X-Planes: Northrop X-21

The Northrop X-21A is the only aircraft in this series that worked and was cancelled for it. Two Douglas WB-66D bombers were given new wings cut with thousands of fine spanwise slots, and pumps inside the wing drew a thin sheet of air through those slots to hold the boundary layer laminar. It did what it was built to do. By the end of the programme the aeroplane was flying with laminar flow over more than ninety-five percent of the area intended for it, in Braslow 1999. Then the programme stopped. This article is the twenty-second in the X-Planes series, following the X-1, the X-2, the X-3, the X-4, the X-5, the X-6, the X-7, the X-8, the X-9, the X-10, the X-11, the X-12, the X-13, the X-14, the X-15, the X-16, the X-17, the X-18, the X-19, and the X-20.

The received account is that laminar flow control failed on maintenance. The slots clogged with debris, insects and rain, and the maintenance the intricate suction system demanded ended the programme. That account is widely repeated, and the Northrop X-21A reference entry gives it. This article is largely about one problem with it. The service-maintenance evaluation was never performed. It was one of the programme’s stated objectives, the programme had not yet been able to address it when the money ran out, and the advisory group that reviewed the programme in October 1965 recommended a major wing modification before any meaningful maintenance data could be gathered. The modification was never made, in Braslow 1999.

So the X-21 is remembered for answering a question it never asked. The standard inventory entry is Jenkins Landis and Miller 2003 American X-Vehicles, An Inventory X-1 to X-50 and the vehicle compilation is Miller 2001 The X-Planes, X-1 to X-45. The indispensable source here is Braslow 1999 A History of Suction-Type Laminar-Flow Control with Emphasis on Flight Research, written by the engineer from the National Aeronautics and Space Administration, or NASA, who served as technical consultant to the Air Force on this programme, and the contemporary engineering account is Kosin 1965 in the Journal of Aircraft. Braslow reproduces the 1965 review group’s report in full as a document appended to the monograph, the Northrop analysis of the cloud problem survives as Hall 1964, the description of the wing written while it was flying is Butz 1963 in Air Force Magazine, and the technical results of the Northrop final report are summarised, with their numbers, in Joslin 1998.

The Research Question

About half the fuel a subsonic transport burns in cruise goes to overcoming skin friction, and on an aircraft of that size the boundary layer is turbulent over nearly all of it, in Butz 1963. A laminar boundary layer exerts roughly a tenth of the shear. The prize is therefore enormous and has been obvious since the 1930s, and the history of the subject is a history of trying to collect it, in Freeman 1935, Quinn 1944, Von Doenhoff 1944, Cliett 1952, Torda 1952, Northrop Aircraft Inc Hawthorne Ca 1956, Raspet 1957, Northrop Corp Hawthorne Ca Norair Div 1964, Northrop Corp Hawthorne Ca Norair Div 1964, Kosin 1965, Anderson et al 1969, Khusnutdinova 1972.

Laminar flow is not stable at these Reynolds numbers. It becomes turbulent through the growth of small disturbances, and suction through the surface thins the boundary layer and makes the profile fuller, which damps those disturbances. That those disturbances exist at all, and grow at the rate theory predicted, was settled by Schubauer and Skramstad 1947, whose measurement of laminar boundary-layer oscillations turned boundary-layer stability from a calculation into an observed phenomenon and made everything the X-21 attempted arguable in advance. The theory and measurement of that mechanism are in Trilling 1950, Ringleb 1952, Schwartzberg and Braslow 1952, Gortler 1957, Hasimoto 1957, Gribben 1959, Dennon 1962, Görtler’s new series method for steady laminar boundary layer flows with suction 1962, Morduchow 1963, Pate and Deitering 1963, Pechau 1963, Kozlov 1965, Kozlov 1966, Messiha 1966, Mills 1966, Singleton 1967, Zamir and Young 1967, Kozlov 1968, Leitsina and Pavlyukevich 1968, Tokhunts 1968, Kozlov and Tsyganyuk 1969, Baldwin 1970, Gershbein 1970, Pushkareva 1970, Kobayashi 1972, Kobayashi 1974, Savoshchik and Tokhunts 1974, Harvey 1975, Carter 1977, Srokowski 1978, Thiede 1978, Mack 1980.

The period record behind that sentence is large, and it supplies about an eighth of this article’s dated research records. The 1930s to the 1960s produced a continuous literature on boundary-layer control, low-drag aerofoils, profile drag, transition and the behaviour of swept wings, gathered in the clusters below on suction laminar flow control, on natural laminar flow and low-drag aerofoils and on boundary-layer stability and transition, and represented here by Von Doenhoff 1941, Holt 1943, Zalovcik 1945, McCullough and Gault 1947, Brennecke 1948, Loftin and Burrows 1949, Pfenninger 1949, Black 1952, Banner et al 1958, Gregory 1960. The programme was built on top of this body of work, and read together it makes the X-21 look less like an experiment and more like an engineering demonstration of a settled result.

The X-21 was not built to discover whether this works. It was known to work, in Braslow 1999. It was built to find out three practical things. Whether a slotted suction wing can be manufactured to the required tolerance at acceptable cost. Whether laminar flow can be held on a large swept wing in real air. And whether such a wing can be operated and maintained by a squadron rather than by a laboratory, in Joslin 1998.

It answered the first two. It never got to the third.

Programme Origin

Werner Pfenninger brought continuous-suction laminar flow work to the Norair division of Northrop after the Second World War, and through the 1950s that group ran an extended series of Air Force contracts on slot arrangements, crossflow stabilisation and wing design, in Butz 1963 and Braslow 1999. The wind tunnel and glove experiments that preceded the aeroplane are in Brennecke 1948, McCullough and Gault 1948, Loftin and Burrows 1949, Pfenninger 1949, Northrop Aircraft Inc Hawthorne Ca 1956, Lachmann 1959, Gregory 1961, Groth 1961, Pfenninger and Bacon 1961, Pfenninger and Groth 1961.

The direct ancestor is a glove on a fighter. In the middle 1950s Northrop flew a slotted suction section built into one wing of a Lockheed F-94, and the flights showed laminar flow over the whole of the suction section, in Butz 1963. Joslin 1998 records that the F-94A glove carried 69 slots, that it held fully laminar flow in level flight, climbs, turns and descents, and that it lost laminar flow in cloud and regained it within 30 seconds of leaving the cloud, the X-21’s cloud result seen a decade early on a smaller wing.

The programme was approved in stages. The review group’s report, reproduced in Braslow 1999, dates the X-21A Laminar Flow Control Demonstration Program from August 1960, following a report of the Aerospace Vehicles Panel of the Scientific Advisory Board on boundary layer control dated 25 June 1959, and lists the reviews that followed it. The Vehicles Panel reexamined the programme in November 1961 and November 1962, an advisory group of the Aeronautical Systems Division reviewed it on 22 and 23 June 1964, and the Air Force held a programme review on 13 January 1965. Braslow adds that the Wright Air Development Division chose the WB-66D for minimum cost, a high degree of safety and a short development time, and that the Federal Aviation Administration contributed money. Butz 1963 gives the sums as 35 million dollars from the Air Force and 2 million from the Federal Aviation Agency.

The Air Force wanted the result for range. An aircraft with the drag of an X-21 wing could loiter far from base or cross an ocean without refuelling, and the 1963 case for the programme was put in those terms in Butz 1963, promising a lift-to-drag ratio raised from about twenty to more than thirty and an airframe more than forty percent smaller for the same mission. The example given was one model of the B-52, reduced from 400,000 pounds to 225,000 pounds without changing its speed, range, endurance or load, which is a reduction of

\[1 - \frac{225{,}000}{400{,}000} = 0.44\]

or 44 percent. Those numbers are for an aircraft designed around laminar flow from the start. They are not what a converted bomber can show, and the difference between the two turns out to matter more than it looks.

The Aircraft and Its Suction System

The X-21A was a WB-66D with the engines moved and the wing replaced. Two General Electric J79 turbojets, which the cutaway drawing in Butz 1963 labels YJ79-GE-13, went to pods on the rear fuselage, which made room under the wing for the suction pumps. The new wing spans 93.5 feet with 1,250 square feet of area and 30 degrees of sweep, taken here as quarter-chord sweep, reduced from the 35 degrees of the bomber, in the Northrop X-21A reference entry. It was scaled from a Douglas DC-8 wing at about forty percent, so that the result would be representative of a large subsonic transport instead of a research shape.

\[\bar{c} = \frac{S}{b} = \frac{1{,}250}{93.5} = 13.37\ \text{ft}\] \[A\!R = \frac{b^{2}}{S} = \frac{93.5^{2}}{1{,}250} = 6.99\]

The suction surface needs careful description, because it is routinely described wrongly. The wing skin carries spanwise slots running from root to tip, between 0.0025 and 0.008 inches wide, comparable with the cutting edge of a razor blade. Beneath each slot a plenum chamber is milled into an inner skin, and from each plenum a large number of small holes is drilled down into deep ducts formed between layers of bonded sandwich structure. Air is drawn through the slots into the plenums, through the holes into the ducts, and along the ducts at speeds reaching about two hundred miles per hour to four small gas turbines, two in each underwing nacelle, one serving the forward wing and one the rear, as Butz 1963 describes the installation. The skin panels, usually about five feet by twenty, were set in the final assembly jig first and the spars and understructure added behind them, so that the wing was built from the outside in and the outer contour did not follow the structure’s tolerances. Butz reports the programme’s estimate that such a wing would cost only about twenty percent more to produce than a conventional wing of the same size and weight.

Two numbers make the surface concrete. With leading-edge slots about 0.0035 inches wide on a pitch of 0.75 inches, the values Joslin 1998 gives for the slots that controlled the flow at the attachment line, the open fraction of the skin is

\[\sigma = \frac{w}{p} = \frac{0.0035}{0.75} = 0.00467\]

The suction surface is 99.5 percent solid. And since the flow through the skin averages 0.145 feet per second, continuity gives the speed in the slots themselves as

\[v_{\text{slot}} = \frac{v_{w}}{\sigma} = \frac{0.145}{0.00467} = 31\ \text{ft/s}\]

or about 21 miles per hour, against the 200 miles per hour reported in the ducts below. The air does not rush through the slot. The surface behaves as a very slightly porous skin.

The figure of eight hundred thousand, which reference works such as the Northrop X-21A entry give as a count of slots, is a count of those metering holes. Butz 1963 is explicit, describing air gathered in small chambers and channelled through pin-size holes, more than 800,000 of them, into the larger ducts. No primary document available here gives a count of the slots themselves, but their spacing bounds it. Joslin 1998 gives typical slot spacings of 1.1, 2.0 and 1.2 inches over 1 to 5, 5 to 40 and 40 to 100 percent of the chord. At the mean chord of 160.4 inches each surface therefore carries about

\[N_{\text{slot}} = \sum_{i} \frac{\Delta (x/c)_{i} \, \bar{c}}{s_{i}} = \frac{0.04 \times 160.4}{1.1} + \frac{0.35 \times 160.4}{2.0} + \frac{0.60 \times 160.4}{1.2} = 5.8 + 28.1 + 80.2 = 114\]

slots, where $\Delta (x/c){i}$ is the chordwise extent of each region and $s{i}$ its slot spacing, so the wing has of the order of a hundred slots on each surface and a little over two hundred in all. The estimate ignores the closer spacing on the attachment line and treats the mean chord as typical. The distinction is not pedantry. A slot is a continuous span-long opening and a hole is a discrete metering orifice, and confusing the two makes the wing sound like a sieve when it is in fact a small number of very fine, very long, very accurately cut openings. The design and construction of such surfaces is treated in Cliett 1952, Gregory 1961, Meade 1977, Klotzsche et al 1984, Lineberger 1984, Fischler 1986, Maddalon and Poppen 1986, Goodyear 1987, Maddalon and Braslow 1990, Maddalon 1991, Traub et al 2021, Prasannakumar et al 2024, Kashaev et al 2026.

Sizing From First Principles

The natural question is whether the suction power exceeds the drag it saves. At the suction coefficient assumed here it does not, by a wide margin, although the local coefficients the X-21 design called for narrow the margin considerably, as the end of this section shows. The calculation deserves care, because its answer removes the obvious explanation for the programme’s fate and forces the search elsewhere.

Take a nominal cruise at Mach 0.75 and 40,000 feet. At that altitude the density is $5.851 \times 10^{-4}$ slugs per cubic foot and the speed of sound is 968.1 feet per second, so the flight speed and dynamic pressure are

\[V = M a = 0.75 \times 968.1 = 726.0\ \text{ft/s}\] \[q = \tfrac{1}{2} \rho V^{2} = 154.2\ \text{lbf/ft}^{2}\]

and at the 83,000 pound gross weight of the Northrop X-21A entry the wing is working at

\[C_{L} = \frac{W}{q S} = \frac{83{,}000}{154.2 \times 1{,}250} = 0.431\]

The mean-chord Reynolds number is

\[Re_{\bar{c}} = \frac{V \bar{c}}{\nu} = \frac{726.0 \times 13.37}{5.075 \times 10^{-4}} = 1.913 \times 10^{7}\]

At that Reynolds number the mean skin-friction coefficients on a smooth flat plate are

\[C_{f,\text{turb}} = \frac{0.455}{(\log_{10} Re)^{2.58}} = 0.002713\] \[C_{f,\text{lam}} = \frac{1.328}{\sqrt{Re}} = 0.000304\]

The ratio of the two sets the size of the prize,

\[\frac{C_{f,\text{lam}}}{C_{f,\text{turb}}} = \frac{0.000304}{0.002713} = 0.112\]

so the laminar boundary layer exerts 11.2 percent of the turbulent shear and the saving available is

\[\Delta C_{f} = 0.002713 - 0.000304 = 0.002409\]

On the area the programme guaranteed to laminarise, 70 percent of the wing in Butz 1963, with the exposed wetted area of the wing taken as about 2,222 square feet, that coefficient is worth

\[\Delta D = q \, \Delta C_{f} \, S_{\text{lam}} = 154.2 \times 0.002409 \times 1{,}555 = 578\ \text{lbf}\]

What the Suction Costs

The suction system ingests a mass flow set by the suction coefficient $C_Q$, the ratio of the velocity through the surface to the freestream velocity,

\[v_{w} = C_{Q} V_{\infty} = 2 \times 10^{-4} \times 726.0 = 0.145\ \text{ft/s}\]

The suction amounts to no more than that. The air is drawn through the skin at about a tenth of a mile per hour, and everything the wing achieves follows from removing it. The mass flow is

\[\dot{m} = \rho_{\infty} V_{\infty} C_{Q} S_{\text{wet}}\]

That air arrives in the ducts at a total pressure well below freestream total pressure, because it has passed through a slot narrower than a razor edge and along a duct. To expel it without a momentum penalty the system must restore it, and the ideal work per unit mass is the isentropic compression work. Writing $\Pi$ for the ratio of freestream total pressure to duct total pressure and $\kappa = (\gamma-1)/\gamma$, the shaft power is

\[P = \frac{\dot{m} c_{p} T_{t}}{\eta} \left[ \Pi^{\kappa} - 1 \right]\]

The equivalent drag is that power divided by flight speed, and dividing again by $q S_{\text{wet}}$ gives a drag coefficient on the same basis as the skin friction,

\[C_{D,\text{suction}} = \frac{2 C_{Q} c_{p} T_{t} \left[ \Pi^{\kappa} - 1 \right]}{\eta V^{2}}\]

Density has cancelled. Substituting $c_p T = a^{2}/(\gamma-1)$ and $V = M a$ removes the atmosphere entirely and leaves

\[C_{D,\text{suction}} = \frac{2 C_{Q} \left[ \Pi^{\kappa} - 1 \right]}{\eta (\gamma-1) M^{2}} \left( 1 + \frac{\gamma-1}{2} M^{2} \right)\]

The suction penalty is a function of Mach number and of the machinery, and of nothing else. It does not depend on altitude at all. An aircraft climbing at fixed Mach number carries exactly the same suction drag coefficient at 20,000 feet as at 50,000, which is not obvious in advance and which removes an entire axis from the design trade.

With a design suction coefficient of $C_Q = 2 \times 10^{-4}$, a duct total-pressure recovery of 0.60 and a combined turbine and compressor efficiency of 0.75,

\[C_{D,\text{suction}} = 4.144 \times 10^{-4}\]

which is 17.2 percent of the friction saving. Five sixths of the prize survives the pumping.

How Far From Break-Even

Setting the suction penalty equal to the friction saving gives the suction coefficient at which the system stops paying,

\[C_{Q,\text{crit}} = \frac{\Delta C_{f} \, \eta (\gamma-1) M^{2}}{2 \left[ \Pi^{\kappa} - 1 \right] \left( 1 + \tfrac{\gamma-1}{2} M^{2} \right)} = 1.163 \times 10^{-3}\]

That is 5.8 times the design value. In absolute terms, at the nominal condition and over the wing area the programme guaranteed to laminarise, the suction system moves 4.25 pounds of air per second and absorbs about 131 horsepower, while the drag it removes is 578 pounds of force, worth 763 horsepower at flight speed. Written as a ratio of powers,

\[\frac{\Delta D \, V}{P} = \frac{578 \times 726.0}{131 \times 550} = 5.8\]

the same 5.8 as the margin in suction coefficient, and necessarily so, since both sides of the comparison are linear in $C_Q$.

The margin is large enough that the interesting question is not whether it is positive but how it varies. It improves with Mach number, because the suction penalty falls as $1/M^{2}$ while the friction saving is nearly constant.

Mach	$C_{D,\text{suction}}$	$\Delta C_{f}$	penalty as a fraction
0.40	0.001351	0.002578	0.524
0.50	0.000880	0.002518	0.350
0.60	0.000624	0.002469	0.253
0.70	0.000469	0.002428	0.193
0.75	0.000414	0.002409	0.172
0.80	0.000369	0.002392	0.154
0.85	0.000332	0.002376	0.140

Laminar flow control is a high-speed technology, and the reason is thermodynamic rather than aerodynamic. The work needed to recompress a unit of captured air is set by temperature, and the kinetic energy available to pay for it goes as the square of speed. For that reason the application was always aimed at fast subsonic transports and why a slow aircraft is a poor candidate however clean its wing.

The Suction the X-21 Drew

The suction coefficient of $2 \times 10^{-4}$ used above is an assumption, and the record of the programme allows it to be tested. The local suction coefficients the X-21 design called for were higher. Joslin 1998, summarising the Northrop final report, gives typical values of $v_{w}/V_{\infty}$ from $5 \times 10^{-4}$ in regions of nearly flat pressure to $10 \times 10^{-4}$ near the leading edge, and adds that at Mach 0.74 and 41,400 feet the measured suction distribution matched the predicted one in shape but exceeded it in level by 50 percent. Since the break-even coefficient above is fixed by the friction saving and the machinery, the margin at any other coefficient is

\[\frac{C_{Q,\text{crit}}}{C_{Q}} = \frac{1.163 \times 10^{-3}}{5 \times 10^{-4}} = 2.3 \qquad \frac{1.163 \times 10^{-3}}{10 \times 10^{-4}} = 1.16\]

Because the penalty is linear in $C_{Q}$, its share of the friction saving rises from 0.172 to $0.172 \times 2.5 = 0.43$ and $0.172 \times 5 = 0.86$. So where the suction was strongest the system was close to paying nothing back, and over the flat-pressure regions that make up most of the chord it returned a little more than twice what it cost.

The same summary reports the system’s flow at one condition, 43,000 feet and Mach 0.75, as ranging from 1.94 to 7.18 pounds per second. At 43,000 feet, 3,000 feet into the isothermal layer whose scale height is $H = R T/g = 20{,}805$ feet, the density is

\[\rho_{43} = \rho_{40} \, e^{-3{,}000/H} = 5.851 \times 10^{-4} \times e^{-0.1442} = 5.065 \times 10^{-4}\ \text{slugs/ft}^{3}\]

and if those flows are totals over the whole wetted wing they imply an area-averaged coefficient of

\[\bar{C}_{Q} = \frac{\dot{m}}{\rho_{43} V_{\infty} S_{\text{wet}}} = \frac{7.18/32.174}{5.065 \times 10^{-4} \times 726.0 \times 2{,}222} = 2.7 \times 10^{-4}\]

at the upper figure and $0.74 \times 10^{-4}$ at the lower. The summary does not say whether those flows are system totals or the flows of individual pumps or regions, and the two readings differ by a factor of several. Read as totals, they bracket the assumed coefficient and leave a margin of about four at the upper figure. Read against the local design values, the margin is between one and two. The record therefore supports the sign of the conclusion, that suction cost less than it saved, and does not support the factor of six as a measured property of this aeroplane.

Why the Testbed Could Not Show What Was Being Claimed

The trade above is per unit of laminarised area. The aircraft-level benefit depends on how much of the aeroplane is laminarised, and here the X-21 was structurally limited by being a conversion.

The wing is about 44 percent of the aircraft wetted area. The programme guaranteed laminar flow over 70 percent of the wing, in Butz 1963, which is 31 percent of the aeroplane.

The zero-lift drag has to be built from its parts, not assumed, because the friction saving must be credited on the same basis the drag is charged. Taking a wetted-area ratio of 4.0 and a form factor of 1.20, which accounts for the thickness of a real wing over a flat plate,

\[C_{D_{0}} = k \, C_{f,\text{turb}} \frac{S_{\text{wet}}}{S} + C_{D,\text{other}} = 1.20 \times 0.002713 \times 4.0 + 0.0050 = 0.01802\]

where the residual 0.0050 is interference, excrescence and everything that is not skin friction. Drag build-up of this kind, and the span-efficiency factor the induced term needs, are period subjects with a settled literature, in Report No. 349. A proof of the theorem regarding the distribution of lift over the span, for minimum induced drag 1930, Oswald 1932, Lockwood Taylor 1942, Sibert 1943, von Kármán and Tsien 1945, Lakshminarayana 1962, Frost and Rutherford 1963, Sanders 1965, Gilman and Burdges 1967, Lundry 1967, Snyder and Zumwalt 1969, Hancock 1971, Morita 1971, Hancock 1972, Letcher 1972, Moriya 1972, Mamada and Ando 1973, Mamada and Ando 1974, Sears 1974, Deyoung 1980, Smith and Kroo 1990, Straussfogel and Maughmer 1991, Smith 1996. The efficiency factor and the performance charts that go with it are Oswald 1932, whose span-efficiency factor is the 0.80 this article uses for the induced drag and is still called by his name. With an Oswald efficiency of 0.80 the induced drag at the cruise lift coefficient is

\[C_{D_{i}} = \frac{C_{L}^{2}}{\pi A\!R\, e} = \frac{0.431^{2}}{\pi \times 6.99 \times 0.80} = 0.01055\] \[C_{D} = C_{D_{0}} + C_{D_{i}} = 0.02857 \qquad \frac{L}{D} = \frac{C_{L}}{C_{D}} = 15.07\]

That decomposition also settles the claim the article opened with. The friction share of total drag is

\[\frac{k \, C_{f,\text{turb}} S_{\text{wet}} / S}{C_{D}} = \frac{0.01302}{0.02857} = 0.456\]

The assertion that about half a transport’s fuel goes to overcoming skin friction came from the literature. This number came from geometry and a flat-plate correlation. They agree to about nine percent, from two directions that share nothing, and that agreement carries more weight than either statement alone.

Now the saving. Over the laminarised wetted area the friction falls by $k \Delta C_{f}$ and the suction charges $C_{D,\text{suction}}$, which is a power and carries no form factor, so

\[\Delta D = \left( k \, \Delta C_{f} - C_{D,\text{suction}} \right) q \, S_{\text{lam}}\] \[\frac{\Delta D}{D} = 0.108 \qquad \frac{L}{D} = 16.89\]

By the Breguet relation,

\[R = \frac{V}{c_{t}} \frac{L}{D} \ln \frac{W_{i}}{W_{f}}\]

range is proportional to lift-to-drag ratio at fixed speed, specific fuel consumption and fuel fraction, so the aeroplane as flown was worth about twelve percent in range. The relation itself is textbook material, not journal material, treated in Anderson 1999 Aircraft Performance and Design, and the cruise-optimisation literature around it is narrow, being Hale 1976, Guynn 2011. That literature is thin under this name, because the word range in an aeronautical archive mostly means the distance to a target or a ballistic test facility rather than the distance an aircraft can cruise.

laminarised	fraction of wetted area	drag reduction	$L/D$	range gain
wing, 70 percent	0.311	0.108	16.89	0.121
wing, 95 percent	0.422	0.146	17.66	0.172
wing entire	0.444	0.154	17.82	0.182
wing and tail	0.618	0.215	19.18	0.273
wing, tail, nacelles	0.758	0.263	20.44	0.356
everything	1.000	0.347	23.07	0.531

What the 1963 Claim Requires

The case for laminar flow control was a doubling of range and a lift-to-drag ratio above thirty, in Butz 1963.

Take the last row of that table seriously for a moment and laminarise the entire aeroplane, fuselage included, which is impossible but bounds the argument. The zero-lift drag becomes

\[C_{D_{0}}^{\ \text{lam}} = 1.20 \times 0.000304 \times 4.0 + 0.0050 = 0.00646\]

and adding induced drag and the suction charge over the whole wetted area,

\[C_{D}^{\ \text{best}} = 0.00646 + 0.01055 + 4.0 \times 4.144 \times 10^{-4} = 0.01866 \quad \Rightarrow \quad \frac{L}{D} = 23.1\]

Twenty-three, not thirty, with every square foot of the aeroplane laminar. The reason is visible in the arithmetic. A lift-to-drag ratio of thirty at this lift coefficient allows a total drag coefficient of

\[C_{D} = \frac{C_{L}}{30} = \frac{0.431}{30} = 0.01435\]

and the induced drag alone is 0.01055, which is 73 percent of the entire budget. Induced drag is set by the planform and is untouched by anything done to the boundary layer. Inverting for the aspect ratio that would leave room,

\[A\!R = \frac{C_{L}^{2}}{\pi e \left( C_{D} - C_{D_{0}}^{\ \text{lam}} - C_{D,\text{suction}} S_{\text{wet}}/S \right)} = 11.8\]

against the X-21’s 6.99.

So laminar flow control does not merely clean a wing. It moves the optimum planform. Once friction is removed, induced drag dominates, and the aircraft that collects the full benefit wants a much longer, thinner wing than the one the X-21 inherited from a bomber. This gives a second and independent reason the testbed could not demonstrate the case, stronger than the wetted-area argument because no amount of suction on this planform can reach the number. Nothing in that table reaches it, and the last row already assumes the impossible. The gap does not mean the case was in error. It measures the difference between laminarising an existing aeroplane and designing one around laminar flow, where the wing shrinks, the fuel load shrinks, the structure lightens and the sizing spiral runs inward. The transport design studies that worked this through are in Marino et al 1975, Povinelli et al 1976, Bonner et al 1977, Gratzer and George-Falvy 1978, Klineberg 1978, Sturgeon 1978, Aircraft energy efficiency program 1979, Klineberg 1979, Swift and Mungur 1979, Wright 1979, Evaluation of laminar flow 1980, Wagner et al 1984, Middleton et al 1985, Wagner 1986, Lange 1988, Parikh and Nagel 1990, Arcara et al 1991, Fischer and Vemuru 1991.

The programme knew this in 1963. Butz 1963 lists among the features of a future laminar flow control aircraft a wing of at least fifty percent more area than a conventional aircraft of the same gross weight, because once friction drag is removed from the sizing the span can be increased to cut the drag due to lift without an undue structural penalty. The programme thus stated the planform argument above while the X-21 was flying, which confirms that the thirty was always a figure for a different wing.

So the X-21 was an instrument that could not measure the quantity being argued about. It could show that the flow stayed laminar, and it did. It could not show the aircraft-level benefit, because the aircraft-level benefit belonged to a different aeroplane.

What Went Wrong at the Attachment Line

On a swept wing the flow divides along a line near the leading edge, and the boundary layer along that line is fed spanwise from the root outward. If it becomes turbulent anywhere, the turbulence can propagate along the line and contaminate everything outboard of it. No amount of suction downstream helps, because the flow arrives already turbulent. This is spanwise contamination, treated in Cumpsty and Head 1967, Cumpsty and Head 1969, Cebeci 1974, Poll 1979, Cumpsty and Head 1980, Hall et al 1984, Hall and Mackerrell 1988, Holmes 1991, Naval K Agatwal et al 1992, Joslin 1996, Joslin 1996, Smith 1996, Poll 2007, Reed and Saric 2013, Yatskih et al 2016, Manuilovich 2017, Manuilovich 2017, Liu 2019, Fedorov and Egorov 2021, Xi et al 2021, Xi et al 2025.

The governing parameter is an attachment-line Reynolds number formed from the spanwise velocity and the boundary-layer scale at the line. It is better assembled than quoted, because each part is a separate physical statement. The flow along the line moves at the spanwise component,

\[W_{e} = V_{\infty} \sin\Lambda = 726.0 \times \sin 33.2^{\circ} = 397.5\ \text{ft/s}\]

The boundary layer there does not grow, because it is continuously stretched by the chordwise acceleration away from the line, which for a leading edge of radius $r$ is

\[\left. \frac{d U_{e}}{ds} \right|_{\text{al}} = \frac{2 V_{\infty} \cos\Lambda}{r} = 7{,}385\ \text{s}^{-1}\]

That balance fixes a constant thickness scale,

\[\ell = \sqrt{\frac{\nu}{d U_{e}/ds}} = 2.62 \times 10^{-4}\ \text{ft} = 0.0031\ \text{in}\]

Three thousandths of an inch, the width of a slot. The Reynolds number on that scale is

\[\bar{R} = \frac{W_{e} \ell}{\nu} = \sin\Lambda \sqrt{\frac{V_{\infty} r}{2 \nu \cos\Lambda}}\]

The sweep in that expression is the leading-edge sweep, not the quarter-chord sweep quoted for a planform. With a taper ratio of 0.30 the X-21 wing carries

\[\tan\Lambda_{LE} = \tan\Lambda_{c/4} + \frac{1-\lambda}{A\!R(1+\lambda)} \quad \Rightarrow \quad \Lambda_{LE} = 33.2^{\circ}\]

Below about 100 an introduced disturbance decays. Above about 245 it is sustained and propagates. Between the two, whether the attachment line stays laminar depends on the size of the disturbance introduced, in the experiments reviewed by Joslin 1998. Evaluating along the span, with leading-edge radius taken as 0.008 of the local chord,

span station	chord, ft	$\bar{R}$
root	20.57	205
0.3	16.25	183
0.5	13.37	166
0.7	10.49	147
tip	6.17	113

The X-21 sat in the ambiguous band along its entire span, and highest at the root. That is the worst place to be. In the safe regime disturbances die whatever the wing does, and in the hopeless regime nothing can be laminar. In the band between, the outcome depends on how large a disturbance the wing root, the fuselage junction or an accumulation of insects happens to present. The programme reported that behaviour, in Braslow 1999, and it is the behaviour hardest to design against.

The result is sensitive to the leading-edge radius, which is assumed here and not measured. At a radius of 0.012 chord the root value rises to 252 and crosses the sustaining threshold outright. The X-21’s wing root therefore sits on the boundary of attachment-line contamination, and which side of it depends on a geometric detail of about one percent of chord. The most visible external feature of the aeroplane, a large fairing over the forward wing-root region, exists to tailor the pressure field there, in Butz 1963.

The X-21’s design rule expresses the same thing in different variables. The analysis behind the wing required the momentum-thickness Reynolds number on the attachment line to stay below 100, according to Joslin 1998. The two parameters are related through the spanwise velocity profile of the swept attachment-line flow, the swept Hiemenz solution, whose spanwise momentum thickness in units of $\ell$ is

\[\frac{\theta}{\ell} = \int_{0}^{\infty} g(\eta) \left[ 1 - g(\eta) \right] d\eta = 0.404 \qquad \Rightarrow \qquad R_{\theta} = 0.404 \, \bar{R}\]

where $g = w/W_{e}$ is the spanwise velocity ratio, obtained by integrating $f’’’ + f f’’ + 1 - f’^{2} = 0$ and $g’’ + f g’ = 0$ numerically, and $\eta = y/\ell$. The design limit is therefore

\[\bar{R} < \frac{100}{0.404} = 247\]

matching the upper threshold of 245 quoted above to within the precision of either. The experimental basis of that threshold, from Cumpsty and Head in 1969 to Poll 1979, is reviewed by Joslin. At the root the table gives $R_{\theta} = 0.404 \times 205 = 83$. The wing met its design criterion at every station along the span, so the contamination it suffered was not a failure of the criterion. It came from large disturbances, the turbulent fuselage boundary layer above all, and for large disturbances the relevant threshold is the lower one near 100, which every station exceeded.

The cure was accordingly to remove the disturbance rather than to change the wing. The review group of October 1965, in its report reproduced in Braslow 1999, concluded that apart from some uncertainty about contamination by turbulence from the fuselage the flights had verified the method of aerodynamic design, that the cause of the leading-edge trouble had been believed found by the middle of 1964, and that two means of controlling contamination from the fuselage had been developed and proven in flight, a special fence and a suction leading-edge portion containing vertical slots. Joslin 1998 lists a fence, vertical slots and a gutter at the wing-fuselage junction.

The related instability, in which the spanwise pressure gradient on a swept wing drives a crossflow profile with an inflection point, is the reason sweep is hostile to laminar flow in the first place, and its treatment is Dagenhart et al 1989, Mousseux et al 1989, Reed 1989.

Surface Tolerance

A slotted suction wing must be smoother than a conventional one, because roughness and waviness trigger transition that suction downstream cannot undo. The tolerance literature is Van Driest and Blumer 1961, Lyons et al 1962, Nagamatsu et al 1964, Merkle et al 1974, Boudreau 1978, Leventhal and Reshotko 1981, Kendall 1982, Bertolotti 1993, Masad 1996, Saric 1997, Stewart and Simpson 2005, Saric 2008, Zhong 2009, Saric 2010, Borodulin et al 2015, Crouch et al 2015, Malik et al 2015, Ustinov and Kachanov 2021, Borodulin et al 2022, Xu and Wu 2022, Zoppini et al 2022, Nakagawa et al 2023, Zheng et al 2026.

The tolerance is computable and it is brutal. A three-dimensional roughness element trips a laminar boundary layer when the Reynolds number formed on its height and the velocity at that height passes a critical value of about 600, in Hall 1964. Near the wall the Blasius profile is linear, $u/U_{\infty} = 0.332\,\eta$, so

\[Re_{k} = \frac{u_{k} k}{\nu} = \frac{0.332\, k^{2} V_{\infty}^{3/2}}{\nu^{3/2} x^{1/2}} = 600\]

Solving for the admissible height at a station one foot behind the leading edge,

\[k_{\text{crit}} = \left( \frac{600 \, \nu^{3/2} x^{1/2}}{0.332 \, V_{\infty}^{3/2}} \right)^{1/2} = 0.0123\ \text{in}\]

About a hundredth of an inch, and it scales as $x^{1/4}$, so it is smaller further forward.

That criterion is not a modern reconstruction. It was established in these same years. Its dense literature is gathered in the cluster on roughness, waviness and surface tolerance below, and the criteria of that period include Tani and Hama 1953, A low-speed experimental investigation of the effect of a sandpaper type of roughness on boundary-layer transition 1959, Dommett 1959, Gibbings 1959, Smith and Clutter 1959, Van Driest and McCAULEY 1960, Potter and Whitfield 1961, Lyons et al 1962, Review of the effect of distributed surface roughness on boundary-layer transition 1962, Holloway and Sterrett 1963. The simplified method for the critical height of distributed roughness particles is Braslow and Knox 1958, so the man who wrote the history this article relies on for its narrative also wrote the criterion it uses for its arithmetic. For comparison the leading-edge slots are 0.0035 inches wide, in Joslin 1998, so the surface had to be held flat to within about three slot widths.

Much of the programme’s first two years went here. The wings were built to close tolerances and the tolerances were still not close enough. Discontinuities at the spanwise wing splices were large enough to trip the boundary layer, putty used to fair them chipped away in flight, and the resulting roughness did the tripping instead. The problem was solved, but it consumed the schedule, in Braslow 1999. Joslin 1998, drawing on Fowell and Antonatos, gives the mechanism. The wings were built from many panels, the splices between them were filled with epoxy to meet the step and waviness tolerances at the high unit Reynolds number, and the epoxy cracked and chipped under the wing loading and the temperature changes of flight. The cause was traced to the bonding process, which was changed until the fills held.

The tolerances the programme arrived at can be set against the criterion above. For Mach 0.8 at 45,000 feet the Northrop final report, as Joslin 1998 gives it, allows forward-facing steps of 0.02 inches, rearward-facing steps of 0.009 inches and spanwise gaps 0.25 inches wide, and requires single waves to have an amplitude less than 0.004 of their wavelength. Five thousand feet into the isothermal layer the density is $\rho_{40} e^{-5{,}000/H} = 4.601 \times 10^{-4}$ slugs per cubic foot, the viscosity is unchanged, and the kinematic viscosity rises in proportion to $6.454 \times 10^{-4}$ square feet per second. Since the admissible height above goes as $\nu^{3/4} V^{-3/4}$,

\[k_{\text{crit},45} = 0.0123 \left( \frac{6.454}{5.075} \right)^{3/4} \left( \frac{726.0}{774.5} \right)^{3/4} = 0.0140\ \text{in}\]

one foot behind the leading edge at $V = 0.8 \times 968.1 = 774.5$ feet per second. The allowed steps bracket the three-dimensional criterion, rearward steps tighter than it and forward steps looser. A step is a two-dimensional disturbance with its own criterion, so the agreement is one of scale rather than an identity, but it shows the arithmetic and the hardware were working on the same thousandths of an inch. The October 1965 review found that a previously suggested criterion for single surface waves had been confirmed in flight, and that vibration of isolated panels at a discrete frequency did not disturb the laminar flow, in Braslow 1999. Joslin 1998 adds that the wing tolerated a gap 0.125 inches wide and 0.18 inches deep at 60 percent chord on the lower outboard wing.

Atmospheric Particles

Flying in or near cirrus, the X-21 lost laminar flow. On emerging from the cloud it regained it immediately, in Braslow 1999. The Northrop study of the problem, Hall 1964, opens with the observation that numerous flights had shown total loss of laminar flow control within or near visible cloud, and erratic performance in light haze when the humidity was relatively high. The mechanism is that ice crystals passing through the boundary layer generate turbulent wedges, and the effect is not a contamination of the surface but a continuous seeding of the flow.

Both halves of that observation follow from one calculation. A point disturbance in a laminar boundary layer does not simply trip the flow at that point. It produces a turbulent spot that grows as it convects, spreading laterally at a roughly constant angle, and the mechanism and its measurement are Schubauer and Klebanoff 1955, Sokolov et al 1980, Glezer et al 1986, Henningson et al 1987, Henningson et al 1988, Seifert et al 1992, Singer and Joslin 1995, Simon and Boyle 1998, LaGraff 2001, Blackwelder et al 2007, Clark et al 2007, Fehrs and Kaiser 2025. The spreading half-angle is about ten degrees, in Schubauer and Klebanoff 1955, so a particle entering at the leading edge contaminates

\[A_{w} = \bar{c}^{\,2} \tan\alpha = 13.37^{2} \times \tan 10^{\circ} = 31.5\ \text{ft}^{2}\]

which means that

\[n = \frac{S}{A_{w}} = \frac{1{,}250}{31.5} \approx 40\]

wedges are enough to blanket the wing. The laminar boundary layer is thin,

\[\delta = \frac{5 \bar{c}}{\sqrt{Re}} = 0.0153\ \text{ft} = 0.18\ \text{in}\]

so the rate at which particles enter it across the span is

\[\dot{N} = N V_{\infty} b \, \delta\]

At the cirrus number density of ten thousand crystals per cubic metre assumed here, which is low for cirrus as shown below, that is about 294,000 encounters per second. The wing is blanketed in roughly 140 microseconds. On leaving the cloud the wedges convect away in

\[t = \frac{\bar{c}}{V_{\infty}} = \frac{13.37}{726.0} = 18\ \text{ms}\]

So the loss is instantaneous with five orders of magnitude to spare, and the recovery is instantaneous because nothing has been left behind on the surface. That distinguishes an ice crystal from an insect, and explains why the two problems needed entirely different answers.

The analysis in Hall 1964 supplies what the assumed density lacks. Working at Mach 0.75 and at 25,000 and 40,000 feet, the same nominal condition used throughout this article, he took cirrus crystals to be hexagonal columns with a length 2.5 times their diameter and found a critical size at 25,000 feet of 17 by 42 micrometres, below which a crystal is too small to start a turbulent spot. Joslin 1998 gives the critical diameters as 17 and 32 micrometres at the two altitudes. Hall’s typical cirrus condition is columns 100 micrometres long and 40 wide at an ice content of 0.01 grams per cubic metre. Treating each column as a cylinder of ice of density $\rho_{\text{ice}} = 917$ kilograms per cubic metre, the number density is

\[N = \frac{\text{IWC}}{\rho_{\text{ice}} \, \tfrac{\pi}{4} d^{2} L} = \frac{0.01\ \text{g/m}^{3}}{917{,}000\ \text{g/m}^{3} \times \tfrac{\pi}{4} \times (40 \times 10^{-6})^{2} \times 100 \times 10^{-6}\ \text{m}^{3}} = 8.7 \times 10^{4}\ \text{m}^{-3}\]

where IWC is the ice water content, nearly nine times the density assumed above, and every one of those crystals is well above the critical size. The conclusion that the wing is blanketed at once is therefore conservative.

The same model shows why haze mattered. Hall defined the threshold of erratic operation as a time-averaged turbulent area of ten percent of the wing. If wedges arrive at random at rate $\dot{N}$, each covering $A_{w}$ for the convection time $t$, the time-averaged turbulent fraction is

\[\phi = 1 - \exp \left( - \frac{\dot{N} A_{w} t}{S} \right)\]

and setting $\phi = 0.10$ gives

\[\dot{N}_{10} = \frac{-\ln 0.9 \times 1{,}250}{31.5 \times 0.0184} = 227\ \text{s}^{-1} \qquad N_{10} = \frac{\dot{N}_{10}}{V_{\infty} b \, \delta} = \frac{227}{29.4\ \text{m}^{3}/\text{s}} = 7.7\ \text{m}^{-3}\]

with the volume swept by the boundary layer, $V_{\infty} b \delta = 726.0 \times 93.5 \times 0.0153 = 1{,}039$ cubic feet per second, converted to cubic metres. About eight crystals of critical size in each cubic metre are enough to make laminar flow erratic on this crude model, four orders of magnitude below a typical cirrus, so air that looks clear can still carry enough of them. Hall reports that pattern, with total loss in cloud where the visibility was several thousand feet and erratic, partially degraded performance in haze through which flight test engineers and pilots could see 30 to 60 miles horizontally.

The recovery observed in flight was slower than the convection time. The review group recorded that laminarity lost in cloud was reestablished a few seconds after return to clear air, in Braslow 1999, against the 18 milliseconds computed above. Since the wing itself clears in milliseconds, the seconds are most plausibly the time taken to leave the diffuse edge of a cloud and the haze around it, which is an inference rather than a finding of the programme. The F-94A glove a decade earlier regained laminar flow within 30 seconds of leaving cloud, in Joslin 1998.

The claim that the data did not exist is checkable, and it holds up. Hall wrote in 1964 that quantitative data from sampling of ice particles in natural clouds was practically nil, and took his typical crystal sizes from three references that disagreed in magnitude. Measuring ice-crystal concentration in cloud is a subject of its own, in Mossop and Ono 1969, Mossop and Hallett 1974, Davis and Fischer 1983, Baumgardner et al 2014, Huang 2021, and that literature amounts to a handful of campaigns, not the climatology an operator would need. The quantity the X-21 required was not a concentration in one cloud on one day. It was a distribution over altitude, season and route, and measuring it would have taken a different and much larger programme than anyone had run.

This literature is very small. Work in the archives and journals connecting atmospheric ice particles to laminar flow is essentially Davis 1982, Davis and Fischer 1983, Jasperson et al 1984, Jasperson et al 1984, Davis et al 1986, Davis et al 1987, Davis et al 1989, Pohya et al 2019, for the reason the programme itself gave. There was no statistical description of particle size and concentration as a function of altitude, season and geography, so the operational significance could not be computed, in Braslow 1999. A theory was produced, in Hall 1964, and the data to apply it did not exist. Hall recommended that further work be predominantly experimental, by controlled laboratory tests and by flight tests carrying instruments to sample the ice particles. Joslin 1998 records that the cloud encounters of NASA’s JetStar leading-edge flight tests in the 1980s, which did carry a particle spectrometer, agreed with Hall’s criteria.

Dependent Systems

The suction pumps are not free of the airframe. The four turbines sit in nacelles under the wing, but their ducting occupies volume inside a wing that must also carry fuel and structure, and the ducts are large because a small duct has its own boundary-layer losses, a pumping penalty on the pumping system, in Butz 1963.

The duct size is a check on the suction coefficient from outside the model. The mass flow over the guaranteed area is 0.132 slugs per second, the contemporary description reports duct speeds reaching 200 miles per hour, and the duct runs at roughly the recovered density, so

\[A_{\text{duct}} = \frac{\dot{m}}{\rho_{d} V_{d}} = \frac{0.132}{3.51 \times 10^{-4} \times 293} = 1.28\ \text{ft}^{2}\]

which is 185 square inches in total, or about 46 square inches in each of four ducts, near enough seven inches square. That fits inside a wing of 13.4 foot chord without difficulty, and the figure comes from a reported duct velocity and a computed mass flow that share no input. The pumping installation and its ducting are treated in Maddalon and Braslow 1990, High Reynolds Number Hybrid 1999, High Reynolds Number Hybrid Lamin, Prasannakumar et al 2024, a small literature for a component that decides whether the scheme is worth installing. Had the assumed suction coefficient been an order of magnitude wrong, the duct would not have fitted in the wing. The turbojets themselves were moved to the rear fuselage to make room under the wing for the suction pumps, in Butz 1963, and that relocation gives the aeroplane its appearance.

The suction was metered finely and the noise was kept out deliberately. Joslin 1998 records 96 suction control valves employed to control the suction in the slots independently, and a system that operated between 85 and 130 percent of the nominal design flow so that the unproven method of estimating the flow could be tested. Engine noise can trip a laminar boundary layer, so a movable plug in each engine inlet set up a standing shock wave that kept the compressor noise from reaching the wing, in Butz 1963. The 1965 reviewers found that laminar flow persisted at external sound levels about 6 decibels above the established criterion, that noise inside the suction ducts did no harm unless a duct was driven at a resonance large enough to disturb the flow outside the slots, that atmospheric turbulence strong enough to change the angle of attack by one degree either way and to produce 0.3 g at the centre of gravity did not disturb the laminar flow, and that small changes of flight condition, of about 0.04 in lift coefficient or 0.03 in Mach number, did not affect it appreciably, in Braslow 1999. Joslin adds that slot Reynolds numbers of 120 to 140 produced disturbances at the slot that dominated any noise from the ducts, so the slots were held below about 100.

The friction the exercise is aimed at is the subject of the cluster on skin friction and the turbulent boundary layer below, with foundations in Vessey 1935, Young 1949, Rubesin and Johnson 1949, Cope 1952, Lee and Faget 1956, Spalding and Chi 1964, White and Christoph 1972. The alternative is to shape a wing so that the pressure gradient holds the flow laminar without any suction at all, which needs no plumbing but constrains the aerodynamics severely. Its early flight experiments are Carmichael 1979, Montoya et al 1981, Holmes et al 1984, Wentz et al 1984, Holmes et al 1985, Runyan et al 1987, Horstmann et al 1990, Holmes and Obara 1992.

The Flight Test Record

First flight was on 18 April 1963 from Edwards Air Force Base, in the Northrop X-21A reference entry, and Braslow 1999 places the start of flight research in April 1963. Two aircraft were converted, serials 55-0408 and 55-0410. By August 1963 the flights had covered altitudes from 8,000 to 45,000 feet and speeds up to Mach 0.8, in Butz 1963. Joslin 1998 gives the envelope as Mach 0.3 to 0.8 at altitudes from 5,000 to 44,000 feet, and the programme as more than 200 flights with the laminar flow control system operating.

The progress was real and slow. When the suction system was first switched on, about the outer twelve percent of the wing went laminar. By the middle of 1963 that had reached roughly the outer half, through adjustments to pumping pressures and to the flow distribution among the leading-edge slots, and the project managers were already persuaded that closer slot spacing would be needed on the inboard leading edges, in Butz 1963. By the end of the programme, in Braslow 1999, laminar flow was being obtained over more than ninety-five percent of the area intended for laminarisation.

The Laminar Run, Year by Year

The chordwise results are better compared as the Reynolds number of the laminar run, the product of the laminar fraction of the chord and the chord Reynolds number,

\[Re_{x} = \frac{x_{\text{lam}}}{c} \, Re_{c}\]

Joslin 1998 gives the progress on the mean aerodynamic chord as follows.

period	laminar chord	$Re_{c}$	$Re_{x}$
end of 1963	nearly 0.60	$20 \times 10^{6}$	$12 \times 10^{6}$
1964	0.70	$20 \times 10^{6}$	$14 \times 10^{6}$
1964	0.30 to 0.55	$30 \times 10^{6}$	up to $16.5 \times 10^{6}$
1965	0.96	$20 \times 10^{6}$	$19.2 \times 10^{6}$
1965	0.81	$30 \times 10^{6}$	$24.3 \times 10^{6}$
1965	0.59	$40 \times 10^{6}$	$23.6 \times 10^{6}$

The run Reynolds number roughly doubled in two years and then stopped growing with chord Reynolds number. In 1965 the run at $30 \times 10^{6}$ and at $40 \times 10^{6}$ is the same to within four percent, the pattern a fixed limit on the run, rather than on the fraction of chord, would produce. The group’s report gave laminar runs up to a length Reynolds number of 46 million on some parts of the wing, against consistently obtained maxima of less than half that, below 23 million, the year before, and judged that design procedures suitable for 60 million were known but would need flight verification. It also recorded that some direct measurements of overall drag by speed-power methods agreed well with prediction, in Braslow 1999.

The chord Reynolds numbers are consistent with the geometry used in this article. With the assumed taper ratio of 0.30 the root chord is $c_{r} = 2S/[b(1+\lambda)] = 20.57$ feet and the mean aerodynamic chord is

\[\bar{c}_{\text{MAC}} = \tfrac{2}{3} c_{r} \frac{1 + \lambda + \lambda^{2}}{1 + \lambda} = \tfrac{2}{3} \times 20.57 \times \frac{1.39}{1.30} = 14.66\ \text{ft}\]

so at Mach 0.70 and 40,000 feet the chord Reynolds number is $1.913 \times 10^{7} \times (0.70/0.75) \times (14.66/13.37) = 1.96 \times 10^{7}$, the $20 \times 10^{6}$ of Joslin’s example flight at that condition, on which 74 percent of the upper surface and 61 percent of the lower were laminar.

Set against the programme guarantee of seventy percent of the wing, that is a success. Set against the purpose of the aeroplane, it is one objective out of three.

The Review Group of October 1965

The review group met on 14 and 15 October 1965 at the Norair Division in Hawthorne, California, and dated its report 8 November 1965. It was chaired by Holt Ashley and signed by members from NASA, the Federal Aviation Agency, Lockheed, Douglas, Boeing, North American Aviation and Colorado State University, and its report is reproduced in Braslow 1999. Braslow’s narrative says that the third objective, service experience comparable to an operational aircraft, had not even been initiated, because the effort had gone into surface tolerance and spanwise contamination. The report’s own wording is milder. It says that the programme had not been able completely to address the operational questions because of difficulties with the basic aerodynamic design, that those difficulties had been overcome, and that the programme should now be directed toward answering them.

Those questions were specific. Can the structure be built to the roughness and waviness criteria, and at what weight and complexity? How far does the surface deteriorate in service? What does it take to keep the wing clean of insects, mud and rain on a routine basis? How tolerant is laminar flow of rain, sleet, ice crystals, cloud and gusts? And what do manoeuvres, off-design lift and non-optimum suction do to it? The group judged an operational suitability evaluation the minimum effort that could justify further funds, recommended that maintainability data be collected to Air Force standard procedures so that availability could be compared with non-laminar aircraft, and asked that the modified aircraft go to the Air Force Flight Test Center well ahead of Northrop’s schedule. It asked for serious consideration of a wholly new set of metal upper wing surfaces needing no filler, judged closing out the flight programme at once particularly undesirable because a large military transport with laminar flow control, perhaps drawn from the C-5A line, was under consideration, and called for systems analyses comparing laminar flow control with better engines, lighter structures, hydrogen fuel and refuelling, in Braslow 1999.

The operational phase never took place. The modification was never made, and the reason recorded by the programme’s NASA consultant, in Braslow 1999, is the resource demands of the war in Vietnam, not any finding about laminar flow.

The Maintenance the Programme Did Record

The programme was not silent on maintenance, and what it recorded points away from insects. In August 1963 Butz 1963 reported that the early flight results had substantiated an estimate that the airframe of a laminar flow control aircraft would cost about ten percent more to maintain than a conventional one, that the wing would probably not need cleaning after every flight, and that steam cleaning once a month would keep it adequately free of dirt and grease. He also reported that insects sticking to the wing had not so far bothered the suction system. The review group confirmed that insect strikes had not disturbed the laminar flow, and warned in the same sentence that the flights had been made over a desert where insects are sparse and small, so the result said nothing about operation worldwide, in Braslow 1999. Joslin 1998, summarising Fowell and Antonatos, shows that the ground maintenance time went mostly to repairing the epoxy-filled splices between wing panels. The maintenance burden the X-21 did record was a manufacturing defect, not an operational one, and the review group’s new metal upper surfaces were meant to remove it.

Comparison With Ground Prediction

The wind tunnel and glove work of the 1950s predicted that suction would hold laminar flow to high chord Reynolds numbers, and the flight article confirmed it. On that axis the ground prediction was good. Joslin 1998 records about 160 hours of high-speed and 1,300 hours of low-speed wind tunnel testing of a model X-21A wing, and good agreement between tunnel and flight once aeroelastic deformation and the flight pressure distributions were allowed for. The suction requirement was the exception, measured 50 percent above prediction at Mach 0.74 and 41,400 feet, so the theory erred on the side that costs power. Companion tunnel tests showed that a sudden loss of laminar flow would not endanger the aircraft, although the lateral-directional and long-period longitudinal motions might need more artificial damping than a turbulent aircraft.

On two axes it was not, and both failures were of the same kind. The severity of spanwise contamination had been seen in British swept-wing experiments and in small-scale tunnel tests, and its significance had gone unrecognised until a full-size swept wing met it. The surface tolerance achievable in manufacture had likewise been explored only on small-scale structural samples rather than on wings, and a wing has splices where a sample does not, in Braslow 1999.

Neither is a failure of theory. Both are failures of scale, in the specific sense that the feature which caused the trouble did not exist on the model. A tunnel model has no spanwise splice and no fuselage junction of the relevant size. The transition-prediction methods that have since been built to carry this kind of analysis are L. Cutrone et al, Transition Prediction in Hyperson, Meelan Choudhari et al, Leslie M Mack 1984, Bushnell et al 1988, Hall and Smith 1988, Hall and Smith 1989, Iyer and Spall 1991, Su 2015, Zhou et al 2015, Kosarev et al 2016, Xia and Chen 2016, Mishriky and Walsh 2018, Zhou et al 2018, Han and Cao 2019, Introduction to the Special Section on Hypersonic Boundary-Layer Transition Prediction 2019, Martín and Paredes 2021, Nie et al 2022, Niu and Su 2023, Nathaniel Hildebrand et al 2025, Yan et al 2025.

What the Data Changed

Laminar flow control stopped being a question of physics and became a question of engineering practice, and the change is due more to this programme than to any other.

Attachment-line contamination went from a problem whose significance had gone unrecognised to a named design constraint with a criterion attached, one that requires careful attention in the design of any large laminar wing. The manufacturing tolerance problem was shown to be solvable, expensively. The atmospheric particle problem was identified and left open for want of data, in Braslow 1999.

What did not change is the thing the programme is remembered for. No service-maintenance evaluation was made, and the maintenance it did log was dominated by repairs to filled splices, so nothing was learned about whether a well-made wing could be kept clean in service, and the widespread belief that the X-21 settled that question is not supported by the record of the programme.

The Contemporary Literature

Cluster	Records
Suction laminar flow control and its flight research	166
Hybrid laminar flow control	70
Natural laminar flow and low-drag aerofoils	200
The attachment line and leading-edge contamination	86
Crossflow instability on swept wings and its control	223
Roughness, waviness and surface tolerance	112
Turbulent spots and wedges	19
Boundary-layer stability and transition	260
Ice crystals, cloud and the icing of aircraft	39
Insect contamination and its mitigation	18
Coatings against ice and erosion	32
Suction systems, perforated surfaces and their manufacture	45
Suction and injection in the boundary layer	102
Skin friction and the turbulent boundary layer	49
Riblets and the reduction of turbulent skin friction	114
Induced drag, span efficiency, cruise and aircraft design	25
Total	1,560

Laminar flow control did not die in 1965. It went quiet for a decade, revived when fuel prices made drag expensive, and has been a continuous field ever since, as the clusters below show, and the revival and its later reviews are recorded in Povinelli et al 1976, Braslow and Muraca 1978, Klineberg 1978, Aircraft energy efficiency program 1979, Wagner et al 1988, Joslin 1998, Saric and Reed 2004, Green 2008, Atkin 2008, Beck et al 2018.

But it is not the same technology. Every major line of work since the X-21 has conceded something the X-21 did not, and each concession buys the same thing, a surface that can be built to tolerance and kept clean in service. The field retreated from the full-chord slotted wing, and it spent the sixty years since on the question the X-21 was cancelled before it could ask.

The First Concession Gives Up Aft Suction

Hybrid laminar flow control applies suction only near the leading edge and lets a favourable pressure gradient carry the flow from there, in Hybrid laminar flow control 1982, Collier et al 1992, Wie et al 1992, High Reynolds Number Hybrid 1999, High Reynolds Number Hybrid Lamin, Young and Fielding 2001, Young et al 2001, Young et al 2003, Eppink and Wlezien 2011, Risse and Stumpf 2014, Vermeersch and Bouteiller 2014, Shahneh 2016, Krishnan et al 2017, Kalarikovilagam Srinivasan and Bertram 2019, Yang et al 2019, Shi et al 2020, Lüdeke and Breitenstein 2021, Schrauf and von Geyr 2021, Sudhi et al 2021, Barklage et al 2022, Jentys et al 2022, Yang et al 2022, Zare Shahnehb 2022, Jentys and Breitsamter 2023, Lobitz et al 2023, Shi et al 2023, Sudhi et al 2023, Karpuk and Mosca 2024, Athar et al 2025, Fröhler et al 2025, Brinza et al 2026, Corelli Grappadelli et al 2026, Wong et al 2026, Zhang et al 2026.

That is a direct response to the X-21’s cost problem. The leading edge is a small fraction of the chord, so the plumbing shrinks greatly, the structural intrusion largely disappears, and the wing behind the front spar becomes an ordinary wing. What is given up is the aft laminar run, and with it a meaningful part of the drag saving. The trade is worse aerodynamically and much better industrially, and every subsequent decision in this field has had that shape.

The Second Concession Gives Up Suction Altogether

Natural laminar flow shapes the aerofoil so that the pressure gradient alone holds the boundary layer laminar, asking nothing of the plumbing because there is none. The cluster on natural laminar flow and low-drag aerofoils below holds its literature, and its flight record includes Montoya et al 1981, Wentz et al 1984, Runyan et al 1987, Horstmann et al 1990, Frederick et al 2014, Tokugawa et al 2026.

The constraint moves from the wing structure to the wing shape, and the shape constraint is severe. A natural laminar flow wing cannot be swept much, because sweep drives the crossflow instability that no pressure gradient can suppress, and the X-21’s thirty degrees is already beyond what natural laminar flow tolerates. That confines it to slower and smaller aircraft.

The Insect Question the X-21 Never Reached

The X-21’s service-maintenance objective was never reached. The literature took it up in the late 1970s and has worked on it since, in Fisher and Peterson 1978, Peterson and Fisher 1978, Maresh and Bragg 1984, Croom and Holmes 1985, Croom and Holmes 1986, Lorenzi et al 2011, Kok et al 2013, Siochi et al 2013, Wohl et al 2013, Kok and Young 2014, Kok et al 2014, Kok et al 2015, Shanahan et al 2015, Wohl et al 2015, Alexander et al 2016, Krishnan et al 2016, Wicke et al 2016, Kok et al 2017.

The modern answer is chemistry and geometry rather than cleaning. Coatings that insect residue cannot adhere to, surfaces engineered at the micrometre scale so that a strike does not leave a critical roughness, and shields deployed over the leading edge during the low-altitude part of the flight where the insects are. Covering the leading edge is not new, since a paper cover protected the F-94 glove from insects in the 1950s, in Joslin 1998, so the last of these is a mechanical answer the X-21 could have carried.

Here the contemporary literature bears most sharply on this programme. The problem widely cited as having killed laminar flow control is the one problem the programme never evaluated, and it has turned out to be tractable, in part by coatings and engineered surfaces developed since.

Ice Crystals Remain Unsettled

The X-21 lost laminar flow in cirrus and recovered it on leaving. The article computes why both happen within milliseconds. What it could not compute in 1965, and what is still not fully answered, is how much of a real route is spent in such cloud, in Pueschel et al 1996, Jensen et al 2008, Oliver 2015, Leroy et al 2016, Leroy et al 2017, Sullivan et al 2017, Norde et al 2018, Bravin and Strapp 2019, Gurevich et al 2019, Haggerty et al 2019, Tan and Lai 2019, Xiao et al 2019, Flegel 2021, Hu et al 2021, Rugg et al 2021, Hu et al 2022, Yang et al 2022, Potts et al 2023, Yang et al 2023, Malik et al 2024, Peters et al 2024, Tatsuta et al 2024, Ma et al 2025, Yang et al 2026.

The modern literature approaches this from a different direction, because high ice water content became an engine problem before it was ever resolved as a laminar flow problem. The measurements now exist because ice crystals were damaging turbofans, an odd route to the data the X-21 needed, and the coverage is still organised around convective cloud at tropical latitudes rather than around cruise-altitude cirrus on transport routes, in Bravin and Strapp 2019 and Leroy et al 2017.

The Crossflow Is Now Controlled Rather Than Suffered

The instability that made a swept laminar wing hard is no longer merely endured. The cluster on crossflow instability below collects its literature, and the control by roughness elements and plasma actuators described next is in Saric 1997, Carpenter et al 2009, Malik et al 2013, Saric et al 2015, Nicholson et al 2019, Duchmann et al 2014, Dörr and Kloker 2015, Wang et al 2018, Yadala et al 2018.

The elegant result is that roughness can be used against itself. Discrete roughness elements placed at a chosen spacing near the leading edge excite a crossflow mode that is not the most amplified one, and that mode then suppresses the dangerous one through nonlinear interaction. A wing is made more laminar by being made rougher in a specific way, which would have been an unwelcome sentence in a programme spending two years chasing thousandths of an inch of waviness. Plasma actuators do the same job with no surface change at all.

Transition Prediction Since the X-21

The X-21 was designed with stability criteria and correlations and flown with trailing-edge total-pressure rakes, in Joslin 1998. The design tool that would have let Northrop compute the growth of disturbances through the full three-dimensional boundary layer did not exist, and the tools that replaced it are Petzold and Radespiel 2015, Pérez et al 2015, Roberts et al 2015, Wedin et al 2015, Wedin and Cherubini 2016, Евстигнеев and Рябков 2016, Egorov et al 2017, Egorov et al 2018, Grigor’ev and Ershov 2019, He et al 2019, Kudryavtsev and Khotyanovsky 2019, Hildebrand et al 2020, Morgan and Davies 2020, Yalcin et al 2021, Toppings and Yarusevych 2023, Chaturvedi et al 2024, Zou and Zhong 2024.

This article’s attachment-line calculation is the first term of what those methods now solve properly. A criterion with a threshold has become a stability computation over the full three-dimensional boundary layer, and the crossflow and attachment-line modes that the X-21 met one at a time and by surprise now appear together in a single analysis before anything is built.

Making the Surface Is Still the Hard Part

Sixty years on, the manufacturing problem the programme spent its first two years on remains the one that decides whether the technology is affordable, in Drake et al 2005, Bender and Drake 2006, Li et al 2024, Dou et al 2026, Kashaev et al 2026.

What has changed is the aperture. The X-21 cut slots and drilled more than eight hundred thousand metering holes with purpose-built tooling, in Butz 1963. The modern surface is laser-drilled microperforated titanium, with holes measured in tens of micrometres and counted in millions, in Joslin 1998. The tolerance requirement has not relaxed at all, and by the roughness criterion in this article it never could have, since it is set by the physics of the boundary layer and not by the state of the art.

Flight Demonstrations Since

What followed was a sequence of smaller and more targeted experiments, in Montoya et al 1981, Fischer et al 1983, Wagner et al 1986, Hastings et al 1987, Fisher and Fischer 1987, Runyan et al 1987, Wagner et al 1988, Wagner et al 1989, Horstmann et al 1990, Wagner et al 1992, Collier 1993, Barry et al 1994, Anders and Fischer 1999, High Reynolds Number Hybrid 1999, Saric et al 2015, Shanahan et al 2015, Tucker et al 2016, Yang et al 2022.

Gloves on transport wings and fins, and a supersonic laminar flow experiment on a delta, in Joslin 1998. Every one of them is narrower than the X-21.

The Installed Cost of Suction

The article charges the suction system a thermodynamic power and states that it does not charge weight, volume or structural penalty. The titles in this survey that name the suction system or its power are Maddalon and Braslow 1990, High Reynolds Number Hybrid 1999, Yasuaki and Egami 1999, Saeed et al 2010, Saeed et al 2011, Perry and Ansell 2017, Baranov et al 2021, Prasannakumar et al 2024, and they are few.

That thinness is informative. The installed cost of a suction system is the quantity a manufacturer needs and also the one hardest to publish, because it is specific to an airframe and commercially sensitive. The public literature is rich on the aerodynamics and poor on the installation, the same imbalance the X-21 programme had.

Riblets as the Competing Route

Riblets and their successors do not prevent turbulence. They reduce the friction of turbulence that is already there. The riblet cluster below gathers that literature, and representative records are Walsh 1982, Walsh and Lindemann 1984, Lazos and Wilkinson 1988, Bixler and Bhushan 2013, Benschop and Breugem 2017, Catalano et al 2020, Ricco et al 2021, Gatti et al 2025.

They give a few percent where laminar flow gives tens of percent, and they ask almost nothing of the aircraft. No plumbing, no power, no structural intrusion, and a surface that is a film rather than a wing. That ratio explains why they keep returning, and it states most clearly the trade inside which this subject sits. The prize scales with how much of the boundary layer you control, and so does the cost of controlling it.

What the Survey Shows

The X-21 was right about the physics and the field has never disputed it. Laminar flow control works, the suction power costs less than the drag it saves, and nothing in sixty years has overturned either statement.

What the field did instead was retreat from the installation. Hybrid laminar flow control gave up the aft suction. Natural laminar flow gave up the suction. Riblets gave up laminar flow. Each step traded aerodynamic benefit for a surface that could be manufactured and maintained, and every one of those trades is a decision about the question the X-21 never got to answer.

And that question is now being answered in the affirmative. Insect residue is being attacked by engineered surfaces and coatings rather than by cleaning, the crossflow is being controlled instead of avoided, and the surface is being made by processes unavailable to the programme. The verdict that came to be passed on this programme was that laminar flow control was impractical to maintain. It was passed without evidence, and the evidence that has accumulated since largely contradicts it.

Suction laminar flow control and its flight research

The keystone. Laminar flow control by suction as a programme, from the Northrop and NACA suction work of the 1940s and 1950s through the X-21 to the leading-edge flight test on NASA’s JetStar, the Langley 8-foot tunnel experiment, the F-16XL supersonic experiment and the transport design studies of the energy efficiency programme. 19 of its 166 titles name flight.

166 records. [Agrawal et al 1989] [Ahmadi-Baloutaki et al 2013] [Aircraft energy efficiency program 1979] [Allison 1983] [Anders and Fischer 1999] [Atkin 2008] [Barklage et al 2023] [Berry 1986] [Berry et al 1987] [Bin Guo et al 2017] [Bobbitt et al 1992] [Bonner et al 1977] [Braslow and Fischer 1985] [Braslow and Muraca 1978] [Brennecke 1948] [Brooks and Harris 1987] [Brooks et al 1991] [Bushnell and Malik 1987] [Bushnell and Tuttle 1979] [Carpenter et al 2018] [Choudhari 1992] [Choudhari et al 2003] [Collier 1993] [Crouch 2008] [Crouch 2015] [Crowley and Atkin 2016] [Bushnell and Malik 1987] [Dodd et al 2001] [Edwards 1961] [Etchberger 1983] [Evaluation of laminar flow 1980] [Fischer 1992] [Fischer and Vemuru 1991] [Fischer et al 1983] [Fisher and Fischer 1987] [Fuel efficiency and laminar] [George et al 1992] [George-Falvy 1978] [Gratzer and George-Falvy 1978] [Green 2008] [Groth 1961] [Haans et al 2004] [Hall 1992] [Harris and Brooks 1988] [Harris et al 1988] [Harvey 1982] [Harvey and Pride 1982] [Harvey et al 1986] [Harvey et al 1989] [Hefner 1992] [Hefner and Bushnell 1979] [Hefner and Wagner 1986] [Hein et al 2009] [Herbert et al 1996] [Influence of Environment in 1980] [Jernell 1978] [Jobe et al 1978] [Jobe et al 1979] [Jones et al 1989] [Joslin 1998] [Joslin 1998] [Joslin and Anders 2008] [Kakkar et al 2026] [Kaups and Cebeci 1977] [Keith and Dewitt 1993] [Klineberg 1978] [Klineberg 1979] [Klotzsche et al 1984] [Kosin 1965] [Kroo and Sturdza 2003] [Lachmann 1959] [Lachmann 1962] [Laminar flow control wing 1999] [Lange 1983] [Lange 1987] [Li Causi et al 2026] [Lineberger 1984] [Lo 1994] [Loftin and Burrows 1949] [Lovell et al 1978] [Maddalon 1982] [Maddalon and Wagner 1986] [Maddalon et al 1987] [Malik 1987] [Marshall 1999] [Marshall 2000] [Masad and Abid 1996] [Masad and Malik 1994] [Masad and Nayfeh 1992] [Mauerer et al 2026] [McQuilkin 1982] [Meade 1977] [Messing and Kloker] [Middleton et al 1985] [Mosca et al 2024] [Muraca 1978] [Nahoui et al 2025] [Northrop Aircraft Inc Hawthorne Ca 1956] [Northrop Corp Hawthorne Ca Norair Div 1964] [Northrop Corp Hawthorne Ca Norair Div 1964] [Ostowari 1992] [Parikh and Nagel 1990] [Pearce 1982] [Pearce 1983] [Pearce et al 1984] [Pfenninger 1949] [Pfenninger 1968] [Pfenninger 1981] [Pfenninger 1987] [Pfenninger and Bacon 1961] [Pfenninger and Groth 1961] [Pfenninger and Syberg 1974] [Pfenninger and Vemuru 1988] [Pfenninger and Vemuru 1988] [Pfenninger and Vemuru 1992] [Pfenninger et al 1980] [Pfenninger et al 1990] [Poll 2000] [Povinelli et al 1976] [Powell 1987] [Powell et al 1989] [Ram et al 1985] [Raspet 1952] [Rogers 1979] [Rogers and King 1984] [Saric and Reed 2004] [Sattarzadeh and Fransson 2016] [Semenov et al 2021] [Shi et al 2015] [Singleton 1967] [Srokowski 1978] [Stability Theory for Laminar 1990] [Sturgeon 1978] [Sturgeon 1978] [Sturgeon and Bennett 1976] [Sturgeon et al 1976] [Swift and Mungur 1979] [Swinford 1976] [Tokhunts 1968] [Tucker et al 2014] [Tucker et al 2014] [Tucker et al 2016] [Tucker et al 2016] [Turriziani et al 1979] [Tuttle and Maddalon 1982] [Tutty et al 2000] [Wagner 1986] [Wagner and Fischer 1983] [Wagner et al 1984] [Wagner et al 1986] [Wagner et al 1988] [Wagner et al 1988] [Wagner et al 1988] [Wagner et al 1989] [Wagner et al 1990] [Wagner et al 1992] [Warner et al 1980] [Watmuff 1992] [Wen 2024] [Wheldon and Whites 1966] [Wie et al 1991] [Woan et al 1991] [Wright 1979] [Wright and Nelson 2000] [Yoshida and Tani 1993] [Zhao et al 2018]

Hybrid laminar flow control

The first concession. Suction confined to the leading edge, with the pressure gradient carrying the flow aft, from the design studies of the 1980s to the high Reynolds number flight experiment, the A320 fin and the current suction panels.

70 records. [Abdul Nour and Mueller 1993] [Arcara et al 1991] [Athar et al 2025] [Atkin 2004] [Barklage et al 2022] [Barry et al 1994] [Bobbitt et al 1992] [Brinza 2024] [Brinza 2025] [Brinza et al 2026] [Collier et al 1992] [Corelli Grappadelli et al 2022] [Corelli Grappadelli et al 2023] [Corelli Grappadelli et al 2026] [Dargel and Thiede 2002] [Egreteau et al 2022] [Fröhler et al 2025] [High Reynolds Number Hybrid 1999] [High Reynolds Number Hybrid 1999] [Hybrid laminar flow control 1982] [Hybrid laminar flow flight 1999] [Eppink and Wlezien 2011] [Jentys and Breitsamter 2023] [Jentys et al 2022] [Kakkar et al 2025] [Kalarikovilagam Srinivasan and Bertram 2019] [Karpuk and Mosca 2024] [Krishnan et al 2017] [Lange 1988] [Lawson et al 2018] [Lobitz et al 2023] [Lüdeke and Breitenstein 2021] [Mery et al 2025] [Mosca and Karpuk 2025] [Mullender and Poll 1995] [Paluch 2001] [Parikh et al 1990] [Pascual and Zingg 2026] [Pohya and Wicke 2019] [Pohya et al 2019] [Pralits and Hanifi] [Prasannakumar et al 2024] [Risse and Stumpf 2014] [Risse et al 2014] [Schrauf and von Geyr 2020] [Schrauf and von Geyr 2021] [Schrauf and von Geyr 2021] [Shahneh 2016] [Shi et al 2020] [Shi et al 2023] [Streett 2003] [Streett 2003] [Sudhi et al 2021] [Sudhi et al 2023] [Lorenzi et al 2011] [Traub et al 2021] [Vermeersch and Bouteiller 2014] [Wahlich et al 2024] [Wie et al 1992] [Wong et al 2026] [Yanagimoto and Rinoie 2019] [Yang et al 2019] [Yang et al 2022] [Young 2012] [Young and Fielding 2001] [Young and Fielding 2001] [Young et al 2001] [Young et al 2003] [Zare Shahnehb 2022] [Zhang et al 2026]

Natural laminar flow and low-drag aerofoils

The second concession. Wings, aerofoils, gloves and nacelles shaped so that the pressure gradient alone holds the flow laminar, with the low-drag aerofoil work of the NACA period that preceded both.

200 records. [Cella et al 2010] [Coder et al 2014] [Kerschen 1987] [McCullough and Gault 1950] [Somers 2012] [Tufts et al 2014] [Ueda et al 2014] [Venkatachari et al 2021] [Barnwell and Hussaini 1992] [Befus et al 1987] [Belisle et al 2012] [Bell and Cornelius 1987] [Bidwell 1991] [Bozeman et al 2024] [Bragg and Gregorek 1989] [Braun et al 2011] [Campbell and Lynde 2017] [Campbell et al 1986] [Carmichael 1979] [Cebeci and Smith 1968] [Chakraborty et al 2022] [Chen and Tang 2017] [Chen et al 2023] [Chen et al 2023] [Clayton] [Coder 2020] [Coder 2021] [Coder 2026] [Coder and Somers 2020] [Collazo Garcia and Ansell 2018] [Collazo Garcia and Ansell 2019] [Colletti et al 2021] [Coutu et al 2011] [Somers and Maughmer 2022] [de Rosa and Donelli 2020] [Deng et al 2026] [Djeddi and Ekici 2024] [Dodbele 1992] [Dodbele et al 1986] [Fan et al 2017] [Fan et al 2021] [Fehrs and Kaiser 2025] [Frederick et al 2014] [Fuhrmann 1993] [Fujino 2002] [Fujino et al 2003] [Garzon and Matisheck 2012] [Garzon et al 2014] [Gibson and Gerhardt 1995] [Gopalarathnam and Selig 2001] [Gratzer 1990] [Guha and Sengupta 2025] [Haines 1957] [Hammer and Shumway 2026] [Han et al 2016] [Han et al 2018] [Hansman and Barsotti 1985] [Hart et al 2026] [Hastings et al 1986] [Hastings et al 1987] [Helm et al 2023] [Hiller et al 2025] [Holmes 1984] [Holmes 1985] [Holmes and Obara 1982] [Holmes and Obara 1992] [Holmes et al 1983] [Holmes et al 1984] [Holmes et al 1985] [Holmes et al 1986] [Holt 1943] [Horstmann et al 1990] [Hosseinverdi and Fasel 2023] [Ide et al 2012] [Ishida et al 2026] [Ishikawa et al 2017] [Iuliano et al 2011] [Hefner and Sabo 1987] [Karpuk and Elham 2023] [Keller 2019] [Kolesar 1987] [Krishnan et al] [Krumbein et al 2022] [Laflin and Papadakis 2001] [Lamb et al 1985] [Lamb et al 1986] [Lampropoulos et al 2025] [Li and Zhong 2016] [Li et al 2020] [Lin et al 2011] [Low Drag Airfoil Design 1980] [Lynde and Campbell 2016] [Lynde and Campbell 2017] [Lynde et al 2021] [Ma et al 2023] [Malik et al 2013] [Malik et al 2015] [Manning and Kroo 1999] [Maughmer et al 2018] [Maughmer et al 2025] [McCullough and Gault 1947] [McCullough and Gault 1948] [Medina et al 2009] [Methel et al 2025] [Metkowski and Maughmer 2021] [Metkowski and Maughmer 2023] [Bright et al 2020] [Moniripiri et al 2023] [Montoya et al 1981] [Mount and Millman 1985] [Murri et al 1987] [Natural laminar flow airfoil 1979] [Negi et al 2018] [Nie et al 2025] [Obara et al 1986] [Ortiz-Melendez et al 2020] [Ortiz-Melendez et al 2022] [Paredes et al] [Peltzer and Nitsche 2004] [Pfenninger and Vemuru 1989] [Pfenninger et al 1986] [Placek and Ruchała 2018] [Quagliarella and Iuliano 2017] [Quagliarella and Iuliano 2018] [Quast et al 1990] [Quinn 1944] [Quinn 1947] [Quinn 1947] [Radespiel et al 1989] [Redeker and Wichmann 1988] [Redeker et al 1990] [Reeh and Tropea 2015] [Riedel and Sitzmann 1997] [Riedel et al 1998] [Riedel et al 1998] [Rizzetta and Garmann 2025] [Rizzetta and Visbal 2016] [Rizzetta and Visbal 2016] [Roberts et al 2012] [Roberts et al 2015] [Runyan et al 1987] [Sabater et al 2022] [Sakakibara et al 2021] [Sapuppo and Archer 1982] [Selig et al 1995] [Sengupta et al 2021] [Shi et al 2021] [Somers 1992] [Somers 2019] [Srilatha et al 1990] [Stalewski and Sznajder 2016] [Steen 1995] [Steffen et al 2024] [Streit et al 2011] [Streit et al 2015] [Sturdza 2007] [Sturdza and Rajnarayan 2014] [Tang et al 2017] [Tang et al 2022] [Tani 1952] [Tao et al 2020] [Tokugawa and Yoshida 2006] [Tokugawa et al 2008] [Tokugawa et al 2026] [Tracy et al 2002] [Twiss et al 2019] [Vandam 1986] [Vermeersch et al 2015] [Von Doenhoff 1941] [Von Doenhoff 1944] [Waggoner et al 1987] [Wang and Guo 2022] [Wang et al 2019] [Wang et al 2020] [Wentz et al 1984] [Wentz et al 1985] [Wylie et al 2021] [Xiong et al 2018] [Xu and Kroo 2014] [Xu et al 2018] [Xu et al 2019] [Xu et al 2021] [Yan et al 2023] [Yang et al 2021] [Yang et al 2022] [Yao et al 2021] [Young and Booth 1951] [Zakharin and Wygnanski 2008] [Zalovcik 1945] [Zauner and Sandham 2020] [Zauner et al 2023] [Zhang and Deng 2025] [Zhang et al 2015] [Zhang et al 2022] [Zhang et al 2023] [Zhao et al 2014] [Zhao et al 2015] [Zhao et al 2017] [Zhong and Li 2017] [Zhu et al 2019]

The attachment line and leading-edge contamination

The problem the X-21 met first. The swept attachment line, its stability, its contamination from the fuselage and its relaminarisation, which the article reduces to a single Reynolds number.

86 records. [Arnal and Juillen 1990] [Balakumar and Trivedi 1998] [Cebeci 1974] [Coleman and Poll 2001] [Cooke 1962] [Cooke and Robins 1970] [Creel 1991] [Cumpsty and Head 1967] [Cumpsty and Head 1969] [Cumpsty and Head 1980] [Davey 1961] [Dietz et al 1999] [Dimas et al 2003] [Fedorov and Egorov 2021] [Flores et al 1991] [Fujii and Hornung 2003] [Gowree and Atkin 2017] [Hall and Mackerrell 1988] [Hall and Malik 1986] [Hall et al 1984] [Heeg] [Heeg and Geurts] [Holden and Kolly 1995] [Holmes 1991] [Itoh 2000] [Itoh 2000] [Joslin 1996] [Joslin 1996] [Joslin 2004] [Juillen and Arnal 1995] [Kazakov 1990] [Kazakov 2002] [Kitzinger et al 2023] [Lasseigne and Jackson 1992] [Li and Choudhari 2008] [Libby and Chen 1970] [Libby and Kassoy 1970] [Lin and Malik 1994] [Lin and Malik 1995] [Lin and Malik 1995] [Lin and Malik 1996] [Lin and Malik 1997] [Lin and Malik 1999] [Liu 2019] [Mabey 1962] [Mangalam et al 1990] [Manuilovich 2017] [Manuilovich 2017] [Methel et al 2022] [Naval K Agatwal et al 1992] [Neyland and Filippov 2004] [Nishizawa et al 2009] [Obrist and Schmid 2003] [Pfenninger and Bacon 1969] [Poll 1979] [Poll 1980] [Poll 1981] [Poll 1983] [Poll 2007] [Poll and Danks 1995] [Poots 1965] [Pérez et al 2012] [Pérez Pérez] [Reed and Saric 2013] [Schofield and Davey 1967] [Sengupta and Dipankar 2005] [Sesterhenn and Friedrich 2004] [Sesterhenn and Friedrich 2005] [Sesterhenn and Friedrich 2006] [Skuratov and Fedorov 1991] [Smith 1996] [Speer et al 2004] [Theofilis 1998] [Theofilis and Dallmann 1997] [Theofilis et al] [Theofilis et al 1990] [Ustinov] [Ustinov 2004] [van Oudheusden 2003] [Wang et al 2017] [Wrona and Tuliszka-Sznitko 2002] [Xi and Fu 2026] [Xi et al 2021] [Xi et al 2025] [Yatskih et al 2016] [Yermolaev et al 2016]

Crossflow instability on swept wings and its control

The reason sweep is hostile to laminar flow. Crossflow vortices, swept-wing transition and their control. 24 of these titles name discrete or distributed roughness elements and 12 name plasma actuators, the two means of control the article describes.

223 records. [Abegg et al 2000] [Ahmed et al 1989] [Allison and Dagenhart 1978] [Allison and Dagenhart 1987] [Ambrosino et al 2025] [Arai et al 2000] [Archambaud et al 2004] [Arnal et al 1990] [Arnal et al 2000] [Azuma 1995] [Babuev et al 1998] [Baranov et al 2021] [Belisle et al 2010] [Bertelrud et al 2000] [Bippes 1990] [Black 1952] [Boiko et al 2024] [Boiko et al 2024] [Boiko et al 2025] [Borodulin et al 2015] [Borodulin et al 2016] [Borodulin et al 2017] [Borodulin et al 2019] [Borodulin et al 2020] [Borodulin et al 2020] [Borodulin et al 2022] [Borodulin et al 2023] [Borodulin et al 2025] [Borodulin et al 2026] [Borodulin et al 2026] [Botter] [Bridges 1994] [Bui 2013] [Bui 2015] [Carpenter et al 2008] [Carpenter et al 2009] [Carpenter et al 2010] [Carrillo et al 1997] [Carter 1977] [Chen et al 2021] [Chernoray et al 2010] [Chernoray et al 2013] [Chernyshev et al 2016] [Choi and Kim 2018] [Choi and Kwon 2015] [Chuvakhov and Fedorov 2022] [Coleman 1998] [Cooke et al 2019] [Crabtree 1954] [Craig and Saric 2015] [Cumpsty and Head 1967] [Cumpsty and Head 1970] [Dagenhart 1992] [Dagenhart and Saric 1999] [Dagenhart et al 1989] [Dagenhart et al 1990] [De Tullio et al 2013] [De Vincentiis et al 2021] [Denison et al 2019] [Denison et al 2019] [Dryasov et al 2016] [Duan et al 2013] [Dörr and Kloker 2015] [Dörr and Kloker 2016] [Eppink 2021] [Eppink 2023] [Eppink and Yao 2017] [Eppink et al 2019] [Fedorov and Novikov 2023] [Friederich and Kloker 2011] [Garcia et al 1995] [Glauser et al 2014] [Govindarajan and Narasimha 2000] [Groot et al 2026] [Haynes and Reed 1996] [Hefner and Bushnell 1980] [Hopkins et al 1960] [Hosseini et al 2013] [Hunt and Saric 2011] [Iyer and Mahesh 2013] [Iyer et al 1992] [Janke 2001] [Eppink 2017] [Eppink et al 2015] [Jisheng and Heng 1998] [Johnson 1988] [Johnston 1970] [Joslin 1995] [Joslin and Streett 1992] [Kachanov and Tararykin 1990] [Kakkar et al 2026] [Kandil and El-Hady 1988] [Kirilovskiy et al 2019] [Kloker 2008] [Kosarev et al 2016] [Kotvitskii et al 2025] [Kozlov 1966] [Kube 2026] [Kurz and Kloker 2014] [Kurz and Kloker 2016] [Lekoudis 1980] [Levchenko and Scherbakov 1995] [Li and Su 2025] [Li et al 2010] [Li et al 2011] [Li et al 2013] [Li et al 2014] [Liu et al 2023] [Lombardi et al 1996] [Lu and Luo 2017] [Lu et al 2026] [Lundry 1967] [Luo and Zhou 2000] [Mack 1980] [Mack et al 2009] [Maddalon et al 1990] [Malik and Li 1993] [Malik et al 1999] [Manuilovich 2021] [Manuilovich 2023] [Denison et al 2023] [Martin 2008] [Martin et al 2008] [Masad 1996] [Matsuno et al 2020] [Meneghello et al 2015] [Messing and Kloker] [Messing and Kloker 2004] [Messing and Kloker 2010] [Method of the Qualitative Description of the Laminar-Turbulent Transition Position on a Swept Wing in the Flow with an Elevated Degree of Free-Stream Turbulence 2020] [Michel et al 1985] [Mistry et al 2012] [Moniripiri et al 2026] [Mori et al 2024] [Mousseux et al 1989] [Mughal and Ashworth 2013] [Malik] [Nakagawa et al 2023] [Newman et al 1982] [Newman et al 1984] [Ng and Crouch 1999] [Nicholson et al 2019] [Nishino and Shariff 2009] [Niu et al 2024] [Niu et al 2025] [Nomura 2006] [Novikov et al 2024] [Owen and Frendi 2018] [Owens et al 2014] [Pate and Brillhart 1963] [Pate and Groth 1966] [Perraud et al 2003] [Placidi et al 2023] [Pruett et al 2000] [Radeztsky et al 1992] [Radeztsky et al 1993] [Radeztsky et al 1995] [Radeztsky et al 1999] [Radzetsky et al 1994] [Reed 1985] [Reed 1989] [Reed 1991] [Reed and Saric 2002] [Reibert et al 1996] [Reibert et al 1997] [Rius-Vidales and Kotsonis 2020] [Rizzetta and Visbal 2016] [Rizzetta et al 2010] [Rizzetta et al 2010] [Rizzo et al 2019] [Romano et al 2017] [Rudenko and Ryzhkova 1968] [Saric 1997] [Saric 1998] [Saric 2008] [Saric 2010] [Saric and Yeates 1985] [Saric et al 2015] [Saric et al 2019] [Savoshchik and Tokhunts 1974] [Schuelein 2008] [Schueltke and Stumpf 2017] [Sedney 1955] [Sequeira et al 2023] [Serpieri and Kotsonis 2016] [Shahriari et al 2018] [Soltani et al 2011] [Spall and Malik 1992] [Sun and Du 2026] [Thomas et al 2017] [Three-dimensional boundary-layer transition on 1994] [Tolkachev et al 2012] [Tolkachev et al 2012] [Tufts et al 2014] [Ustinov 2013] [Ustinov 2017] [Ustinov 2018] [Ustinov 2018] [Ustinov and Kachanov 2021] [Viken et al 1990] [Wang et al 1994] [Wang et al 2018] [Wang et al 2023] [Wang et al 2025] [Wang et al 2026] [Warren et al 1997] [White and Saric 2000] [White et al 2001] [Wu et al 2024] [Xu and Wu 2022] [Xu et al 2019] [Xu et al 2021] [Yadala et al 2018] [Yadala Venkata et al 2018] [Yakeno and Obayashi 2021] [Yasuaki and Egami 1999] [Ye 2025] [Zhang et al 2017] [Zhou et al 2015] [Zhou et al 2018] [Zoppini et al 2022] [Zoppini et al 2022] [Zoppini et al 2023]

Roughness, waviness and surface tolerance

The tolerance the programme spent its first two years on. Critical roughness heights, steps, gaps, waviness and the transition they cause, from the period criteria to the present.

112 records. [A low-speed experimental investigation of the effect of a sandpaper type of roughness on boundary-layer transition 1959] [Al-Malki et al 2022] [Atkin et al 2006] [Balakumar 2003] [Balakumar and Kegerise 2013] [Bandettini and Isler 1959] [Beguet et al 2017] [Bender and Drake 2006] [Beratlis et al 2016] [Bertin et al 1981] [Bertin et al 1996] [Bertolotti 1993] [Borodulin et al 2021] [Boudreau 1978] [Bouslog et al 1997] [Braslow and Knox 1958] [Cebeci and Smith 1968] [Dimond et al 2019] [He et al 2018] [Hollis 2020] [Clutter and Smith 1960] [Coleman et al 1989] [Crouch et al 2015] [de Bruin 1990] [Dommett 1959] [Drake et al 2005] [Dunn 1964] [Effect of uniformly distributed roughness on turbulent skin-friction drag at supersonic speeds 1959] [Finishing and inspection of model surfaces for boundary-Layer-Transition tests 1959] [Gibbings 1959] [Harris 1964] [Harvey 1975] [Hein et al 2018] [Hildebrand et al 2020] [Hollis 2017] [Holloway and Sterrett 1963] [Hosseini et al 2015] [Iannuzzi 1968] [Eppink 2018] [James 1959] [Joseph and Diwan 2022] [Joseph et al 2022] [Kegerise et al 2010] [Keller and Kloker 2015] [Kendall 1982] [Kerho and Bragg 1995] [Khotyanovsky and Kudryavtsev 2016] [Klebanoff 1966] [Klebanoff and Tidstrom 1972] [Kozlov 1968] [Kozlov et al 2009] [Kubota and Berg 1977] [Leventhal and Reshotko 1981] [Li et al 2009] [Lyons et al 1962] [Ma and Mahesh 2023] [Maddalon and McMillin 1983] [Mačák and Fürst 2023] [McCAULEY et al 1966] [Choudhari et al 2018] [McCroskey 1969] [Merkle et al 1974] [Mishriky and Walsh 2018] [Mochizuki 1961] [Morrisette et al 1986] [Nagamatsu et al 1964] [Obara and Holmes 1986] [Pate 1970] [Pate 1971] [Peterson and Horton 1959] [Phinney and Baltakis 1974] [Pinson and Wang 1999] [Potter and Whitfield 1961] [Potter and Whitfield 1962] [Potter and Whitfield 1964] [Prieto Rueda] [Pérez et al 2015] [Qadri and Schmid 2017] [Radespiel et al 2018] [Reda et al 2012] [Review of the effect of distributed surface roughness on boundary-layer transition 1962] [Rius-Vidales and Kotsonis 2022] [Rizzetta and Visbal 2015] [Rizzetta and Visbal 2015] [Schneider and Beckwith 1992] [Schwartzberg and Braslow 1952] [Smith and Clutter 1959] [Spaid 1969] [Stewart and Simpson 2005] [Stüer et al 2000] [Suryanarayanan et al 2017] [Swigart 1972] [Tani 1961] [Tani and Hama 1953] [Tani and Sato 1956] [Thompson 1970] [Transonic flight test of 1994] [Van Driest and Blumer 1961] [Van Driest and Blumer 1968] [Van Driest and McCAULEY 1960] [Van Driest et al 1967] [Von Doenhoff and Braslow 1961] [White 2005] [White and Reshotko 2008] [Whitfield and Iannuzzi 1969] [Wie and Malik 1998] [Wilder et al 2015] [Young 1965] [Zahn and Rist 2016] [Zhao and Dong 2020] [Zheng et al 2026] [Zhong 2009]

Turbulent spots and wedges

The mechanism of the cloud calculation. The growth and spreading of turbulent spots and wedges, which the article uses to show why an ice crystal blankets the wing at once.

19 records. [Blackwelder et al 2007] [Clark et al 2007] [Coles and Savas 1980] [Glezer et al 1986] [Gregory 1960] [Henningson et al 1987] [Henningson et al 1988] [Kazakov 1995] [Krishnan and Sandham] [LaGraff 2001] [Matsui 1980] [Seifert et al 1992] [Shaikh and Gaster 1995] [Simon and Boyle 1998] [Singer and Joslin 1995] [Smith 2007] [Sokolov et al 1980] [Thompson and Emrich 1967] [Van Atta and Helland 1980]

Boundary-layer stability and transition

The theory behind everything else. Linear and nonlinear stability, receptivity, transition prediction and active transition control in two-dimensional and high-speed boundary layers.

260 records. [Algaidy and El Shrif 2023] [Anderson and Meyer 1990] [Anderson et al 1988] [Antunes Dantas] [Arnal et al 1978] [Arumugam and Durairaj 2011] [Asai et al 1997] [Azzazy et al 1985] [Banks et al 2012] [Berry et al 2008] [Campbell et al 2012] [Dodbele 1990] [Dougherty and Fisher 1980] [Drake et al 1996] [Eppink et al 2023] [Ferziger and Reed 1988] [Fisher and Dougherty 1982] [Fisher and Dougherty 1982] [Guo et al 2022] [Holmes et al 1986] [Hornung 2002] [Hsu 1989] [Johnson et al 1972] [Leonard 1979] [Li and Luo 2012] [Lozano-Durán et al 2018] [Marineau 2017] [Nakagawa et al 1978] [Obara and Holmes 1985] [Obara et al 1990] [Plate boundary layer transition 2023] [Poll and Paisley 1985] [Reshotko 1990] [Rouviere et al 2023] [Sohn and Reshotko 1991] [Venkatachari et al 2021] [Balakumar 2005] [Banner et al 1958] [Blackwell 1968] [Blackwell 1969] [Bloom 1951] [Bloom 1952] [Bloom 1954] [Bloxsom 1965] [Boltz et al 1960] [Boundary layer receptivity, stability] [Bradfield et al 1954] [Braslow 1966] [Bulat et al 2021] [Bushnell et al 1988] [Channabasav and Hampali 2022] [Chapman 1961] [Chaturvedi et al 2024] [Cheng 1953] [Chiu and Norton 1990] [Clutter and Smith 1963] [Collier et al 1987] [Craik 1980] [Creel et al 1987] [Crouch 2021] [Damaren 2018] [Damaren 2023] [Demetriades 1958] [Demetriades 1977] [Demetriades 1985] [Doyle 2001] [Drake et al 1994] [Duchmann et al 2014] [Durante et al 2024] [Dörr and Kloker 2017] [Effects of extreme surface cooling on boundary layer transition 1958] [Egorov and Novikov 2016] [Egorov and Palchekovskaya 2025] [Egorov et al 2017] [Egorov et al 2018] [Egorov et al 2021] [Egorov et al 2022] [Egorov et al 2024] [Egorov et al 2026] [Esaki et al 2018] [Evstigneev and Magnitskii 2020] [Fasel 1990] [Fasel 2000] [Fedorov 1985] [Fedorov 2015] [Flight survey of the 1987] [Freeman 1935] [Freymuth 1966] [Gan 2022] [Gan 2023] [Gaponov and Smorodsky 2000] [Garaev 2017] [Garbaruk et al 2021] [Gaudet 1990] [Gogineni 2014] [Goldstein and Sescu 2008] [Grabe et al 2018] [Grigor’ev and Ershov 2019] [Guo and Kloker 2020] [Görtler and Velte 1967] [Hall and Smith 1988] [Hall and Smith 1989] [Hammachi] [Han and Cao 2019] [Hanks 1969] [He et al 2019] [Henderson and Maddalon 1967] [High Efficiency from Wing 1985] [Holman and Fürst 2021] [Holmes 1986] [Hopkins and Keener 1968] [Horstmann 2008] [Horstmann et al 1990] [Hosseinverdi and Fasel 2018] [Introduction to the Special Section on Hypersonic Boundary-Layer Transition Prediction 2019] [Iyer and Spall 1991] [Jahanbakhshi and Zaki 2019] [Jiang 2026] [Jillie and Hopkins 1961] [Joussot et al 2013] [Kanda 2019] [Kanner and Schetz 1999] [Khanjari et al 2023] [Kiley and Narsipur 2024] [Kim et al 2025] [Kimmel et al 2015] [Kline et al 2019] [Kobashi and Hayakawa 1980] [Kohama et al 2000] [Korotkin 1966] [Korotkin 1966] [Kroo 2008] [Kudryavtsev and Khotyanovsky 2019] [Cutrone et al] [LaGraff 2007] [Landahl 1962] [Landahl 1990] [Lange and Gieseler 1953] [Lauchie and Park 2000] [Laufer and Vrebalovich 1960] [Lee 1953] [Lee 1954] [Leehey and Shapiro 1980] [Lees 1956] [Lees and Reshotko 1962] [Leslie M Mack 1984] [Lessen 1953] [Li and Choudhari 2011] [Lipatov and Ustinov 2018] [Lohse et al 2016] [Low 1955] [Low 1956] [Lyons et al 1961] [Lysenko 1995] [Mack 1967] [Malik et al 2007] [Mangiarotty 1981] [Mankbadi 1992] [Manuel and Doty 1990] [Martín and Paredes 2021] [Maslennikova and Zelman 1985] [Mason and Gatlin 2015] [Matsuda et al 2020] [McElderry 1973] [Choudhari et al] [Choudhari et al 2016] [Meier and Zhou 1989] [Miró Miró 2020] [Moralev et al 2020] [Moralev et al 2021] [Moralev et al 2022] [Morgan and Davies 2020] [Morrisette and Creel 1987] [Nangia et al 2003] [Nark 1963] [Nathaniel Hildebrand et al 2025] [Nie et al 2018] [Nie et al 2022] [Nitsche and Szodruch 1993] [Niu and Li 2022] [Niu and Su 2023] [Osmokrovic et al 2015] [Owens et al 2019] [Parveen and Chowdhury 1970] [Pate 1980] [Pavlyuchenko et al 1985] [Paredes and Choudhari 2018] [Persh 1956] [Persh 1957] [Petzold and Radespiel 2015] [Pisharoti and Brizzolara 2025] [Potter 1968] [Potter 1975] [Powers et al 1963] [Probstein and Elliott 1954] [Qin and Yu 2020] [Raper 1981] [Rarata et al 2023] [Raspet 1957] [Reda 1977] [Redchyts et al 2025] [Reed 1990] [Rioual et al 1994] [Robison and Loth 2017] [Rosenbaum 1966] [Russo and Brehm 2023] [Saeed and Graham 2015] [Saric 1998] [Saric 2004] [Satish et al 2017] [Schneider and Collicott 2000] [Schrauf et al 1998] [Schubauer and Klebanoff 1955] [Schubauer and Skramstad 1947] [Shadloo and Hadjadj 2017] [Shalaev 2018] [Sharma] [Sheetz 1965] [Shepherd 2014] [Shimmy 2026] [Simon et al 2015] [Skripachev 1969] [Smith 1990] [Song et al 2024] [Sternberg 1954] [Struminskii 1967] [Su 2015] [Szulga et al 2015] [Takematsu 1970] [Tang et al 2026] [Tani 1980] [Tatar et al 2019] [Tatsumi 1952] [Theofilis and Poll 1995] [Toppings and Yarusevych 2023] [Ustinov 2016] [Ustinov et al 2022] [Vaglio-Laurin 1978] [van Dam and Holmes 1988] [Van Dam et al 1993] [Vandam 1993] [Vlahostergios et al 2019] [Weilmuenster 1974] [Wilcox 1979] [Wood 1968] [Wu et al 2026] [Wygnanski 1977] [Xia and Chen 2016] [Xiao and Papadakis 2017] [Yan et al 2025] [Yan et al 2025] [Yang and Hu 2008] [Yang and Huano 1969] [Yates et al 2020] [Yih et al 1952] [Yu et al 2026] [Zhigulev 1985] [Zore et al 2023] [Zou and Zhong 2024] [Евстигнеев and Рябков 2016]

Ice crystals, cloud and the icing of aircraft

The atmospheric problem. Ice crystals and cloud particles, their measurement and their effect on laminar flow, with the icing of aircraft surfaces kept as a neighbouring subject of the same leading edge. 7 of the 39 titles name laminar flow, and 12 name an engine or high ice water content, which is the route by which most of the modern measurement arrived.

39 records. [Baumgardner et al 2014] [Bravin and Strapp 2019] [Davis 1982] [Davis and Fischer 1983] [Davis et al 1986] [Davis et al 1987] [Davis et al 1989] [Flegel 2021] [Gurevich et al 2019] [Haggerty et al 2019] [Hayashi and Yamamoto 2014] [Hu et al 2021] [Hu et al 2022] [Huang 2021] [Jasperson et al 1984] [Jasperson et al 1984] [Jensen et al 2008] [Leroy et al 2016] [Leroy et al 2017] [Ma et al 2025] [Malik et al 2024] [Mossop and Hallett 1974] [Mossop and Ono 1969] [Norde et al 2018] [Oliver 2015] [Palacios et al 2012] [Peters et al 2024] [Potts et al 2023] [Pueschel et al 1996] [Rugg et al 2021] [Sullivan et al 2017] [Tan and Lai 2019] [Tatsuta et al 2024] [Wu et al 2026] [Xiao et al 2019] [Xu et al 2024] [Yang et al 2022] [Yang et al 2023] [Yang et al 2026]

Insect contamination and its mitigation

The question the X-21 never reached. Insect residue on laminar leading edges and the means of preventing it, by washing, by protection systems on the leading edge, by aerofoil geometry and by engineered surfaces. The earliest of these 18 records is from 1978, and 7 of them are on the NASA Technical Reports Server.

18 records. [Alexander et al 2016] [Croom and Holmes 1985] [Croom and Holmes 1986] [Fisher and Peterson 1978] [Kok and Young 2014] [Kok et al 2013] [Kok et al 2014] [Kok et al 2015] [Kok et al 2017] [Krishnan et al 2016] [Lorenzi et al 2011] [Maresh and Bragg 1984] [Peterson and Fisher 1978] [Shanahan et al 2015] [Siochi et al 2013] [Wicke et al 2016] [Wohl et al 2013] [Wohl et al 2015]

Coatings against ice and erosion

A neighbouring literature, kept as doubtful. Superhydrophobic, photothermal and protective coatings for leading edges. 29 of the 32 titles name ice, icing, frost or freezing and none names an insect, so this literature is about ice rather than about the contamination the X-21 programme feared.

32 records. [Allahdini et al 2022] [Chen and Guo 2025] [De Pauw and Dolatabadi 2017] [Du et al 2023] [Ferrari and Cirisano 2023] [Gao et al 2019] [Hasegawa et al 2021] [Hoksbergen et al 2023] [Hoksbergen et al 2023] [Hou and Choy 2022] [Kapustin et al 2022] [Li et al 2025] [Liu et al 2022] [Liu et al 2026] [Pan et al 2016] [Peng et al 2024] [Piscitelli et al 2020] [Psarski et al 2022] [Song and Liu 2025] [Song et al 2025] [Sullivan et al 2022] [Tempelis and Mishnaevsky 2025] [Wang and Chen 2012] [Wang et al 2024] [Wang et al 2024] [Wang et al 2026] [Wu et al 2025] [Yang et al 2025] [Zeng et al 2021] [Zhang et al 2023] [Zhang et al 2025] [Zhou et al 2024]

Suction systems, perforated surfaces and their manufacture

The hardware. Slots, holes, perforated and porous skins, suction panels, valves and the structure and manufacture of the suction surface.

45 records. [Arakeri and Narasimha 1983] [Ball 1969] [Bishara et al 2018] [Brandeis 1983] [Cliett 1952] [Danilov 2011] [Dannenberg et al 1957] [Egreteau et al 2024] [Fischler 1986] [Fischler et al 1989] [Flax et al 1953] [Goglia and Wilkinson 1980] [Goodyear 1987] [Gregory 1961] [Hybrid Suction Surface for 1980] [Kashaev et al 2026] [Khan et al 1982] [Kozlov 1965] [Li et al 2024] [MacManus and Eaton 1996] [Maddalon 1991] [Maddalon and Braslow 1990] [Maddalon and Poppen 1986] [Mansy and Faruque 2023] [Meitz and Fasel 1995] [Mills 1971] [Pearce 1978] [Pearson 1967] [Perry and Ansell 2017] [Petley et al 1985] [Pfenninger et al 1967] [Roache 1979] [Saeed et al 2010] [Saeed et al 2011] [Thiede 1978] [Thomas and Cornelius 1981] [Thomas and Cornelius 1982] [Traub 2026] [Traub 2026] [Traub 2026] [van Rijn 1997] [Wadia 1983] [Watmuff 1994] [Wilson 1980] [Wlezien 1989]

Suction and injection in the boundary layer

The mechanism outside the aircraft. Suction and injection through a wall as boundary-layer theory, in porous channels and tubes, at separation and for high lift, kept because each shares the mechanism of wall suction with this subject.

102 records. [Adcock et al 1966] [Albacete and Glowacki 1967] [Album 1969] [Asai et al] [Atik and van Dommelen 2007] [Baldwin 1970] [Ball 1970] [Barrow 1958] [Braslow et al 1951] [Brynjell-Rahkola et al 2019] [Cherubini et al 2015] [Chuang 1974] [Coleman 1957] [Corelli Grappadelli et al 2025] [Curle 1960] [Dennon 1962] [Dershin et al 1966] [Dershin et al 1967] [Dorrance and Dore 1954] [El-Gamal and Barclay 1975] [Elkouh 1971] [Fedorov and Soudakov 2014] [Fukusako et al 1971] [Gershbein 1970] [Gershbein 1975] [Gortler 1957] [Gray and Maus 1981] [Gribben 1959] [Görtler’s new series method for steady laminar boundary layer flows with suction 1962] [Hasimoto 1957] [Head 1961] [Hughes and Reid 1965] [Improving Wind Turbine Power with Boundary Layer Suction 2025] [Investigation of Laminar Flow of Fluid with One Porous Bounding Wall 2016] [Khusnutdinova 1972] [Kinney 1967] [Kobayashi 1972] [Kobayashi 1974] [Koh and Hartnett 1961] [Kozlov and Tsyganyuk 1969] [Laouer et al 2016] [Leitsina and Pavlyukevich 1968] [Lipatov and Neiland 1974] [Lysenko et al 2024] [MacDermott et al 1965] [MacDermott et al 1966] [Maddaus and Shanebrook 1983] [Maeda 1968] [Messiha 1966] [Mills 1966] [Morduchow 1963] [Morduchow and Reyle 1967] [Ness 1961] [Northrop Aircraft Inc Hawthorne Ca 1952] [Northrop Aircraft Inc Hawthorne Ca 1953] [Nouri-Borujerdi and Seyyed-Hashemi 2016] [Pappas and Okuno 1960] [Pasamanick and Proterra 1948] [Passamanick 1948] [Pate 1969] [Pechau 1963] [Pfenninger 1961] [Poulain et al 2024] [Powers 1966] [Prochukhaev 1973] [Pushkareva 1970] [Quaile and Levy 1972] [Quaile and Levy 1975] [Racisz and Quinn 1948] [Reed and Nayfeh 1981] [Reynolds et al 1982] [Rhudy 1970] [Ricco et al 2013] [Ringleb 1952] [Roberts et al 1966] [Roy and Saikrishnan 2003] [Schlichting 1949] [Schlichting and Gersten 2000] [Seebaugh and Childs 1970] [Sellars 1955] [Seraudie et al 1999] [Simpson et al 1969] [Strike and Pate 1961] [Swean and Inger 1974] [Terrill 1960] [Terrill 1965] [Thummar et al 2023] [Thummar et al 2024] [Torda 1952] [Torda 1953] [Trilling 1950] [Wahidi and Bridges 2008] [Wedin and Cherubini 2016] [Wedin et al 2015] [Wu 1973] [Wuest 1952] [Wuest 1961] [Yalcin et al 2021] [Yuan 1955] [Zamir and Young 1967] [Ze-yan 1981] [Zolotov and Khodorkovsky 1963]

Skin friction and the turbulent boundary layer

The friction that sets the prize. Laminar and turbulent skin friction, its measurement and its prediction, including the flat-plate correlations the sizing uses.

49 records. [Allen and Monta 1965] [Bertram and Neal 1967] [Brown and Joubert 1969] [Bruno and Risher 1968] [Chang 1960] [Cope 1952] [Cornish 1963] [Eckert 1950] [Eckert et al 1961] [Granville 1960] [Hebbar and Paranjpe 1969] [Hopkins and Keener 1972] [Hopkins et al 1970] [Hopkins et al 1972] [Howell 1970] [Kelnhofer 1969] [Komar 1966] [Korkegi 1956] [Krzywoblocki 1953] [Lee and Faget 1956] [Li and Nagamatsu 1953] [Li et al 2023] [Liu 1967] [Matsui 1963] [Moulic 1963] [Nark and Lee 1963] [Nicoll and Ramaprian 1969] [Reggiori 1972] [Roberts 1963] [Rubesin and Johnson 1949] [Shanebrook and Sumner 1970] [Sirovich and Karlsson 1999] [Spalding and Chi 1963] [Spalding and Chi 1964] [Sparrow et al 1963] [Spence 1960] [Sternberg 1961] [Swafford 1979] [The determination of local turbulent skin friction from observations in the viscous sub-layer 1960] [Uram 1960] [Vessey 1935] [White and Christoph 1972] [White et al 1973] [Wilson 1962] [Winkler 1961] [Winkler and Cha 1959] [Yamaguchi et al 2020] [Yoshizawa 1969] [Young 1949]

Riblets and the reduction of turbulent skin friction

The competing route. Riblets, compliant walls, polymers, spanwise forcing and other means of reducing the friction of a turbulent layer. 63 of the 114 titles name riblets.

114 records. [Abbassi et al 2017] [Advisory Group for Aerospace Research and Development 1992] [Allen and Bridges 2003] [Amico and Cafiero 2026] [Anderson et al 1969] [Ao et al 2021] [Bai et al 2016] [Beck et al 2018] [Benschop and Breugem 2017] [Bixler and Bhushan 2013] [Bliamis et al 2022] [Blick and Walters 1968] [Boomsma and Sotiropoulos 2015] [Bushnell et al 1984] [Cacciatori et al 2022] [Cafiero and Iuso 2022] [Carpenter et al 2001] [Catalano et al 2020] [Ceccio et al 2007] [Chamorro et al 2013] [Chehrghani et al 2025] [Chen et al 2015] [Choi 2004] [Cipelli et al 2026] [Cornish 2024] [Cornish III and Boatwright 1960] [Cui et al 2023] [Dimotakis et al 2003] [Dou et al 2026] [Drag Reduction Performance of Triangular (V-groove) Riblets with Different Adjacent Height Ratios 2023] [First Laminar Flow Flight 1992] [Gad-El-hak and Blackwelder 1987] [Gatti and Quadrio 2016] [Gatti et al 2025] [Geng et al 2021] [Gordon et al 2024] [Gowree et al 2019] [Granville 1973] [Guo et al 2020] [Hendrikse] [Howard et al 1975] [Hu and Yao 2023] [Hu and Yao 2024] [Hwang and Biesiadny 1998] [Inasawa et al 2024] [Kaneko et al 2026] [Koo and Kang 2020] [Kousik Kumaar. R and Dr. P. Maniiarasan 2015] [Landahl and Widnall 1979] [Landahl and Widnall 1980] [Landahl and Widnall 1981] [Landahl and Widnall 1982] [Lazos and Wilkinson 1988] [Leadon and Bartle 1960] [Lee and Kim 2017] [Li et al 2022] [Li et al 2024] [Liang et al 2022] [Lin and Mao 2024] [Ma et al 2025] [Mahfoze and Laizet 2017] [Mamori and Kondo 2019] [Marino et al 1975] [Mawignon et al 2023] [Mele 2022] [Mele et al 2020] [Mignanwande 2024] [Olivucci et al 2021] [Pang et al 2024] [Pate 1964] [Pate and Deitering 1963] [Pate and Deitering 1963] [Perlin et al 2016] [Poreh and Hsu 1972] [Raayai-Ardakani and McKinley 2019] [Ran et al 2020] [Rastegari and Akhavan 2018] [Ricco et al 2021] [Rothstein 2010] [Rydalch 2013] [Safari et al 2025] [Saito et al 2021] [Saito et al 2022] [Sareen et al 2013] [Scarano et al 2022] [Selvanose et al 2024] [Sharma and Dutta 2022] [Sharma et al 2025] [Shimizu et al 2021] [Sinha and Ravande 2006] [Soleimani and Eckels 2021] [Stroh et al 2016] [Takahashi et al 2019] [Vlachos et al 2007] [Walsh 1982] [Walsh and Lindemann 1984] [Walsh et al 1986] [Wang et al 2013] [Wei et al 2013] [Wilkinson et al 1987] [Wong et al 2024] [Wu 1972] [Wu and Tulin 1970] [Ye et al 2019] [Yonezawa et al 2017] [Youngren 2008] [Yu 2017] [Zaki 2014] [Zeng et al 2024] [Zhang and Yin 2019] [Zhang et al 2020] [Zhou et al 2020] [Zhou et al 2022] [Zong et al 2022]

Induced drag, span efficiency, cruise and aircraft design

The aircraft-level accounting. Induced drag and span efficiency, cruise and range, and the design of aircraft around laminar flow.

25 records. [Deyoung 1980] [Frost and Rutherford 1963] [Gilman and Burdges 1967] [Guynn 2011] [Hale 1976] [Hancock 1971] [Hancock 1972] [Kehayas 2023] [Lakshminarayana 1962] [Letcher 1972] [Lockwood Taylor 1942] [Mamada and Ando 1973] [Mamada and Ando 1974] [Morita 1971] [Moriya 1972] [Report No. 349. A proof of the theorem regarding the distribution of lift over the span, for minimum induced drag 1930] [Sanders 1965] [Sears 1974] [Sibert 1943] [Smith 1996] [Smith and Kroo 1990] [Snyder and Zumwalt 1969] [Straussfogel and Maughmer 1991] [von Kármán and Tsien 1945] [Oswald 1932]

Where the Framing Breaks Down

The calculation above treats the suction system as a compressor restoring a total pressure, and charges the aircraft the ideal work divided by an efficiency. That is a thermodynamic accounting, not an installation. It does not include the weight of the ducts and turbines, the volume they take from the fuel tanks, the drag of the nacelles that house them, or the structural penalty of a skin perforated by some two hundred slots. A real trade would carry all of those, and they act against the technology, not for it.

The whole-aircraft figures rest on an assumed wetted-area breakdown, an assumed form factor and an assumed non-friction drag increment. They are the weakest numbers in the article and are offered as an order of magnitude. They are also the numbers most sensitive to consistent bookkeeping. A zero-lift drag coefficient that implies a form factor, set against a laminarised area credited with a flat-plate saving that carries none, understates the benefit by about a quarter, so the coefficient is built here from its parts and the charge and the credit use the same basis.

The aspect-ratio argument inverts a drag polar that already contains three assumptions, so the figure of 11.8 should be read as evidence that the required planform is very different rather than as a specification for it.

The attachment-line calculation assumes a leading-edge radius, a taper ratio and that the quoted sweep is measured at the quarter chord. The conclusion drawn from it, that the wing root sits near the contamination threshold, survives across the plausible range, but the precise value does not.

The suction coefficient is the least well constrained input of all. The value used is an assumed design figure rather than a measured X-21 quantity, and the entire margin over break-even scales inversely with it. The local values the programme reports are two and a half to five times higher, and the reported flow rates can be read either way, so the margin of 5.8 is an upper estimate for the nominal machinery and the true figure for the X-21 lies somewhere between about one and about four.

The cloud calculation assumes a number density of ten thousand crystals per cubic metre, which no source states. Hall’s typical cirrus gives nearly nine times as many, so the assumption errs toward understating the effect, and the threshold of about eight crystals per cubic metre for erratic operation inherits every crudeness of the wedge model, including the assumption that every wedge starts at the leading edge and spans the mean chord.

The Source Base

The programme is unusually well documented for one that ended without a production application, and the reason is that a NASA engineer who consulted on it wrote its history thirty years later.

Braslow 1999 is the load-bearing source for the programme narrative, including the Vietnam attribution, and it reproduces the primary document behind the rest, the report of the review group that met in October 1965. Kosin 1965 is the contemporary engineering description in the Journal of Aircraft, and Wheldon and Whites 1966 is a paper on the flight testing of the aeroplane. The most useful description of the wing’s internal construction is Butz 1963, a trade-press account published while the aeroplane was flying, which is where the distinction between the slots and the metering holes is stated plainly. Hall 1964 is the Northrop report on the cloud problem, now on the NASA Technical Reports Server. The quantitative results of the Northrop final report, its suction coefficients, flow rates, tolerances and year-by-year laminar extents, are known here through the summary in Joslin 1998, because the Northrop reports themselves, held by the Defense Technical Information Center, were not retrievable. Where this article quotes those numbers it quotes a careful secondary summary of a primary document rather than the document.

A caution about the secondary literature. The figure of eight hundred thousand slots is repeated by reference works, encyclopaedias and aviation histories alike. The 1963 description in Butz 1963 makes clear that the number counts drilled holes connecting plenums to ducts, and the slot spacings Joslin reports put the slots at of the order of a hundred on each surface. The error is not consequential for the physics and it is very consequential for the impression the reader forms, since it converts a precisely engineered surface into something that sounds like a colander.

The research survey holds 1,560 research records drawn from the NASA Technical Reports Server, the Defense Technical Information Center, the Office of Scientific and Technical Information and the journal literature indexed by Crossref, of which 456 are report primaries, being 29.2 percent, by the identifier rule set out under The Definition of Primary. The median publication year is 2000. Of the 1,536 records that carry a year, 186, or 12.1 percent, are dated 1965 or earlier, the period in which the aircraft was designed and flown, and 639, or 41.6 percent, are dated 2011 or later, so the reference base leans toward the contemporary literature.

The Definition of Primary

Primacy is a property of the document, and this survey derives it from the identifier. A record is a report primary when its address is on the NASA Technical Reports Server or the Office of Scientific and Technical Information, or when it carries a digital object identifier under the 10.21236 prefix, which Crossref registers for Defense Technical Information Center reports. The Northrop summaries of laminar boundary layer control research and the reports of NASA’s leading-edge flight test are report primaries under that rule, and so is Braslow and Knox’s simplified method for the critical roughness height. The rule does not reach first-hand papers published in journals, such as Kosin 1965 on the suction system of the X-21A and Schubauer and Skramstad 1947 on laminar boundary-layer oscillations, so 29.2 percent is a lower bound on how much of the survey is first-hand.

Thin Subjects

Six subjects that this article’s derivations depend on have thin literatures. Each count below is the number of research records in this survey whose title carries at least one of the listed terms, with a hyphen read as a space. Where a row joins two groups of terms with the word together, a title must carry a term from each group.

Subject	Title terms counted	Records
the X-21 programme itself	X-21, X-21A, X 21, X 21A	2
maintenance and service experience of a laminar wing	maintenance, maintainability, airline service, operational experience, operational assessment, operational considerations, durability, in service	6
cruise performance and the range of a transport	cruise, long range, best range, global range, Breguet	14
ice particles acting on laminar flow	ice, ice crystal, ice crystals, cloud, clouds, particle, particles, cloud encounter, cloud encounters together with laminar, laminar flow, LFC	11
turbulent spots and wedges	turbulent spot, turbulent spots, turbulent wedge, turbulent wedges	19
the power, weight and installation of the suction system	suction power, suction system, suction systems, pumping power, power requirement, power requirements, suction pump, suction pumps	8

The aeroplane has almost no literature of its own. Two titles name the X-21, Kosin 1965 on its suction system and Wheldon and Whites 1966 on its flight testing, so the programme’s record rests on Braslow’s history and Joslin’s summary rather than on a body of papers.

The question the programme never answered has six titles. The earliest is from 1986, two decades after the review group asked for an operational evaluation, and three of them report simulated airline service of laminar flow control leading-edge systems.

The particle problem is mostly one decade’s work. Of the eleven titles that join ice, cloud or particles to laminar flow, seven date from 1982 to 1989, the decade of NASA’s leading-edge flight test, and three of the rest concern ice accretion or ice protection rather than particles in the boundary layer.

The wedge model rests on the spot literature. One of the 19 titles on spots and wedges names a turbulent wedge, so the spreading angle the cloud calculation uses comes from studies of the spot.

The installed suction system is named in eight titles and its power in one. That is the quantity a manufacturer needs and the one this article does not compute.

What the Survey Admits

A record is admitted only when a person reading its title finds it on this article’s subject, the subject being suction laminar flow control, the stability theory behind it, the surfaces it needs and the insects and ice that contaminate them. The phrase boundary layer belongs to meteorology as much as to aerodynamics, and a whole journal is devoted to the atmospheric boundary layer. Operating-room laminar air flow, crossflow filtration and a Cretaceous stick-insect fossil share this subject’s vocabulary in the same way and none of them is part of it.

One homonym in this subject lives inside aerodynamics rather than outside it. Boundary layer control means suction to hold laminar flow and reduce drag, and it also means suction or blowing to delay separation and increase lift. They share a name, a mechanism and often an author, and they are different technologies answering different questions, so a title that names boundary layer control does not by itself place a record in this subject. Records that control separation or stall by plasma, jets, fluidic oscillators, tubercles or flexible surfaces, without suction through the wall, are excluded on that ground.

Many words this subject depends on also name something else, and a record that shares only such a word is excluded. Perforated also names panels made to absorb sound, laminar also names flames and the internal flow and heat transfer of tubes, channels and fins, and coating also names protection against corrosion and the surfacing of pavements, insulators and face masks. Non-Newtonian, magnetohydrodynamic and porous-medium flows, chemical mixers, fuel cells, surgery, seaweed and power grids share the vocabulary of the boundary layer without its subject. Transition also names the conversion of a tilting aircraft from hover to forward flight, noise also names the sound an aerofoil radiates, and attachment also names separated flows reattaching behind a step or a ramp. Bluff bodies and their wakes, turbomachinery blades and cascades, the chemistry of re-entry boundary layers and the interaction of shock waves with separated flow are excluded for the same reason. Notices of erratum, correction, retraction or withdrawal, figure, table and supplementary-file records, peer-review reports and journal front matter are excluded as well, because a survey counts research works and those are parts of works or editorial events rather than works, and a second deposit of one document is counted once. A review of a book, a dictionary or short encyclopedia entry and a notice such as an editorial, an announcement or a news item are not research works either and are excluded.

Some records of doubtful relevance are kept. Laminar flow with suction or injection through the porous wall of a tube or channel, magnetohydrodynamic boundary-layer control with suction, suction for high lift, anti-icing and leading-edge erosion coatings including those for wind turbine blades, the icing of aircraft surfaces, and drag reduction by polymers and riblets in liquids each share with this subject either the mechanism of wall suction, the problem of a contaminated leading edge or the aim of reducing skin friction, so each is kept and a reader may weigh it accordingly. Transition at supersonic and hypersonic speeds is kept because it is the same stability theory.

Every research title has been read for relevance, so the off-topic share that remains is a matter of reading judgement. A reading of titles cannot show that a work is relevant in substance, so this is a statement about titles only.

Epistemic State

Established by the programme and not in dispute. Laminar flow was obtained and held on a large swept wing over more than ninety-five percent of the intended area. Surface tolerance and spanwise contamination were the two problems that consumed the schedule, and both were overcome, in Braslow 1999.

Computed here, and dependent on stated assumptions. The suction penalty is about a sixth of the friction saving at the nominal cruise condition, the break-even suction coefficient is about six times the design value, and the suction drag coefficient is independent of altitude. The last of those is exact given the model. The first two carry the uncertainty of the suction coefficient and the duct recovery. Against the local suction coefficients of the Northrop final report, the margin falls to between about one and two, so what survives the primary record is that suction cost less than it saved, not that it cost a sixth.

Inferred, and offered as inference. That the X-21 as a converted bomber could not have demonstrated more than about twelve percent in range follows from the wetted-area accounting, and the wetted areas are estimated. That the 1963 claim of a lift-to-drag ratio above thirty is unreachable on this planform at any level of laminarisation follows from induced drag alone consuming 73 percent of the drag budget such a ratio allows, and that argument is stronger than the wetted-area one because it does not depend on how much of the aeroplane is laminarised.

Asserted widely and not supported by the record. That the X-21 demonstrated laminar flow control to be impractical on maintenance grounds. The maintenance evaluation was not performed. The maintenance the programme did record went mostly to repairing filled splices, and insects did not disturb the laminar flow in the desert where it flew, in Joslin 1998 and Braslow 1999.

Read in the record, with a caveat on wording. Braslow 1999 writes that the service-experience objective had not even been initiated by October 1965. The review group’s report says the programme had not been able completely to address the operational questions. Both agree that no operational evaluation took place.

Unknown then and largely unknown still. The operational significance of atmospheric ice particles, for want of statistics on particle size and concentration along real routes, in Braslow 1999.

Out of Scope

Supersonic laminar flow control, which the programme discussed and did not test, and which later became its own subject. The detailed stability theory of the slotted surface, including the disturbances the slots themselves introduce. The structural and weight penalties of the suction wing, which are named as a limitation above and not computed. The economics of laminar flow control for commercial operators, which belongs to the transport design studies, not to this aeroplane.

Conclusion

The X-21A did the hard part. It held laminar flow on a large swept wing in real air, having first beaten a surface-tolerance problem and an attachment-line problem that nobody had properly anticipated, and it did so with a suction system that, on the assumptions used here, cost about a sixth of what it returned, and on the local coefficients the programme reports cost between about two fifths and most of it, since the penalty fraction of 0.172 scales with the coefficient to 0.43 at $5 \times 10^{-4}$ and 0.86 at $10 \times 10^{-4}$.

The power trade was never the difficulty. The margin over break-even is a factor of six at the assumed suction coefficient and nearer two at the reported local values, the penalty falls with Mach number, and it does not depend on altitude at all. An aeroplane designed around this from the start would have collected far more of it than a converted bomber ever could, and the case for the technology was accordingly always made about an aircraft that did not exist.

What the programme was for, in the end, was to find out whether such a wing could be lived with. It never found out. The maintenance objective had not been reached when the funding stopped, the recommended wing modification that would have made the evaluation meaningful was never made, and the aeroplane was retired to Edwards to be photographed, in the Northrop X-21A reference entry. The verdict that laminar flow control is impractical because of maintenance was reached anyway.

That makes this aircraft unusual in a series full of failures. The X-19 fell out of the sky. The X-20 was cancelled without flying. The X-21 succeeded at what it was asked to do and was then remembered for failing at something it was never allowed to attempt.

References

Books

• Anderson 1999 Aircraft Performance and Design
• Braslow 1999 A History of Suction-Type Laminar-Flow Control with Emphasis on Flight Research
• Jenkins Landis and Miller 2003 American X-Vehicles, An Inventory X-1 to X-50
• Miller 2001 The X-Planes, X-1 to X-45

Reference

• Butz, Laminar Flow Control, Air Force Magazine, August 1963
• Douglas DC-8
• Douglas WB-66D
• Hall, On the Mechanics of Transition Produced by Particles Passing Through an Initially Laminar Boundary Layer and the Estimated Effect on the LFC Performance of the X-21 Aircraft, Northrop, 1964
• Joslin, Overview of Laminar Flow Control, NASA/TP-1998-208705, 1998
• Northrop X-21A

Related Post

• X-1
• X-10
• X-11
• X-12
• X-13
• X-14
• X-15
• X-16
• X-17
• X-18
• X-19
• X-2
• X-20
• X-3
• X-4
• X-5
• X-6
• X-7
• X-8
• X-9
• X-Planes series

Research

• A low-speed experimental investigation of the effect of a sandpaper type of roughness on boundary-layer transition 1959
• Abbassi et al 2017
• Abdul Nour and Mueller 1993
• Abegg et al 2000
• Adcock et al 1966
• Advisory Group for Aerospace Research and Development 1992
• Agrawal et al 1989
• Ahmadi-Baloutaki et al 2013
• Ahmed et al 1989
• Aircraft energy efficiency program 1979
• Al-Malki et al 2022
• Albacete and Glowacki 1967
• Album 1969
• Alexander et al 2016
• Algaidy and El Shrif 2023
• Allahdini et al 2022
• Allen and Bridges 2003
• Allen and Monta 1965
• Allison 1983
• Allison and Dagenhart 1978
• Allison and Dagenhart 1987
• Ambrosino et al 2025
• Amico and Cafiero 2026
• Anders and Fischer 1999
• Anderson and Meyer 1990
• Anderson et al 1969
• Anderson et al 1988
• Antunes Dantas
• Ao et al 2021
• Arai et al 2000
• Arakeri and Narasimha 1983
• Arcara et al 1991
• Archambaud et al 2004
• Arnal and Juillen 1990
• Arnal et al 1978
• Arnal et al 1990
• Arnal et al 2000
• Arumugam and Durairaj 2011
• Asai et al
• Asai et al 1997
• Athar et al 2025
• Atik and van Dommelen 2007
• Atkin 2004
• Atkin 2008
• Atkin et al 2006
• Azuma 1995
• Azzazy et al 1985
• Babuev et al 1998
• Bai et al 2016
• Balakumar 2003
• Balakumar 2005
• Balakumar and Kegerise 2013
• Balakumar and Trivedi 1998
• Baldwin 1970
• Ball 1969
• Ball 1970
• Bandettini and Isler 1959
• Banks et al 2012
• Banner et al 1958
• Baranov et al 2021
• Barklage et al 2022
• Barklage et al 2023
• Barnwell and Hussaini 1992
• Barrow 1958
• Barry et al 1994
• Baumgardner et al 2014
• Beck et al 2018
• Befus et al 1987
• Beguet et al 2017
• Belisle et al 2010
• Belisle et al 2012
• Bell and Cornelius 1987
• Bender and Drake 2006
• Benschop and Breugem 2017
• Beratlis et al 2016
• Berry 1986
• Berry et al 1987
• Berry et al 2008
• Bertelrud et al 2000
• Bertin et al 1981
• Bertin et al 1996
• Bertolotti 1993
• Bertram and Neal 1967
• Bidwell 1991
• Bin Guo et al 2017
• Bippes 1990
• Bishara et al 2018
• Bixler and Bhushan 2013
• Black 1952
• Blackwelder et al 2007
• Blackwell 1968
• Blackwell 1969
• Bliamis et al 2022
• Blick and Walters 1968
• Bloom 1951
• Bloom 1952
• Bloom 1954
• Bloxsom 1965
• Bobbitt et al 1992
• Bobbitt et al 1992
• Boiko et al 2024
• Boiko et al 2024
• Boiko et al 2025
• Boltz et al 1960
• Bonner et al 1977
• Boomsma and Sotiropoulos 2015
• Borodulin et al 2015
• Borodulin et al 2016
• Borodulin et al 2017
• Borodulin et al 2019
• Borodulin et al 2020
• Borodulin et al 2020
• Borodulin et al 2021
• Borodulin et al 2022
• Borodulin et al 2023
• Borodulin et al 2025
• Borodulin et al 2026
• Borodulin et al 2026
• Botter
• Boudreau 1978
• Boundary layer receptivity, stability
• Bouslog et al 1997
• Bozeman et al 2024
• Bradfield et al 1954
• Bragg and Gregorek 1989
• Brandeis 1983
• Braslow 1966
• Braslow and Fischer 1985
• Braslow and Knox 1958
• Braslow and Muraca 1978
• Braslow et al 1951
• Braun et al 2011
• Bravin and Strapp 2019
• Brennecke 1948
• Bridges 1994
• Bright et al 2020
• Brinza 2024
• Brinza 2025
• Brinza et al 2026
• Brooks and Harris 1987
• Brooks et al 1991
• Brown and Joubert 1969
• Bruno and Risher 1968
• Brynjell-Rahkola et al 2019
• Bui 2013
• Bui 2015
• Bulat et al 2021
• Bushnell and Malik 1987
• Bushnell and Malik 1987
• Bushnell and Tuttle 1979
• Bushnell et al 1984
• Bushnell et al 1988
• Cacciatori et al 2022
• Cafiero and Iuso 2022
• Campbell and Lynde 2017
• Campbell et al 1986
• Campbell et al 2012
• Carmichael 1979
• Carpenter et al 2001
• Carpenter et al 2008
• Carpenter et al 2009
• Carpenter et al 2010
• Carpenter et al 2018
• Carrillo et al 1997
• Carter 1977
• Catalano et al 2020
• Cebeci 1974
• Cebeci and Smith 1968
• Cebeci and Smith 1968
• Ceccio et al 2007
• Cella et al 2010
• Chakraborty et al 2022
• Chamorro et al 2013
• Chang 1960
• Channabasav and Hampali 2022
• Chapman 1961
• Chaturvedi et al 2024
• Chehrghani et al 2025
• Chen and Guo 2025
• Chen and Tang 2017
• Chen et al 2015
• Chen et al 2021
• Chen et al 2023
• Chen et al 2023
• Cheng 1953
• Chernoray et al 2010
• Chernoray et al 2013
• Chernyshev et al 2016
• Cherubini et al 2015
• Chiu and Norton 1990
• Choi 2004
• Choi and Kim 2018
• Choi and Kwon 2015
• Choudhari 1992
• Choudhari et al
• Choudhari et al 2003
• Choudhari et al 2016
• Choudhari et al 2018
• Chuang 1974
• Chuvakhov and Fedorov 2022
• Cipelli et al 2026
• Clark et al 2007
• Clayton
• Cliett 1952
• Clutter and Smith 1960
• Clutter and Smith 1963
• Coder 2020
• Coder 2021
• Coder 2026
• Coder and Somers 2020
• Coder et al 2014
• Coleman 1957
• Coleman 1998
• Coleman and Poll 2001
• Coleman et al 1989
• Coles and Savas 1980
• Collazo Garcia and Ansell 2018
• Collazo Garcia and Ansell 2019
• Colletti et al 2021
• Collier 1993
• Collier et al 1987
• Collier et al 1992
• Cooke 1962
• Cooke and Robins 1970
• Cooke et al 2019
• Cope 1952
• Corelli Grappadelli et al 2022
• Corelli Grappadelli et al 2023
• Corelli Grappadelli et al 2025
• Corelli Grappadelli et al 2026
• Cornish 1963
• Cornish 2024
• Cornish III and Boatwright 1960
• Coutu et al 2011
• Crabtree 1954
• Craig and Saric 2015
• Craik 1980
• Creel 1991
• Creel et al 1987
• Croom and Holmes 1985
• Croom and Holmes 1986
• Crouch 2008
• Crouch 2015
• Crouch 2021
• Crouch et al 2015
• Crowley and Atkin 2016
• Cui et al 2023
• Cumpsty and Head 1967
• Cumpsty and Head 1967
• Cumpsty and Head 1969
• Cumpsty and Head 1970
• Cumpsty and Head 1980
• Curle 1960
• Cutrone et al
• Dagenhart 1992
• Dagenhart and Saric 1999
• Dagenhart et al 1989
• Dagenhart et al 1990
• Damaren 2018
• Damaren 2023
• Danilov 2011
• Dannenberg et al 1957
• Dargel and Thiede 2002
• Davey 1961
• Davis 1982
• Davis and Fischer 1983
• Davis et al 1986
• Davis et al 1987
• Davis et al 1989
• de Bruin 1990
• De Pauw and Dolatabadi 2017
• de Rosa and Donelli 2020
• De Tullio et al 2013
• De Vincentiis et al 2021
• Demetriades 1958
• Demetriades 1977
• Demetriades 1985
• Deng et al 2026
• Denison et al 2019
• Denison et al 2019
• Denison et al 2023
• Dennon 1962
• Dershin et al 1966
• Dershin et al 1967
• Deyoung 1980
• Dietz et al 1999
• Dimas et al 2003
• Dimond et al 2019
• Dimotakis et al 2003
• Djeddi and Ekici 2024
• Dodbele 1990
• Dodbele 1992
• Dodbele et al 1986
• Dodd et al 2001
• Dommett 1959
• Dorrance and Dore 1954
• Dou et al 2026
• Dougherty and Fisher 1980
• Doyle 2001
• Drag Reduction Performance of Triangular (V-groove) Riblets with Different Adjacent Height Ratios 2023
• Drake et al 1994
• Drake et al 1996
• Drake et al 2005
• Dryasov et al 2016
• Du et al 2023
• Duan et al 2013
• Duchmann et al 2014
• Dunn 1964
• Durante et al 2024
• Dörr and Kloker 2015
• Dörr and Kloker 2016
• Dörr and Kloker 2017
• Eckert 1950
• Eckert et al 1961
• Edwards 1961
• Effect of uniformly distributed roughness on turbulent skin-friction drag at supersonic speeds 1959
• Effects of extreme surface cooling on boundary layer transition 1958
• Egorov and Novikov 2016
• Egorov and Palchekovskaya 2025
• Egorov et al 2017
• Egorov et al 2018
• Egorov et al 2021
• Egorov et al 2022
• Egorov et al 2024
• Egorov et al 2026
• Egreteau et al 2022
• Egreteau et al 2024
• El-Gamal and Barclay 1975
• Elkouh 1971
• Eppink 2017
• Eppink 2018
• Eppink 2021
• Eppink 2023
• Eppink and Wlezien 2011
• Eppink and Yao 2017
• Eppink et al 2015
• Eppink et al 2019
• Eppink et al 2023
• Esaki et al 2018
• Etchberger 1983
• Evaluation of laminar flow 1980
• Evstigneev and Magnitskii 2020
• Fan et al 2017
• Fan et al 2021
• Fasel 1990
• Fasel 2000
• Fedorov 1985
• Fedorov 2015
• Fedorov and Egorov 2021
• Fedorov and Novikov 2023
• Fedorov and Soudakov 2014
• Fehrs and Kaiser 2025
• Ferrari and Cirisano 2023
• Ferziger and Reed 1988
• Finishing and inspection of model surfaces for boundary-Layer-Transition tests 1959
• First Laminar Flow Flight Tests Prove Drag Reduction Potential 1992
• Fischer 1992
• Fischer and Vemuru 1991
• Fischer et al 1983
• Fischler 1986
• Fischler et al 1989
• Fisher and Dougherty 1982
• Fisher and Dougherty 1982
• Fisher and Fischer 1987
• Fisher and Peterson 1978
• Flax et al 1953
• Flegel 2021
• Flight survey of the 757 wing noise field and its effects on laminar boundary layer transition. Volume 1 Program description and data analysis 1987
• Flores et al 1991
• Frederick et al 2014
• Freeman 1935
• Freymuth 1966
• Friederich and Kloker 2011
• Frost and Rutherford 1963
• Fröhler et al 2025
• Fuel efficiency and laminar flow control
• Fuhrmann 1993
• Fujii and Hornung 2003
• Fujino 2002
• Fujino et al 2003
• Fukusako et al 1971
• Gad-El-hak and Blackwelder 1987
• Gan 2022
• Gan 2023
• Gao et al 2019
• Gaponov and Smorodsky 2000
• Garaev 2017
• Garbaruk et al 2021
• Garcia et al 1995
• Garzon and Matisheck 2012
• Garzon et al 2014
• Gatti and Quadrio 2016
• Gatti et al 2025
• Gaudet 1990
• Geng et al 2021
• George et al 1992
• George-Falvy 1978
• Gershbein 1970
• Gershbein 1975
• Gibbings 1959
• Gibson and Gerhardt 1995
• Gilman and Burdges 1967
• Glauser et al 2014
• Glezer et al 1986
• Gogineni 2014
• Goglia and Wilkinson 1980
• Goldstein and Sescu 2008
• Goodyear 1987
• Gopalarathnam and Selig 2001
• Gordon et al 2024
• Gortler 1957
• Govindarajan and Narasimha 2000
• Gowree and Atkin 2017
• Gowree et al 2019
• Grabe et al 2018
• Granville 1960
• Granville 1973
• Gratzer 1990
• Gratzer and George-Falvy 1978
• Gray and Maus 1981
• Green 2008
• Gregory 1960
• Gregory 1961
• Gribben 1959
• Grigor’ev and Ershov 2019
• Groot et al 2026
• Groth 1961
• Guha and Sengupta 2025
• Guo and Kloker 2020
• Guo et al 2020
• Guo et al 2022
• Gurevich et al 2019
• Guynn 2011
• Görtler and Velte 1967
• Görtler’s new series method for steady laminar boundary layer flows with suction 1962
• Haans et al 2004
• Haggerty et al 2019
• Haines 1957
• Hale 1976
• Hall 1992
• Hall and Mackerrell 1988
• Hall and Malik 1986
• Hall and Smith 1988
• Hall and Smith 1989
• Hall et al 1984
• Hammachi
• Hammer and Shumway 2026
• Han and Cao 2019
• Han et al 2016
• Han et al 2018
• Hancock 1971
• Hancock 1972
• Hanks 1969
• Hansman and Barsotti 1985
• Harris 1964
• Harris and Brooks 1988
• Harris et al 1988
• Hart et al 2026
• Harvey 1975
• Harvey 1982
• Harvey and Pride 1982
• Harvey et al 1986
• Harvey et al 1989
• Hasegawa et al 2021
• Hasimoto 1957
• Hastings et al 1986
• Hastings et al 1987
• Hayashi and Yamamoto 2014
• Haynes and Reed 1996
• He et al 2018
• He et al 2019
• Head 1961
• Hebbar and Paranjpe 1969
• Heeg
• Heeg and Geurts
• Hefner 1992
• Hefner and Bushnell 1979
• Hefner and Bushnell 1980
• Hefner and Sabo 1987
• Hefner and Wagner 1986
• Hein et al 2009
• Hein et al 2018
• Helm et al 2023
• Henderson and Maddalon 1967
• Hendrikse
• Henningson et al 1987
• Henningson et al 1988
• Herbert et al 1996
• High Efficiency from Wing Laminar Flow 1985
• High Reynolds Number Hybrid 1999, High Reynolds Number Hybrid Lamin
• High Reynolds Number Hybrid 1999, High Reynolds Number Hybrid Lamin
• Hildebrand et al 2020
• Hiller et al 2025
• Hoksbergen et al 2023
• Hoksbergen et al 2023
• Holden and Kolly 1995
• Hollis 2017
• Hollis 2020
• Holloway and Sterrett 1963
• Holman and Fürst 2021
• Holmes 1984
• Holmes 1985
• Holmes 1986
• Holmes 1991
• Holmes and Obara 1982
• Holmes and Obara 1992
• Holmes et al 1983
• Holmes et al 1984
• Holmes et al 1985
• Holmes et al 1986
• Holmes et al 1986
• Holt 1943
• Hopkins and Keener 1968
• Hopkins and Keener 1972
• Hopkins et al 1960
• Hopkins et al 1970
• Hopkins et al 1972
• Hornung 2002
• Horstmann 2008
• Horstmann et al 1990
• Horstmann et al 1990
• Hosseini et al 2013
• Hosseini et al 2015
• Hosseinverdi and Fasel 2018
• Hosseinverdi and Fasel 2023
• Hou and Choy 2022
• Howard et al 1975
• Howell 1970
• Hsu 1989
• Hu and Yao 2023
• Hu and Yao 2024
• Hu et al 2021
• Hu et al 2022
• Huang 2021
• Hughes and Reid 1965
• Hunt and Saric 2011
• Hwang and Biesiadny 1998
• Hybrid laminar flow control 1982
• Hybrid laminar flow flight tests underway 1999
• Hybrid Suction Surface for Turbulent Boundary Layer Flow 1980
• Iannuzzi 1968
• Ide et al 2012
• Improving Wind Turbine Power with Boundary Layer Suction 2025
• Inasawa et al 2024
• Influence of Environment in Laminar Boundary Layer Control 1980
• Introduction to the Special Section on Hypersonic Boundary-Layer Transition Prediction 2019
• Investigation of Laminar Flow of Fluid with One Porous Bounding Wall 2016
• Ishida et al 2026
• Ishikawa et al 2017
• Itoh 2000
• Itoh 2000
• Iuliano et al 2011
• Iyer and Mahesh 2013
• Iyer and Spall 1991
• Iyer et al 1992
• Jahanbakhshi and Zaki 2019
• James 1959
• Janke 2001
• Jasperson et al 1984
• Jasperson et al 1984
• Jensen et al 2008
• Jentys and Breitsamter 2023
• Jentys et al 2022
• Jernell 1978
• Jiang 2026
• Jillie and Hopkins 1961
• Jisheng and Heng 1998
• Jobe et al 1978
• Jobe et al 1979
• Johnson 1988
• Johnson et al 1972
• Johnston 1970
• Jones et al 1989
• Joseph and Diwan 2022
• Joseph et al 2022
• Joslin 1995
• Joslin 1996
• Joslin 1996
• Joslin 1998
• Joslin 1998
• Joslin 2004
• Joslin and Anders 2008
• Joslin and Streett 1992
• Joussot et al 2013
• Juillen and Arnal 1995
• Kachanov and Tararykin 1990
• Kakkar et al 2025
• Kakkar et al 2026
• Kakkar et al 2026
• Kalarikovilagam Srinivasan and Bertram 2019
• Kanda 2019
• Kandil and El-Hady 1988
• Kaneko et al 2026
• Kanner and Schetz 1999
• Kapustin et al 2022
• Karpuk and Elham 2023
• Karpuk and Mosca 2024
• Kashaev et al 2026
• Kaups and Cebeci 1977
• Kazakov 1990
• Kazakov 1995
• Kazakov 2002
• Kegerise et al 2010
• Kehayas 2023
• Keith and Dewitt 1993
• Keller 2019
• Keller and Kloker 2015
• Kelnhofer 1969
• Kendall 1982
• Kerho and Bragg 1995
• Kerschen 1987
• Khan et al 1982
• Khanjari et al 2023
• Khotyanovsky and Kudryavtsev 2016
• Khusnutdinova 1972
• Kiley and Narsipur 2024
• Kim et al 2025
• Kimmel et al 2015
• Kinney 1967
• Kirilovskiy et al 2019
• Kitzinger et al 2023
• Klebanoff 1966
• Klebanoff and Tidstrom 1972
• Kline et al 2019
• Klineberg 1978
• Klineberg 1979
• Kloker 2008
• Klotzsche et al 1984
• Kobashi and Hayakawa 1980
• Kobayashi 1972
• Kobayashi 1974
• Koh and Hartnett 1961
• Kohama et al 2000
• Kok and Young 2014
• Kok et al 2013
• Kok et al 2014
• Kok et al 2015
• Kok et al 2017
• Kolesar 1987
• Komar 1966
• Koo and Kang 2020
• Korkegi 1956
• Korotkin 1966
• Korotkin 1966
• Kosarev et al 2016
• Kosin 1965
• Kotvitskii et al 2025
• Kousik Kumaar. R and Dr. P. Maniiarasan 2015
• Kozlov 1965
• Kozlov 1966
• Kozlov 1968
• Kozlov and Tsyganyuk 1969
• Kozlov et al 2009
• Krishnan and Sandham
• Krishnan et al
• Krishnan et al 2016
• Krishnan et al 2017
• Kroo 2008
• Kroo and Sturdza 2003
• Krumbein et al 2022
• Krzywoblocki 1953
• Kube 2026
• Kubota and Berg 1977
• Kudryavtsev and Khotyanovsky 2019
• Kurz and Kloker 2014
• Kurz and Kloker 2016
• Lachmann 1959
• Lachmann 1962
• Laflin and Papadakis 2001
• LaGraff 2001
• LaGraff 2007
• Lakshminarayana 1962
• Lamb et al 1985
• Lamb et al 1986
• Laminar flow control wing 1999
• Lampropoulos et al 2025
• Landahl 1962
• Landahl 1990
• Landahl and Widnall 1979
• Landahl and Widnall 1980
• Landahl and Widnall 1981
• Landahl and Widnall 1982
• Lange 1983
• Lange 1987
• Lange 1988
• Lange and Gieseler 1953
• Laouer et al 2016
• Lasseigne and Jackson 1992
• Lauchie and Park 2000
• Laufer and Vrebalovich 1960
• Lawson et al 2018
• Lazos and Wilkinson 1988
• Leadon and Bartle 1960
• Lee 1953
• Lee 1954
• Lee and Faget 1956
• Lee and Kim 2017
• Leehey and Shapiro 1980
• Lees 1956
• Lees and Reshotko 1962
• Leitsina and Pavlyukevich 1968
• Lekoudis 1980
• Leonard 1979
• Leroy et al 2016
• Leroy et al 2017
• Leslie M Mack 1984
• Lessen 1953
• Letcher 1972
• Levchenko and Scherbakov 1995
• Leventhal and Reshotko 1981
• Li and Choudhari 2008
• Li and Choudhari 2011
• Li and Luo 2012
• Li and Nagamatsu 1953
• Li and Su 2025
• Li and Zhong 2016
• Li Causi et al 2026
• Li et al 2009
• Li et al 2010
• Li et al 2011
• Li et al 2013
• Li et al 2014
• Li et al 2020
• Li et al 2022
• Li et al 2023
• Li et al 2024
• Li et al 2024
• Li et al 2025
• Liang et al 2022
• Libby and Chen 1970
• Libby and Kassoy 1970
• Lin and Malik 1994
• Lin and Malik 1995
• Lin and Malik 1995
• Lin and Malik 1996
• Lin and Malik 1997
• Lin and Malik 1999
• Lin and Mao 2024
• Lin et al 2011
• Lineberger 1984
• Lipatov and Neiland 1974
• Lipatov and Ustinov 2018
• Liu 1967
• Liu 2019
• Liu et al 2022
• Liu et al 2023
• Liu et al 2026
• Lo 1994
• Lobitz et al 2023
• Lockwood Taylor 1942
• Loftin and Burrows 1949
• Lohse et al 2016
• Lombardi et al 1996
• Lorenzi et al 2011
• Lorenzi et al 2011
• Lovell et al 1978
• Low 1955
• Low 1956
• Low Drag Airfoil Design Utilizing Passive Laminar Flow and Coupled Diffusion Control 1980
• Lozano-Durán et al 2018
• Lu and Luo 2017
• Lu et al 2026
• Lundry 1967
• Luo and Zhou 2000
• Lynde and Campbell 2016
• Lynde and Campbell 2017
• Lynde et al 2021
• Lyons et al 1961
• Lyons et al 1962
• Lysenko 1995
• Lysenko et al 2024
• Lüdeke and Breitenstein 2021
• Ma and Mahesh 2023
• Ma et al 2023
• Ma et al 2025
• Ma et al 2025
• Mabey 1962
• MacDermott et al 1965
• MacDermott et al 1966
• Mack 1967
• Mack 1980
• Mack et al 2009
• MacManus and Eaton 1996
• Maddalon 1982
• Maddalon 1991
• Maddalon and Braslow 1990
• Maddalon and McMillin 1983
• Maddalon and Poppen 1986
• Maddalon and Wagner 1986
• Maddalon et al 1987
• Maddalon et al 1990
• Maddaus and Shanebrook 1983
• Maeda 1968
• Mahfoze and Laizet 2017
• Malik
• Malik 1987
• Malik and Li 1993
• Malik et al 1999
• Malik et al 2007
• Malik et al 2013
• Malik et al 2015
• Malik et al 2024
• Mamada and Ando 1973
• Mamada and Ando 1974
• Mamori and Kondo 2019
• Mangalam et al 1990
• Mangiarotty 1981
• Mankbadi 1992
• Manning and Kroo 1999
• Mansy and Faruque 2023
• Manuel and Doty 1990
• Manuilovich 2017
• Manuilovich 2017
• Manuilovich 2021
• Manuilovich 2023
• Maresh and Bragg 1984
• Marineau 2017
• Marino et al 1975
• Marshall 1999
• Marshall 2000
• Martin 2008
• Martin et al 2008
• Martín and Paredes 2021
• Masad 1996
• Masad and Abid 1996
• Masad and Malik 1994
• Masad and Nayfeh 1992
• Maslennikova and Zelman 1985
• Mason and Gatlin 2015
• Matsuda et al 2020
• Matsui 1963
• Matsui 1980
• Matsuno et al 2020
• Mauerer et al 2026
• Maughmer et al 2018
• Maughmer et al 2025
• Mawignon et al 2023
• Mačák and Fürst 2023
• McCAULEY et al 1966
• McCroskey 1969
• McCullough and Gault 1947
• McCullough and Gault 1948
• McCullough and Gault 1950
• McElderry 1973
• McQuilkin 1982
• Meade 1977
• Medina et al 2009
• Meier and Zhou 1989
• Meitz and Fasel 1995
• Mele 2022
• Mele et al 2020
• Meneghello et al 2015
• Merkle et al 1974
• Mery et al 2025
• Messiha 1966
• Messing and Kloker
• Messing and Kloker
• Messing and Kloker 2004
• Messing and Kloker 2010
• Methel et al 2022
• Methel et al 2025
• Method of the Qualitative Description of the Laminar-Turbulent Transition Position on a Swept Wing in the Flow with an Elevated Degree of Free-Stream Turbulence 2020
• Metkowski and Maughmer 2021
• Metkowski and Maughmer 2023
• Michel et al 1985
• Middleton et al 1985
• Mignanwande 2024
• Mills 1966
• Mills 1971
• Miró Miró 2020
• Mishriky and Walsh 2018
• Mistry et al 2012
• Mochizuki 1961
• Moniripiri et al 2023
• Moniripiri et al 2026
• Montoya et al 1981
• Moralev et al 2020
• Moralev et al 2021
• Moralev et al 2022
• Morduchow 1963
• Morduchow and Reyle 1967
• Morgan and Davies 2020
• Mori et al 2024
• Morita 1971
• Moriya 1972
• Morrisette and Creel 1987
• Morrisette et al 1986
• Mosca and Karpuk 2025
• Mosca et al 2024
• Mossop and Hallett 1974
• Mossop and Ono 1969
• Moulic 1963
• Mount and Millman 1985
• Mousseux et al 1989
• Mughal and Ashworth 2013
• Mullender and Poll 1995
• Muraca 1978
• Murri et al 1987
• Nagamatsu et al 1964
• Nahoui et al 2025
• Nakagawa et al 1978
• Nakagawa et al 2023
• Nangia et al 2003
• Nark 1963
• Nark and Lee 1963
• Nathaniel Hildebrand et al 2025
• Natural laminar flow airfoil 1979
• Naval K Agatwal et al 1992
• Negi et al 2018
• Ness 1961
• Newman et al 1982
• Newman et al 1984
• Neyland and Filippov 2004
• Ng and Crouch 1999
• Nicholson et al 2019
• Nicoll and Ramaprian 1969
• Nie et al 2018
• Nie et al 2022
• Nie et al 2025
• Nishino and Shariff 2009
• Nishizawa et al 2009
• Nitsche and Szodruch 1993
• Niu and Li 2022
• Niu and Su 2023
• Niu et al 2024
• Niu et al 2025
• Nomura 2006
• Norde et al 2018
• Northrop Aircraft Inc Hawthorne Ca 1952
• Northrop Aircraft Inc Hawthorne Ca 1953
• Northrop Aircraft Inc Hawthorne Ca 1956
• Northrop Corp Hawthorne Ca Norair Div 1964
• Northrop Corp Hawthorne Ca Norair Div 1964
• Nouri-Borujerdi and Seyyed-Hashemi 2016
• Novikov et al 2024
• Obara and Holmes 1985
• Obara and Holmes 1986
• Obara et al 1986
• Obara et al 1990
• Obrist and Schmid 2003
• Oliver 2015
• Olivucci et al 2021
• Ortiz-Melendez et al 2020
• Ortiz-Melendez et al 2022
• Osmokrovic et al 2015
• Ostowari 1992
• Oswald 1932
• Owen and Frendi 2018
• Owens et al 2014
• Owens et al 2019
• Palacios et al 2012
• Paluch 2001
• Pan et al 2016
• Pang et al 2024
• Pappas and Okuno 1960
• Paredes and Choudhari 2018
• Paredes et al
• Parikh and Nagel 1990
• Parikh et al 1990
• Parveen and Chowdhury 1970
• Pasamanick and Proterra 1948
• Pascual and Zingg 2026
• Passamanick 1948
• Pate 1964
• Pate 1969
• Pate 1970
• Pate 1971
• Pate 1980
• Pate and Brillhart 1963
• Pate and Deitering 1963
• Pate and Deitering 1963
• Pate and Groth 1966
• Pavlyuchenko et al 1985
• Pearce 1978
• Pearce 1982
• Pearce 1983
• Pearce et al 1984
• Pearson 1967
• Pechau 1963
• Peltzer and Nitsche 2004
• Peng et al 2024
• Perlin et al 2016
• Perraud et al 2003
• Perry and Ansell 2017
• Persh 1956
• Persh 1957
• Peters et al 2024
• Peterson and Fisher 1978
• Peterson and Horton 1959
• Petley et al 1985
• Petzold and Radespiel 2015
• Pfenninger 1949
• Pfenninger 1961
• Pfenninger 1968
• Pfenninger 1981
• Pfenninger 1987
• Pfenninger and Bacon 1961
• Pfenninger and Bacon 1969
• Pfenninger and Groth 1961
• Pfenninger and Syberg 1974
• Pfenninger and Vemuru 1988
• Pfenninger and Vemuru 1988
• Pfenninger and Vemuru 1989
• Pfenninger and Vemuru 1992
• Pfenninger et al 1967
• Pfenninger et al 1980
• Pfenninger et al 1986
• Pfenninger et al 1990
• Phinney and Baltakis 1974
• Pinson and Wang 1999
• Piscitelli et al 2020
• Pisharoti and Brizzolara 2025
• Placek and Ruchała 2018
• Placidi et al 2023
• Plate boundary layer transition 2023
• Pohya and Wicke 2019
• Pohya et al 2019
• Poll 1979
• Poll 1980
• Poll 1981
• Poll 1983
• Poll 2000
• Poll 2007
• Poll and Danks 1995
• Poll and Paisley 1985
• Poots 1965
• Poreh and Hsu 1972
• Potter 1968
• Potter 1975
• Potter and Whitfield 1961
• Potter and Whitfield 1962
• Potter and Whitfield 1964
• Potts et al 2023
• Poulain et al 2024
• Povinelli et al 1976
• Powell 1987
• Powell et al 1989
• Powers 1966
• Powers et al 1963
• Pralits and Hanifi
• Prasannakumar et al 2024
• Prieto Rueda
• Probstein and Elliott 1954
• Prochukhaev 1973
• Pruett et al 2000
• Psarski et al 2022
• Pueschel et al 1996
• Pushkareva 1970
• Pérez et al 2012
• Pérez et al 2015
• Pérez Pérez
• Qadri and Schmid 2017
• Qin and Yu 2020
• Quagliarella and Iuliano 2017
• Quagliarella and Iuliano 2018
• Quaile and Levy 1972
• Quaile and Levy 1975
• Quast et al 1990
• Quinn 1944
• Quinn 1947
• Quinn 1947
• Raayai-Ardakani and McKinley 2019
• Racisz and Quinn 1948
• Radespiel et al 1989
• Radespiel et al 2018
• Radeztsky et al 1992
• Radeztsky et al 1993
• Radeztsky et al 1995
• Radeztsky et al 1999
• Radzetsky et al 1994
• Ram et al 1985
• Ran et al 2020
• Raper 1981
• Rarata et al 2023
• Raspet 1952
• Raspet 1957
• Rastegari and Akhavan 2018
• Reda 1977
• Reda et al 2012
• Redchyts et al 2025
• Redeker and Wichmann 1988
• Redeker et al 1990
• Reed 1985
• Reed 1989
• Reed 1990
• Reed 1991
• Reed and Nayfeh 1981
• Reed and Saric 2002
• Reed and Saric 2013
• Reeh and Tropea 2015
• Reggiori 1972
• Reibert et al 1996
• Reibert et al 1997
• Report No. 349. A proof of the theorem regarding the distribution of lift over the span, for minimum induced drag 1930
• Reshotko 1990
• Review of the effect of distributed surface roughness on boundary-layer transition 1962
• Reynolds et al 1982
• Rhudy 1970
• Ricco et al 2013
• Ricco et al 2021
• Riedel and Sitzmann 1997
• Riedel et al 1998
• Riedel et al 1998
• Ringleb 1952
• Rioual et al 1994
• Risse and Stumpf 2014
• Risse et al 2014
• Rius-Vidales and Kotsonis 2020
• Rius-Vidales and Kotsonis 2022
• Rizzetta and Garmann 2025
• Rizzetta and Visbal 2015
• Rizzetta and Visbal 2015
• Rizzetta and Visbal 2016
• Rizzetta and Visbal 2016
• Rizzetta and Visbal 2016
• Rizzetta et al 2010
• Rizzetta et al 2010
• Rizzo et al 2019
• Roache 1979
• Roberts 1963
• Roberts et al 1966
• Roberts et al 2012
• Roberts et al 2015
• Robison and Loth 2017
• Rogers 1979
• Rogers and King 1984
• Romano et al 2017
• Rosenbaum 1966, Turbulent compressible boundary l
• Rothstein 2010
• Rouviere et al 2023
• Roy and Saikrishnan 2003
• Rubesin and Johnson 1949
• Rudenko and Ryzhkova 1968
• Rugg et al 2021
• Runyan et al 1987
• Russo and Brehm 2023
• Rydalch 2013
• Sabater et al 2022
• Saeed and Graham 2015
• Saeed et al 2010
• Saeed et al 2011
• Safari et al 2025
• Saito et al 2021
• Saito et al 2022
• Sakakibara et al 2021
• Sanders 1965
• Sapuppo and Archer 1982
• Sareen et al 2013
• Saric 1997
• Saric 1998
• Saric 1998
• Saric 2004
• Saric 2008
• Saric 2010
• Saric and Reed 2004
• Saric and Yeates 1985
• Saric et al 2015
• Saric et al 2019
• Satish et al 2017
• Sattarzadeh and Fransson 2016
• Savoshchik and Tokhunts 1974
• Scarano et al 2022
• Schlichting 1949
• Schlichting and Gersten 2000
• Schneider and Beckwith 1992
• Schneider and Collicott 2000
• Schofield and Davey 1967
• Schrauf and von Geyr 2020
• Schrauf and von Geyr 2021
• Schrauf and von Geyr 2021
• Schrauf et al 1998
• Schubauer and Klebanoff 1955
• Schubauer and Skramstad 1947
• Schuelein 2008
• Schueltke and Stumpf 2017
• Schwartzberg and Braslow 1952
• Sears 1974
• Sedney 1955
• Seebaugh and Childs 1970
• Seifert et al 1992
• Selig et al 1995
• Sellars 1955
• Selvanose et al 2024
• Semenov et al 2021
• Sengupta and Dipankar 2005
• Sengupta et al 2021
• Sequeira et al 2023
• Seraudie et al 1999
• Serpieri and Kotsonis 2016
• Sesterhenn and Friedrich 2004
• Sesterhenn and Friedrich 2005
• Sesterhenn and Friedrich 2006
• Shadloo and Hadjadj 2017
• Shahneh 2016
• Shahriari et al 2018
• Shaikh and Gaster 1995
• Shalaev 2018
• Shanahan et al 2015
• Shanebrook and Sumner 1970
• Sharma
• Sharma and Dutta 2022
• Sharma et al 2025
• Sheetz 1965
• Shepherd 2014
• Shi et al 2015
• Shi et al 2020
• Shi et al 2021
• Shi et al 2023
• Shimizu et al 2021
• Shimmy 2026
• Sibert 1943
• Simon and Boyle 1998
• Simon et al 2015
• Simpson et al 1969
• Singer and Joslin 1995
• Singleton 1967
• Sinha and Ravande 2006
• Siochi et al 2013
• Sirovich and Karlsson 1999
• Skripachev 1969
• Skuratov and Fedorov 1991
• Smith 1990
• Smith 1996
• Smith 1996
• Smith 2007
• Smith and Clutter 1959
• Smith and Kroo 1990
• Snyder and Zumwalt 1969
• Sohn and Reshotko 1991
• Sokolov et al 1980
• Soleimani and Eckels 2021
• Soltani et al 2011
• Somers 1992
• Somers 2012
• Somers 2019
• Somers and Maughmer 2022
• Song and Liu 2025
• Song et al 2024
• Song et al 2025
• Spaid 1969
• Spalding and Chi 1963
• Spalding and Chi 1964
• Spall and Malik 1992
• Sparrow et al 1963
• Speer et al 2004
• Spence 1960
• Srilatha et al 1990
• Srokowski 1978
• Stability Theory for Laminar Flow Control Design 1990
• Stalewski and Sznajder 2016
• Steen 1995
• Steffen et al 2024
• Sternberg 1954
• Sternberg 1961
• Stewart and Simpson 2005
• Straussfogel and Maughmer 1991
• Streett 2003
• Streett 2003
• Streit et al 2011
• Streit et al 2015
• Strike and Pate 1961
• Stroh et al 2016
• Struminskii 1967
• Sturdza 2007
• Sturdza and Rajnarayan 2014
• Sturgeon 1978
• Sturgeon 1978
• Sturgeon and Bennett 1976
• Sturgeon et al 1976
• Stüer et al 2000
• Su 2015
• Sudhi et al 2021
• Sudhi et al 2023
• Sullivan et al 2017
• Sullivan et al 2022
• Sun and Du 2026
• Suryanarayanan et al 2017
• Swafford 1979
• Swean and Inger 1974
• Swift and Mungur 1979
• Swigart 1972
• Swinford 1976
• Szulga et al 2015
• Takahashi et al 2019
• Takematsu 1970
• Tan and Lai 2019
• Tang et al 2017
• Tang et al 2022
• Tang et al 2026
• Tani 1952
• Tani 1961
• Tani 1980
• Tani and Hama 1953
• Tani and Sato 1956
• Tao et al 2020
• Tatar et al 2019
• Tatsumi 1952
• Tatsuta et al 2024
• Tempelis and Mishnaevsky 2025
• Terrill 1960
• Terrill 1965
• The determination of local turbulent skin friction from observations in the viscous sub-layer 1960
• Theofilis 1998
• Theofilis and Dallmann 1997
• Theofilis and Poll 1995
• Theofilis et al
• Theofilis et al 1990
• Thiede 1978
• Thomas and Cornelius 1981
• Thomas and Cornelius 1982
• Thomas et al 2017
• Thompson 1970
• Thompson and Emrich 1967
• Three-dimensional boundary-layer transition on a swept wing at Mach 3.5 1994
• Thummar et al 2023
• Thummar et al 2024
• Tokhunts 1968
• Tokugawa and Yoshida 2006
• Tokugawa et al 2008
• Tokugawa et al 2026
• Tolkachev et al 2012
• Tolkachev et al 2012
• Toppings and Yarusevych 2023
• Torda 1952
• Torda 1953
• Tracy et al 2002
• Transonic flight test of a laminar flow leading edge with surface excrescences 1994
• Traub 2026
• Traub 2026
• Traub 2026
• Traub et al 2021
• Trilling 1950
• Tucker et al 2014
• Tucker et al 2014
• Tucker et al 2016
• Tucker et al 2016
• Tufts et al 2014
• Tufts et al 2014
• Turriziani et al 1979
• Tuttle and Maddalon 1982
• Tutty et al 2000
• Twiss et al 2019
• Ueda et al 2014
• Uram 1960
• Ustinov
• Ustinov 2004
• Ustinov 2013
• Ustinov 2016
• Ustinov 2017
• Ustinov 2018
• Ustinov 2018
• Ustinov and Kachanov 2021
• Ustinov et al 2022
• Vaglio-Laurin 1978
• Van Atta and Helland 1980
• van Dam and Holmes 1988
• Van Dam et al 1993
• Van Driest and Blumer 1961
• Van Driest and Blumer 1968
• Van Driest and McCAULEY 1960
• Van Driest et al 1967
• van Oudheusden 2003
• van Rijn 1997
• Vandam 1986
• Vandam 1993
• Venkatachari et al 2021
• Venkatachari et al 2021
• Vermeersch and Bouteiller 2014
• Vermeersch et al 2015
• Vessey 1935
• Viken et al 1990
• Vlachos et al 2007
• Vlahostergios et al 2019
• Von Doenhoff 1941
• Von Doenhoff 1944
• Von Doenhoff and Braslow 1961
• von Kármán and Tsien 1945
• Wadia 1983
• Waggoner et al 1987
• Wagner 1986
• Wagner and Fischer 1983
• Wagner et al 1984
• Wagner et al 1986
• Wagner et al 1988
• Wagner et al 1988
• Wagner et al 1988
• Wagner et al 1989
• Wagner et al 1990
• Wagner et al 1992
• Wahidi and Bridges 2008
• Wahlich et al 2024
• Walsh 1982
• Walsh and Lindemann 1984
• Walsh et al 1986, Combined Devices for Turbulent-Dr
• Wang and Chen 2012
• Wang and Guo 2022
• Wang et al 1994
• Wang et al 2013
• Wang et al 2017
• Wang et al 2018
• Wang et al 2019
• Wang et al 2020
• Wang et al 2023
• Wang et al 2024
• Wang et al 2024
• Wang et al 2025
• Wang et al 2026
• Wang et al 2026
• Warner et al 1980
• Warren et al 1997
• Watmuff 1992
• Watmuff 1994
• Wedin and Cherubini 2016
• Wedin et al 2015
• Wei et al 2013
• Weilmuenster 1974
• Wen 2024
• Wentz et al 1984
• Wentz et al 1985
• Wheldon and Whites 1966
• White 2005
• White and Christoph 1972
• White and Reshotko 2008
• White and Saric 2000
• White et al 1973
• White et al 2001
• Whitfield and Iannuzzi 1969
• Wicke et al 2016
• Wie and Malik 1998
• Wie et al 1991
• Wie et al 1992
• Wilcox 1979
• Wilder et al 2015
• Wilkinson et al 1987
• Wilson 1962
• Wilson 1980
• Winkler 1961
• Winkler and Cha 1959
• Wlezien 1989
• Woan et al 1991
• Wohl et al 2013
• Wohl et al 2015
• Wong et al 2024
• Wong et al 2026
• Wood 1968
• Wright 1979
• Wright and Nelson 2000
• Wrona and Tuliszka-Sznitko 2002
• Wu 1972
• Wu 1973
• Wu and Tulin 1970
• Wu et al 2024
• Wu et al 2025
• Wu et al 2026
• Wu et al 2026
• Wuest 1952
• Wuest 1961
• Wygnanski 1977
• Wylie et al 2021
• Xi and Fu 2026
• Xi et al 2021
• Xi et al 2025
• Xia and Chen 2016
• Xiao and Papadakis 2017
• Xiao et al 2019
• Xiong et al 2018
• Xu and Kroo 2014
• Xu and Wu 2022
• Xu et al 2018
• Xu et al 2019
• Xu et al 2019
• Xu et al 2021
• Xu et al 2021
• Xu et al 2024
• Yadala et al 2018
• Yadala Venkata et al 2018
• Yakeno and Obayashi 2021
• Yalcin et al 2021
• Yamaguchi et al 2020
• Yan et al 2023
• Yan et al 2025
• Yan et al 2025
• Yanagimoto and Rinoie 2019
• Yang and Hu 2008
• Yang and Huano 1969
• Yang et al 2019
• Yang et al 2021
• Yang et al 2022
• Yang et al 2022
• Yang et al 2022
• Yang et al 2023
• Yang et al 2025
• Yang et al 2026
• Yao et al 2021
• Yasuaki and Egami 1999
• Yates et al 2020
• Yatskih et al 2016
• Ye 2025
• Ye et al 2019
• Yermolaev et al 2016
• Yih et al 1952
• Yonezawa et al 2017
• Yoshida and Tani 1993
• Yoshizawa 1969
• Young 1949
• Young 1965
• Young 2012
• Young and Booth 1951
• Young and Fielding 2001
• Young and Fielding 2001
• Young et al 2001
• Young et al 2003
• Youngren 2008
• Yu 2017
• Yu et al 2026
• Yuan 1955
• Zahn and Rist 2016
• Zakharin and Wygnanski 2008
• Zaki 2014
• Zalovcik 1945
• Zamir and Young 1967
• Zare Shahnehb 2022
• Zauner and Sandham 2020
• Zauner et al 2023
• Ze-yan 1981
• Zeng et al 2021
• Zeng et al 2024
• Zhang and Deng 2025
• Zhang and Yin 2019
• Zhang et al 2015
• Zhang et al 2017
• Zhang et al 2020
• Zhang et al 2022
• Zhang et al 2023
• Zhang et al 2023
• Zhang et al 2025
• Zhang et al 2026
• Zhao and Dong 2020
• Zhao et al 2014
• Zhao et al 2015
• Zhao et al 2017
• Zhao et al 2018
• Zheng et al 2026
• Zhigulev 1985
• Zhong 2009
• Zhong and Li 2017
• Zhou et al 2015
• Zhou et al 2018
• Zhou et al 2020
• Zhou et al 2022
• Zhou et al 2024
• Zhu et al 2019
• Zolotov and Khodorkovsky 1963
• Zong et al 2022
• Zoppini et al 2022
• Zoppini et al 2022
• Zoppini et al 2023
• Zore et al 2023
• Zou and Zhong 2024
• Евстигнеев and Рябков 2016