Camber curve: how much lean, and where it comes from

Camber gain is one divide away from the swing-arm length you found last chapter. Here's the equation, why it lands within a few percent of the solver, and how to work out how much you actually need.

Last chapter you found the instant center and the swing-arm length it implies. Now the payoff, because FVSA buys you exactly one thing and it’s the most important angle on the car.

Camber is the wheel’s lean from vertical, viewed from the front. Negative camber tips the top of the wheel inward. A tire makes its best lateral grip at a small negative camber — usually −1° to −3° depending on construction — and falls off a cliff either side of that. So the entire game is: keep the loaded outside tire near its happy angle while the body is leaning over at 2 or 3 degrees and the suspension is compressed 30 mm.

Static alignment can’t do that, because static alignment is one number and the car has many positions. The linkage has to do it. That’s camber gain.

The equation

Your wheel is on an imaginary arm of length $L_{\text{FVSA}}$ pivoting at the instant center. Rotate that arm a little and the wheel rises along an arc while tipping by that same angle — so the camber change per millimeter of travel is just one over the arm length, in radians. Multiply by 57.3 to get degrees, and you have the only equation this chapter needs:

\text{camber gain} \;\approx\; \frac{57.3}{L_{\text{FVSA}}} \quad \left[\frac{\text{deg}}{\text{mm}}\right]

$L_{\text{FVSA}}$ is the swing-arm length in mm you read off the screen last chapter. One divide, and you have the number that decides whether your tire stays flat in a corner.

Everything else below is this same equation pointed in different directions: forwards to predict a car’s behavior, and backwards to size a swing arm from a camber target you choose.

Does it actually work?

Let’s check it against the solver rather than assuming. Take the app’s double wishbone layout, which has an FVSA of 2999 mm at rest, so the prediction is $57.3 / 2999 = 0.0191$ deg/mm.

Now the measured values, 0 mm to 20 mm of bump: camber goes from −0.461° to −0.853°. That’s 0.392° over 20 mm, or 0.0196 deg/mm.

Prediction 0.0191, reality 0.0196. Within 3%, from a one-line formula. Not bad for something you can do on a napkin.

Better still, change the geometry and check that the formula tracks it. Drop the upper arm’s inner mount by 20 mm — one point, nothing else touched — and the swing arm shortens to 1607 mm:

	FVSA	predicted gain	measured gain
as designed	2999 mm	0.0191 °/mm	0.0196
upper inner 20 mm lower	1607 mm	0.0357 °/mm	0.0360

The swing arm got 1.87× shorter and the camber gain got 1.84× steeper. That is the relationship doing exactly what it claims, on the same suspension, with one variable changed.

When it seems to fail (and doesn’t)

Now try it on the strut layout, whose FVSA is 3473 mm at rest. Predicted gain: $57.3 / 3473 = 0.0165$ deg/mm. Measured from 0 to +20 mm: camber goes −1.277° → −1.575°, which is 0.298° over 20 mm = 0.0149 deg/mm.

That’s 10% adrift. Did the formula break?

No — we used the wrong FVSA. From chapter 2 you already know this swing arm doesn’t sit still: it’s 3473 mm at static but 4300 mm by 20 mm of bump. The formula wants the swing-arm length over the interval, not at one end of it. Use the midpoint, roughly 3850 mm, and $57.3 / 3850 = 0.0149$ deg/mm — exactly the measured value.

The actionable rule: always evaluate camber gain at the middle of the travel range you care about, not at ride height. Which direction the static figure errs depends on which way the swing arm moves through travel — a lengthening arm makes static optimistic, a shortening one makes it pessimistic.

The equation is fine. The trap is treating FVSA as a constant when it is a function of ride height. Which brings us to the useful question: how much gain should you be aiming at in the first place?

How much gain do you actually need?

Here’s the part most guides skip. Camber gain isn’t good in itself; it’s good insofar as it cancels body roll.

When the body rolls by $\phi$ , the outside wheel — which was vertical relative to the body — is now leaning positive (top outward) by $\phi$ relative to the road. That’s the wrong way. Meanwhile that wheel has compressed, so camber gain is pulling it negative. The tire ends up at:

\gamma_{\text{road}} \approx \gamma_{\text{static}} + \phi - \frac{d\gamma}{dz}\, \Delta z

with $\phi$ the roll angle in degrees and $\Delta z$ the outside wheel’s compression. Set that equal to your tire’s preferred angle and you have a design target instead of a guess.

Worked example. A track-day car rolls 2.5° at 1.0 g, and at that roll the outside wheel compresses 30 mm. The tire wants −2.0°. You’re running −1.5° static.

Substituting: $-2.0 = -1.5 + 2.5 - \text{gain} \times 30$ , so the gain you need is $3.0 / 30 = 0.100$ deg/mm — which implies a swing arm of $57.3 / 0.100 \approx 573$ mm. That’s a very short swing arm — far shorter than anything in the rules of thumb from chapter 2, and it would come with brutal track change and a roll center that never sits still.

That’s not a calculation error. It’s the actual trade, stated honestly: you cannot fix a lot of body roll with geometry alone. Real cars split the job — some static camber, some gain, and crucially less roll to begin with via springs and anti-roll bars. Anyone promising to solve roll with camber curve alone is selling something.

Go look at it

Open the wishbone layout — the rear axle of the app’s Mini preset.

Read FVSA — Swing Arm: about 2999 mm. Divide 57.3 by it. You should get about 0.019 deg/mm. Write that down as your prediction.

Read Camber at rest: −0.5°.

Hit ▲▼ Bump and watch the Camber readout sweep. Deep in bump it reaches roughly −1.3°; at full droop it goes positive, to about +0.3°.

Check the prediction over the first 20 mm: camber moves −0.46° → −0.85°, which is 0.39° over 20 mm — 0.020 deg/mm against your napkin figure of 0.019.

Now change one thing. Drag the Upper Chassis Mount — the inner end of the upper arm — down about 20 mm. FVSA should fall to roughly 1607 mm. Re-run ▲▼ Bump: the camber sweep is visibly steeper, and 57.3 ÷ 1607 ≈ 0.036 deg/mm predicts it.

Watch what you paid. RC — Roll Center moved too, and the Setup Advisor will have opinions. A short swing arm buys camber gain and pays for it in roll-center position and track change. There is no free camber gain.

The wheel rides an arc centered on the IC, so rising and leaning are the same motion. A short arm means a tight arc and fast camber gain.

Sweep a camber curve — wishbone preset →

Rules of thumb

Camber gain: road car 0.01–0.02 deg/mm; track car 0.03–0.05 deg/mm; formula car 0.05–0.08 deg/mm. The wishbone layout above sits at 0.020 as designed, and reaches 0.036 once you shorten its swing arm with a 20 mm pivot move.
Rally-raid / desert truck: roughly 0.003–0.01 deg/mm, an order of magnitude below a circuit car — or exactly zero, if there’s a beam axle under it. Not because camber matters less, but because travel is 5–10× larger and gain gets multiplied by every millimeter of it.
Target tire camber under load: −1° to −3° for most road and track tires. Slicks often want less than people assume. Off-road tires on a rough surface want to stay near flat: the contact patch is already being abused by the terrain, and adding camber to that is a puncture waiting to happen.
Front vs rear: more gain at the front than the rear is normal — the front does the steering and sees the bigger roll contribution. Matching them exactly is rarely right.
Sanity check: if your design needs more than about 0.06 deg/mm to hit its camber target, the real problem is roll angle, not geometry. Go stiffen something.

Run it backwards to size a long-travel car

The worked example above sized camber gain from roll angle. For a desert or rally-raid car you size it from travel instead, because travel is what dominates — and it’s the same equation rearranged, not a new one.

Decide how much total camber swing the tire will tolerate across the whole stroke — commonly 2–3° for a big off-road tire — then rearrange the same equation: $L_{\text{FVSA}} = 57.3 \times \text{travel} \div \text{camber budget}$ .

A Dakar T1+ car at 350 mm of travel with a 2.5° budget needs 8000 mm. An unlimited desert truck at 600 mm needs 13,700 mm — nearly fourteen meters of virtual swing arm.

Those two numbers are the actionable output of this whole chapter for a long-travel builder: they tell you immediately that short arms cannot work, which is why these cars have such visibly wide, flat control arms, and why the beam-axle answer — where the swing arm is infinite for free — refuses to die.

The general lesson is worth taking beyond off-road: camber gain is a per-millimeter quantity, and its consequences scale with how far your suspension actually moves. A number that’s conservative on a formula car is violent on a Dakar car. Always ask “gain times what travel?” before deciding whether a figure is sane.

Try this

Set the rear FVSA as long as you can — arms as parallel as the linkage allows — and run the bump sweep. How flat can you make the camber curve, and what does the roll center do while you’re doing it?
Compute the required camber gain for your own car using the equation above. Then check it against the rules of thumb. Is your roll angle the real problem?
Compare front and rear camber curves on the same car. The front is a strut, the rear a wishbone. Which one holds its gain more consistently through travel — and does that match what chapter 2 predicted?

Next: the roll center. It’s the point that decides how much your car leans in the first place, it’s constructed from the instant center you already know how to find, and almost everything commonly said about it is wrong.

The Mini R56 preset is an illustrative sketch with approximate hardpoints, not a measured factory spec. Every number above came from the same solver the app runs, at the same hardpoints you’ll see on screen.