Airfoil aerodynamics — camber, chord, thickness: Question bank

The three descriptors we keep returning to: the chord line (nose-to-tail ruler), the mean camber line (the halfway skeleton), and the thickness (top-to-bottom fatness). Almost every trap below comes from confusing two of these, or from pushing one to a limiting value.

Symbols and conventions (read this first)

Before any trap, we pin down every symbol used on this page, each anchored to a picture.

Definition Surface parameterisation (where

$x_{U}, x_{L}, y_{U}, y_{L}$ come from) To draw the actual skin we start on the camber line and step a distance $y_{t}$ along its normal (the direction perpendicular to the skeleton). Look at the second figure: the camber line has local slope angle $θ_{s} = arctan (\frac{d y _{c}}{d x}),$ so its unit normal points along $(- sin θ_{s}, cos θ_{s})$ . Stepping $+ y_{t}$ that way gives the upper surface point $(x_{U}, y_{U})$ ; stepping $- y_{t}$ gives the lower point $(x_{L}, y_{L})$ : $x_{U} = x - y_{t} sin θ_{s}, y_{U} = y_{c} + y_{t} cos θ_{s},$ $x_{L} = x + y_{t} sin θ_{s}, y_{L} = y_{c} - y_{t} cos θ_{s} .$ Here $θ_{s}$ is the camber-slope angle — it changes from point to point along the airfoil. (Later we meet a different angle $ϕ$ , the change-of-variable used in thin-airfoil theory; keep them separate.) Throughout, $y_{c} (x)$ is the camber line — some texts call the same curve $z (x)$ ; on this page $z \equiv y_{c}$ , so $d z / d x$ and $d y_{c} / d x$ are the identical slope. We use $y_{c}$ everywhere.

Definition Force symbols and sign conventions

Angle of attack $α$ : the angle between the chord line and the oncoming flow. Sign convention: == $α > 0$ means nose-up== (chord tilted so the nose rises into the flow).
Degrees vs radians. The lift line $c_{ℓ} = 2 π (α - α_{L 0})$ needs $α$ in radians. When a trap below writes $α = 0^{\circ}$ or $α < 0^{\circ}$ we mean the same physical angle, just quoted in the everyday degree scale; convert with $α_{rad} = α_{deg} \times π /180$ before plugging into the formula. The sign is unit-independent, so " $α < 0$ " says the same thing in either scale.
Lift $L$ : the force component perpendicular to the flow, counted positive when it points upward (away from the belly). $Γ$ is the circulation and $ρ V_{\infty}$ the density times free-stream speed in $L = ρ V_{\infty} Γ$ .
Lift coefficient $c_{ℓ}$ : lift made dimensionless, $c_{ℓ} = L / (\frac{1}{2} ρ V_{\infty}^{2} c)$ per unit span — a pure number that strips out size and speed so two airfoils can be compared fairly.
Zero-lift angle $α_{L 0}$ : the special $α$ at which $c_{ℓ} = 0$ . For a symmetric airfoil $α_{L 0} = 0$ ; camber makes $α_{L 0} < 0$ . Thin-airfoil theory gives the working line $c_{ℓ} = 2 π (α - α_{L 0})$ (with $α$ in radians).

Intuition Why the slope is exactly

$2 π$ per radian — and where the $π$ in $Γ$ comes from Picture replacing the thin airfoil by a sheet of tiny spinning vortices lying on the camber line. Their total spin is the circulation $Γ$ , and $Γ$ is what the flow "sees" as lift via $L = ρ V_{\infty} Γ$ . The sharp trailing edge forces one specific $Γ$ (the Kutta condition: flow must leave the tail smoothly).

Where the $π$ comes from. For a flat plate the vortex strength per length that satisfies the Kutta condition is $γ (ϕ) = 2 V_{\infty} α \frac{1 + cos ϕ}{sin ϕ}$ , using the same $x = \frac{c}{2} (1 - cos ϕ)$ substitution. Integrating this over the chord, $Γ = \int_{0}^{c} γ d x$ , the change of variable turns $d x = \frac{c}{2} sin ϕ d ϕ$ , cancelling the $sin ϕ$ in the denominator and leaving $Γ = c V_{\infty} α \int_{0}^{π} (1 + cos ϕ) d ϕ$ . Now $\int_{0}^{π} 1 d ϕ = π$ and $\int_{0}^{π} cos ϕ d ϕ = 0$ , so the surviving $π$ is exactly that first integral — the "average height" of the vortex distribution across the chord. Hence $Γ = π c V_{\infty} α$ : the $π$ is a definite integral over half a period of a cosine, nothing more exotic.

Feed that into $L = ρ V_{\infty} Γ$ and divide by $\frac{1}{2} ρ V_{\infty}^{2} c$ : the $ρ$ , $V_{\infty}^{2}$ and $c$ cancel, and the extra factor $\frac{1}{2}$ in the denominator doubles the $π$ to give $c_{ℓ} = 2 π α$ . So the final $2 π$ = (the $π$ from the integral) $\times$ (the $2$ from the $\frac{1}{2} ρ V^{2}$ normalisation). Camber adds a fixed extra $Γ$ (independent of $α$ ), which only slides the line sideways to give $α_{L 0}$ .

True or false — justify

A symmetric airfoil produces zero lift at every angle of attack.

False — it produces zero lift only at

α = 0^{\circ}

; tilt it nose-up (

α > 0

) and it turns air down like any wing, so

c_{ℓ} = 2 π α

grows with

α

A symmetric airfoil at a negative angle of attack produces downward (negative) lift.

True — with

α_{L 0} = 0

c_{ℓ} = 2 π α

, so

α < 0

(nose-down, e.g.

α = - 3^{\circ}

) makes

c_{ℓ} < 0

; the flow is turned upward and the reaction pushes the wing down.

A cambered airfoil can produce lift even when its chord line is exactly level with the flow.

True — camber pre-bends the flow downward, so

α_{L 0} < 0

and

c_{ℓ} = 2 π (0 - α_{L 0}) > 0

α = 0

The chord line and the mean camber line are the same line for a NACA 0012.

True — the leading "00" means zero camber, so the halfway skeleton

y_{c} (x)

lies exactly on the nose-to-tail chord everywhere.

Increasing thickness always increases the maximum lift an airfoil can make.

False — beyond a point extra thickness (or camber) triggers early boundary-layer separation, so peak

c_{ℓ}

drops and drag rises; see Boundary Layer & Flow Separation.

Camber shifts the thin-airfoil lift line left without changing its slope.

True — camber only sets

α_{L 0}

; the slope stays

2 π

per radian because that comes from the vortex-sheet/Kutta cancellation (see the sketch above), not from the shape.

Two airfoils with the same chord always have the same Reynolds number.

False —

Re = V_{\infty} c / ν

also needs the same free-stream speed

V_{\infty}

and same kinematic viscosity

ν

; equal chord alone is not enough. See Reynolds Number.

The thickness is measured perpendicular to the chord line in the NACA construction.

False — it is laid down perpendicular to the mean camber line (the normal offset above); only when the camber-slope angle

θ_{s} \to 0

do the two directions coincide.

A flat plate (zero camber, zero thickness) makes no lift at any angle.

False — a flat plate at

α > 0

still turns flow downward, and thin-airfoil theory gives it exactly

c_{ℓ} = 2 π α

; it is the reference case, not a null case.

Spot the error

"The top surface is longer, so air must speed up to rejoin at the trailing edge, and that's the lift."

The equal-transit-time claim is false — top air actually arrives sooner. Lift is net downward turning of the flow (circulation + Kutta), not a rejoin rule.

"This airfoil is very fat and curvy on both surfaces, so it has a lot of camber."

Curvature of each surface is not camber; camber is only the offset of the mean line

y_{c}

from the chord. A thick symmetric airfoil is very curvy yet has zero camber.

"NACA 2412 means 2% thick, 4% cambered, at 12% chord."

The digits are (max camber %, its position in tenths of chord, max thickness %): so

2%

camber at

40%

chord,

12%

thick.

"Camber changes the slope of the

c_{ℓ}

α

line."

It doesn't — the slope is fixed at

2 π

/rad by Thin-Airfoil Theory; camber only moves the whole line horizontally via

α_{L 0}

"Since

L = ρ V_{\infty} Γ

, lift needs zero viscosity — it's a purely inviscid effect."

The Kutta–Joukowski Theorem formula is inviscid, but viscosity is what sets the circulation

Γ

by enforcing smooth flow off the sharp trailing edge (the Kutta condition). Without viscosity you couldn't pick

Γ

"Chord length affects how much lift the shape makes per unit angle."

Chord is just the reference scale; the coefficient

c_{ℓ}

(a dimensionless number) depends on shape and

α

, not on how long the chord is.

Why questions

Why do we decompose an airfoil into a camber line plus a thickness envelope instead of just two surface curves?

Because it cleanly separates the lift-producing asymmetry (camber

y_{c}

) from the structural, symmetric part (thickness

y_{t}

), which is exactly how NACA generates and how theory analyses airfoils.

Why does thin-airfoil theory get away with modelling the airfoil as just the camber line (ignoring thickness)?

Thickness is symmetric about the camber line, so the extra vortices it adds above and below cancel in pairs to first order; the net circulation — hence all the lift — comes only from camber and angle of attack, both of which live on the mean line.

Why is the surface offset made along the normal to the camber line using

sin θ_{s}

and

cos θ_{s}

, rather than straight up and down?

Thickness is defined perpendicular to the skeleton, so we step along the unit normal

(- sin θ_{s}, cos θ_{s})

; the trig components rotate that "up" direction to match the local camber slope

θ_{s} = arctan (d y_{c} / d x)

— exactly the second figure.

Why does the naive surface formula reduce to

y_{U} \approx y_{c} + y_{t}

for a thin airfoil?

When camber slope is small,

θ_{s} \to 0

, so

cos θ_{s} \to 1

and

sin θ_{s} \to 0

; the normal direction becomes nearly vertical and the offset

y_{U} = y_{c} + y_{t} cos θ_{s}

collapses to a simple up/down addition.

Why does a symmetric airfoil have

α_{L 0} = 0

Its camber slope

d y_{c} / d x

is zero everywhere, so the zero-lift-angle integral

α_{L 0} = - \frac{1}{π} \int_{0}^{π} \frac{d y _{c}}{d x} cos ϕ d ϕ

vanishes term by term (with

ϕ

the change-of-variable angle from Thin-Airfoil Theory defined above).

Why can an aircraft with a cambered wing still fly upside down?

Inverted, the camber now points the "wrong" way, but the pilot simply increases

α

(nose-up relative to the flow) until angle-of-attack turning overpowers the adverse camber and net lift

L

is again positive (upward).

Why does thickness improve stall behaviour even though it adds drag?

A fatter nose keeps the boundary layer attached to a higher

α

, so separation and stall come on more gently — a robustness-versus-drag trade-off.

Edge cases

What is the camber of a perfectly symmetric airfoil, and what does the mean camber line look like?

Camber is exactly zero and the mean camber line lies right on top of the chord line — the "smile" is flattened into the ruler.

At the leading and trailing edges the half-thickness

y_{t} \to 0

; what happens to the upper and lower surface points there?

With

y_{t} = 0

the offsets vanish, so

x_{U} = x_{L} = x

and

y_{U} = y_{L} = y_{c}

— both surfaces pinch onto the single camber-line point (the LE and TE are shared points).

If camber slope

d y_{c} / d x

were somehow very large (steep skeleton), does the simple thin-airfoil "add

y_{t}

vertically" approximation still hold?

No — with large

θ_{s}

sin θ_{s}

is no longer negligible, so the normal offset

(- y_{t} sin θ_{s}, y_{t} cos θ_{s})

genuinely differs from vertical and you must use the full

sin / cos

surface formulas.

In the limit

α \to α_{L 0}

, what is the lift coefficient, for both a symmetric and a cambered airfoil?

c_{ℓ} = 2 π (α - α_{L 0}) \to 0

in both cases — that's the definition of the zero-lift angle; for symmetric it happens at

α = 0

, for cambered at some

α < 0

What happens to lift if you keep raising

α

well past the linear range predicted by thin-airfoil theory?

The linear

c_{ℓ} = 2 π (α - α_{L 0})

breaks down: the flow separates,

c_{ℓ}

peaks and then drops sharply (stall), because the inviscid vortex-sheet model no longer describes the real, separated flow.

Two airfoils have the same chord and the same

12%

thickness ratio. Does that force them to share the same Reynolds number?

No — equal chord and thickness do not fix

Re = V_{\infty} c / ν

; the free-stream speed

V_{\infty}

and kinematic viscosity

ν

must match too before

Re

agrees.

Recall One-line self-test before you leave

Cover this and answer: what single gap defines camber, and what does making it zero do to lift at $α = 0$ ? The gap between mean camber line and chord line ::: is the camber; setting it to zero makes $α_{L 0} = 0$ , so $c_{ℓ} = 0$ when the chord is level.

Connections

Airfoil aerodynamics — camber, chord, thickness — parent note these traps drill.
Thin-Airfoil Theory — source of $c_{ℓ} = 2 π (α - α_{L 0})$ and $α_{L 0}$ .
Kutta–Joukowski Theorem — the $L = ρ V_{\infty} Γ$ engine behind every lift item.
Bernoulli's Equation — the pressure–velocity link the "equal transit time" trap abuses.
Boundary Layer & Flow Separation — why "bigger is always better" fails.
Reynolds Number — the chord-as-reference-length trap.
NACA Airfoil Series — the 4-digit naming trap.
Lift and Drag Coefficients — why $c_{ℓ}$ is scale-free.