# Downforce, drag & air density

> Downforce and drag both scale with dynamic pressure ½ρV². Engineers work with the products C_L·A and C_D·A, and must correct for air density (temperature, pressure, altitude, humidity) before comparing sessions.

Area: Aerodynamics · Starting level: L1 Foundation · Updated 2026-11-16
Source page: https://race-engineer-academy.vercel.app/lesson/aero-downforce-drag-basics

## L1 · Foundation

Air flowing over, under and around the car creates a pressure field on every surface. Integrated over the car it gives one resultant force, which we split into **drag** $D$ (opposing the direction of travel) and **downforce** $L$ (pushing the car onto the road; aerodynamicists keep the symbol $L$ for "lift", downforce being negative lift).

Both forces scale with the **dynamic pressure** $q$:

$$q = \tfrac12 \rho V^2, \qquad D = q\,C_D A, \qquad L = q\,C_L A$$

The frontal area $A$ of a race car is awkward to define, so trackside we quote the products $C_LA$ and $C_DA$ (in m²) — "ClA" and "CdA". A car with $C_LA = 1.5$ m² simply makes 1.5 N of downforce for every pascal of dynamic pressure.

| Speed | $q$ at $\rho = 1.20$ kg/m³ |
|---|---|
| 100 km/h | ~460 Pa |
| 200 km/h | ~1850 Pa |
| 250 km/h | ~2890 Pa |
| 300 km/h | ~4170 Pa |

**Example (GT3, 250 km/h):** $C_LA \approx 1.5$ m² gives $L \approx 4300$ N (~440 kgf, about a third of the car's weight). $C_DA \approx 1.0$ m² gives $D \approx 2900$ N, and the power to overcome it is $D\cdot V \approx 200$ kW — roughly half of a GT3's engine output spent just pushing air.

Two consequences an engineer must feel instinctively:
- **Aero grip is a high-speed effect.** At 100 km/h the car has only $(100/250)^2 = 16\%$ of its 250 km/h downforce. Slow corners are mechanical-grip corners.
- **Downforce adds load without adding mass.** Tyre grip rises with vertical load, but the inertia the tyres must accelerate does not. That is why downforce is so valuable — and why it is worth relatively more on a light car.

**Air density** $\rho$ multiplies everything. Standard sea-level air at 15 °C is 1.225 kg/m³; a 35 °C afternoon gives ~1.15 kg/m³ (−6 %), and a track at 750 m altitude loses another ~8 %. Same car, same setup, less downforce and less drag.

Indicative values (high-downforce trim, order of magnitude only — always use your manufacturer's aero map):

| Category | $C_LA$ (m²) | $C_DA$ (m²) |
|---|---|---|
| F4 | ~0.9–1.3 | ~0.65–0.8 |
| GT3 | ~1.2–1.8 | ~0.85–1.1 |
| Formula Regional / F3 | ~2–3 | ~0.9–1.2 |
| F2 / Super Formula / IndyCar road course | ~3–4.5 | ~1.1–1.5 |
| LMP2 / Hypercar (sprint trim) | ~3–4 | ~0.8–1.1 |

## L2 · Practitioner

Trackside, aero numbers are only useful if you compare like with like. The first job is therefore **air density**, logged from a weather station (ambient temperature, absolute pressure, relative humidity) every session.

For dry air the ideal gas law gives $\rho = p/(R_d T)$ with $R_d = 287.05$ J/(kg·K) and $T$ in kelvin. Rules of thumb around standard conditions:

| Change | Effect on $\rho$ (and on $L$, $D$ at a given speed) |
|---|---|
| +1 °C air temperature | −0.34 % |
| +10 hPa pressure | +1 % |
| +100 m altitude | about −1.2 % |
| Hot & humid (30 °C, 80 % RH) vs dry | about −1.3 % |

Altitude examples at 20 °C: sea level ~1.20 kg/m³, Interlagos (~760 m) ~1.10, Mexico City (~2240 m) ~0.92 — a **quarter of the downforce and drag gone** at the same speed.

What changes for the engineer:
- **Downforce and drag both scale with $\rho$**, so $C_LA$ and $C_DA$ are the quantities to compare between sessions, not forces or top speeds.
- **Ride height at speed changes too.** Less downforce compresses the springs less, so the car runs higher on the straight. On a ride-height-sensitive car this also shifts aero balance (see the ride-height lesson).
- **Top speed depends on the engine as well.** A naturally aspirated engine loses power roughly in proportion to density, the same as drag, so top speed barely moves while corner speed drops. A boost-controlled turbo holds power until the turbo runs out of margin, so in thin air it gains top speed. Turbo GT3 cars run BoP boost tables (vs rpm and ambient pressure) partly because of this. In Hypercar, power is capped directly (torque sensors, power and energy-per-stint limits), so altitude affects those cars mainly through drag and downforce.
- **Wind.** The car feels *airspeed*, not ground speed. A 15 km/h headwind on a 250 km/h straight raises $q$ by ~12 %: more drag (lower trap speed) but more downforce in the following braking zone and corner. A tailwind does the opposite. Check the wind direction before blaming a setup change for a speed-trap difference.

**Session comparisons:** normalise speed-trap and corner-speed comparisons to the same density, or at least note $\rho$ on the run sheet. A cold morning test (10 °C, $\rho \approx 1.24$) against a hot race (35 °C, $\rho \approx 1.14$) is an 8 % downforce difference — larger than most wing changes.

**Misconception:** "the car lost downforce because the track was hot." Track temperature affects the tyres; *air* temperature affects aero. Keep them separate when you debrief.

## L3 · Advanced

**Humid-air density.** Water vapour ($M \approx 18$ g/mol) is lighter than dry air ($\approx 29$ g/mol), so humid air is *less* dense. Splitting into partial pressures:

$$\rho = \frac{p - p_v}{R_d T} + \frac{p_v}{R_v T}, \qquad p_v = \phi\, p_{sat}(T)$$

with $R_v = 461.5$ J/(kg·K) and a Magnus-type fit $p_{sat} \approx 610.78\,\exp\!\big(17.27\,T_c/(T_c + 237.3)\big)$ Pa ($T_c$ in °C). At 30 °C and 80 % RH: $p_{sat} \approx 4240$ Pa, $p_v \approx 3390$ Pa, $\rho \approx 1.150$ vs 1.164 kg/m³ dry — a 1.3 % effect, small but not negligible in a 0.5 %-resolution correlation exercise (Segers §13.2).

**Sensitivities.** Differentiating $\rho = p/(RT)$: $\delta\rho/\rho = \delta p/p - \delta T/T$. With $T \approx 293$ K one kelvin is 0.34 %; with $p \approx 1013$ hPa one hectopascal is 0.1 %. A weather station good to ±0.5 °C and ±1 hPa therefore gives density to ~±0.2 % — better than the repeatability of any on-track aero measurement.

**Corner speed with aero.** Steady-state cornering at the limit with a simple constant-$\mu$ tyre:

$$\frac{mV^2}{R} = \mu\left(mg + \tfrac12\rho C_LA V^2\right) \;\Rightarrow\; V^2 = \frac{\mu g R}{1 - \dfrac{\mu \rho C_LA R}{2m}}$$

The dimensionless group $\mu\rho C_LA R/(2m)$ is the **aero dependency** of a corner. GT3 example ($m = 1350$ kg, $\mu = 1.4$, $C_LA = 1.6$, $R = 150$ m): the group is 0.15, giving 177 km/h vs 163 km/h without aero. For a formula car with half the mass and twice the $C_LA$ the group is ~4× larger, ~0.6: the denominator is 0.4 and the corner speed is ×1.58 the no-aero value. The theoretical singularity (group = 1) is reached at $R^* = 2m/(\mu\rho C_LA) \approx 250$ m; above that radius the simple model says the car could corner at any speed. In reality tyre load sensitivity (µ falls with load), platform limits and power stop that, but the formula shows why lighter cars and faster corners are where aero pays.

**Top speed.** If drag-limited, $P_{wheel} = \tfrac12\rho C_DA V^3 + F_{roll}V$. Ignoring rolling resistance, $V_{max} \propto (P/\rho C_DA)^{1/3}$: +3 % $C_DA$ costs ~1 % top speed, ~3 km/h at 300 km/h.

**The coefficients are not constants.** $C_LA$ and $C_DA$ depend on front and rear ride height, roll, yaw, steer and (weakly) Reynolds number. Because ride height depends on downforce, which depends on $V^2$, the "constant coefficient" picture is only a local linearisation. A useful math channel is therefore an *aero-load estimate* $L(t) = q(t)\,C_LA(h_F(t), h_R(t))$ using an aero-map lookup, not a single number.

**Misconception:** "dimensionless coefficients make density corrections unnecessary." They remove the *direct* $\rho$ scaling, but the platform (ride heights, tyre deflection) still responds to the absolute load, so on a sensitive car a low-density day can genuinely change $C_LA$ and balance.

## Key formulas

**Dynamic pressure**

$$q = \tfrac12 \rho V^2$$

- q: dynamic pressure (Pa)
- \rho: air density (kg/m³)
- V: airspeed (m/s)

**Downforce and drag**

$$L = q\,C_LA, \qquad D = q\,C_DA$$

- C_LA: downforce coefficient × area (m²)
- C_DA: drag coefficient × area (m²)

**Density of humid air**

$$\rho = \frac{p - p_v}{R_d T} + \frac{p_v}{R_v T}$$

- p: absolute pressure (Pa)
- p_v: vapour partial pressure = RH · p_sat (Pa)
- R_d: 287.05 J/(kg·K)
- R_v: 461.5 J/(kg·K)
- T: air temperature (K)

**Altitude pressure (ISA)**

$$p \approx 101325\,(1 - 2.2558\times10^{-5}\,h)^{5.256}$$

- h: altitude (m)
- p: pressure (Pa)

**Limit corner speed with downforce**

$$V^2 = \frac{\mu g R}{1 - \mu \rho C_LA R/(2m)}$$

- R: corner radius (m)
- m: mass (kg)
- \mu: tyre friction coefficient

## Trackside takeaways

- Downforce and drag scale with ½ρV²: double the speed, four times the force; 100 km/h has only 16 % of the 250 km/h downforce.
- Compare C_L·A and C_D·A between sessions, not forces or trap speeds.
- Density: −0.34 %/°C air temperature, about −1.2 % per 100 m altitude, +1 % per 10 hPa; humidity lowers density slightly.
- The car responds to airspeed: always check wind direction before judging a top-speed or high-speed-corner change.
- NA engines lose power with density roughly as drag does; boosted engines don't — top-speed trends differ.
- Downforce adds load, not mass: worth more on light cars and in fast corners.

## By car category

### GT3

GT3 aero is moderate (~1/3 of weight in downforce at 250 km/h) but very BoP-relevant: wing range, ride height and boost tables are set by the organiser. Hot races cost several percent downforce and the rear usually feels it first in fast corners.

### Formula 4

F4 cars make modest downforce (C_LA around 1 m²); most corners are mechanical-grip corners. Still, wing and rake changes are visible in the fastest corners and in trap speed.

### FIA F3

F3 and F2 cars run at high dynamic pressure and are light: the aero-dependency term μρC_LA·R/2m is large, so density changes between morning and afternoon sessions are clearly visible in fast-corner minimum speeds.

### Hypercar (LMH/LMDh)

LMH/LMDh aero performance is held inside a regulated window (downforce and drag/efficiency homologated, BoP adjusts weight and power). Density matters for both energy-limited stint planning and top speed at Le Mans.

### IndyCar

IndyCar uses distinct aero configurations for road/street, short oval and superspeedway. At high-altitude or hot ovals, density changes the whole trim decision because cars run near the downforce limit flat-out.

## References

- Segers, Analysis Techniques for Racecar Data Acquisition, 2nd ed. (SAE 2014): §13.1 Aerodynamic Measurements, §13.2 Air Density, §13.3 Dynamic Pressure
- Milliken & Milliken, Race Car Vehicle Dynamics (SAE 1995): Ch. 3 Aerodynamic Fundamentals
- Milliken & Milliken, Race Car Vehicle Dynamics (SAE 1995): Ch. 15 Applied Aerodynamics
