, or more simply clamped and releasable at the lever.  A more sophisticated approach is the use of thin, wide blades as levers.  When the tin section is horizontal it operates as a spring member and decreases the swaybar stiffness.  As the blade is rotated toward the vertical, it become stiffer as does the swaybar.  Even more sophisticated push rod suspension dispense with the torsion bar entirely and just use a blade to serve the sway bar function.

Front wheel drive and swaybars is a whole different cat.  Since the driven front wheels do all the work except for holding up the rear of the vehicle, the objective is to get the rear wheels to do a bit more work.  A pretty good hunker of a swaybar only at the rear is usually the better move.

Tires are basically pretty simple items.  For best tire adhesion, they like to be hot but within appropriate temperature range, and flat on the ground with even pressure across the contact patch.  Traction is produced by distending the rubber into the small opening of the pavement surface roughness.  This allows friction coefficients greater than one.

When turning, the tire sort of squirms along with the contact patch elastically displaced from the rest of the tire.  Depending upon the turning load, the tire does not actually move in the direction it is pointed.  Instead, it moves through a wider curve as a result of the distended contact patch.  At any instant, the angle between the pointing direction and the actual direction of travel is defined as the slip angle.  If the front tires run at greater slip angles than the rear, the vehicle understeers.  When the front slip angle becomes too large, the contact patch loses traction, and the tires wash, i.e. slide ineffectively.  Without the distended contact patch there is no slip angle, and very little turning.  Oversteer occurs when the rear tires run at a larger slip angle than the front tires.  When the rear tires reach the extreme situation analogous to washing at the front, the vehicle spins!

Slip angles tend to increase with greater turning force, more weight on the tire, and higher or lower tire pressure than the optimum.  Also for a given weight on an axle, unequal load distribution between the tires on the axle increases slip angle and decreases overall traction.  And greater roll couple such as from a stiffer sway bar increases the outer tire load and decreases the inner tire load.  Which is a roundabout way of saying that vehicle balance, i.e. understeer or oversteer can be adjusted by sway bars or weight jacking.  But always keep in mind that balance can be gained by throwing away tire traction on the end that sticks rather than shifting the traction to the loose end.

Tires can stop, accelerate or produce turning force at fully capacity, but cannot turn at fully capacity while stopping  or accelerating.  This is the concept of the traction circle.  Too much throttle out of the corner and the rear end comes out.  Too much brake on turn-in and the front end washes.  Of course some trailing throttle or brake can aid in weight transfer to help stick the stressed tires.  I don't offer an answer here.  Only a description of the compromises that make driving interesting.

Since there is so much going on with the tires, perhaps the most useful tuning aid is tire temperatures.  Actually tire temperature profile is a more accurate designation. The profile is determined by taking temperatures at the center and inner and outer shoulders of the tire using a thermocouple probe or an infrared pyrometer.  I like the probe since my tires often travel some distance from the track before the temperatures are taken.  Interior temperatures are maintained more accurately than surface temperatures under such conditions.  By I'd really like to have a data logger and a pyrometer monitoring my tires on the track.

The first thing to check is the average temperature.  It should be in the appropriate range -which varies for different tires.  160 to 230 degrees F would be typical.  If you're turning in one direction, of course the outer tires will be hotter.  Differences between the center and shoulder temperatures suggest pressure problems.  Too much pressure causes the center to bulge and run hot, and too little pressure causes the shoulders to run hot.  One shoulder running hotter than the other (with the center about half way between) indicates an incorrect camber setting/curve.  Too much camber heats the inside shoulder and too little the outside edge.

Tire temperatures will reflect an entire lap or more.  If you brake hard for the pits, the front end may dive and produce the symptoms of too much camber.  Be sure to use tire temperatures to optimize the important portions of a lap.  It's not a bad idea to run hard only during turn-in, or some other sector of a turn, for focused tire temperature tuning

Changing the ride height at one wheel only changes the static weight at the wheels by increasing the weight of one front to rear diagonal pair of wheels while decreasing the weight of the other diagonal pair.  These diagonal weights comprise wedge.  For instance, increasing the ride height at the left rear will add static weight the left rear and the right front while removing weight from the right rear and left front.  Handling will respond with less understeer and decreased right rear traction.  If large ride height changes are made, other unwanted changes in camber, roll centers and roll couple may also be made.  In moderation, wedge is an effective way to change vehicle handling balance.
 

Since the Cg is some distance above the ground, and since the tires generate forces at the ground, braking and accelerating, as well as turning, dynamically transfers weight from one end or side of the vehicle to the other.  This has been discussed with regard to turning under the roll.  And it has been alluded to under dive and squat with regard to braking or acceleration.  For the most part these are not result we want to emphasize.  But weight transfer can be very helpful.

Turn-in and acceleration out of a corner requires heroic effort of one or two tires.  For once physics is on our side (assuming the driven wheel are in the rear) in that large amounts of weight is transferred to these tires when most needed.  During turn-in the outer front wheel is heavily loaded by weight transfer induced by turning (which is technically an acceleration) and usually some braking.  This is rather brute force performance and can toast the tire if overdone.  Similarly, when the rear tires are asked to cope with getting the power to the ground as well as keeping the rear end behind the front, weight transfers rearward and outward to facilitate the task.  While we don't much actively tune this phenomenon, it is important to understanding why things work, or otherwise.

Placement of springs in suspension is tricky.  If the spring is located immediately adjacent the wheel, it would be expected that the unsprung end of the spring would move with the wheel and the spring rate, as in pounds per inch, would have its full effect.  And it would.

But if the spring was mount on a support arm half way between the attachment of the support arm to the chassis at one and the wheel at the other, it would seem that since the spring moves half the distance of the wheel, which it does, that the effective spring rate would be half the rated value.  But, since the spring must control the wheel oscillations, spring rate at the wheel would be only a quarter of the rated value .  And you can buy a lot of springs before things start working right.

While I've tried to use a qualitative approach to most topics, this one involves some simple math.  The way to size a spring is to determine the length of the suspension arm which carries the spring from the chassis mounting to the wheel.  Then measure the distance from the chassis mounting to the spring.  Square both quantities and divide the squared distance to the spring by the squared arm length.  This produces the factor which compensates the spring wheel rate.  Also, if the spring is not perpendicular to the arm, its rate must be further diminished.  The cosine of the angle between actual mounting and a position perpendicular to the suspension arm is a reasonable place to start.

Dampers do not require the "squaring" correction but are effectively softer when mounted inboard on a suspension arm.  Use the actual ratio of the spring to damper position on the arm for a correction factor.  Dampers seem to be mounted at angles even more often than are than springs.