# Torque, power & tractive force: power-limited vs traction-limited

> Engine torque becomes wheel force through the total gear ratio; power sets how much force is available at a given speed. Below a crossover speed the tyres, not the engine, limit acceleration.

Area: Engine & Hybrid · Starting level: L1 Foundation · Updated 2026-11-16
Source page: https://race-engineer-academy.vercel.app/lesson/engine-torque-power-tractive-force

## L1 · Foundation

**Torque** is what the engine produces at the crank; **force at the contact patch** is what accelerates the car. The link is the total gear ratio (gearbox ratio × final drive) and the loaded rolling radius:

$$F_x = \frac{T_e \, i_{tot} \, \eta}{r}$$

**Power** is force times speed, $P = F_x v$. At a given road speed, the maximum force you can put down is set by power, not by peak torque: the gearbox lets you trade engine rpm for wheel force, and the best you can do at speed $v$ is $F_x = P\eta / v$. That is why engineers talk about power and gearing together, and why "a torquey engine" is mostly a statement about the *shape* of the power curve.

There is a second ceiling: the **tyres**. The driven tyres can only produce $\mu F_{z}$ of longitudinal force. At low speed, $P/v$ is huge, so the engine could deliver far more force than the tyres accept — the car is **traction-limited**. As speed rises, $P/v$ falls until it drops below the tyre limit; from there the car is **power-limited**.

Typical GT3 numbers (≈1300 kg with driver, ≈400 kW at the crank, ~57 % rear weight, slicks μ ≈ 1.5):

| | Value |
|---|---|
| Max rear-tyre drive force (static rear load) | ≈ 10.9 kN |
| Max rear-tyre drive force (with load transfer, see L3) | ≈ 14.5 kN |
| Power available at the wheels (η ≈ 0.9) | ≈ 360 kW |
| Crossover speed $v^* = P\eta / F_{max}$ | ≈ 33 m/s (119 km/h) static, ≈ 25 m/s (89 km/h) with load transfer |
| Accel at 250 km/h (drag ≈ 2.9 kN, $C_dA$ ≈ 1.0 m²) | ≈ 0.16–0.18 g |

So in a straight line a GT3 is traction-limited in 1st and much of 2nd gear, and power-limited above roughly 90–120 km/h; on a corner exit, where the rears still carry lateral force, it stays traction-limited to higher speed. A Hypercar (≈ 470 kW/t, BoP-capped) or an F2 car (≈ 580 kW/t, and no TC to help) stays traction-limited to considerably higher speed; a Tatuus F4 (≈ 230 kW/t) becomes power-limited soon after slow exits.

Why it matters trackside: in the traction-limited zone, more power does nothing — **traction control, differential, rear grip and driver throttle application** decide lap time. In the power-limited zone, **power, drag and gearing** decide it.

## L2 · Practitioner

**Measuring wheel power from data (Segers §4.1).** You do not need a chassis dyno: speed and longitudinal acceleration give it. The power used to accelerate the mass is

$$P_{mass} = m_{eff} \, a_x \, v$$

and the total wheel power adds drag and rolling resistance, $P_w = (m_{eff} a_x + \tfrac12\rho C_dA v^2 + C_{rr} m g)\,v$. Build it as a math channel and plot it against rpm on full-throttle samples only (throttle > 98 %, gear stable, no shift, |lat g| < 0.2). Segers' Viper example recovered the flywheel torque to within ~6 % — the rest being driveline loss.

**Pitfalls that fake a "down on power" engine:**
- **Gradient**: 1 % uphill costs ≈ 0.01 g, i.e. ≈ 128 N, or ≈ 3.5 kW at 100 km/h (≈ 9 kW at 250 km/h) for a 1300 kg car — 1–2.5 % of a GT3's wheel power, the same size as the effect you are trying to detect. Compare the same straight, not different ones.
- **Wind and tow**: always compare several laps; a tow at 250 km/h is worth several percent of drag.
- **Air conditions**: NA and restricted power scales ≈ $p/\sqrt{T}$ (choked flow for restrictors; the SAE J1349-type correction for NA engines), not with density: +20 °C costs ≈ 3 %, a 30 hPa lower barometer ≈ 3 %, 1000 m altitude ≈ 10 %. Turbo engines compensate up to the turbo limit (minus some knock/charge-temperature loss), restricted engines do not.
- **Mass factor**: rotating inertia. Segers uses $M_f = 1 + 0.04 + 0.0025\, i_{tot}^2$ (road-car derived); in low gears this is 10–25 % of the mass, so ignoring it **underestimates** engine power in 1st/2nd and makes the power curve look as if it rises with gear.

**Recognising which regime you are in.** Plot $a_x$ vs $v$ for full-throttle sections:
- Traction-limited: $a_x$ roughly flat (or rising with downforce) while driven-wheel slip ratio sits at the TC target (5–10 %) or TC is cutting.
- Power-limited: $a_x \cdot v \approx$ constant (a hyperbola), minus drag. Slip ratio falls to 1–3 %.

**What to change.** Traction-limited corners respond to rear grip (tyre pressures/temps, rear roll stiffness, diff drive locking, TC map, exit line). Power-limited sections respond to drag (wing, rake), shift points and gearing. A common mistake is to chase exit traction with setup in a corner where the car is already power-limited by second gear — the data will show low slip and no TC activity; the time is in the straight, not the exit.

**Top speed sensitivity.** Drag power rises with $v^3$: +10 % top speed needs ≈ +33 % power. Conversely an engine 3 % down only loses ≈ 1 % top speed, but it loses 3 % of tractive force over the whole power-limited range. Net acceleration is $(P\eta/v - D)/m$, so that is ≈ 3 % of acceleration at low speed but 5–10 % of net acceleration near the end of the straight (≈ 7 % at 250 km/h for a GT3). Check the mid/late-straight speed gradient, not just the speed trap.

## L3 · Advanced

**Traction limit with load transfer.** For a rear-wheel-drive car of mass $m$, wheelbase $L$, CG height $h$, CG-to-front-axle distance $a$ (static rear fraction $a/L$) and driven-tyre friction $\mu$, longitudinal load transfer adds $m a_x h/L$ to the rear axle:

$$m a_x = \mu\left(m g \frac{a}{L} + m a_x \frac{h}{L} + F_{z,aero,r}\right) \Rightarrow a_x = \frac{\mu\,(g\,a/L + F_{z,aero,r}/m)}{1 - \mu h/L}$$

With $a/L = 0.57$, $\mu = 1.5$, $h = 0.45$ m, $L = 2.7$ m and no aero, $a_x \approx 1.14$ g in pure straight-line, against 0.86 g from static load alone. The higher traction ceiling (≈ 14.5 kN instead of 10.9 kN for a 1300 kg car) *lowers* the straight-line crossover from ≈ 119 km/h to ≈ 89 km/h with 360 kW at the wheels. For **front-wheel drive** (TCR) the sign flips — load leaves the driven axle — so $a_x = \mu g (b/L)/(1 + \mu h/L)$; with 62 % front weight that gives ≈ 0.75 g. That structural deficit is why TCR cars live on front diff locking and why their exit traction is so sensitive to kerbs.

On a **corner exit** the tyre is also carrying lateral force: with a friction-ellipse approximation the available drive force is $F_x = F_{x,max}\sqrt{1 - (F_y/F_{y,max})^2}$. At 80 % of lateral capacity only 60 % of the drive force remains, and the *inside* rear is lighter still. Since $v^* = P\eta/F_{x,max}$, a lower available drive force means a *higher* crossover speed (×1/0.6 ≈ 1.7 in that example). That is why the crossover speed you calculate in a straight line is a **lower bound** for corner exits: with the tyre still carrying lateral force, the car stays traction-limited up to a considerably higher speed and gear.

**Aero moves the crossover.** Rear downforce grows with $v^2$, so the traction ceiling rises with speed. For a high-downforce prototype the traction curve and the $P\eta/v$ curve can run nearly parallel through 3rd–4th gear — the car is "on the edge of traction" for a long time, which is where driver throttle modulation and TC calibration earn the most.

**Equivalent mass and gear choice.** The rotating inertia reflected to the wheel is $J_e i_{tot}^2/r^2$ (engine + flywheel + clutch + gearbox input), plus wheel and drive-shaft inertia. Because it scales with $i^2$, a short 1st gear loses a significant part of its torque advantage to spinning up the engine. When comparing gearing options in a sim, model equivalent mass per gear; otherwise you will systematically over-predict the benefit of short ratios.

**Data method: power curve from the lap.** Bin $P_w/\eta$ by rpm (200 rpm bins), keep the upper envelope (the 90th percentile of each bin rejects TC cuts and shifts), and overlay with the dyno curve. A consistent offset = driveline loss or air conditions (p/√T); a shape change at a particular rpm = intake/exhaust or a restrictor reaching choke; a change only in hot laps = knock-control or IAT-related ignition retard.

## Key formulas

**Wheel drive force**

$$F_x = \frac{T_e\, i_{tot}\, \eta}{r}$$

- T_e: engine torque (Nm)
- i_{tot}: total ratio = gear × final drive (–)
- \eta: driveline efficiency (≈0.88–0.93)
- r: loaded rolling radius (m)

**Power-limited force**

$$F_x = \frac{P\,\eta}{v}$$

- P: engine power (W)
- v: vehicle speed (m/s)

**Crossover speed**

$$v^* = \frac{P\,\eta}{\mu\, F_{z,driven}}$$

- \mu: driven-tyre longitudinal friction
- F_{z,driven}: driven-axle load incl. load transfer and aero (N)

**Wheel power from data (Segers Eq. 4.8)**

$$P_w = \left(M_f\, m\, a_x + \tfrac12 \rho C_d A v^2 + C_{rr} m g\right) v$$


**Segers mass factor (Eq. 4.6)**

$$M_f = 1 + 0.04 + 0.0025\, i_{tot}^2$$


**RWD traction-limited acceleration**

$$a_x = \frac{\mu\, g\, a/L}{1 - \mu h / L}$$


## Trackside takeaways

- At a given speed, power — not peak torque — sets the maximum drive force: F = Pη/v.
- Below the crossover speed (~90–120 km/h straight-line for a GT3, higher on a combined-slip exit) the tyres limit acceleration; more power is wasted there.
- +10 % top speed needs ~+33 % power; an engine down 3 % mainly shows as slower mid/late-straight acceleration.
- Wheel power can be measured from speed + long g; compare the same straight, several laps, and correct for air conditions (≈ p/√T).
- In traction-limited corners work on rear grip, diff and TC; in power-limited sections work on drag and gearing.

## By car category

### GT3

≈300 kW/t. Traction-limited in 1st/2nd on slow exits; power-limited from ~90–120 km/h in a straight line, somewhat higher on exits that still carry lateral g. BoP power changes therefore matter mostly on medium/high-speed exits and straights.

### TCR

Front-wheel drive: longitudinal load transfer *unloads* the driven axle, so the traction limit is much lower (~0.7–0.8 g straight-line) and front-diff locking dominates exit performance.

### Hypercar (LMH/LMDh)

LMH cars with a front MGU become AWD above the deployment speed threshold, raising the traction limit; below it they are rear-driven like an LMDh. Total power is capped by BoP.

### FIA F2

≈ 620 hp for ≈ 800 kg incl. driver — very high power-to-weight with modest mechanical grip: traction-limited well into 3rd gear on slow exits, no TC allowed, so the driver is the traction controller.

### Formula 4

≈ 180 hp, ≈ 570 kg (≈ 230 kW/t): power-limited almost everywhere except the first metres of slow exits. Minimum-speed and momentum dominate; drag and rolling resistance are a large share of resistance.

### IndyCar

Boost level differs between oval and road-course spec; with high power and no TC, slow road-course exits are strongly traction-limited.

## References

- Segers, Analysis Techniques for Racecar Data Acquisition, 2nd ed. (SAE 2014): §4.1 Torque and Horsepower (Wheel torque/power from logged data, mass factor, Viper example)
- Segers, Analysis Techniques for Racecar Data Acquisition, 2nd ed. (SAE 2014): §4.1.1 Aerodynamic Drag (v³ power–speed relation)
- Segers, Analysis Techniques for Racecar Data Acquisition, 2nd ed. (SAE 2014): §4.2 Traction and Longitudinal Slip
- Milliken & Milliken, Race Car Vehicle Dynamics (SAE 1995): Ch. 20 Driving and Braking
