Drag Coefficient Cd: Streamlined Shapes
A flat plate, a sphere and a teardrop moving at the same speed through the same air can experience drag forces that differ by a factor of 20. The drag coefficient Cd is the single number that captures all of that shape-dependence — here's what it actually measures and why dimples on a golf ball make it faster, not slower.
1. The drag equation
Every object moving through a fluid experiences a resisting force that depends on its speed, the fluid's density, its cross-sectional area, and — critically — its shape. All of that shape-dependence is bundled into one dimensionless number, the drag coefficient Cd:
where: FD — drag force (N), ρ — fluid density (kg/m³), A — reference frontal area (m²), v — relative velocity (m/s)
Because FD grows with the square of velocity, drag becomes the dominant force at high speed even for objects with a small Cd — this is why aerodynamics matters far more at 30 m/s than at 3 m/s, whether you're talking about a cyclist, a car or a cricket ball.
2. Pressure drag vs skin friction
Total drag on a real object is the sum of two physically distinct contributions:
- Pressure (form) drag: caused by the pressure difference between the front (high pressure, stagnation) and back (low pressure, wake) of the object. Dominates for bluff bodies — flat plates, spheres, cylinders — where flow separates early and leaves a large turbulent wake.
- Skin friction drag: caused by viscous shear stress in the thin boundary layer of fluid clinging to the surface. Dominates for well-streamlined bodies — teardrops, airfoils — where flow stays attached almost to the trailing edge.
Streamlining a shape (elongating it into a teardrop) reduces pressure drag by delaying separation, but increases the wetted surface area and therefore skin friction. The optimal teardrop-length-to-thickness ratio for minimum total drag is around 3:1 to 4:1 — beyond that, added skin friction outweighs the pressure-drag savings.
3. Cd table: common shapes
All values below are referenced to frontal cross-sectional area and typical Reynolds numbers for that object's real-world speed:
| Shape | Typical Cd |
|---|---|
| Flat plate (perpendicular to flow) | 1.28 |
| Sphere (smooth, subcritical Re) | 0.47 |
| Sphere (smooth, supercritical Re) | 0.10 |
| Golf ball (dimpled) | 0.24–0.30 |
| Cylinder (long, cross-flow) | 0.82–1.17 |
| Cyclist, upright position | 0.9–1.1 |
| Cyclist, aero tuck | 0.5–0.7 |
| Modern sedan car | 0.25–0.35 |
| Teardrop / airfoil (optimal ratio) | 0.04–0.05 |
Notice that the sphere's Cd nearly halves between subcritical and supercritical flow regimes at the same shape — proof that shape alone doesn't determine drag; the flow regime matters just as much (see next section).
4. Reynolds number and the drag crisis
The Reynolds number Re compares inertial forces to viscous forces in the flow, and determines whether the boundary layer around an object is laminar or turbulent:
where: L — characteristic length (e.g. diameter), μ — dynamic viscosity of the fluid
For a smooth sphere, something counterintuitive happens around Re ≈ 3 × 10⁵: the boundary layer transitions from laminar to turbulent before separating, which lets it cling to the surface longer, shrinks the turbulent wake, and causes drag to drop suddenly — this is the drag crisis. Cd falls from ~0.5 to ~0.1 over a narrow speed range.
5. Dimple aerodynamics: why golf balls have dimples
Dimples deliberately trigger the drag crisis earlier, at the much lower Reynolds numbers a golf ball actually flies at (Re ≈ 10⁵ at typical drive speed). By roughening the surface, dimples force the boundary layer to trip into turbulence early, which delays separation and shrinks the wake — a smooth ball at the same speed would separate earlier and experience roughly 2× more drag.
A rough surface reducing drag seems backwards — surely rough = more friction? The catch is that the small increase in skin friction from the dimples is vastly outweighed by the pressure drag saved by the smaller wake. This is the opposite mechanism to why golf balls also spin — see our Magnus effect article for the lift side of the story.
6. Cd across sports
Every endurance and speed sport is, in part, an optimization problem against Cd·A (drag area), since power needed to overcome drag scales as Cd·A·v³:
- Cycling: aero tuck position and drafting (riding in another cyclist's wake) can cut effective Cd·A by 20–30%.
- Speed skating and skiing: the characteristic low crouch minimizes frontal area A directly, at the cost of muscular efficiency — a real biomechanical trade-off.
- Sprint running: upright posture keeps Cd·A relatively fixed, but see our sprint force model article for how the same drag equation applies there.
7. Measuring Cd in a simple simulation
In a numerical drag simulation, Cd is usually a lookup constant (or, for accuracy, a small function of Reynolds number), applied directly in the equation of motion:
function dragForce(v, Cd, rho, A):
// Standard drag equation, direction opposes velocity
speed = magnitude(v)
F_mag = 0.5 * Cd * rho * A * speed ** 2
return -F_mag * normalize(v) // vector opposing motion
function CdFromReynolds(Re):
// Simplified drag-crisis curve for a smooth sphere
if Re < 2e5:
return 0.47 // subcritical, laminar separation
else if Re < 4e5:
// interpolate through the drag crisis
t = (Re - 2e5) / 2e5
return 0.47 - t * (0.47 - 0.10)
else:
return 0.10 // supercritical, turbulent separation
Combined with a semi-implicit Euler or RK4 integrator, this handful of lines is enough to model everything from a falling raindrop reaching terminal velocity to a cyclist's power curve against a headwind.
🔗 Related Simulations
💨 Try the drag coefficient simulation
Compare drag forces across shapes interactively — sphere, teardrop, flat plate — and watch how Reynolds number shifts Cd in real time.
Open simulation →