# Steering ratio: steering-wheel angle vs road-wheel angle

> The logged steering channel measures the steering wheel, not the tyres. Knowing the true, non-linear and compliant relationship to road-wheel angle is the basis of every steering-related math channel.

Area: Steering & Ackermann · Starting level: L1 Foundation · Updated 2026-11-16
Source page: https://race-engineer-academy.vercel.app/lesson/st-steering-ratio

## L1 · Foundation

**Overall steering ratio** = steering-wheel angle / road-wheel angle:

$$SR = \frac{\delta_{SW}}{\delta_{road}}$$

For a rack-and-pinion system with rack travel $c$ per steering-wheel revolution (the "C-factor", mm/rev) and steering-arm length $a$:

$$SR \approx \frac{2\pi\,a}{c}$$

Example: $a$ = 120 mm, $c$ = 60 mm/rev → SR ≈ 12.6:1. 90° at the steering wheel → ~7° at the road wheels.

**Typical ratios (indicative)**
| Car | Overall ratio | Comment |
|---|---|---|
| Road car | 14–17:1 | comfort, low effort |
| GT3 / GT4 | ~11–15:1 | power assisted |
| TCR | ~12–15:1 | road-derived rack, assisted |
| Formula (F4–F2) | ~8–12:1 | unassisted, very little lock |
| Prototype | ~9–13:1 | assisted |

**What ratio choice changes**
- **Quicker (lower) ratio:** less hand movement, faster corrections – but more steering *effort* (torque at the wheel scales with 1/SR) and a twitchier feel around centre.
- **Slower (higher) ratio:** lighter, calmer, but more hand movement – at tight hairpins the driver may run out of comfortable hand position (hand-over-hand is unacceptable with a race wheel full of buttons).

**Important:** the steering-angle sensor on the column measures steering-wheel angle. To get what the tyres actually do you need the road-wheel angle – via the ratio, which you must measure on your car.

## L2 · Practitioner

**Measuring the ratio (Segers' method)**
1. Car on turn plates (or laser alignment) at ride height with load, wheels straight.
2. Steer in steps (e.g. 10° or 20° at the steering wheel, read from the logged steering channel itself, not a protractor) both directions to full lock.
3. Record **outside** wheel angle for each direction (left wheel for right turns, etc.).
4. Fit a curve: road-wheel angle = f(steering angle). Often near-linear for small angles; polynomial if the rack/arm geometry is non-linear.
5. Put that function in the logger as a math channel – it is the basis for understeer angle and Ackermann-based analysis.

Why the *outside* wheel: it is the more loaded wheel and the one that dominates the axle's lateral force; it is also the convention Segers uses for the understeer angle. Note that with Ackermann, the inside wheel angle is larger – using the average (or inside) gives a different "ratio".

**Non-linearities you will find**
- **Ackermann** makes inside and outside ratios diverge with steer.
- **Steering-arm/track-rod angles:** the ratio changes with lock (often quicker at larger angles on race geometry).
- **Variable-ratio racks** on some road-derived cars (TCR, GT4).
- **Bump steer:** the wheel angle at a given steering-wheel angle changes with ride height.

**Compliance – the on-track ratio is not the static one.** Under aligning torque the column, U-joints, rack mounts and track rods twist. The driver must add steering-wheel angle to get the same road-wheel angle: the effective ratio *rises* with load. On an unassisted formula car this can be several degrees of steering wheel at the limit; on assisted cars the column twist is reduced but rack-mount compliance remains.

**Calibration pitfalls**
- Steering sensor zero drifts after a steering-wheel removal or column work: recalibrate on the alignment pad, and check in data – the straight-line steering average should be ~0 (or a small, consistent offset from crosswind/camber).
- Toe changes also offset the zero relationship.
- Swapping the steering rack (spares) can change the C-factor: check part numbers and remeasure.

## L3 · Advanced

**Choosing a ratio as an engineer.** The ratio sets two driver-relevant gains:
- **Yaw-rate gain per steering-wheel degree:** $\frac{r}{\delta_{SW}} = \frac{1}{SR}\cdot\frac{V/L}{1 + K\,V^2/(g L)}$ (linear bicycle model, $K$ = understeer gradient in rad/g). At high speed this falls; quick racks keep the car responsive at high speed but make it nervous on straights.
- **Torque gain:** steering-wheel torque ≈ kingpin moment / (SR·η) minus assist. Changing ratio from 12 to 10:1 raises effort by 20 % on an unassisted car.

**Ratio and driver workload in data.** Because $\delta_{SW} = SR\cdot\delta_{road}$, steering-wheel velocity and acceleration (deg/s, deg/s²) scale **with SR** for the same road-wheel motion. On a quicker rack (lower SR) a driver looks smoother in the steering-wheel channels without being any smoother at the tyres; on a slower rack the same tyre inputs show as bigger, faster hand movements. Convert to road-wheel angle (divide by SR, or use the measured map) before comparing steering-speed channels or their lap averages across cars, drivers on different racks, or rack changes (Segers §14.6 builds the steering-speed smoothness channel; the ratio correction is our addition).

**Kinematic vs effective ratio from data.** You can estimate the *effective* on-track ratio by regression: in steady, low-g corners (where tyre slip and compliance are small), road-wheel angle ≈ $L\,a_y/V^2$ + small understeer; plotting $\delta_{SW}$ against $L\,a_y/V^2$ for gated low-g data gives a slope ≈ SR (plus understeer gradient contribution). Divergence at higher g mixes compliance and understeer – which is exactly why understeer-angle channels on high-compliance cars read high.

**Steering ratio in simulators.** Drivers adapt to ratio quickly but sim correlation requires the column compliance and assist model; otherwise the sim car feels too direct and drivers over-steer the real car in the first laps after a rack change.

## Key formulas

**Overall steering ratio**

$$SR = \frac{\delta_{SW}}{\delta_{road}}$$


**Rack-and-pinion ratio**

$$SR \approx \frac{2\pi\,a}{c}$$

- a: effective steering-arm length (mm)
- c: rack travel per steering-wheel revolution (mm/rev)

**Steering-wheel torque**

$$T_{SW} \approx \frac{M_{kp,L} + M_{kp,R}}{SR\cdot\eta} - T_{assist}$$

- M_kp: moment about each kingpin (N·m)
- η: steering efficiency

## Trackside takeaways

- The steering channel is the steering wheel – convert to road-wheel angle with a measured map.
- Measure on turn plates, outside wheel, both directions, fit a function, use it as a math channel.
- Quicker ratio = less hand movement but more effort and twitchier centre feel.
- Compliance makes the effective on-track ratio higher than the static one.
- Steering-wheel velocity scales with SR: convert to road-wheel angle before comparing drivers or rack changes.

## By car category

### FIA F3

Unassisted with a quick ratio: effort is high and column compliance is noticeable in fast corners. Driver fitness and steering weight complaints are common.

### Formula 4

Unassisted; good car to see how column twist changes the steering trace vs lateral g.

### GT3

Power-assisted; some cars offer alternative rack ratios or assist maps. Re-calibrate the steering sensor after any steering-wheel quick-release work.

### TCR

Road-derived variable-ratio racks are possible; measure the full curve, not one point.

### IndyCar

Ratio differs between oval and road/street configurations; street courses need enough lock for hairpins.

### Hypercar (LMH/LMDh)

Assisted steering with relatively quick ratios; hybrid front-axle torque (LMH) adds steering torque effects under deployment.

## References

- Milliken & Milliken, Race Car Vehicle Dynamics (SAE 1995): Ch. 19 Steering Systems (overall ratio, steering compliance)
- Milliken & Milliken, Race Car Vehicle Dynamics (SAE 1995): Ch. 23 Compliances
- Segers, Analysis Techniques for Racecar Data Acquisition, 2nd ed. (SAE 2014): §7.5 The Understeer Angle (measuring steering-wheel vs wheel angle on turn plates)
