Transonic & Subsonic Aero Review

Maya: Welcome to Transonic & Subsonic Aero Review. I’m Maya. Ethan: And I’m Ethan. Episode one is the question that gets waved away until it absolutely cannot be waved away anymore: why does compressibility matter? Maya: Right. Air is a fluid, but it is not always safe to treat it as a constant-density fluid. The practical question is not, “Can air be compressed?” Of course it can. The question is, “Will density changes be large enough to change the answer we care about?” Ethan: Lift, drag, pressure loads, control effectiveness, stall behavior... the stuff that ruins your flight-test day. Maya: Exactly. And the organizing parameter is Mach number, M, the flow speed divided by the local speed of sound. Ethan: So Mach is not just a speedometer label. It tells us whether pressure disturbances can communicate through the flow quickly enough for density changes to stay modest. Maya: That’s a good physical picture. At low Mach number, pressure information moves through the air much faster than the aircraft moves through it. The flow can adjust with relatively small density variation. As Mach rises, that adjustment becomes less gentle, and compressibility enters the aerodynamic result. Ethan: Let’s put a boundary on it, because “low” is doing a lot of work there. Maya: A common engineering rule is that below about Mach 0.3, incompressible flow is often an adequate approximation, provided you do not need extremely fine accuracy. From roughly Mach 0.3 up toward Mach 0.8, we are still subsonic, but compressibility effects increasingly matter. Ethan: And then around Mach 0.8, things start getting interesting in the bad sense. Maya: Often, yes. The transonic regime is roughly Mach 0.8 to 1.2. Even if the free stream is below Mach 1, air accelerates around a wing, inlet, or other curved surface. Local regions can reach much higher Mach number than the aircraft’s free-stream Mach number. Ethan: Like driving at seventy on a highway, but the air over the top of the wing has found a shortcut and is doing ninety. Maya: A loose analogy, but yes. Local acceleration matters. In transonic flow, subsonic and supersonic regions can coexist. Significant density changes, shocks, and flow separation can alter lift and drag substantially. Ethan: So if someone says, “We’re subsonic, so compressibility is negligible,” your answer is... Maya: “Subsonic” is too broad a category for that conclusion. Mach 0.15 and Mach 0.75 are both subsonic, but they are not aerodynamically equivalent. Ethan: Good. That’s the boundary I want fixed in my head. Low-Mach subsonic can often use incompressible thinking. High-subsonic needs a compressibility check. Near Mach 0.8, you should be very suspicious of constant-density assumptions. Maya: Precisely. Ethan: Okay, equations. What changes when we stop pretending density is constant? Maya: Start with continuity, conservation of mass. In compressible form, it says the time rate of density change plus the divergence of mass flux is zero. Spoken less formally: if more mass enters a small volume than leaves it, density rises. Ethan: Whereas in the standard incompressible shortcut... Maya: We set density approximately constant. Then continuity reduces to divergence of velocity equal to zero. The velocity field may change direction and magnitude, but the fluid does not significantly pile up or thin out. Ethan: That’s the familiar “what goes in must come out” picture from a wind-tunnel streamtube. Maya: Yes. But for compressible flow, velocity and density can both change. If a passage narrows, the fluid can accelerate, but it can also change density. So geometry alone no longer tells the whole story. Ethan: And momentum? Maya: The Navier, Stokes momentum equation still expresses Newton’s second law for fluid motion. Pressure forces, viscous stresses, and acceleration are all there. But density now varies, so inertia varies too. You cannot solve momentum cleanly without knowing how pressure, density, and temperature relate. Ethan: Which brings in the energy equation. Maya: Exactly. For an ideal-gas model, pressure depends on density and temperature. Temperature affects the speed of sound. The speed of sound affects Mach number. And Mach number tells us whether compressibility is important. So continuity, momentum, and energy are coupled. Ethan: Let me try the flight-test version. In incompressible flow, I can often think, “Pressure pushes the air around, and density mostly stays out of the argument.” In compressible flow, pressure changes also squeeze or expand the air, that changes temperature and density, and those changes feed back into the motion. Maya: That is the essential loop. And it is why compressible aerodynamics is not just incompressible aerodynamics with a different density value plugged in. Ethan: Quick worked check. We have an aircraft or a test article moving at 250 meters per second at sea level. Do we need compressible treatment? Maya: Use Mach number. Near sea level, the speed of sound is roughly 340 meters per second. So M is about 250 divided by 340, which is about 0.74. Ethan: Well above the Mach 0.3 rule of thumb. Maya: Yes. At Mach 0.74, incompressible treatment is not a safe default for aerodynamic prediction. You should account for compressibility, especially for pressure distributions, lift and drag estimates, and any geometry where the local flow accelerates further. Ethan: Because that 0.74 is only the free-stream number. A wing’s upper surface could push local Mach toward the transonic range. Maya: Correct. The practical answer is not merely, “Compressibility exists.” It is, “The flow is high enough in Mach number that ignoring density variation can materially distort the result.” Ethan: One caveat, though. If I see buffet or unsteady behavior, I shouldn’t automatically yell, “Compressibility!” Maya: Absolutely. Boundary-layer transition, free-stream turbulence, separation, and structural effects can all matter. Transonic buffet can even occur at subsonic speeds as low as Mach 0.3, so Mach alone does not diagnose every unsteady phenomenon. Ethan: But it tells you when the constant-density mental model is starting to expire. Maya: Well put. Below about Mach 0.3, incompressible flow is often a useful first model. By high subsonic Mach numbers, check compressibility explicitly. And as the flow approaches transonic conditions, treat density changes as central to the problem. Ethan: That’s episode one. Next time, we’ll take that Mach-number framework and ask what pressure really does around an airfoil. Maya: Thanks for listening. See you next time.

Maya: Last time, in “Why Compressibility Matters,” we defined compressibility as density responding to pressure, set the Mach regimes, and started from continuity, momentum, and energy. Today, uh, we make that useful below Mach one. Ethan: Right, because “subsonic” can sound like, you know, the easy setting. Smooth streamlines, no shocks, life is good. But then you run a Mach 0.6 test point and the incompressible numbers start looking... a little too confident. Maya: Exactly. Subsonic means \(M\) less than one. It does not mean density is constant. The practical question is, when is constant density accurate enough for the decision we need to make? Ethan: And the old rule is Mach 0.3, right? Below that, treat air as incompressible unless you have a reason not to. Maya: That’s the rule of thumb, yes. Not a physical cliff. At Mach 0.3, compressibility corrections are often small for preliminary loads, low-speed tunnel work, or a light aircraft cruise estimate. Above about 0.3, density changes increasingly affect pressure, velocity, and instrument interpretation. Ethan: So let’s rebuild Bernoulli without pretending density is frozen. Maya: Good. For steady, inviscid flow along a streamline, the momentum equation gives \(V\,dV + dp/\rho = 0\). For incompressible flow, rho is constant, so integration gives \(p + \tfrac{1}{2}\rho V^2\) equals a constant. Ethan: The familiar pressure-plus-dynamic-pressure budget. Maya: Yes. But for compressible flow, we keep the density inside the integral. Then the result is \(h + V^2/2\) equals a constant, for adiabatic, no-work flow. That is the compressible Bernoulli idea, or more precisely the steady-flow energy equation. Ethan: So instead of pressure directly trading against velocity, it’s enthalpy trading against velocity. Maya: Exactly. Velocity rises, static enthalpy falls. For a calorically perfect gas, \(h=c_pT\), so static temperature falls too. And because pressure and density are tied to temperature through the equation of state, density changes with it. Ethan: Like an airplane accelerating through a shallow descent in altitude, except this happens inside the flow itself. The kinetic-energy account gets funded by the thermal account. Maya: That’s a fair analogy. For isentropic flow, the compact relation is \(T_0/T = 1 + (\gamma-1)M^2/2\). For air, gamma is about 1.4, so that becomes \(T_0/T = 1 + 0.2M^2\). Ethan: And stagnation quantities, just to be clear, are what you’d get if you slowed the flow to rest reversibly. Maya: Right. No shock, no dissipative loss, just an ideal deceleration. Now, here’s a useful Mach 0.6 estimate. Suppose a streamline accelerates locally from Mach 0.6 to Mach 0.7, with essentially constant stagnation conditions. Ethan: Like flow speeding up over the upper surface, not a whole airplane changing flight condition. Maya: Precisely. At Mach 0.6, the temperature factor is \(1 + 0.2 times 0.6 squared\), which is 1.072. At Mach 0.7, it’s 1.098. Ethan: Small-looking numbers. That’s usually where engineers get ambushed. Maya: Uh, yes. Using the isentropic density relation, density scales as that factor to the minus 2.5 power. So the density ratio from Mach 0.6 to 0.7 is roughly \((1.098/1.072)^{-2.5}\), about 0.94. Ethan: So, about a 6% density drop over that local acceleration. Maya: Correct. That’s right in the range where “constant density” has become a visible approximation, not an invisible one. Ethan: Now let’s connect that to dynamic pressure, because people see \(q=\tfrac{1}{2}\rho V^2\) and can accidentally treat rho as a label from the freestream. Maya: Good distinction. The local dynamic-pressure-like quantity scales with local rho times local velocity squared. Going from Mach 0.6 to 0.7, velocity rises by about 17%, so velocity squared rises about 36%. Ethan: If I held density fixed, I’d say local q rose 36%. Maya: But density fell about 6%. So the actual local rho-\(V^2\) product rises by roughly \(0.94 times 1.36\), or about 1.28. That’s a 28% increase, not 36%. Ethan: That’s a nice flight-test sanity check. If your pressure interpretation assumes freestream density everywhere over a fast wing section, you can be several percent off in the local kinetic term. Maya: Yes, with one caveat. Aerodynamic coefficients conventionally use freestream dynamic pressure, \(q_\infty\), and that is still a useful reference quantity. The mistake is not using \(q_\infty\). The mistake is assuming the local flow state obeys incompressible pressure-velocity relations. Ethan: So the pressure coefficient story changes too. Maya: It does. In incompressible flow, more local velocity means lower static pressure through the simple Bernoulli form. In compressible subsonic flow, the direction of that tradeoff remains, but the magnitude changes because density and temperature participate. The pressure response is stronger than incompressible theory predicts as Mach rises. Ethan: Which is why, at Mach 0.6, I would not use incompressible panel intuition as the final answer, even if there are no shocks and the whole flow is subsonic. Maya: Exactly. Use it for a first sketch, maybe. Then apply a compressibility correction or use a compressible analysis. Anderson is a solid derivation source, and FAA material on compressibility corrections is useful for the practical instrument side. Ethan: The practical rule, then, is, below roughly Mach 0.3, incompressible is often fine. Around 0.6, density changes of several percent can alter local pressure and q interpretations enough to matter. Maya: Nicely put. Next time, we’ll turn that pressure-velocity intuition into pressure coefficients and lift, uh, where the engineering bookkeeping gets more interesting. Ethan: And slightly less forgiving. See you then.

Maya: Last time, in “Subsonic Flow Essentials,” we translated Bernoulli into compressible subsonic flow and pulled out practical rules. Ethan: And today those rules start showing their limits. You can be flying well below Mach 1, look at the freestream Mach number, and still have a local patch of supersonic flow sitting on the wing. Maya: Exactly. Transonic flow is usually the broad regime around Mach 0.8 to 1.2, where subsonic and supersonic regions can coexist. But the important physical transition can begin below Mach 0.8 for a particular airfoil, especially one with appreciable thickness, camber, or lift. Ethan: So the airplane does not need to be at Mach 1 for the wing to encounter Mach 1. Maya: Right. The freestream Mach number is an upstream condition. The airfoil accelerates the flow, especially over the upper surface. If the local velocity rises enough, the local Mach number can reach one even while the surrounding flow remains subsonic. Ethan: Like a local speed peak in a flight-test trace. The average vehicle condition looks benign, but one sensor location is already telling you the interesting part is happening. Maya: That is a good analogy. For an airfoil, the first freestream Mach number at which any point on the surface reaches local Mach one is the critical Mach number, usually written M-critical. Ethan: And that is not automatically the drag-divergence Mach number. Maya: Correct. Critical Mach marks first local sonic flow. Drag divergence is a later practical threshold, often defined by a specified rapid increase in drag. They are related, but they are not interchangeable. Ethan: So if I am looking at a pressure plot, what changes at critical Mach? Maya: Before critical Mach, the upper-surface pressure distribution is smooth. Pressure drops as the flow accelerates over the forward part of the airfoil, then recovers downstream. As freestream Mach rises, compressibility intensifies that acceleration. Eventually, a small sonic point appears near the suction peak. Ethan: And then a supersonic pocket forms just downstream of it? Maya: Yes. The flow can accelerate through Mach one into a local supersonic region, then it has to return to subsonic flow farther aft. In transonic flow, that recompression commonly occurs through a shock on the surface. Ethan: Which is the part that makes pilots, test engineers, and performance teams pay attention. Maya: Very much so. Across the shock, pressure rises abruptly. On a pressure-coefficient plot, you see a sharp recovery rather than a gradual one. That abrupt rise is an adverse pressure gradient for the boundary layer. Ethan: So the boundary layer has already spent its momentum climbing the suction side, then it gets asked to recover pressure very quickly. Maya: Precisely. If it cannot tolerate that recovery, it thickens or separates. The result can be increased drag, altered lift and pitching moment, and, in some conditions, buffet. Even before major separation, the shock itself creates wave drag because the compression is irreversible. Ethan: The practical symptom is often a drag-polar knee. You do not need a dramatic change in angle of attack. A modest Mach increase can suddenly cost you a lot of thrust. Maya: Yes. And that is why critical Mach is useful as an early warning metric. It tells you when the assumptions behind simple, smoothly compressible subsonic flow are about to fail locally. Ethan: Let’s put a number on that. Not a design prediction, just a quick engineering estimate. Maya: Good. Take a representative cambered airfoil at a moderate lift coefficient. Suppose its incompressible minimum pressure coefficient on the upper surface is about minus 0.7. Ethan: Meaning the strongest suction peak, before we apply compressibility correction. Maya: Right. The simple Prandtl-Glauert correction says, approximately, that the compressible pressure coefficient is the incompressible value divided by the square root of one minus M-infinity squared. So our estimate is C-p-min equal to minus 0.7 over square root of one minus M-infinity squared. Ethan: As Mach rises, that denominator shrinks, so the predicted suction peak becomes more negative. Maya: Exactly. At Mach 0.65, the square root term is about 0.76. That gives a corrected minimum C-p near minus 0.92. Ethan: But how do we know whether minus 0.92 means local sonic flow? Maya: We compare it with the pressure coefficient required for a local Mach number of one, using the standard isentropic pressure relation. At Mach 0.65, that sonic pressure coefficient is roughly minus 1.01. Our predicted suction peak, minus 0.92, is not yet low enough. So, approximately, no sonic pocket. Ethan: Then try Mach 0.67. Maya: At Mach 0.67, the Prandtl-Glauert estimate gives about minus 0.94. The sonic pressure coefficient is about minus 0.91. Now the predicted suction is slightly stronger than the sonic condition. Ethan: So the crossing is around Mach 0.66 to 0.67. That is our estimated critical Mach. Maya: Yes, for this deliberately simple example. Physically, it says this airfoil’s upper-surface acceleration can produce local Mach one while the aircraft is still only around Mach 0.66. A different thickness, camber, lift coefficient, or surface pressure distribution could shift that substantially. Ethan: And the estimate becomes least trustworthy exactly where we want it most, because Prandtl-Glauert assumes small disturbances and no shocks. Maya: That is the key limitation. It is useful for trend and first-cut sizing, not for certifying a transonic pressure distribution. Near critical Mach, nonlinear compressibility matters. Once shocks appear, the linear correction has exhausted its jurisdiction. Ethan: So the rule of thumb is: use the estimate to identify risk early, then verify with wind-tunnel data, higher-fidelity analysis, and eventually flight test. Maya: Exactly. And from a design perspective, the response is to manage the peak acceleration and the pressure recovery. Supercritical airfoils are one well-known route. For whole aircraft, shaping the volume distribution also matters. Ethan: Next time, we can stay with what happens after that first sonic pocket appears, because the shock is really where the clean subsonic story stops being clean. Maya: Agreed. See you then.

Maya: Last time, in Entering Transonic Regimes, we got to critical Mach and that slightly unnerving fact that a mostly subsonic airplane can carry little supersonic pockets. Today, uh, we ask what those pockets do when geometry tells them to compress or expand. Ethan: Right, because on a pressure plot it can look almost rude. Smooth acceleration, smooth acceleration, then, bang, a sharp pressure rise. And the practical question is, is that a shock, how strong is it, and is it about to make my boundary layer unhappy? Maya: Exactly. Start with the distinction. A shock occurs when supersonic flow is compressed. The change is abrupt: pressure rises, temperature rises, density rises, and velocity falls. Ethan: So, airplane-engineer version, the flow runs into a compression it cannot politely communicate upstream through the supersonic region. It has to reorganize very quickly. Maya: Yes, with one important qualifier. In subsonic flow, disturbances can propagate in all directions, including upstream. In supersonic flow, they cannot travel upstream against the flow. So compression waves pile together into a shock. Ethan: And an expansion is the opposite geometry, but not the opposite kind of jump. Maya: Correct. When supersonic flow turns away from itself, around a convex corner, say, it expands through a Prandtl-Meyer expansion fan. That process is continuous, spread across many weak waves. Pressure falls, temperature falls, and velocity rises. Ethan: I like the hand-test analogy. Imagine streamlines as lanes of traffic. Turn the lanes toward each other, they bunch up, pressure goes up, shock. Turn them apart, they spread out, pressure drops, and the flow accelerates. Maya: Good analogy, as long as we remember compressibility is doing the important work. Now, a short number check. Take a normal shock with upstream Mach number about 1.2. Assume air with gamma equal to 1.4, the usual simple engineering approximation. Ethan: This is the hypothetical shock standing perpendicular to the incoming flow, right? Not necessarily what sits on a swept wing. Maya: Right, it is the cleanest calculation. The normal-shock pressure ratio is Maya: p two over p one equals one plus two gamma over gamma plus one, times M one squared minus one. Maya: Put in 1.2. Mach squared is 1.44, so the excess over one is 0.44. Two gamma over gamma plus one is 2.8 over 2.4, about 1.17. Multiply 1.17 by 0.44, roughly 0.51. Add one. Maya: So p two over p one is about 1.5. Ethan: Meaning, approximately a fifty-percent static-pressure rise across a normal shock at Mach 1.2. That is, um, not a tiny disturbance. Maya: No. And notice the scaling. Mach 1.2 does not sound dramatically above one, but shock strength grows quickly enough that even these local transonic pockets can produce consequential pressure jumps. Ethan: Which is why a wing at a freestream Mach below one can still have a shock on its upper surface. The local flow accelerates over the curved surface, crosses Mach one in a pocket, then has to get compressed back toward the downstream condition. Maya: Precisely. And usually the shock is oblique, not normal. Here is the verbal sketch. Draw flow moving left to right. Now put a surface below it that turns slightly upward into the flow, like a shallow ramp. From the corner, draw a slanted line leaning downstream. That line is the oblique shock. Ethan: So the incoming flow meets the shock at an angle, then downstream the flow has turned to follow the ramp. It is not hitting a flat wall head-on. Maya: Yes. For the shock physics, the component of Mach number normal to that slanted shock controls the compression. A more nearly normal shock has a larger normal component and tends to be stronger. A more swept, more oblique shock can be weaker, all else equal. Ethan: That gives a useful flight-test instinct. If the pressure trace shows a very abrupt recovery over a short chord distance, I should worry less about the label and more about the local compression strength and what sits just downstream. Maya: Exactly. The downstream boundary layer is often the practical issue. A shock imposes a rapid pressure rise. The near-wall flow has low momentum compared with the outer flow, so it can struggle against that adverse pressure change. Separation can follow, which increases drag and can change forces and moments. Ethan: And then stability enters because the shock is not nailed to one spot. Change angle of attack, Mach number, or loading a little, and the shock can move. If it moves over a sensitive part of the wing or tail, the pressure distribution changes with it. Maya: Right. Be careful not to claim every shock causes separation or instability. It does not. But shock strength and shock-boundary-layer interaction are central design concerns because they can affect drag, separation, and stability. Ethan: Where do expansion fans join this picture? Because in a real transonic flow, I do not just see one compression event. Maya: You often see a sequence. Flow accelerates over a curved or convex region, expands and cools as its pressure falls, then later encounters geometry or a required pressure recovery that compresses it through a shock. The fan and shock are linked by the geometry and by the need to match the overall flow field. Ethan: So, uh, a fan can make the later shock problem worse if it accelerates the local flow more before that compression has to happen. Maya: That is a fair practical statement. Stronger local acceleration can create a higher local Mach number, and subsequent compression can then be more severe. Conversely, shaping geometry to avoid excessive local acceleration can help manage shock strength. Ethan: Same logic in transonic machinery, then. If the flow passages create strong expansions followed by strong compressions, performance suffers, and the boundary layer may be the first thing to complain. Maya: Yes. The rule of thumb is not “eliminate every shock,” because transonic configurations may have them. It is, manage the local Mach peaks, avoid unnecessarily strong compression, and pay close attention to where a shock meets a boundary layer. Ethan: And if you want the formal version after this, Anderson, standard compressible-flow texts, and the old NACA shock notes are still good places to check the relations rather than trusting a napkin calculation. Maya: Next time, we will stay with the practical consequences and look more directly at wave drag, pressure recovery, and why small geometry changes can matter so much near Mach one. Ethan: Yep, bring a pencil, and maybe a very suspicious attitude toward any “small” transonic pressure jump.

Maya: Last time, in “Shocks and Expansion Fans in Practice,” we followed how those features form and interact in transonic flow. Today, let’s put one on an airfoil and ask the design question: when Mach rises, why can the same airfoil suddenly feel like a different airfoil? Ethan: Right. Same shape, same angle of attack, same tunnel model... and then the pressure taps start telling a less comfortable story. Maya: Exactly. At low Mach, roughly below 0.3, treating air as incompressible is usually adequate. The pressure field responds to geometry and angle of attack, but density change is small enough that our familiar incompressible picture works. Ethan: Like a slow wind-tunnel run. You increase angle, suction gets stronger on the upper surface, pressure stays higher underneath, and lift follows. Maya: Yes. In compact form, lift comes from integrating the pressure difference over the surface. More upper-surface suction, or more lower-surface pressure, means more lift. As freestream Mach increases, though, density changes within that accelerating flow become important. Ethan: And the key is that the air over the top is not moving at freestream speed. Maya: Precisely. The airfoil accelerates the flow locally. On the upper surface, especially near the leading edge, local velocity can be substantially above the freestream value. So a freestream Mach that sounds comfortably subsonic can still produce a local Mach of one somewhere on the airfoil. Ethan: This is one of those flight-test traps. The airplane says, “I’m below Mach one.” The wing says, “Well, one small patch of me is not.” Maya: Nicely put. Before that local sonic condition appears, compressibility tends to amplify the pressure response. The classic Prandtl, Glauert idea says, very roughly, that an incompressible pressure coefficient is strengthened by a factor involving one over the square root of one minus freestream Mach squared. Ethan: So as Mach rises, the same geometric turning produces a stronger pressure signal. Maya: That is the basic trend, yes, within its range of usefulness. At fixed angle of attack, lift slope can increase. But the correction is not a license to extrapolate indefinitely. Its singular behavior near Mach one is a warning that the underlying smooth, fully subsonic assumption is breaking down. Ethan: In practical terms, the pressure trace gets deeper on the upper surface. More suction. More lift... until it doesn’t behave politely anymore. Maya: Exactly. The important limit is the critical Mach number. That is the freestream Mach at which the first point on the airfoil reaches local sonic speed. Above it, a small locally supersonic pocket can form, usually on the upper surface. Ethan: And then that pocket has to get back to the downstream subsonic flow somehow. Maya: Through a shock. The shock compresses and decelerates the flow abruptly. On a pressure plot, instead of a gradual recovery from the suction peak, you see a sharp pressure rise. That abrupt recovery produces shock-induced drag rise, often called wave drag in this context. Ethan: And a schlieren image makes this almost unfairly obvious. You can actually see the shock sitting over the upper surface, then shifting as Mach or angle changes. Maya: Yes. Wind tunnels, pressure measurements, schlieren, and flight test all support that picture. The shock is not merely a mathematical artifact. It changes the boundary layer’s job. The flow downstream must recover against a steep adverse pressure gradient. Ethan: Which is like asking a tired boundary layer to climb a hill that suddenly turns into a wall. Maya: A fair analogy. If the boundary layer cannot sustain that recovery, it separates. Then drag rises further, and lift may stop increasing as expected or even decrease. That is why increasing Mach can reduce maximum lift coefficient, even though modest compressibility initially strengthened the lift response at fixed angle. Ethan: So the rule of thumb is not “Mach gives you free lift.” It is “Mach steepens the pressure distribution, and eventually the recovery becomes expensive.” Maya: Exactly. Now let’s connect that to airfoil shape. Thickness-to-chord ratio matters because a thicker airfoil generally forces stronger local acceleration around its surface. Stronger acceleration means the local sonic condition arrives at a lower freestream Mach, all else comparable. Ethan: Let’s do the hangar-floor comparison. Same general airfoil family, similar loading and camber. One is 12 percent thick. The other is 18 percent thick. Maya: The 18 percent thick airfoil has a fuller contour. Its streamlines must bend and accelerate more strongly over the upper surface. As Mach rises, its suction peak tends to reach the local sonic condition sooner. So its critical Mach is generally lower. Ethan: Meaning the thicker section is more likely to show that upper-surface supersonic pocket, then a shock, while the 12 percent section is still behaving more smoothly. Maya: Correct. The 12 percent section is not immune. At sufficiently high Mach, it too develops local sonic flow and shocks. But its thinner shape usually delays the onset and moderates the acceleration. Ethan: Although, uh, “thinner is better” is too simple, right? Maya: Much too simple. Camber matters. A strongly cambered airfoil, or one carrying high lift, can create a very strong upper-surface suction peak even if it is relatively thin. Where the camber is placed also matters. Forward loading can deepen that early suction peak and reduce critical Mach. Ethan: So if I’m looking at pressure data, I should not just stare at thickness ratio. I should ask: where is the suction peak, how deep is it, and how abruptly does the pressure recover? Maya: That is an excellent checklist. Thickness affects how much the flow accelerates. Camber and angle of attack affect the loading. Together, they determine whether the pressure distribution stays recoverable as Mach rises. Ethan: And when the shock appears, the practical symptoms are drag rise, possible separation, and maybe a lift curve that starts disappointing you right when you wanted margin. Maya: Yes. The physics is sequential: increasing Mach strengthens pressure effects, local acceleration creates sonic pockets, shocks force abrupt recovery, and the boundary layer may separate. Airfoil geometry determines how early and how severely that sequence unfolds. Ethan: Next time, we’ll stay with the pressure picture and talk about what the shock does to the boundary layer and the aircraft’s response when that interaction gets serious. Maya: See you then.

Maya: Last time, we tracked how an airfoil’s pressure peak can create local sonic flow long before the airplane reaches Mach one. Today, the question is... how do we shape an entire wing so that peak arrives later, and hurts less? Ethan: Right. Because once you’ve seen a shock appear on a pressure plot, the obvious flight-test question is, “Can I just sweep the wing and make the problem go away?” Maya: You can reduce it. You cannot make it go away. Sweep, aspect ratio, and airfoil shape work together. The basic goal is simple: avoid a very sharp upper-surface suction peak, delay local sonic flow, and, if a shock forms, make it weak. Ethan: So this is not “pick a swept wing, problem solved.” It’s more like arranging the wing so the flow gets fewer opportunities to accelerate itself into trouble. Maya: Exactly. Start with sweep. For a swept leading edge, the first approximation is that compressibility effects depend mainly on the Mach number normal to that leading edge. Maya: We write that as M normal equals M infinity times cosine of sweep angle. Ethan: The old cosine rule. A nice first-pass tool. The air is still coming at the airplane at the full freestream Mach number, but only part of that velocity is driving flow directly across the leading edge and over the section. Maya: Yes. The normal component is what a two-dimensional airfoil section “feels,” approximately. Let’s do the useful estimate. Suppose the airplane cruises at Mach 0.85 with 35 degrees of sweep. Maya: Cosine of 35 degrees is about 0.82. So the normal Mach number is 0.85 times 0.82... about 0.70. Ethan: Which is a big change. Mach 0.85 outside the airplane, but roughly Mach 0.70 in the simple normal-flow picture. Maya: Right. Now turn that around. Suppose an unswept wing section, at its particular lift coefficient and Reynolds number, reaches its critical condition around Mach 0.70. With 35 degrees of sweep, the elementary estimate for the corresponding freestream Mach is 0.70 divided by 0.82. Maya: That is about Mach 0.86. Ethan: So, in the idealized calculation, sweep shifts that threshold from 0.70 to roughly 0.86. That is why sweep became such a powerful transonic design lever. Maya: With a large caveat. Real swept-wing flow is three-dimensional. Lift level matters. Surface curvature matters. The boundary layer matters. And critical Mach, where local sonic flow first appears, is not identical to drag-divergence Mach, where drag rises sharply. Ethan: Still, as a rule of thumb, if I’m looking at a transport-like cruise Mach and I add sweep, I should expect the sectional compressibility problem to behave more like a lower-Mach problem... at least as a first screening estimate. Maya: Correct. But sweep has costs. It can complicate structure, low-speed behavior, and aeroelastic response. And it does not remove induced drag. Ethan: Which brings us to aspect ratio. Because the tempting move is, “Fine, make the wing swept and very long.” But then the structure starts sending angry emails. Maya: A fair summary. At a given lift coefficient, induced-drag coefficient scales as C L squared divided by pi times efficiency factor e times aspect ratio A R. Maya: So higher aspect ratio reduces induced drag. If everything else stayed equal, going from aspect ratio 8 to 10 cuts induced drag by about 20 percent, because the term scales roughly with one over aspect ratio. Ethan: That’s the long-wing payoff. Less energy spent making trailing vortices. Like spreading the same lift over a longer lever arm, so the downwash penalty is smaller. Maya: Good intuition. But “everything else equal” is doing a lot of work. A longer span raises bending loads. Thickness, fuel volume, landing gear integration, and root structure all matter. For efficient high-aspect-ratio transport wings, those root constraints can be decisive. Ethan: So the planform trade is not just aerodynamic purity. You are buying lower induced drag with structural weight and integration complexity. Maya: Exactly. And in transonic design, we add another requirement: shape the pressure distribution so the wing does not create a strong shock and then suffer a severe pressure recovery behind it. Ethan: Okay, now give us the supercritical airfoil without making it sound mystical. Maya: A supercritical airfoil is shaped to delay wave-drag onset in transonic flow. Compared with older laminar-style sections, it typically has a larger leading-edge radius, a relatively flattened upper surface, and substantial aft camber. Ethan: Flattened on top, more camber toward the back. That sounds backwards at first. I’d expect camber to mean more acceleration and more trouble. Maya: It depends where the camber is placed. The design aims to avoid an intense leading-edge suction peak. Instead of accelerating the flow very rapidly near the front, the upper surface supports a broader, more controlled region of high-speed, sometimes locally supercritical flow. Maya: Then that region is terminated by a weaker shock, usually farther aft than it would be on a conventional section. Ethan: So the pressure trace is less like driving into a curb and more like rolling over a shallow ramp. You still may get a shock, but you are trying not to get a violent pressure jump followed by boundary-layer separation. Maya: Precisely. The large leading-edge radius helps moderate the front-end acceleration. The flatter upper surface avoids excessive curvature-driven suction. The aft camber recovers the lift that might otherwise be lost. Ethan: And the practical test is not “does a shock exist?” It is, “Where is it, how strong is it, and how much drag and separation does it trigger?” Maya: Yes. A good transonic wing can sustain a substantial supercritical-flow region and end it with a weak shock. That is the design target, not shock-free flow at all costs. Ethan: Put the pieces together for me. Say I’m doing an early design review. Maya: First, choose sweep based on the intended cruise Mach, using the cosine rule as a sanity check, not a certification-grade prediction. Second, retain as much aspect ratio as mission and structure allow, because induced drag remains real. Third, pair the planform with a supercritical section and wing loading distribution that avoid a sharp front-loaded suction peak. Ethan: And then check the pressure distributions across the span, because the root, midspan, and tip are not all living the same aerodynamic life. Maya: Exactly. Especially on a swept wing. The planform is not merely a top-view outline. It determines how the wing loads, how the pressure field develops, and how the wing behaves structurally in the transonic regime. Ethan: For deeper reading, Hoerner is still useful for the practical planform instincts, Anderson gives the clean theory, and the seminal supercritical-airfoil papers are worth reading for the pressure-distribution logic. Maya: Next time, we’ll keep following that logic from wing shape toward the operating limits that make these design choices matter. Ethan: See you then.

Maya: Last time, in “Wing Planform and Supercritical Designs,” the takeaway was that sweep, aspect ratio, and supercritical shaping manage the transonic pressure field, not just the geometry. Today we get to the part that decides whether that pressure field is actually survivable near the wall. Ethan: Right, because a nice-looking pressure recovery on the airfoil plot can hide a fairly ugly reality six millimeters above the surface. Maya: Exactly. The boundary layer is thin relative to chord, but it is not a small effect. It is the region where viscosity has reduced the streamwise velocity from the outer-flow value down to zero at the wall. As it grows, it consumes momentum near the surface. Ethan: Which is the flight-test version of saying, “The wing has less energy in reserve than the freestream Mach number suggests.” Maya: Nicely put. For a zero-pressure-gradient boundary layer, the characteristic thickness grows downstream because viscous diffusion has had more time to act. A useful engineering thickness is the point where velocity reaches roughly ninety-nine percent of the external velocity. But for aerodynamic coupling, displacement thickness is often more revealing. Ethan: Because the outer flow sees the boundary layer as a shape change. Maya: Yes. Displacement thickness is the effective outward displacement of the streamlines caused by the momentum deficit. So even before separation, a thickening boundary layer can alter the local pressure field and move or strengthen a transonic shock. Ethan: Let’s separate two things people sometimes lump together: transition and separation. Maya: Important distinction. Transition is a change in boundary-layer state, broadly from laminar to turbulent. Separation is a failure of the near-wall flow to keep moving downstream against an adverse pressure gradient. At incipient separation, wall shear stress approaches zero. Beyond it, there is local reverse flow. Ethan: And turbulence is not automatically bad here. It costs skin friction, but a turbulent layer has a fuller velocity profile. More momentum lives closer to the wall. Maya: Correct. All else equal, that makes a turbulent boundary layer more resistant to separation than a laminar one. But “all else equal” does a lot of work. Surface roughness, pressure gradient, Reynolds number, and transition location all matter. A turbulent layer can also be thicker, so the interaction with the outer flow is not simply better or worse. Ethan: So if I’m looking at a wing section in a tunnel, I should not ask, “Is the layer turbulent?” and stop there. I should ask, “How thick is it, how much near-wall momentum does it have, and what pressure rise is it about to face?” Maya: That is the practical sequence. Let’s put scale on it with a deliberately simple estimate. Assume a chord Reynolds number of five million, and look at mid-chord, so \(x/c = 0.5\). Then the local Reynolds number is about \(2.5 \times 10^6\). Ethan: We are assuming a smooth, attached, incompressible-style flat-plate estimate here. No shock yet, no strong pressure gradient. Maya: Precisely. For a turbulent zero-pressure-gradient boundary layer, a common estimate is delta over x equal to about 0.37 times local Reynolds number to the minus one-fifth. At \(Re_x = 2.5\) million, that gives delta over x of roughly 0.02. Ethan: So at half chord, delta is about two percent of that half-chord. Put it in chord units and it is about one percent chord. Maya: Roughly, yes. If the chord were one meter, that is on the order of ten millimeters. Not huge geometrically, but very consequential if a shock sits inside or just downstream of that layer. Ethan: And if the layer stayed laminar in the same idealized calculation? Maya: The classical laminar estimate, delta over x about five over square root of local Reynolds number, gives roughly 0.0032. At mid-chord, that is about 0.16 percent of chord. Much thinner, but also generally much less robust under a strong adverse pressure gradient. Ethan: That is a useful sanity check. A layer can be thin and still be fragile. Thickness alone is not the separation criterion. Maya: Exactly. There is no universal “one-percent-chord means separation” threshold. Shock-interaction risk depends on thickness, state, and pressure gradient. The shock matters because it imposes a rapid pressure rise. The external flow decelerates abruptly across it, and the boundary layer must respond with limited near-wall momentum. Ethan: Like asking a loaded aircraft to flare aggressively after it has already bled off too much energy. The outer flow may make the maneuver, but the near-wall flow cannot. Maya: Good analogy. In a transonic shock-boundary-layer interaction, the adverse pressure gradient can thicken the layer rapidly. If the wall shear collapses, separation begins near the shock foot, often with the interaction extending upstream through the boundary layer. Ethan: So the shock is not just a line on a schlieren image. It is coupled to a viscous region with its own effective thickness and pressure response. Maya: Right. And that coupling can be sensitive. A slightly more aft shock location, a stronger pressure recovery, or a thicker incoming layer can shift the system toward separation. This is why supercritical sections aim for controlled pressure recovery, not merely a delayed shock. Ethan: Practical rule of thumb: if your analysis predicts a shock over a region with a thick, low-momentum layer and a steep recovery, treat attached flow as an assumption to verify, not a result to celebrate. Maya: Yes. Check surface pressure distributions, look for evidence of separation in tunnel data, and compare with established boundary-layer correlations. Schlichting remains foundational, Anderson’s boundary-layer sections are a useful engineering bridge, and wind-tunnel separation studies are valuable because the interaction is strongly physical. Ethan: Next time, we’ll keep following that chain from local flow physics to the behaviors that show up on the airplane.

Maya: Last time, in “Boundary Layers and Separation,” we followed the flow right to the point where it can no longer stay attached. Today, uh, we let the structure answer back. Ethan: Right, because at transonic speed, you can look at a wing and think, “That shock is moving around a lot,” and then realize the wing itself may be moving too. Maya: Exactly. And that is aeroelastic coupling. The aerodynamic loads deform the structure. That deformation changes the flow. The changed flow creates new loads. So the fluid and structure are not separate problems anymore. Ethan: Like a flight-test control surface that flexes a little under load, changing its angle, which changes the load that made it flex in the first place. Maya: Yes. The key question is whether that loop damps out or reinforces itself. Maya: In the simplest structural picture, we have mass, damping, and stiffness. A displacement, call it q, produces a restoring force roughly proportional to kq. Inertia scales with mass times acceleration. Damping resists velocity. Maya: Then the aerodynamic force depends on the motion too, not just on the commanded angle of attack. If that aerodynamic force acts in the wrong phase, it can feed energy into the structural motion. Ethan: So, uh, the practical version is: a flexible thing gets pushed, it moves, and the next push arrives at a timing that either calms it down or makes it worse. Maya: Precisely. Phase is the whole story, in a compact form. Ethan: And transonic is where that timing gets especially interesting because we have mixed flow regions, right? Roughly Mach 0.8 to 1.2, with subsonic and supersonic regions coexisting. Maya: Right. A shock can form within that flow field. Now, a shock is not automatically an aeroelastic problem. Shock movement alone is just shock movement. Maya: But when the shock interacts with the boundary layer in an unsteady way, the pressure distribution can oscillate. That periodic oscillation is transonic buffet. Ethan: Let me try the cockpit-window version. The shock moves a little. The boundary layer responds. The loading changes. That changes the conditions around the shock, and then the shock moves again. Not necessarily smoothly, either. Maya: That is the physics-first picture. We should be careful not to claim one universal sequence for every geometry. But the defining point is clear: transonic buffet is periodic oscillation driven by shock-wave interaction with the boundary layer. Ethan: And the wing feels that as unsteady normal force. Maybe a shudder, maybe a repeated loading signal, depending on what you are measuring. Maya: Yes. The structure can respond to that forcing. But forced response is not the same thing as flutter. Ethan: Important distinction. People hear “oscillation” and immediately say flutter. Maya: And that can be misleading. Flutter is a dynamic instability caused by coupling between aerodynamic forces and structural modes. It is not merely a wing being shaken by an unsteady flow. Ethan: So buffet can be the flow shaking the structure. Flutter is the flow and structure, uh, joining forces in a self-reinforcing instability. Maya: Well put. And transonic buzz is different again. Buzz is a high-frequency oscillation of control surfaces caused by shock-induced unsteady aerodynamic forces. Ethan: So we have four things that can look vaguely similar in a noisy test record. Shock movement, which is the flow feature moving. Buffet, which is periodic shock-boundary-layer-driven oscillation. Buzz, focused on a control surface at high frequency. And flutter, the coupled instability. Maya: Exactly. Related, potentially interacting, but not interchangeable labels. Ethan: Let’s make this concrete. Suppose we are reviewing a flight-test trace. Horizontal axis is incremental alpha, just small steps upward. Vertical axis is normal force, and we also have its unsteady component. Maya: At the lower alpha steps, the mean normal force rises with alpha, and the unsteady signal is comparatively small. Then, at a particular increment, a periodic fluctuation appears around the mean force. Ethan: So the mean still tells us the aircraft is carrying more load, but now the trace has this repeated ripple riding on it. Maya: Yes. That is a reasonable buffet-onset signature to investigate, especially in a transonic condition. The interpretation is not “we have proven flutter.” It is: unsteady aerodynamic loading has emerged, consistent with shock-wave and boundary-layer interaction. Ethan: What would make me worry that this is not just the shock moving on the wing? Maya: I would ask where the motion is strongest and what moves with it. If the unsteady normal force is broad across the lifting surface, and the signature appears as alpha is incremented into the onset region, that supports a buffet interpretation. Maya: If instead a control surface shows a prominent high-frequency response under shock-induced unsteady loading, then transonic buzz becomes the more relevant concern. Ethan: And if the structure’s own motion starts participating in a way that reduces the effective damping, then we have to ask the flutter question. Maya: Correct. We would examine the structural modes and the aerodynamic response together. The essential test is not simply, “Is there a frequency?” Every structure has frequencies. The concern is whether the coupled system is losing damping. Ethan: That is a good rule of thumb. A peak on a plot is a clue. A growing, poorly damped coupled response is the real alarm. Maya: Exactly. And the mitigation paths follow directly from that model. Increase stiffness where appropriate, so deformation is reduced or the relevant structural modes are placed away from adverse interaction. Ethan: Mass balance is another practical lever, especially for control surfaces. You are trying to manage how readily the surface wants to rotate or vibrate. Maya: Yes. Mass distribution matters, as does damping. The goal is not one magic modification. It is to ensure the structure, its modes, and the unsteady aerodynamic forcing do not combine unfavorably. Ethan: In test terms, we are not asking, “Can a shock move?” Of course it can. We are asking, “What does that motion excite, how strongly, and does the aircraft dissipate or gain energy?” Maya: That is the compact framework. Flow feature, unsteady load, structural response, feedback. Ethan: And, uh, don’t diagnose flutter just because the airplane is shaking. Maya: Nor dismiss a repeated force trace as mere noise. Identify the flow condition, observe the frequency and location of the response, and check whether the structure is participating. Ethan: Next time, we’ll keep building from that foundation and look at how these ideas show up in the wider transonic design problem. Maya: Until then, keep the loop in view: flow changes structure, structure changes flow.

Maya: Last time, we looked at shocks coupling to structure, and how that can become buffet, buzz, or outright instability. Today, um, we ask a more basic question, how do you know the shock you see in a tunnel is the airplane’s shock, not the tunnel talking back? Ethan: Right, because a transonic tunnel can look wonderfully controlled, and then you realize the walls, the support sting, the Reynolds number, even a little transition trip, have all voted on your result. Maya: Exactly. A transonic tunnel gives you a controlled Mach number, nominal dynamic pressure, model attitude, and instrumentation. It does not give you flight automatically. The central job is similarity. Ethan: So, the two headline knobs are Mach and Reynolds number. Maya: Yes. Mach number, M equals V over a, controls compressibility. Near transonic conditions, it strongly affects local acceleration, shock position, and wave drag. Reynolds number, rho V c over mu, controls the boundary-layer state relative to the model chord c. It affects skin friction, transition, separation, and therefore shock, boundary-layer interaction. Ethan: And in flight test terms, Mach tells me whether I’m in the right speed regime. Reynolds tells me whether the flow has the right, uh, “stickiness history” over the surface. Maya: That is a fair shorthand. At fixed Mach, a small model usually has lower Reynolds number than the aircraft. You can raise tunnel density, often by raising total pressure. Some facilities also reduce temperature. But matching full-scale Mach and Reynolds simultaneously may be impractical. Ethan: So if I match Mach but miss Reynolds low, what do I watch for? Maya: A thicker, more separation-prone boundary layer, generally. The shock may move, shock-induced separation can appear earlier, maximum lift may fall, and buffet onset may shift. But do not assume every shift goes in one universal direction. Geometry, transition, and the pressure recovery all matter. Ethan: Which is why, if somebody hands me one beautiful polar at Mach point eight two, I ask, “At what Reynolds number, with what transition state, and how repeatable was it?” Maya: Precisely. And, uh, whether Mach was actually held. In transonic testing, a change of a few thousandths in Mach can matter near drag rise or buffet onset. The tunnel control system may report a stable free-stream Mach while the model flow is fluctuating because the shock is unsteady. Ethan: Like a flight-test airspeed trace that looks calm, while the tufts on the wing are having a disagreement. Maya: Very much so. That is why force data alone are not enough. You want a force and moment balance, surface pressures, and at least one flow-visualization method. Schlieren can show shock location. Oil flow or tufts can reveal separation. Pressure-sensitive paint can help, if its calibration and temporal response suit the test. Ethan: And the old NASA and NACA visualization guides are still useful here. Not because the pictures are nostalgic, but because they teach you what each technique can and cannot prove. Maya: Right. A schlieren image shows density gradients, so it is excellent for shocks. It does not, by itself, prove surface separation. Oil flow shows near-wall behavior, but it is not a precise pressure measurement. Use complementary evidence. Ethan: Let’s build a practical lift polar. Say we need a wing-body polar at cruise-like transonic conditions. Where do we start? Maya: First, define the intended comparison point. Suppose the target is Mach zero point eight two and a chord Reynolds number of, say, fifteen million. We choose the largest feasible model, because larger chord helps Reynolds number and reduces relative balance and geometry errors. Ethan: Then we check whether the facility can actually reach that Mach and Reynolds combination without running into pressure limits, model loads, or tunnel constraints. Maya: Yes. If it cannot reach fifteen million, we might test at, say, eight, twelve, and the maximum attainable Reynolds number, while holding Mach zero point eight two. That gives a Reynolds trend rather than pretending one low-Re point is flight truth. Ethan: Like doing a few altitude points in flight test instead of declaring victory from one pass. Maya: Exactly. For the polar, we run angle of attack from below the expected linear range through the onset of nonlinearity. Perhaps minus two to five degrees, with fine increments near the expected shock movement or buffet boundary. At each point, record lift, drag, pitching moment, tunnel Mach, total pressure, temperature, and model pressures. Ethan: I’d also repeat a few points while coming back down in angle. If the curve does not retrace, that’s a clue, right? Maya: A very useful clue. Hysteresis can indicate separated flow, unsteady shock motion, thermal drift, or simply inadequate settling time. Never average away a repeatability problem without identifying it. Ethan: What about interference? Because this is where people, um, get burned. Maya: We check blockage and wall interference before believing the polar. Solid blockage comes from the model displacing flow area. Wake blockage comes from momentum loss in the wake. Both can alter the effective stream conditions. In transonic flow, walls can also reflect pressure disturbances and shift the shock. Ethan: So the tunnel can make the model think it is at a slightly different Mach number or angle of attack than the instrumentation says. Maya: Yes. The correction depends on the facility, test section, wall treatment, and model. Use the tunnel’s validated correction method, not a generic spreadsheet formula. And compare pressure distributions or shock positions at more than one model size or wall condition when the result is important. Ethan: Practical rule, if the lift curve slope or drag rise changes suspiciously when I move the model, change the support, or alter tunnel settings that should be equivalent, I have an interference question. Maya: Good rule. Also inspect the support interference. A sting or strut can alter base pressure, pitching moment, and local flow. Drag is especially fragile, because it is the small difference between larger forces and includes tare, alignment, and wall effects. Ethan: Then reading the polar, the initial linear section gives us lift-curve slope. A rounded knee might mean shock-induced separation, but it might also mean Reynolds mismatch, wall interference, or a transition issue. Maya: Correct. Correlate it with the pressure data. If the upper-surface shock moves aft with increasing lift, then strengthens, and the pressure recovery becomes less effective as the polar bends, that is coherent transonic evidence. If the force break appears with no supporting pressure or visualization change, investigate the measurement chain. Ethan: So, don’t ask the tunnel for one magic coefficient. Ask whether the force balance, pressures, and flow picture tell the same story. Maya: That is the habit. Wind-tunnel technique handbooks give the correction frameworks. NASA and NACA visualization references teach the diagnostic logic. The engineer’s job is to connect them, carefully. Ethan: Next time, we’ll take those measured trends and ask how to turn them into design decisions without, you know, fooling ourselves on margins. Maya: See you then, and bring your suspicious-looking polar.

Maya: Last time, “Transonic Experimental Techniques,” we used wind-tunnel and flow-visualization methods to diagnose transonic flow. Ethan: And today is the part I always want before a flight card review, which is, okay, uh, what are the fast checks? When do I stop treating a drag trend as ordinary airplane mess and start suspecting compressibility? Maya: Start with Mach number, but do not stop there. Transonic behavior is roughly Mach 0.8 to 1.2, and compressibility effects grow noticeably as you approach Mach 0.8. The key is that the wing can accelerate flow locally above the freestream Mach number. Ethan: Right. The airplane might be below Mach one, while a patch over the upper surface is, you know, locally supersonic. Maya: Exactly. Then that patch has to return to subsonic flow, often through a shock. The practical consequences are pressure-recovery loss, drag rise, possible boundary-layer separation, buffet, and changes in control effectiveness. Ethan: So the first rule of thumb is not, “We’re subsonic, therefore shocks are impossible.” Maya: Correct. A better checklist starts with three questions. What is the freestream Mach? What is the lift coefficient, meaning how hard is the wing being asked to work? And what is the geometry, especially thickness, camber, and sweep? Ethan: Because a clean low-lift cruise point and a maneuver point at the same Mach are not the same aerodynamic problem. Maya: Precisely. Higher loading generally means stronger local acceleration. More camber or thickness can also make the local peak velocity higher. Sweep helps because the component of freestream Mach normal to the leading edge is reduced, approximately as Mach times cosine of sweep. Ethan: The flight-test version is, if I’m sweeping Mach, I want to hold configuration tightly, hold altitude or dynamic pressure as planned, and compare points at matched lift coefficient where I can. Otherwise I may call it drag rise when I’ve really just changed trim or loading. Maya: Yes. Normalize before diagnosing. At each stabilized point, examine required thrust or inferred drag, lift coefficient, trim setting, control position, vibration, and, if available, surface-pressure or flow-visualization evidence. Ethan: And if the drag curve starts bending upward faster than the lower-Mach trend, that’s the first flag. Maya: A flag, not proof. Drag rise is consistent with shock formation, but flight data alone can have confounders. Engine installation effects, trim drag, Reynolds-number changes, instrumentation uncertainty, and atmospheric variation all deserve a check. Ethan: Okay, let’s do a worked example. Say we fly a Mach sweep in clean cruise configuration. At the lower points, required thrust increases smoothly. Then, near the upper end of the sweep, the required thrust begins climbing more sharply than expected. At the same time, we need a little more stabilizer trim, and the crew reports a faint onset of buffet. Maya: That is a credible transonic pattern. The sharper thrust requirement suggests drag rise. The trim change suggests the pressure distribution, and therefore pitching moment, is moving. The buffet suggests unsteady separation or shock motion may be involved. Ethan: But we still should not say, “Confirmed shock,” based on buffet and thrust alone. Maya: Right. We would say, “The data are consistent with shock-induced effects.” Next, repeat carefully. Make an up-sweep and a down-sweep. Repeat at a second lift coefficient. Check whether the onset shifts when the wing loading changes. Ethan: That’s a nice practical discriminator. If the feature gets worse when we ask for more lift, it supports the local-acceleration story. Maya: Exactly. Then use a targeted diagnostic. A few pressure measurements, oil flow, tufts, or suitable visualization can show whether a shock is present and whether separation develops behind it. Keep the test question narrow. Ethan: Not, “Let’s collect everything.” More like, “Does the shock move aft with Mach, and does the boundary layer stay attached?” Maya: Yes. And for design action, separate aerodynamic mitigation from structural mitigation. Aerodynamically, sweep reduces the normal Mach component. Camber and loading changes can reduce the local suction peak or move loading away from a problematic region. Ethan: Though that is never free. Less aggressive camber or a different loading distribution may cost you somewhere else, like low-speed performance or trim. Maya: Correct. It is a trade. Also, structural tuning can matter when buffet or shock motion couples into a flexible surface. More stiffness, altered mass balance, or a changed mode shape may improve the response, but structural tuning does not remove the underlying shock. Ethan: That distinction matters. Don’t use a structural fix to pretend the aerodynamic limit vanished. It may make the airplane more tolerable, but you still need to understand the drag and control behavior. Maya: Good rule. Another quick check, if the onset is sharp, repeatable, and linked to Mach and lift, treat it as a compressibility candidate. If it changes mainly with power setting, configuration, or data reduction choices, investigate those first. Ethan: And for planning, don’t run the first test point right at the suspected knee. Approach it in small increments, watch the trend, and leave margin for buffet, control changes, and the possibility that the onset comes earlier than predicted. Maya: That is the compact toolkit. Match the conditions, look for a nonlinear drag trend, cross-check trim and buffet, confirm with focused flow evidence, then choose the mitigation that addresses the actual mechanism. Ethan: For deeper reference, I’d keep the FAA and industry flight-test guidance nearby, then use Anderson for the physics and Hoerner for the practical sanity checks. Maya: Nicely put. That closes our review, and, uh, thanks for working through the checklist with us. Ethan: Thanks for listening. Fly the sweep carefully, trust the data, and we’ll see you next time.