Official Problem Statement
The experimental tests of an airfoil in a wind tunnel indicate that the velocity profile upstream remains uniform, whereas behind the airfoil there exists a wake region with significant velocity deficit whose thickness increases downstream, resulting in the flow field sketched in the figure of CL1.pdf. For the airfoil tested, the velocity measurements at a given downstream section reveal that the velocity in the wake varies linearly from the value $u = U/2$ at the centerline to reach the value $u = U$ at the edge of the wake, located at $y = y_0$. To analyze the problem, consider a control volume limited laterally by the streamlines that extend from $x \to -\infty$, $y = \pm y_{-\infty}$ ($y_{-\infty} < y_0$) to $x = x_0$, $y = \pm y_0$ and internally by the surface of the airfoil. Assuming that the air behaves as an incompressible fluid with density $\rho$ and that viscous stresses are negligible both in the wake and along the bounding streamlines, obtain the value of $y_{-\infty}$ as a function of $y_0$, as well as the value of the drag on the airfoil as a function of $\rho$, $y_0$ and $U$. In the analysis, assume that on the outer boundary of the control volume the pressure differences from the ambient value $p_\infty$ are negligible, and so is the effect of gravity on the air flow.
Source: CL1.pdf (page 1). Notes.pdf, Chapter 3: integral mass balance Eqs. 3.10-3.11, integral momentum balance Eqs. 3.35-3.36, force of the fluid on a body Eq. 3.32.
Worked solutionexpand / collapse
Hypotheses, Data and Unknowns
- Steady, planar (everything per unit span), incompressible ($\rho$ constant); neglect $\bar{g}$.
- Viscous stresses negligible in the wake and on the bounding streamlines; $p - p_\infty \simeq 0$ on the whole outer boundary of the control volume (given). On the airfoil surface pressure and viscous stresses are not neglected: they make the drag.
- Wake profile at $x = x_0$ (symmetric about $y = 0$): $v_x = \tfrac{U}{2}\left(1 + \tfrac{|y|}{y_0}\right)$ for $|y| \le y_0$, and $v_x = U$ outside.
- Data: $[\rho,\ U,\ y_0]$. Unknowns: $[y_{-\infty},\ D]$, with $D$ the drag per unit span.
| Symbol | Meaning | SI unit |
|---|---|---|
| $U$ | uniform upstream velocity | m/s |
| $y_0$ | half-width of the wake at $x = x_0$ | m |
| $y_{-\infty}$ | half-width of the stream tube far upstream | m |
| $\rho$ | air density | kg/m$^3$ |
| $D$ | drag per unit span | N/m |
Dimensional check of the unknown: $[\rho U^2 y_0] = (\text{kg/m}^3)(\text{m}^2/\text{s}^2)\,\text{m} = \text{N/m}$, a force per unit length, as required for a planar problem.
Frames, Control Volume and Surfaces
- Inertial frame fixed to the airfoil and the tunnel; basis $(\bar{e}_x, \bar{e}_y, \bar{e}_z)$ with $\bar{e}_x$ along the stream. The control volume is fixed: $\bar{v}_c = 0$.
- Surfaces of $\Sigma_c$ (outward normal $\bar{n}$): $\Sigma_1$, far upstream, $|y| < y_{-\infty}$, $\bar{n} = -\bar{e}_x$; $\Sigma_2$, the section $x = x_0$, $|y| < y_0$, $\bar{n} = +\bar{e}_x$; $\Sigma_l$, the two bounding streamlines ($\bar{v}\cdot\bar{n} = 0$); $\Sigma_w$, the airfoil surface ($\bar{n}$ pointing into the airfoil).
- Change of variable for the wake integrals: $s = 1 + y/y_0$, $dy = y_0\,ds$; $y \in [0, y_0]$ maps to $s \in [1, 2]$.
Force of the air on the airfoil (Notes Eq. 3.32), written on the wetted surface with the normal $\bar{n}$ of the control volume:
Hence the $\Sigma_w$ terms of the momentum balance add up to $-\bar{F}$.
Balances, Surface by Surface
Each surface integral, surface by surface (the vertical line drawn under each integral in class):
| Surface | $\int\rho\,\bar{v}\cdot\bar{n}\,d\sigma$ |
|---|---|
| $\Sigma_1$ | $\rho\,U\,(-1)\,2y_{-\infty} = -2\rho U y_{-\infty}$ |
| $\Sigma_2$ | $2\rho\int_0^{y_0}\tfrac{U}{2}\left(1 + \tfrac{y}{y_0}\right)dy = \tfrac{3}{2}\rho U y_0$ |
| $\Sigma_l$ | $0$ (streamlines, $\bar{v}\cdot\bar{n} = 0$) |
| $\Sigma_w$ | $0$ (impermeable wall) |
The $\Sigma_2$ entry in detail: the antiderivative of $1 + y/y_0$ is $y + y^2/(2y_0)$, and by Barrow's rule between $y = 0$ and $y = y_0$, $\rho U\left[y + \frac{y^2}{2y_0}\right]_0^{y_0} = \rho U\left(y_0 + \frac{y_0}{2}\right) = \frac{3}{2}\rho U y_0$ (the factor 2 of the table counts both halves of the wake).
Each surface integral, surface by surface (the vertical line drawn under each integral in class):
| Surface | $\int\rho\bar{v}\,(\bar{v}\cdot\bar{n})\,d\sigma$ | $-\int(p - p_\infty)\,\bar{n}\,d\sigma$ | $\int\bar{\bar{\tau}}'\cdot\bar{n}\,d\sigma$ |
|---|---|---|---|
| $\Sigma_1$ | $\rho U\,(-U)\,2y_{-\infty}\,\bar{e}_x = -2\rho U^2y_{-\infty}\,\bar{e}_x$ | $0$ ($p \simeq p_\infty$) | $0$ (uniform flow) |
| $\Sigma_2$ | $2\rho\int_0^{y_0}v_x^2\,dy\;\bar{e}_x = \tfrac{7}{6}\rho U^2 y_0\,\bar{e}_x$ | $0$ ($p \simeq p_\infty$) | $0$ (given) |
| $\Sigma_l$ | $0$ ($\bar{v}\cdot\bar{n} = 0$) | $0$ ($p \simeq p_\infty$) | $0$ (given) |
| $\Sigma_w$ | $0$ (impermeable) | $-\int_{\Sigma_w}(p - p_\infty)\,\bar{n}\,d\sigma$ | $\int_{\Sigma_w}\bar{\bar{\tau}}'\cdot\bar{n}\,d\sigma$ |
Boxed $\Sigma_w$ terms: the force of the airfoil on the air, $-\bar{F}$. The $\Sigma_1$ entry uses (1): $-2\rho U^2y_{-\infty} = -\tfrac{3}{2}\rho U^2 y_0$.
The $\Sigma_2$ momentum entry in detail, with $s = 1 + y/y_0$, $dy = y_0\,ds$ and Barrow's rule between $s = 1$ and $s = 2$: $2\rho\frac{U^2}{4}\int_0^{y_0}\left(1 + \frac{y}{y_0}\right)^2dy = \frac{\rho U^2y_0}{2}\int_1^2 s^2\,ds = \frac{\rho U^2y_0}{2}\left[\frac{s^3}{3}\right]_1^2 = \frac{\rho U^2y_0}{2}\cdot\frac{8 - 1}{3} = \frac{7}{6}\rho U^2y_0$. The sum of the flux column is $\left(-\frac{9}{6} + \frac{7}{6}\right)\rho U^2y_0 = -\frac{1}{3}\rho U^2 y_0$.
For a wake symmetric about $y = 0$ (a symmetric airfoil at zero incidence, as the symmetric profile implies) the $y$ component of the momentum flux through $\Sigma_2$ cancels between the two halves, so $F_y = 0$: no lift.
Interpretation, Limits and Dimensional Check
- The wake carries less mass flux ($\tfrac{3}{2}\rho U y_0$) than the undisturbed stream through the same width ($2\rho U y_0$): the bounding streamlines are pushed outwards, from $y_{-\infty} = \tfrac{3}{4}y_0$ to $y_0$. The momentum deficit left in the wake is the drag.
- Limit $y_0 \to 0$: no wake, $y_{-\infty} \to 0$ and $D \to 0$ (d'Alembert's paradox for a frictionless attached flow).
- Dimensions (SI): $[y_{-\infty}] = \text{m}$; $[D] = [\rho U^2 y_0] = \text{N/m}$.
- Numerical cross-check (midpoint quadrature of the defining integrals with $2\times10^5$ cells, $U = 1$ m/s, $\rho = 1.2$ kg/m$^3$, $y_0 = 0.35$ m): $y_{-\infty}/y_0 = 0.75000$ and $D = 0.140000$ N/m $= \rho U^2y_0/3$.
- Agreement with the official solution: CL1.pdf obtains $y_{-\infty} = \tfrac{3}{4}y_0$ and $\bar{F}_{\text{airfoil}} = \tfrac{1}{3}\rho U^2 y_0\,\bar{e}_x$.
With the momentum thickness $\theta = \int_{-y_0}^{y_0}\frac{v_x}{U}\left(1 - \frac{v_x}{U}\right)dy = 2y_0\int_0^1\frac{1 - \sigma^2}{4}\,d\sigma = 2y_0\left[\frac{\sigma - \sigma^3/3}{4}\right]_0^1 = 2y_0\cdot\frac{1}{6} = \frac{y_0}{3}$ ($\sigma = y/y_0$), the result reads $D = \rho U^2\theta$, the classical momentum-deficit formula for the drag from a wake survey.



