# Ackermann geometry: 100 %, parallel and anti-Ackermann

> Ackermann sets how much more the inside front wheel steers than the outside. It is exact for slow parking-speed turns, nearly irrelevant in fast corners, and at the limit it is really a tool for distributing slip angle between a loaded and an unloaded tyre.

Area: Steering & Ackermann · Starting level: L1 Foundation · Updated 2026-11-17
Source page: https://race-engineer-academy.vercel.app/lesson/st-ackermann

## L1 · Foundation

**The geometry.** In a turn the inside front wheel follows a tighter radius than the outside wheel. For both wheels to roll without sideways scrub at low speed, their axes must meet at a common point on the extension of the rear axle. That needs the inside wheel to steer more:

$$\delta_o = \arctan\frac{L}{R + t/2},\qquad \delta_i = \arctan\frac{L}{R - t/2}$$

with $L$ wheelbase, $t$ front track and $R$ the turn radius at the rear-axle midpoint. This is **100 % Ackermann**. Equivalent condition: $\cot\delta_o - \cot\delta_i = t/L$.

**Other settings**
- **Parallel steer (0 %)** – both wheels steer the same angle.
- **Anti-Ackermann (< 0 %)** – the outside wheel steers more than the inside.
- **Percent Ackermann** – the actual inside-minus-outside difference divided by the 100 % difference, at a given steer angle.

**How big is it?** GT3-sized car, L = 2.75 m, t = 1.65 m:
| Corner radius | $\delta_o$ | $\delta_i$ | 100 % difference |
|---|---|---|---|
| 15 m hairpin | 9.86° | 10.98° | **1.1°** |
| 50 m | 3.10° | 3.20° | 0.10° |
| 100 m fast corner | 1.56° | 1.59° | **0.03°** |

The difference scales roughly with $L\,t/R^2$. In a fast corner it is far smaller than the 3–6° of slip angle the tyres run at – Ackermann does almost nothing there. In a hairpin it is comparable to the slip angles, so it matters.

**How it is built.** The steering arms are angled so their extensions point roughly towards the rear-axle centre (the classic Ackermann trapezoid). Changing the steering-arm pickup or moving the rack fore/aft changes the percentage. Many GT3 and prototype manuals offer two or three steering-arm positions.

**See it on the turn plates.** Below, the front end of a car stands on two turn plates. The steering wheel moves the rack, the track rods turn the steering arms, and the two wheels end up at different angles. The bars compare the inner, outer and ideal (100 %) outer angles; the chart plots toe-out on lock (inner minus outer angle) against the inner angle.

**Try it:**
- Pick **0 mm parallel** and steer to full lock: both wheels turn by almost the same angle – the curve stays on the zero (parallel) line.
- Pick **+60 mm ≈100 %**: the inner wheel turns more. At **26.7°** inside the outer wheel is at **20.7°**, about **6° of toe-out on lock**, following the 100 % curve.
- Pick **−20 mm anti**: the *outer* wheel now turns more than the inner (toe-out goes negative, **−1.3°** at full lock).
- Steer only a little: all the curves start together near zero – Ackermann hardly matters at small steering angles.

**Takeaway:** the steering-arm geometry, not the rack, sets the Ackermann percentage, and its effect grows with lock. It matters in hairpins, much less in fast corners where the tyres' slip angles dominate.

**Race practice:** most race cars run well below 100 % – often near parallel, sometimes anti – except for very tight street circuits.

## L2 · Practitioner

**Why less than 100 %? The slip-angle argument.** At racing speed the wheels are *not* rolling without slip – each tyre runs a slip angle. Relative to 100 % Ackermann, a car with less Ackermann gives the outside front a larger slip angle than the inside front by the "Ackermann deficit". With parallel steer and the turn centre on the rear-axle line (low speed, rear slip ≈ 0):

$$\alpha_o - \alpha_i \approx (\delta_i - \delta_o)_{100\%} \approx \frac{L\,t}{R^2}$$

This is an upper bound. At racing speed the rear tyres also run slip angle, the turn centre moves forward of the rear axle and the split shrinks – roughly halved in a 20 m corner at the limit (Level 3).

**The load-sensitivity argument.** At the limit, the outside front may carry 2–3× the load of the inside front. A tyre's slip angle for peak lateral force *increases* with load. So the heavily loaded outside tyre wants more slip angle than the lightly loaded inside tyre. Parallel or anti-Ackermann gives exactly that. With 100 % Ackermann in a slow corner, the inside tyre is dragged beyond its (lower) peak slip angle – it scrubs, heats, generates drag and adds little lateral force.

**What the driver feels / what you see**
- Too much Ackermann at the limit in slow corners: front feels like it "pushes and scrubs", inside front overheats (inside front temps spike after a street circuit), steering may feel heavy or "fighting" at high lock, tyres squeal at low speed in the pit lane – a harmless symptom.
- Too little (anti) at low speed: car reluctant to turn at the slowest hairpin, tyre scrub in the paddock and pit lane, and in slow corners the inside tyre does too little work.
- Fast corners: practically no change – a driver who reports a big high-speed balance change after an Ackermann change is reporting something else (or the change also moved bump steer).

**Measuring:** on turn plates, steer until the outside wheel reads 10°, 15°, 20°; read the inside wheel. Compare with the 100 % values. Toe-out on turns (inside minus outside at 20°) is the quick number in many manuals.

**Pitfalls**
- Static toe is *not* Ackermann: toe-out offsets both wheels at all steer angles; Ackermann grows with steer angle. Don't measure Ackermann without zeroing toe first.
- Changing steering-arm position also changes the steering ratio and can change bump steer: re-check both.

## L3 · Advanced

**Slip-angle distribution, more rigorously.** Take a car at speed $V$, yaw rate $r = V/R$ and body sideslip $\beta$ at the CG (front axle a distance $a$ ahead, rear axle $b$ behind, $L = a + b$). For small angles the velocity direction of a front wheel at lateral offset $\pm t/2$ is

$$\theta \approx \left(\beta + \frac{a}{R}\right)\left(1 \pm \frac{t}{2R}\right)$$

so the inside wheel's path points further into the turn than the outside wheel's by $(\beta + a/R)\,t/R$. The rear slip angle fixes the sideslip: $\alpha_r = b/R - \beta$, so $\beta + a/R = L/R - \alpha_r$. Subtracting the geometric steer difference the linkage provides, $A\,L\,t/R^2$:

$$\alpha_o - \alpha_i \approx \frac{t}{R}\left(\frac{L}{R} - \alpha_r\right) - A\,\frac{L\,t}{R^2} = (1 - A)\,\frac{L\,t}{R^2} - \alpha_r\,\frac{t}{R}$$

where $A$ is the Ackermann fraction (1 = 100 %, 0 = parallel, negative = anti). At parking speed ($\alpha_r$ = 0) you recover the familiar $L\,t/R^2$. At racing speed the rear slip moves the turn centre forward and shrinks the split. L = 2.75 m, t = 1.65 m, parallel steer:
- $R$ = 15 m, $\alpha_r$ = 5°: 1.16° kinematic, but only **≈0.6°** in the corner.
- $R$ = 20 m, $\alpha_r$ = 4°: 0.65° kinematic, **≈0.32°** in the corner.
- $R$ = 100 m, $\alpha_r$ = 3°: ≈ −0.02°. The sign has reversed (because $\alpha_r > L/R$), but the value is noise next to toe and compliance – another reason Ackermann does nothing in fast corners.

**Is anti-Ackermann worth it?** Take a tyre whose peak slip angle rises with load from ~5° at 2 kN to ~6.5° at 6 kN. At a 20 m corner with heavy lateral load transfer and 4° rear slip, parallel steer gives only ~0.3° slip difference, anti-Ackermann (say −50 %) ~0.65°. Both are well short of the ~1.5° the load-sensitivity argument asks for, so moving towards anti brings the two tyres closer to their individual peaks. The gain is real but small: a few tenths of a percent of front axle grip, mostly in slow corners. Against that, anti-Ackermann increases low-speed scrub and can make the steering heavier at large angles.

**Dynamic and compliance effects.** Effective Ackermann at the limit is modified by:
- **Compliance steer:** lateral force and aligning torque steer both front wheels out of the turn, and the heavily loaded outside wheel most (it toes out). That takes steer off the outside wheel relative to the inside – a **pro-Ackermann-like** shift. It is one reason a static setting near parallel or anti can make sense: compliance adds some Ackermann back at the limit (see compliance lesson).
- **Toe change in roll (bump steer)** – shifts both wheels' slip angles differently.
- **Static toe** – a fixed offset; front toe-out acts like adding pro-Ackermann at zero steer.
So the *measured* static Ackermann is only part of the slip angle split the tyres see.

**Steering torque side-effect.** With strong pro-Ackermann at high lock, the inside wheel runs beyond its peak and its pneumatic trail collapses (or reverses), while the outside wheel's aligning torque dominates; this changes how steering torque builds in hairpins and can produce the "steering goes light then heavy" sensation at large angles.

## L4 · Expert

**When Ackermann is not about slip angles at all.** In the slowest corners of street circuits (Monaco, Macau, Long Beach, Pau) steer angles of 20–30° road wheel occur. Here kinematic differences are 2–4°, comparable to or larger than the slip angles, and too little Ackermann causes the inside wheel to *fight* the outside one – substantial drag, a stalled turn-in and a car that physically won't rotate at walking pace. That's why street-circuit kits on formula cars often switch to more Ackermann even though the high-speed theory favours parallel.

**Load sensitivity is not the only factor.** The optimum split also depends on camber per wheel (caster gives the outside negative camber, the inside positive camber with steer), on tyre temperature (an overheating inside front reduces its own peak), on drive/brake torques (the LSD-loaded inside front of a FWD TCR car under power), and on the inside tyre's role in transient yaw. Treat Ackermann as a slow-corner tuning variable, validated with inside/outside front tyre temperatures and slow-corner understeer angle, rather than as a theory-derived constant.

**Simulation note:** a bicycle model has no Ackermann. To evaluate it you need a four-wheel model with load-sensitive tyres and correct steering kinematics tables (wheel angle vs rack travel per wheel), including compliance. Otherwise the model will tell you Ackermann does nothing – which is only true in fast corners.

## Key formulas

**100 % Ackermann wheel angles**

$$\delta_o = \arctan\frac{L}{R + t/2},\quad \delta_i = \arctan\frac{L}{R - t/2}$$

- L: wheelbase (m)
- t: front track (m)
- R: turn radius at rear-axle centre (m)

**Ackermann condition**

$$\cot\delta_o - \cot\delta_i = \frac{t}{L}$$


**Percent Ackermann**

$$\%A = \frac{(\delta_i - \delta_o)_{actual}}{(\delta_i - \delta_o)_{100\%}}\cdot 100$$


**Slip-angle split, front axle**

$$\alpha_o - \alpha_i \approx (1 - A)\,\frac{L\,t}{R^2} - \alpha_r\,\frac{t}{R}$$

- A: Ackermann fraction (1 = 100 %)
- α_r: rear axle slip angle (rad)

## Trackside takeaways

- Ackermann difference scales with L·t/R²: ~1° in a hairpin, ~0.03° in a fast corner.
- Loaded tyres peak at higher slip angles, so race cars run less than 100 % – often near parallel.
- Static toe is not Ackermann; measure toe-out on turns at 10–20° with toe zeroed.
- Changing steering-arm position also changes steering ratio and bump steer.
- Tight street circuits are the exception where more Ackermann helps.

## By car category

### GT3

Manuals often offer alternative steering-arm/track-rod positions. Most teams stay near the base setting except at street tracks; inside-front temperatures in slow corners are the best tell-tale.

### TCR

FWD with road-derived steering: Ackermann interacts with traction on the inside front out of slow corners. Usually not adjustable.

### FIA F3

Spec steering geometry; Ackermann is fixed, but note it when comparing slow-corner understeer between tracks.

### Formula 4

Fixed geometry; slow-corner pit-lane scrub and squeal are normal and not a fault.

### IndyCar

Street-course packages differ from oval: ovals use near-parallel or anti; street circuits with 20°+ road-wheel lock benefit from more Ackermann.

### Super Formula

High-downforce car where most lap time is in fast corners: Ackermann is a minor tool, set for the slowest corners.

## References

- Milliken & Milliken, Race Car Vehicle Dynamics (SAE 1995): Ch. 19 Steering Systems (Ackermann geometry, percent Ackermann, race-car practice)
- Milliken & Milliken, Race Car Vehicle Dynamics (SAE 1995): Ch. 2 Tire Behavior (load sensitivity of peak slip angle)
- Segers, Analysis Techniques for Racecar Data Acquisition, 2nd ed. (SAE 2014): §7.5 The Understeer Angle (kinematic (Ackermann) steering angle – not to be confused with Ackermann steering)
