Roll stiffness distribution: the balance knob

The total load transfer is fixed. How it splits between the axles is not — and that split is the most powerful understeer/oversteer adjustment on the car. It's also why the anti-roll bar is the cheapest tuning part you own.

Last chapter: total lateral load transfer depends on mass, grip, CG height and track, and springs have no vote.

This chapter: the split of that transfer between the front and rear axles is decided almost entirely by springs and bars — and it’s the primary balance control on any car.

How the split works

Both axles are connected by the same chassis, so they roll together through the same angle. Whichever axle resists that roll more takes a larger share of the total transfer.

\text{front share} = \frac{K_{\phi,\text{front}}}{K_{\phi,\text{front}} + K_{\phi,\text{rear}}}

where $K_\phi$ is roll stiffness in N·m per radian. And the roll stiffness of an axle comes from its wheel rate and its track:

K_\phi = \frac{1}{2} \, k_{\text{wheel}} \, t^2

That $t^2$ matters: roll stiffness goes as track squared, so a wider axle resists roll far more for the same spring.

The layout in the app:

	Wheel rate	ARB rate	Roll stiffness
Front	23.21 N/mm	25.00 N/mm	51 942 N·m/rad
Rear	15.88 N/mm	15.00 N/mm	33 270 N·m/rad

That’s a 61% front / 39% rear split — while the car’s weight is 64% front / 36% rear.

The rule that follows

Here’s the whole of balance tuning in one line:

The axle with the larger share of roll stiffness gets the larger share of load transfer, loses more grip to load sensitivity, and therefore slides first.

More front roll stiffness → the front loses grip first → understeer.
More rear roll stiffness → the rear loses grip first → oversteer.

Now compare the two splits above. Roll stiffness is 61% front against a weight distribution of 64% front — the front is carrying less stiffness share than its weight share, which nudges the car toward oversteer. And sure enough, the app reports this setup as rear-limited, oversteering at the limit, with a negative understeer gradient. The numbers agree with the rule.

Why the anti-roll bar is the tool of choice

You could change balance with springs, but springs do three jobs at once: ride frequency (chapter 12), flat ride (chapter 13), and roll stiffness. Change a spring to fix balance and you’ve moved the other two.

An anti-roll bar changes only roll stiffness. It does nothing in pure heave — both wheels rising together just rotate the bar, storing no energy — so ride is untouched. That separation is why it exists, and why it’s the first thing a race engineer reaches for.

The catch is the sensitivity. Bar stiffness goes as diameter to the fourth power:

Bar diameter	Wheel rate contribution
18 mm	7.73 N/mm
22 mm	17.25 N/mm
26 mm	33.66 N/mm
30 mm	59.66 N/mm

From 18 mm to 30 mm — a 67% increase in diameter — is a 7.7× increase in stiffness. That’s $(30/18)^4 = 7.7$ , exactly. Which is why bars come in small increments and why “the next size up” is often a much bigger change than it sounds.

Go look at it

Open the strut layout and go to the Handling tab.

Find Sway bars and the Balance readout. Note the current state: this car reads oversteer at the limit, limited by the rear.

Fix it the direct way. Increase the front bar and watch Balance move toward understeer, and Understeer grad move toward positive. You’re giving the front a bigger share of a fixed total.

Now do the same thing from the other end: put the front bar back and soften the rear bar. Same direction of change in balance, opposite part touched.

Watch Max grip as you go. There’s usually a balance point where total grip peaks, and it isn’t always where the car feels most neutral.

Now the honesty check. Note Body roll before and after a big bar change, then look at Load transfer — the total hasn’t changed at all, exactly as chapter 18 said. You moved the split.

Finally, try to reach the same balance change using spring rates instead of bars. It works — and now go back to the Springs/Shocks tab and see what happened to your ride frequency and flat-ride ratio.

The total is fixed; the split is not. Whichever axle resists roll more takes a bigger share of the transfer, loses more grip, and gives up first.

Move the balance with the bars →

What it costs you

Roll stiffness is not free grip. Every increase in an axle’s roll stiffness takes grip away from that axle, so tuning balance is always a matter of deciding which end you’d rather compromise. You are choosing where the car gives up, not whether it gives up.

There’s a second cost that catches people on rough roads: a stiff bar ties the two wheels of an axle together. Hit a one-wheel bump and the bar transmits it across the car, so a big bar makes single-wheel inputs worse and can lift the inside wheel entirely. On a smooth circuit that’s acceptable. On a bumpy road or a rally stage it’s a real loss of grip, and it’s why rally cars run softer bars than their cornering stiffness would suggest.

And a limit worth knowing: once the inside wheel lifts, adding more stiffness to that axle does nothing at all. The axle can’t transfer more than all of its load, so the adjustment saturates — a car that’s already lifting a wheel won’t respond to a bigger bar, and the fix has to come from somewhere else.

Rules of thumb

Roll stiffness split usually sits within a few percent of the weight distribution. Front share above weight share → understeer; below → oversteer.
Road cars ship with understeer built in — safer for untrained drivers, since lifting off in a corner tightens the line rather than spinning the car.
Bar diameter goes as d⁴. Changes of 1–2 mm are meaningful; 4 mm is a different car.
Change one end at a time, and expect the front bar and the rear bar to be interchangeable tools for the same job in opposite directions.
Rally-raid and desert: bars are softer than cornering alone would want, or disconnected entirely, because articulation matters more than roll control. A bar that ties the wheels together is actively harmful when one wheel needs to drop half a meter into a rut. Balance is instead handled with spring split and damping, and the roll angles these cars accept would be alarming on a circuit.

Try this

Find the bar combination that makes this car neutral at the limit rather than oversteering. How much did total grip change?
Get the same balance change twice — once with the front bar, once with the rear. Are the two cars identical in every other respect?
Soften both bars to zero. What happens to balance, roll angle, and max grip? Is the direction what you predicted?

Next: the fact underneath all of this. Tires don’t give grip in proportion to load, and once you’ve seen the curve, every chapter in this module makes sense at once.

Preset values are illustrative starting points, not a measured setup for any particular car.