A clearance offers one more level, the weight permits it on paper, and the recommended maximum altitude shown on board falls exactly on that level. Do you accept? The question is not whether the aircraft holds up there in level flight at one g: it does. It is how much load factor remains before buffet onset, because that remainder — not the knot gap on the speed tape — is what a weather-avoidance bank angle, a mountain-wave bump, or a few tonnes of error on the assumed weight will spend.
Buffet is a symptom, not a limit
Buffet is neither a structural limit nor a flight state: it is the vibratory signature of a flow separation. The airflow partially leaves the upper surface, the unsteady wake beats against the wing and the tailplane, and the structure carries that beating through to the seats. On a swept wing, two paths lead there: too much angle of attack, or too much local Mach number — in which case separation is born at the foot of a shock wave, under the effect of its pressure jump.
In high cruise, the limit retained is not the stall but the onset of buffet, defined in flight test by a measured vibration level and not by a loss of lift. Two reasons: degradation of flying qualities begins there — pitch damping that deepens, a tendency to roll departure — and recovery is paid for in altitude, a currency not always available. The usable envelope is therefore not bounded by the boundary itself, but by a margin ahead of it.
The trap. Buffet does not tell you which side of the envelope you are on: the vibration is of the same nature at both ends, and the corrections are opposite. Every reflex action turns against its author on the wrong side — extending a drag device near the low-speed branch, pulling up near the compressibility branch. The diagnosis is read from the indications, not from the seat: Mach and its trend, angle of attack or pitch attitude, current thrust, the event of the past minute.
Buffet is a warning that the wing is about to lose lift. It happens when you’re flying too slow or too fast for the altitude. The key is to know how much extra load (like turning or turbulence) the wing can handle before buffet starts. If you’re too close to the limit, even a small bank angle could trigger it.
An Airbus A330 at FL410 with a true airspeed of 450 knots and a load factor of 1.0g. The buffet margin is 0.3g. A 30-degree bank angle (1.15g) would consume the entire margin, risking buffet. If a sudden gust adds 0.2g, the aircraft would enter buffet immediately.
A single boundary, seen from its two ends
We speak of two limits drawing closer together; there is only one. Plot the lift coefficient on the vertical axis and Mach on the horizontal axis: the onset-of-buffet boundary there is a single curve, roughly flat at moderate Mach numbers, then collapsing where shocks assert themselves. What remains is to locate the aircraft within that plane. In level flight, n · W = CL · q · S, and dynamic pressure can be rewritten, in air, as q = 0.7 · p · M²: since the speed of sound already contains temperature, temperature drops out of the equation. Hence CL = n · W / (0.7 · p · M² · S).

At a given weight and level, the operating locus is not a point but a curve: required lift varies as the inverse square of Mach. This hyperbola cuts the boundary in two places — the low-speed branch to the left, the compressibility branch to the right — and one flies between the two. Climb at constant weight: static pressure falls, the hyperbola shifts upward, the intersections draw closer together. The famous corner is the level where they merge; beyond it, none exists any longer, meaning no Mach number permits level flight at one g without buffet.
The common statement — the low-speed limit rises in Mach, the high-speed limit barely moves — only describes this shift. In calibrated airspeed, that same low-speed limit barely moves at all: at a given pressure altitude, the two quantities are in a fixed relationship, both derived from impact pressure and static pressure. Temperature does not enter into this conversion; it enters only for true airspeed.
| Aspect | Low-speed branch | Compressibility branch |
|---|---|---|
| Origin of separation | High angle of attack, at high lift | Foot of a shock wave, by pressure jump |
| What sets the limit | Usable lift before measurable vibration | Shock-separation Mach number, lower the higher the lift |
| Climbing at constant weight | The Mach limit rises quickly | The limit barely moves, and rather downward |
| Cues on the flight deck | Increasing pitch attitude and angle of attack, thrust at the stop, slowly decreasing speed | Increasing Mach, nose-down tendency, degraded pitch damping |
| What widens the margin | Descend, reduce weight, reduce load factor | Reduce Mach, reduce load factor |
| Classic mistake | Holding level flight by pulling, thus loading the wing | Correcting pitch attitude by pulling, which lowers the Mach number at which buffet appears |
Weight relative to pressure: the governing variable
Only the load n · W and the product p · M² appear. The boundary is therefore plotted once and for all, per type, as a function of Mach and the ratio of weight to relative pressure — that is the entry point for onset-of-buffet charts. First consequence: two weight-level pairs that give the same ratio give the same margin. Burning fuel means buying altitude, and step climbing follows the rebuilding of a margin that the initial weight had denied.
Second consequence, less intuitive: temperature does not appear in it. At a given pressure altitude and weight, the buffet margin is the same in extreme cold and in hot weather. Temperature changes true airspeed for a given Mach, hence the drag-thrust balance, and above all available thrust: it shifts the thrust ceiling and leaves the buffet ceiling where it was. If a buffet chart offers a temperature-deviation entry, look for what it actually corrects — a thrust assumption, a combination of criteria, a speed conversion. Not the aerodynamic boundary.
| Parameter | Buffet margin | Thrust ceiling | Mechanism |
|---|---|---|---|
| Fuel consumed | Increases | Rises | Lower load to balance at a given pressure and Mach |
| One level higher at constant weight | Decreases | Draws closer | Static pressure is in the denominator of required lift |
| Positive temperature deviation | Unchanged | Lowers | The boundary depends only on pressure altitude and weight |
| Bleed air, engine anti-ice | Unchanged | Lowers | Reduced thrust, with no aerodynamic effect |
Three ceilings, and the lowest one governs
One aircraft, one instant, three ceilings. The certified ceiling is a type limit — structure, pressurisation, emergency descent and oxygen requirements — independent of weight and of the day. The thrust ceiling is the level where cruise thrust leaves only a conventional residual vertical speed, on the order of a few hundred feet per minute depending on the manufacturer’s criterion. The buffet ceiling is the one where the manoeuvre margin falls to the criterion, typically a capability on the order of 1.3 g before buffet onset — that is, a budget of 0.3 g above level flight.
The lowest of the three governs, and it is not always the same one. At high weight and in hot weather, it is thrust: the aircraft refuses to climb. In very cold weather, the thrust ceiling rises while the buffet ceiling does not move a foot, and the aircraft accepts without effort a level where the margin has become thin. A cold atmosphere favours the climb, not the margin. The optimum and maximum altitudes computed on board blend these logics, and knowing which one produced the number changes the confidence one places in it.
The trap. The recommended maximum altitude is not a margin: it is a calculation whose assumptions escape anyone who has not read them. Depending on the implementation, it blends a thrust criterion and a buffet criterion, sometimes retains the instantaneous weight rather than the level-off weight, takes the temperature deviation felt now rather than that of the layer being aimed for, and assumes all engines, symmetric flight, no anti-ice. A level held at that value is flyable at one g; it promises nothing about the bank angle available ten minutes later to skirt a cell.
Reading the margin in load factor, not in knots
If the criterion is expressed in g, it is not out of a taste for abstraction: what consumes the margin is an increment of load factor. Bank angle, pull-up, and an ascending gust all have the same effect on required lift. The gap in knots between the two branches, however, is misleading: its correspondence with load-factor capacity differs from one level to another, and a comfortable-looking corridor on the speed tape may contain only a tenth of a g. In a turn at constant altitude, load factor equals the inverse cosine of the bank angle: the table is short and worth memorising.
| Bank angle | Load factor | Share of budget consumed |
|---|---|---|
| 15 degrees | 1.04 | about an eighth |
| 25 degrees | 1.10 | about a third |
| 30 degrees | 1.15 | a little more than half |
| 40 degrees | 1.31 | the whole budget |
| 45 degrees | 1.41 | beyond the budget |
More insidious than an avoidance turn: with Mach and level held constant, the margin degrades without any speed parameter moving. An underestimated weight, a late loading change not properly accounted for, fuel that is not burned as planned, all shift required lift without touching the speed tape; the only witness then is angle of attack.
The right reflex. Translate the margin into bank angle before accepting the level. Enter the chart with the weight you will have at level-off, read the load-factor capability, convert it into a sustainable angle, then compare that figure to what the sector’s weather is likely to demand. If the gap amounts to a few degrees, the question is no longer whether the level is authorised: it is whether to request the lower level, or to defer the climb.
At the edge of the envelope, the aircraft no longer flies the same way
Near the low-speed branch, the aircraft is on the wrong side of the drag curve: a loss of speed increases drag, which worsens the loss. At low altitude, thrust breaks the spiral; at high altitude it is no longer enough, and the only abundant resource is potential energy. Automation diagnoses nothing: an autopilot holding altitude, with a Mach target that thrust can no longer sustain, lets speed decay, compensates in pitch attitude, and trims nose-up. Nothing flashes for several minutes; the anomaly is read in the speed trend, an unusual pitch attitude, thrust at the stop.
On the other branch, shock displacement alters the flow seen by the tailplane and the pitch balance, with a nose-down tendency that calls for a correction — that is, the very action that loads the wing. Hence an asymmetry to be settled in advance: on this side, reducing Mach is possible in level flight, by thrust and, if the documentation permits it at this level, by a drag device; on the other, neither of these means helps, and the second one hurts. A mountain wave, finally, can present both branches within the same minute and on the same heading.
The right reflex. On the low-speed branch, the first action is to release angle of attack, and the currency of exchange is altitude, not thrust: at this level, thrust alone does not pull the aircraft out of an established deceleration. So decide in advance that the level is negotiable, announce it, let the speed recovery happen through a gentle descent with symmetric control inputs, then re-establish the level. The report to control comes after the action, it does not condition it.
What to check at home, and what remains grey
Three checks, made once in one’s own documentation, change the reading of every subsequent flight. Knowing whether the type’s onset-of-buffet chart is plotted at one g or already at the manoeuvre criterion: both exist, and confusing them amounts to claiming a margin one does not have. Knowing how to enter it, and if a temperature entry appears, what it is for. Knowing, finally, which criteria the onboard computation combines, and under what assumptions of weight, temperature, bleed air, and engine count. Added to this are the level penalties published by the Minimum Equipment List and the Configuration Deviation List: a missing panel is not handled by guesswork.
The grey zone does not close. The margin criterion is a fixed number, whereas the gust environment it covers is not: no turbulence forecast converts into a load factor, and moderate turbulence forecast in a stable layer does not carry the same content as the same forecast downwind of high terrain. Added to this is an operational tension, since the most economical levels are those where the margin is smallest, and the trade-off is made in flight, on partial information.
For anyone preparing a type rating or a technical interview, the expectation has shifted: naming the coffin corner demonstrates nothing. What must be demonstrated is the relationship that governs the operating point, the reason the boundary ignores temperature while the thrust ceiling depends on it, the translation of a g capability into usable bank angle, and how to handle the first ten seconds of a buffet whose side is not yet known. This reasoning is worked through in training, and marginal cruise cases are compared on the forum.
The type’s certification specifications, its approved flight manual, the operator’s documentation, and published procedures prevail in all circumstances over this article, whose purpose is to explain a mechanism and not to serve as a procedure or a chart.
Questions fréquentes
Why does the low-speed limit rise in Mach number with altitude when the corresponding calibrated airspeed barely moves?
Because these are two ways of describing the same point. The buffet boundary sits at a given lift, which fixes, for a given weight, the product of static pressure and the square of Mach. Now, calibrated airspeed and Mach are in a fixed relationship at a given pressure altitude, both derived from impact pressure and static pressure: as altitude increases, the same calibrated airspeed corresponds to an ever-higher Mach number. The aerodynamics do not change, only the conversion shifts.
Does a very cold atmosphere improve the buffet margin?
No. The boundary depends only on pressure altitude and weight, because dynamic pressure can be rewritten, up to a constant factor, as the product of static pressure and the square of Mach, where temperature does not appear. Cold weather improves available thrust, thus raising the thrust ceiling, and leaves the buffet ceiling exactly where it was. Hence a useful inversion to keep in mind: in cold weather, the aircraft climbs easily to a level where the manoeuvre margin is thin.
What does a capability of 1.3 g before buffet onset concretely represent?
It means the wing can carry about thirty percent more than the aircraft’s weight before the vibration appears. In a turn at constant altitude, load factor equals the inverse cosine of the bank angle: thirty degrees consume a little more than half of this budget, forty degrees consume it entirely. What remains must still absorb gusts and any elevator input.
How do you distinguish low-speed buffet from compressibility buffet?
Not by the sensation, which is of the same nature in both cases. The diagnosis is made from the indications: displayed Mach and the direction of its change, angle of attack or pitch attitude, thrust setting, and recent events — a climb, a deceleration, entry into a mountain wave. Since the corrections are opposite, this diagnosis must precede the action, and it is better to have rehearsed it in the simulator than to discover it on the day.
Does the recommended maximum altitude computed on board guarantee the manoeuvre margin?
Not by design, and that is the problem: the content of that number depends on the implementation. Some combine a thrust criterion and a buffet criterion, others favour one of the two; assumptions about weight, temperature deviation, bleed air, and the number of engines running also vary. One must have read in the type’s documentation what that number actually contains before treating it as a margin.
Should you climb to the highest available level for fuel economy?
Optimum altitude and maximum altitude do not answer the same question, and flying very close to the latter costs manoeuvre margin for often modest fuel savings. The sound approach is to climb in steps as the aircraft grows lighter, checking at each stage the margin obtained at the level-off weight, and keeping in mind the weather deviation the sector might require. The nature of the flight shifts the trade-off: on a short sector, the fuel saving from one extra level is small compared with the energy cost of the climb, whereas the loss of manoeuvre margin is immediate and lasts until the descent.

Leave a Reply
You must be logged in to post a comment.