Working memory training repeatedly practices tasks requiring temporary maintenance and manipulation of information, often adjusting difficulty as performance changes. Improving a practiced task, improving a different working memory task and improving classroom learning are separate outcomes.

A study using randomization of children with low working memory compared adaptive working memory training with comparison conditions and tested outcomes beyond the practiced exercises. Gains extended to some untrained working memory measures but did not establish improved academic progress or nonverbal reasoning.^[ref] This distinction motivates controls for practice and expectations, independent outcome tasks and follow-up measurements. A higher training score alone cannot show that a general cognitive capacity has increased; the claim depends on which additional tasks improve and for how long.

The winglet cant angle specifies the inclination of a winglet away from a stated reference direction. Here $\varphi =0$ means an upright winglet, with positive $\varphi$ tilting outward from vertical in the transverse plane. A convention measured upward from the main wing plane would instead give $90{}^{\circ }-\varphi$ for this ideal geometry. A quoted angle therefore needs its reference plane. Winglet toe angle concerns a different orientation, involving the chord and incoming flow.

For a straight unswept winglet of length $\ell$, the purely geometric projections areIf symmetric winglets attach at the tips of a main wingspan $b$, their outer-tip projected separation is $b+2\ell \sin \varphi$. With stipulated $b=10\mathrm{m}$, $\ell =0.60\mathrm{m}$ and $\varphi =20{}^{\circ }$, the vertical height is $0.5638\mathrm{m}$, each tip moves outward $0.2052\mathrm{m}$, and the total projected separation is about $10.4104\mathrm{m}$. None of these projections alone supplies a new span efficiency factor or a prediction of lift-induced drag.

Cant also changes how loads act on the supporting wing. In a right-handed frame with $y$ outward, $z$ upward and $x$ completing the triad, the rolling-moment contribution of an applied force is $M_x=y{F}_z-z{F}_y$. For a stipulated $200\mathrm{N}$ vertical force applied at the outer tip, the moment about the vehicle centreline is $1000\mathrm{N}\mathrm{m}$ when the winglet is upright and about $1041.0\mathrm{N}\mathrm{m}$ at the stated outward cant. This isolates the changed moment arm; an actual aerodynamic load need not stay vertical or keep the same magnitude when cant changes.

Richard T. Whitcomb's primary design discussion explicitly balances lift-induced drag, skin friction and wing bending moments, and discusses outward cant separately from incidence and twist. It also identifies interference at the upper-winglet root as part of the design problem.^[ref] A controlled wind tunnel comparison must therefore record forces and moments at matched conditions, rather than infer an optimum cant solely from the projected span.

The wing area $S$ is the area of an aircraft wing's projected wing planform. It is a reference area for such quantities as lift coefficient and wing aspect ratio. It is not the sum of the upper and lower curved surface areas.^[ref]

If the projected airfoil chord at spanwise position $y$ is $c(y)$, thenFor a symmetric trapezoidal wing planform this becomes twice the area under a linear half-wing chord distribution:With wingspan $b=10\mathrm{m}$, root chord $3\mathrm{m}$ and tip chord $1\mathrm{m}$, the result is $20{\mathrm{m}}^2$ for both halves together. Each half has area $10{\mathrm{m}}^2$. This factor of two matters when converting a half-wing drawing into a whole-wing reference area.^[ref]

The choice of reference area also affects reported coefficients. From $L=qS{C}_L$, where $q$ is dynamic pressure, a measured lift of $4000\mathrm{N}$ at $q=500\mathrm{Pa}$ and $S=20{\mathrm{m}}^2$ gives $C_L=0.4$. Using the half-wing area in that same whole-wing equation would incorrectly produce $C_L=0.8$. When comparing experiments or calculations, preserve the reference-area convention along with the measured force.

A wing sweep construction that shifts both projected edges aft by the same amount at each $y$ preserves $c(y)$ and therefore preserves wing area. The original comparison in wing planform illustrates this invariance. Changes to airfoil camber or thickness can likewise alter the three-dimensional surface and aerodynamic behavior without changing the projected outline.

The wing taper ratio $\lambda$ is the tip airfoil chord $c_t$ divided by the root airfoil chord $c_r$:It describes the relative narrowing of a wing planform toward its tips. A rectangular wing has $\lambda =1$; a straight-edged triangular wing has $\lambda =0$. The ratio specifies two chord values, not the entire spanwise chord distribution.^[ref]

For a symmetric trapezoidal wing planform with wingspan $b$, the chord varies linearly on each half-wing:Integrating this expression gives wing areaFor a prescribed wing area and wingspan, the required chords are thereforeFor example, $S=20{\mathrm{m}}^2$, $b=10\mathrm{m}$ and $\lambda =1/3$ give $c_r=3\mathrm{m}$ and $c_t=1\mathrm{m}$. At $y=2.5\mathrm{m}$ the chord is $2\mathrm{m}$. A rectangular wing with the same wing area and wingspan instead has $2\mathrm{m}$ chord everywhere.

Changing wing sweep while preserving these chord lengths leaves the wing taper ratio unchanged. Curved or compound edges can also share the same root and tip chords while enclosing different wing areas. Consequently the trapezoidal area formula requires the stated straight-edge assumption; a taper ratio alone cannot supply a general area or a prediction of lift.

Wing sweep describes the inclination of a chosen spanwise line in a wing planform relative to the direction perpendicular to the aircraft's longitudinal axis. It must name the reference line: a leading-edge sweep angle and a quarter-chord sweep angle can differ on the same aircraft wing. A swept wing therefore needs a reference convention, not just an unqualified angle.^[ref]

Let $x$ increase aft, let $y$ increase from root toward one tip, and define the straight leading edge and linearly varying airfoil chord byThe line at chord fraction $f$ has $x_f(y)=x_{\mathrm{LE}}(y)+fc(y)$. Differentiating gives the exact geometric relationFor $b=10\mathrm{m}$, $c_r=2\mathrm{m}$, $c_t=1\mathrm{m}$ and ${\Lambda }_{\mathrm{LE}}=20{}^{\circ }$, the quarter-chord sweep angle is $\arctan (\tan 20{}^{\circ }-0.05)\simeq 17.43{}^{\circ }$. This reference-line dependence of wing sweep is purely geometric.

There is also a useful high-speed approximation. Resolving the incoming velocity perpendicular to a swept straight leading edge gives the normal Mach number for a swept leading edge, $M_n=M\cos {\Lambda }_{\mathrm{LE}}$. Thus $M=0.9$ and ${\Lambda }_{\mathrm{LE}}=30{}^{\circ }$ give $M_n\simeq 0.779$. This component is a guide to compressibility effects, not an exact prediction of shock wave formation on a finite wing with thickness and a fuselage.

In a 1947 experiment, Richard T. Whitcomb compared unswept wings with forward and backward sweep in the Langley 8-foot high-speed wind tunnel. Pressure measurements and wake surveys over Mach numbers from 0.60 to 0.96 showed shock-associated changes occurring later for the swept configurations. The reported shifts were not exactly the simple cosine prediction.^[ref]

A variable-density wind tunnel changes the density of the test gas, typically by pressurizing the circulating flow system, to control aerodynamic similarity without relying only on higher speed or larger models. It is especially useful for matching Reynolds number when the model is smaller than the full-size object.

For characteristic length $L$, speed $U$, gas density $\rho$ and dynamic viscosity $\mu$,Under an ideal-gas approximation at fixed temperature and composition, density scales with absolute pressure. If a model has one-tenth the full-size length, using ten times the density at the same speed and approximately the same viscosity restores the same Reynolds number.

An original calculation uses full-size $L=2$ m, $U=30$ m/s, $\rho =1.2$ kg/m${}^3$ and $\mu =1.8\times 10^{-5}$ Pa·s. This gives $\mathrm{Re}=4.0\times 10^6$. A 0.2 m model at density 12 kg/m${}^3$ gives the same result at 30 m/s. At ordinary density, that model would instead require 300 m/s. If the speed of sound is stipulated as 340 m/s, the two test speeds correspond to Mach numbers about 0.088 and 0.88. Density control can therefore help match viscous scaling while avoiding a large change in compressibility effects.

The historical NACA investigation of 78 related airfoil sections used a variable-density wind tunnel to study changes in thickness and mean-line form at a large Reynolds number. Its primary report describes the models, apparatus and force/moment measurements.^[ref] The systematic geometry variation illustrates why controlling flow conditions makes comparisons interpretable.

Matching Reynolds number alone is not a guarantee of complete similarity. Mach number, relative surface roughness, incoming fluctuations and model blockage remain separate issues. Changing density also changes dynamic pressure, $q=\frac{1}{2}\rho U^2$, and hence the forces a balance must measure. The pressure vessel and circulation machinery are therefore parts of both the similarity strategy and the experimental loading problem.

Wind-tunnel turbulence intensity describes velocity fluctuations in a stated test region relative to its mean flow. For the streamwise velocity $u(t)$, define $U=⟨u⟩>0$ and $u^{\prime }=u-U$. The streamwise turbulence intensity isIts numerator is the root mean square of the fluctuation, equivalently its standard deviation when the same averaging convention is used. It is not the raw root mean square of $u$, which still contains the mean. Component direction, averaging duration, probe position and retained frequency band belong with the reported value.

A hot-wire anemometer converts changes in convective cooling into an electrical signal; measurement calibration relates that signal to speed. Dan H. Neuhart and Catherine B. McGinley's NASA study used a survey rake and additional probes, including pressure and temperature measurements for hot-wire calibration. It separated mean and fluctuating channels and used reference wires to identify coherent disturbances. Its frequency-band comparisons show why filtering can change a reported intensity.^[ref]

For an original independent-noise example, suppose measured fluctuations have standard deviation $0.30m/s$, additive sensor noise has $0.18m/s$, and $U=30m/s$. If their covariance vanishes, the physical fluctuation variance is $0.30^2-0.18^2=0.0576{\mathrm{m}}^2/{\mathrm{s}}^2$, giving $I_u=0.8\%$, rather than the uncorrected $1.0\%$. Correlated noise requires an additional covariance term; an estimated noise variance larger than the measured variance signals an inconsistent subtraction model.

Slow drift is another distinct contribution. A deterministic mean change $a(t-T/2)$ over an interval of length $T$ has time-averaged variance $a^2{T}^2/12$. With $a=0.02m/{\mathrm{s}}^2$ and $T=60\mathrm{s}$, its apparent fluctuation standard deviation is $0.3464m/s$, despite no random turbulence in this example. This mean-drift contribution to a turbulence-intensity estimate cannot be distinguished from turbulence by a single unqualified RMS number.

A tunnel survey should therefore retain the signal-processing convention and spatial positions, rather than only a facility-wide percentage. Changes in turbulence intensity can affect boundary-layer transition; equal Reynolds numbers do not guarantee equal fluctuation environments.

A wind-tunnel model support correction estimates the aerodynamic contribution of a model's mounting system to measured forces and torques. It can include the support's direct load and the altered fluid flow produced when support and model are together. An empty-tunnel support tare addresses the first effect but need not identify the second.

For a single load channel, write the defined measurement model asHere $F_M$ is the desired isolated-model load, $F_S$ is the support-only load and $I_{MS}$ is the wind-tunnel support interaction residual. The equations assume the same channel zero, conditions and support configuration, with other loads already removed. For stipulated values $F_M=20\mathrm{N}$, $F_S=3\mathrm{N}$ and $I_{MS}=2\mathrm{N}$, simple tare subtraction reports $22\mathrm{N}$, retaining a 10% bias relative to the desired load.

This limitation affected real design data. In their 1939 report, Eastman N. Jacobs and Ira H. Abbott found that support interference had made published airfoil drag coefficients too large and had exaggerated their increase with section thickness. Their correction therefore changed more than a single constant offset: it affected the inferred trend used to compare airfoil shapes.^[ref]

A proposed diagnostic repeats measurements with altered supports or attachment positions while maintaining a common target condition. Dependence on the mounting arrangement demonstrates a setup sensitivity; extrapolating to an unmounted model requires a justified mathematical model, not merely averaging the mounts. Wind-off orientation tares, support-only wind-on loads and model–support interactions answer different questions. Record each separately and propagate its measurement uncertainty through the reported correction.