Use third springs to separate heave from roll

Generated from content/lms/suspension-and-chassis-design/04-advanced-geometry-strategies/03-third-springs-and-heck-springs.md; edit the source file, not this page.

Source path: content/lms/suspension-and-chassis-design/04-advanced-geometry-strategies/03-third-springs-and-heck-springs.md

Course: Design suspension geometry that actually wins races

Module: Use geometry to solve handling problems

Estimated duration: 60 minutes

Skill goal

You use a third spring when one end of the car needs more support in heave than it needs in roll. That is the whole lesson. The normal left and right ride springs, plus the anti-roll bar, are already trying to manage tire loading, roll balance, bump compliance, and the driver's low-speed confidence. When a winged or underbody car also asks those same springs to hold up a large high-speed aero load, the setup becomes a compromise. If you stiffen the two corner springs enough to hold the platform at speed, you may give away the slow-corner grip that the car needed from softer springs. A third spring gives you a separate load path for the motion where both wheels at one end move together.

This is not a magic spring. It is not a cure for a poor roll-center path, a binding anti-roll bar, bad toe settings, or a damper mismatch. It is a specific tool for a specific conflict: the car wants one stiffness in common bump or heave, and a different stiffness in roll. The skill is learning to identify that conflict, keep the roll system and heave system conceptually separate, and test the change without fooling yourself.

What heave and roll mean here

At one axle, roll is opposite wheel travel. One wheel moves into bump while the other moves into rebound as the body rolls. Heave is parallel wheel travel. Both wheels at that end move upward relative to the chassis together, as they do when the car is pressed down by aerodynamic load or meets a smooth vertical input. A pushrod and bell-crank car makes the distinction visible. The anti-roll bar can be linked so it rocks freely during parallel wheel travel, but resists when the two sides move oppositely. That is why the bar is a roll device first, not simply another ride spring.

A third spring, or heave element, is arranged so it responds mainly to that common motion. In the Blundell and Harty example of a student race car, the anti-roll bar mechanism rocks during parallel wheel travel and resists during opposite wheel travel. The same discussion notes that a small spring-damper can add heave stiffness and damping, and that alternative three-spring systems allow more independent control of damping in roll compared with damping in heave. That is the mechanical core of the lesson: the suspension linkage decides which motion the spring sees.

The third spring belongs in that common-motion path. The ride springs and anti-roll bar still define the basic mechanical platform. The third spring adds support when both wheels at that end move together. If you keep that mental model clear, you stop making the common setup mistake of treating every suspension symptom as a reason to increase all four corner spring rates.

Why not just fit stiffer ride springs

Aero load is speed dependent. Haney describes the problem directly: if the normal coil-over springs had to support the large loads from wings and ground-effects underwings, they would need to be much stiffer than would be desirable in slow corners. The third spring is the solution in that context because the normal coil-over spring and damper system can continue doing its low-speed work while the third spring becomes important at higher speeds as the aero load compresses the main springs.

That matters because mechanical grip has not gone away just because the car has aero. Carroll Smith's point is useful for the intermediate driver or engineer: the basis of cornering power and balance is still a linear car with good mechanical grip, and aerodynamic grip is additive to that base. The same source notes that the apex speed of the average racing corner is below the range where aerodynamic download dominates. So if you stiffen the whole car for the fastest part of the lap, you may damage the grip and balance that matter in the slower parts.

The third spring is a way to respect both truths. At low speed, you want the ride springs, bars, geometry, and tires to keep the car compliant and predictable. At high speed, if aero load is driving the chassis into a poor ride-height range or toward bottoming, you need extra platform support. The third spring gives you a place to put that extra support without making every roll and one-wheel bump event pay the full price.

The load paths you are trying to separate

Think in three load paths. The roll path is dominated by the two corner springs, their motion ratios, the anti-roll bar, and the geometry around the roll centers. That path determines how lateral load transfer is distributed front to rear and how the car feels in steady and transient roll. The heave path is the common vertical motion path at an axle. That is where a third spring can add stiffness and damping. The aero platform path is the practical result: as speed rises, the car sees more vertical load, the suspension compresses, and ride height changes can affect the downforce the car generates.

Do not blur those paths during diagnosis. If the car rolls too much in a medium-speed corner and the balance is wrong, the third spring may not be the first tool. You still need to understand roll stiffness at the front and rear, and Blundell and Harty specifically frame roll stiffness and damping as front and rear suspension properties to be determined separately. If the car is fine in roll but compresses too far on the high-speed straight or through a fast aero-loaded section, the heave path becomes the suspect.

The test is not whether a stiffer part makes the car feel more serious. The test is whether the car gains high-speed platform control without losing low-speed mechanical grip or confusing the roll balance. That is a narrower and more useful standard.

When the third spring is the right tool

Start by asking whether the problem is speed dependent and common to both wheels at one end. A proper third-spring problem usually shows up when speed and aero load rise. The driver may feel the car get closer to the ground, the data may show damper displacement increasing with speed, or the car may lose consistency in high-speed sections while still needing softer mechanical setup in slow corners. If the problem exists just as strongly in a slow hairpin, a one-wheel bump, or a curb strike, you may be looking at ride spring, bar, damper, geometry, bump-steer, or compliance work instead.

The car also has to be eligible for the device. Blundell and Harty describe three-spring systems as common in higher formula motorsports events when rules allow. That phrase matters for a club racer. If your class rules do not allow a third element, the right engineering answer is still illegal. In that case, the skill becomes recognizing the same roll-versus-heave conflict and solving as much of it as possible inside the allowed tools.

Before you tune the third spring, clean up the ordinary suspension. Haney's setup sequence around springs and bars is blunt about removing weird toe settings, assuming no severe bump-steer or roll-steer, and doing those measurements carefully in the shop before testing begins. That is not a side note. A third spring will not make a crooked or binding suspension honest. If the car has bump steer, roll steer, or a side-to-side damper mismatch, a heave-element test can become a very polished way to chase the wrong problem.

Build the baseline first

The baseline is not the setup you happened to unload from the trailer. It is the known reference condition you can return to. For this lesson, the baseline includes ride spring rates, anti-roll bar settings, damper settings, ride heights, alignment, tire condition, fuel load range, and the data channels or driver notes you will use to judge the result. If you have damper potentiometers, ride-height sensors, or load cells, the baseline also includes the channel scaling and known spring or damper load relationship.

McBeath's aerodynamic test discussion is directly useful here. The recommended test discipline is to change only one configuration, run a repeatable number of laps, average the results, discard abnormal times, and return periodically to the baseline because weather, track condition, and tire deterioration can move the target. That discipline is just as important for a third-spring test as it is for a wing test. If you change the third spring, the rear bar, the wing angle, and tire pressures in the same run, you have made a faster or slower car, but you have not learned what the third spring did.

For an intermediate driver, this is where the engineering skill becomes a driving skill. You have to drive the test consistently enough that the car can speak. A sloppy out lap, different traffic, and different braking confidence can hide the platform effect. The goal is not to set a heroic lap. The goal is to produce comparable laps that reveal whether the high-speed heave behavior improved without harming the slow sections.

What you adjust and what you do not invent

The bonded material supports the principle and the test method, but it does not provide exact third-spring rate formulas, packer gaps, bump-rubber stacks, or damper valving procedure. So the practical rule is conservative: make one modest, documented heave-platform change within the car builder's or engineer's normal range, then test it against the baseline. Do not pretend there is a universal rate or gap that works across cars.

If the car already has a third spring or heave element, the adjustment may be a spring-rate change, a preload or engagement change, or a damping change depending on the hardware. The mechanism in this lesson is the same either way: you are changing the common-motion support at one end while trying not to disturb the roll balance work that belongs to the corner springs and anti-roll bar. If you cannot identify which hardware changes heave and which hardware changes roll, stop and map the linkage before you tune.

This is also why variable-rate springs require caution. Van Valkenburgh notes that variable-rate springs are hard to deal with on a race car because they also create variable roll rate. That warning applies to the mental habit as much as the part. Any device that changes stiffness with travel can make the car feel like it has one balance early in the corner and another balance later. A third spring is useful because it can be separated from roll by linkage design. If your adjustment destroys that separation, you have made the car harder to read.

How to read the result from data

The cleanest data signature is not simply a lower lap time. A lower lap time is useful, but it is downstream of many causes. For a third-spring test, first look at ride-height or damper-displacement behavior in the fast sections where aero load is expected to matter. McBeath's discussion of downforce measurement describes using damper potentiometers calibrated to spring displacement, then multiplying displacement by spring constant to estimate spring force. The same source notes that load cells at damper tops can be used for correlation, and laser ride-height sensors can account for suspension and tire compression, though wet conditions can contaminate the readings.

Your third-spring question is simpler than a full downforce map. You want to know whether the chassis is using less unwanted heave travel at comparable speed and whether the ride-height window is more repeatable. If you have four damper potentiometers, compare the high-speed sections at the same throttle and speed range. If the heave support is working, the common compression at the tuned end should be better controlled in the targeted part of the lap.

Then check the parts of the lap that should not have changed much. Haney's reason for the third spring is that it does not engage at low speeds in the same way, allowing softer ride springs to preserve grip in slow corners. Your slow-corner apex speeds, exit consistency, and driver comments should not show a new penalty. If the car becomes harsh, nervous, or slower in the slow sections, you may have let the heave element intrude into the range where the ride springs should have remained in charge.

Finally, compare sectors rather than worshipping the whole lap. McBeath's wing-test example values sector performance and handling balance, not just the headline lap. That is exactly right here. A successful third-spring change may help the high-speed sector, leave the slow sector unchanged, and produce a modest total lap gain. A bad change may help one fast straight but cost more in a slow complex. The sector split tells you whether you solved the intended problem or merely moved the compromise.

How to read the result from the driver's seat

You should feel the car become more consistent at the speed where the platform problem existed. On a high-speed straight or fast bend, the car should stop settling deeper and deeper into the suspension in a way that makes the next input uncertain. The steering should not suddenly require a new roll-balance correction in the slow parts, because the roll path was not supposed to be the target.

That is the key driver cue: targeted calm, not global stiffness. If the whole car simply feels harder everywhere, you may have used the wrong lever. A good third-spring change can make the fast section feel more held up while the slow corner still lets the tire work. A poor change makes the car feel impressive in the paddock and less cooperative at the apex.

The driver should also report whether the car's platform changes predictably with speed. If the car feels fine on the out lap and first timed lap but changes as tires, fuel, or track conditions evolve, remember that developed race cars can be sensitive over the length of a race, and that baseline conditions can drift. This is why returning to baseline matters. Without the return, you may confuse tire deterioration or changing track grip with a third-spring effect.

Heave damping versus roll damping

The spring rate is only part of the third-element story. Blundell and Harty point out that three-spring arrangements can allow independent control of damping in roll compared with damping in heave. That is the subtle part many drivers miss. If the heave path has its own damper element, you can influence the rate at which the platform moves under common vertical load without making the car resist roll motion the same way.

For the intermediate driver, the practical takeaway is simple. Do not describe every platform problem as needing more spring. Sometimes the car needs different support over time, not just more support at displacement. The corpus does not give a damper tuning recipe, so this lesson should not invent one. But it does support the distinction: roll damping and heave damping can be treated separately in the right linkage. Your notes should preserve that distinction when you debrief the engineer.

Side-to-side equality still matters. Van Valkenburgh warns that road-racing damping rate should be equal from side to side at both ends of the car. If one side of the car has a different damper behavior, the heave test is contaminated because common motion is no longer truly common. The third spring is not a substitute for a matched left and right suspension.

A disciplined setup sequence

First, verify that the car's ordinary suspension is worthy of the test. Alignment, toe, bump steer, roll steer, damper equality, and bar freedom matter. Haney's warning about bump-steer and roll-steer being critical shop processes belongs before the third-spring day, not after. Second, define the current roll package. Know the ride springs, anti-roll bar settings, and basic front/rear roll-stiffness intent. Blundell and Harty describe roll stiffness as something to determine separately for front and rear suspension elements, with roll-center positions defining the roll axis in a full vehicle model.

Third, define the heave problem. Which end of the car compresses too far, in which speed range, and by what evidence? Fourth, make one heave-path change only. Fifth, run a disciplined comparison, ideally with a baseline return. Sixth, accept or reject the change by the combined evidence: high-speed ride-height behavior, sector performance, low-speed grip preservation, and driver repeatability.

This sequence keeps you from using the third spring as a guessing device. It also keeps the lesson distinct from the sibling lessons. Roll-center migration is about where the body rolls and how lateral load transfer changes as the geometry moves. Anti-dive and anti-squat are pitch-control geometry tools. A third spring is about common vertical support at an axle, especially when aero load creates a high-speed requirement that the slow-corner ride springs should not have to satisfy alone.

What good looks like

A good third-spring setup makes the car easier to keep in its intended platform window at high speed without making the low-speed mechanical platform worse. The driver can repeat fast sections with less change in attitude. The data shows controlled common compression where speed and aero load are highest. The slow corners do not lose the compliance that made the tires work. The sector times support the story. A baseline return confirms the gain was not track evolution.

Good also looks boring in the notebook. One change. One reason. One expected signature. One comparison. If the car improved, you know why. If it got worse, you know what to undo. That is the standard for any advanced geometry strategy: the setup becomes more understandable, not merely more complicated.

Worked example: winged road-course car with fast straights and slow corners

Imagine a club formula or sports-racing car that is strong on a fast straight but poor when the setup is stiffened enough to prevent high-speed compression. The driver reports that slow corners become nervous and traction-limited when the ride springs are increased, yet the car still needs more support at speed. This is the exact conflict Haney's third-spring discussion is built around: the normal coil-over springs would need to be too stiff for the slow corners if they had to carry the full aero load alone.

The wrong response is to keep increasing the left and right ride springs until the fastest part of the lap looks tidy. That may hold ride height, but it asks the slow corners to pay for a high-speed problem. The better response is to keep the ride springs and anti-roll bars in the range that gives the car mechanical grip and balance, then use the third spring at the affected end to add common-motion support as speed and aero load rise.

The test should be narrow. Run a baseline on the current setup for five laps if traffic and session length allow. Change only the heave element within the car's known safe range. Run another five-lap set. Compare the fast-sector speed, damper displacement or ride-height trace if available, and the slow-corner apex and exit behavior. Then return to the baseline if the session allows. If the fast section is more stable and the slow section is not worse, the third spring is doing the job it exists to do. If the slow section gets worse, you have probably let the heave solution leak into the mechanical-grip problem.

Worked example: student race car pushrod linkage

The Blundell and Harty student race car example is useful because it exposes the load path. During parallel wheel travel, the anti-roll bar mechanism can rock with the bell cranks. During opposite wheel travel, the bar resists because one side is pushed while the other is pulled. That is why a pushrod car can separate common vertical motion from roll motion if the linkage is designed for it.

Suppose this car has a balance complaint in a fast section and a separate balance complaint in a slow corner. Before touching parts, classify the complaint. If the slow corner has entry understeer or exit oversteer that tracks with body roll and lateral load transfer, the third spring is not the first suspect. You look at roll stiffness, anti-roll bar setting, ride springs, and the geometry around the roll centers. If the fast section shows both front wheels compressing together as speed rises, the heave path is now in play.

The useful engineering question is not whether the car needs to be stiffer. It is which motion needs more support. The pushrod linkage gives you the vocabulary. Opposite motion belongs to roll. Parallel motion belongs to heave. The third spring belongs in the heave conversation. When you keep those categories separate, the setup meeting stops turning into a pile of spring-rate guesses.

Worked example: aero measurement day with damper potentiometers

McBeath's aero test material gives you a practical way to keep third-spring work honest. For aerodynamic force measurement, the setup can use damper potentiometers calibrated to spring displacement, known spring constants, load cells for correlation, and laser ride-height sensors where conditions allow. The same passage warns that softer springs can improve measurement resolution, but going too soft changes ride height and therefore can change downforce. That warning applies directly to third-spring development.

On a test day, you may be tempted to soften the car dramatically so the damper pots show a large signal. That can make the data look clean while changing the thing you meant to measure. A more disciplined plan is to use a consistent setup, the same dampers with known loads, and the same instrumentation. If you are measuring aerodynamic load directly, disconnecting anti-roll bars can help isolate one-wheel bumps, binding, or hysteresis. If you are testing the third spring on track, you may keep the race bar package installed, but you still need to understand whether a bar or linkage issue is contaminating the result.

The useful output is not a decorative plot. It is a comparison: at similar speed, throttle, and track position, did the tuned end use less unwanted heave travel, and did that happen without a new slow-corner penalty? If the answer is yes and a baseline return supports it, you have evidence. If the answer is hidden by a wing change, pressure change, tire age change, or traffic, you only have a story.

Common mistakes

Mistake 1 is using the ride springs as the only platform tool. This is the classic failure. The car compresses too much at speed, so you stiffen the two corner springs. The high-speed attitude may improve, but the slow corner loses the softer mechanical setup that gave the tires grip. Good looks like asking whether the unwanted motion is common heave from aero load before you punish every roll and bump event.

Mistake 2 is tuning the third spring before the basic suspension is honest. Haney's sequence insists that bump-steer and roll-steer work must be done carefully before testing. Weird toe settings, binding, or severe steer effects can make a third-spring change look guilty or brilliant for the wrong reason. Good looks like a clean baseline: alignment known, bars free, dampers matched, and the car repeatable.

Mistake 3 is changing too many things in the same run. McBeath's test discipline is valuable because it is simple: one configuration change, repeatable laps, averages, abnormal laps removed, and periodic baseline returns. Good looks like a notebook where the third-spring change is the only meaningful setup change between the two comparisons.

Mistake 4 is ignoring the slow corners after a high-speed gain. The third spring exists partly so the normal ride springs can remain useful in slow corners. If the fast section improves but the car gives away apex speed or exit grip in the slow section, the test is not automatically a win. Good looks like a high-speed platform gain with no slow-corner mechanical-grip penalty.

Mistake 5 is forgetting that variable stiffness can create variable balance. Van Valkenburgh's warning about variable-rate springs creating variable roll rate is a reminder that stiffness changing with travel can make the car's balance move around. Good looks like preserving separation: heave support changes the common-motion problem, while roll balance remains readable.

Mistake 6 is trusting ride-height data without understanding the sensor. Damper potentiometers can be useful when calibrated to spring displacement. Load cells can correlate damper-top loads. Lasers can account for suspension and tire compression but can be contaminated in wet conditions. Good looks like knowing what your channel measures, what it misses, and whether the conditions made it trustworthy.

Drill: three-run heave split test

Use this drill only on a car that already has a legal third spring or heave element and only with a safe, approved adjustment range. The drill count is three runs. The intended duration is one test session block or three comparable sessions if traffic prevents clean running. The success criterion is a documented high-speed heave improvement with no measurable slow-corner penalty and a baseline return that supports the conclusion.

Run 1 is the baseline. Warm the car normally, then run five comparable laps. Record lap times, sector times, tire condition notes, fuel state, bar settings, damper settings, ride heights, and driver comments. If you have damper potentiometers or ride-height sensors, mark the fast sections where aero load is expected to compress the car.

Run 2 is the heave change. Make one approved third-spring or heave-element change at one end of the car. Do not change wing angle, tire pressure target, anti-roll bar, alignment, or ride spring. Run five comparable laps. The driver should focus on repeatability, not hero laps. Afterward, compare the targeted fast section first, then the slow sections that should have been protected.

Run 3 is the baseline return. Put the heave element back to the original setting and repeat the same five-lap structure if the day allows. This run tells you whether the apparent gain survived changing track and tire conditions. If the car improves in Run 2 and returns toward the original behavior in Run 3, you have a much stronger case. If all three runs drift in the same direction, the track, tires, or driver probably moved the baseline under you.

Write the result in one sentence: the heave change did or did not reduce high-speed common compression, and it did or did not preserve slow-corner grip. If you cannot write that sentence from the evidence, repeat the drill with cleaner conditions rather than making a second setup change.

When this principle breaks down

The third-spring principle breaks down when the problem is not a heave problem. A curb strike, one-wheel bump, midcorner roll-balance issue, or geometry-driven steer change should not be forced into the third-spring box. Those are real problems, but they live in other parts of the suspension system.

It also breaks down when the car does not have enough aero load to create the conflict. The bonded sources support third springs in the context of wings, ground-effects underwings, higher formula-style systems, and high-speed road-course demands. They do not support adding a third spring to an ordinary low-aero HPDE car just to make the setup more exotic. In that car, the better work is usually mechanical grip, kinematics, damping equality, and clean testing.

The principle also weakens when rules or hardware prevent true separation. If a class forbids the device, the setup has to be solved with legal springs, bars, ride heights, and geometry. If the linkage makes the heave element affect roll more than expected, you no longer have the clean separation this lesson depends on. Finally, it breaks down when the test method is loose. Changing several variables, ignoring tire deterioration, or failing to return to baseline can make any setup theory look better than it is.

Author Review

No quiz questions are attached to this lesson.

Sources

#	Document	Chunk	Pages	Score	Collection
1	The Racing and High-Performance Tire Paul Haney	bed6afd2-b117-684f-79cf-4ed27416f06e	251	1	uio_books_raw_v1
2	The Multibody Systems Approach to Vehicle Dynamics Michael Blundell Damian Harty	39c4b92a-d196-c932-a400-4bb5e68da569	368	1	uio_books_raw_v1
3	Racing Chassis and Suspension Design Carroll Smith	148524fa-62af-201e-6dff-3b729c84477a	8	1	uio_books_raw_v1
4	Competition Car Aerodynamics 3rd Edition McBeath Simon	3577d4db-719a-058d-b657-780df1923f4b	354	1	uio_books_raw_v1
5	Competition Car Aerodynamics 3rd Edition McBeath Simon	c0cd0f54-6d9c-7f08-e9af-37c31b3421d3	345	1	uio_books_raw_v1
6	Race Car Engineering Mechanics Paul Van Valkenburgh	d28a129d-788b-edbc-abf5-07f679587b4d	43	1	uio_books_raw_v1