From the book "First Light" by Geoffrey Wellum, where the author describes dog-fight manoeuvres against an Me 109:

If you want to shake someone off your tail you have to fly your Spitfire to its limits. In a tight turn you increase the G loading to such an extent that the wings can no longer support the weight and the plane stalls, with momentary loss of control. However, in a Spitfire, just before the stall, the whole aircraft judders, it’s a stall warning, if you like. With practice and experience you can hold the plane on this judder in a very tight turn. You never actually stall the aircraft and you don’t need to struggle to regain control because you never lose it. A 109 can’t stay with you.

What is the physical explanation for the pre-stall judder? Is it some sort of unstable linear-turbulent flow transition across the wings, or something else?

  • 6
    $\begingroup$ Flying to the buffet is a common method of achieving a maximum-rate turn in an aircraft where the g limit is high enough to allow it. Select maximum power to enable the highest g loading, roll, and then pull progressively until you feel the whole aircraft buffeting, then hold it. You can go even tighter if you aren’t concerned about maintaining your level by rolling to 90°. $\endgroup$
    – Arkhem
    Jul 11, 2021 at 6:28
  • 1
    $\begingroup$ The 109 was a pig to fly. About 30% of the 109s lost in WWII were not shot down in combat, but crashed on takeoff or landing. $\endgroup$
    – alephzero
    Jul 11, 2021 at 10:43
  • 7
    $\begingroup$ It wasn't a pig to fly at all. The slats allowed you to pull very high G without it departing in a snap roll if you went past the limit. It was a pig to take off and land being ground loop prone. In spite of that, many German pilots preferred it to the FW190, because experienced pilots could handle the exciting ground handling. $\endgroup$
    – John K
    Jul 11, 2021 at 13:01
  • 4
    $\begingroup$ You thermal a Schweitzer 1-26 glider exactly like this. The optimal min sink condition is achieved with stall buffet starting, and the buffet was quite strong, so you go around with the thing shuddering all the way. $\endgroup$
    – John K
    Jul 11, 2021 at 13:04
  • 3
    $\begingroup$ @Arkhem: It's also quite common in GA training planes, or at least the ones I've flown (C-150 and PA-28-180, mostly). In training, I would have to fly just to the edge of stall, where the buffeting starts, and then maintain that speed & attitude. $\endgroup$
    – jamesqf
    Jul 11, 2021 at 16:29

3 Answers 3


It's because aircraft don't stall all at once.

A stall occurs when the airflow over the upper surface of the wing separates from the entire upper surface, causing loss of lift. This airflow separation starts at the trailing edge (some separation right at the trailing edge often occurs even in normal flight). As the aircraft's angle of attack increases, the point on the wing where the airflow separates from the upper surface moves further and further forwards.

Up to a point, the lift generated by the wing continues to increase with increasing angle of attack, despite the growing separation bubble, as the wing deflects air more and more downwards. Past a certain critical angle, though, the forward edge of the area of separated flow moves forwards very rapidly until it's nearly at the leading edge, lift drops off, drag increases sharply, and the wing's stalled.

The separated airflow is highly turbulent, and, once the pocket of separated airflow spreads far enough forwards and becomes big enough, this turbulence can sometimes be felt by the pilot as buffeting or juddering as the large amounts of turbulent air flow rearwards and strike the aircraft's horizontal tail. Whether this can be felt before the aircraft actually stalls depends on the aircraft:

  • Many aircraft, including (apparently) the Spitfire, have fairly-docile stall behaviour, with the separated flow producing noticeable buffet over a considerable angle-of-attack range before the wing finally stalls. These are mostly aircraft built with at least one of the following:1
    • Thick wings, which cause the area of separated flow to grow slowly and gradually rather than all-at-once.
    • Low-mounted horizontal tails, which allow the turbulent airflow coming off the wings in the area of separated flow to strike, and buffet, the horizontal stabilisers and elevators.
    • Washout (a slight twist built into the wing going from root to tip, so that the angle of incidence - and, thus, the angle of attack - at the wing root is slightly higher than at the wingtip), which causes the airflow over the wing roots (the portion of the wing whence comes the turbulent airflow responsible for striking the horizontal tail and producing pre-stall buffet) to separate earlier than that over the wingtips, which provides a warning (in the form of said pre-stall buffet) earlier than would otherwise be the case.2
  • Other aircraft (generally those with fairly-thin wings, high tails, and little-to-no washout, such as - rather infamously - first-generation Learjets) stall abruptly, with the area of separated flow growing almost instantly from too-small-to-noticeably-buffet-the-tail to encompassing-the-entire-upper-surface (i.e., stalled). Since there's no significant warning buffet before the actual stall, these aircraft need stickshakers to warn the pilot of an impending stall.

1: In the specific case of the Spitfire, the washout was the deciding factor (thanx @Mark and @Guy Inchbald).

2: Another benefit of wing washout is that the lower angle of attack at the wingtips - where the aircraft's ailerons are mounted - enhances the ailerons' control authority (ailerons become less and less effective with increasing angle of attack), preserving roll control at high angles of attack.

  • 3
    $\begingroup$ Good answer, but a separation bubble happens in laminar flow, followed by transition and reattachment. Here you talk of turbulent flow separating without a chance of reattachment, so there is no bubble involved. Also, what you feel is mostly that separated, highly turbulent flow hitting the tail. High-frequency lift variations on the tail cause that judder. The T-tail of the Learjet is too high to be hit by separation turbulence; therefore it needs the stick shaker. $\endgroup$ Jul 11, 2021 at 12:50
  • 1
    $\begingroup$ Separation bubbles that trigger leading edge stall that cascades down the wing all at once require a pusher. $\endgroup$
    – John K
    Jul 11, 2021 at 13:11
  • 3
    $\begingroup$ Back-to-front stall might be the cause of buffet in the general case, but in the specific case of the Spitfire, the buffet comes from the wings being designed to stall from the root outwards. $\endgroup$
    – Mark
    Jul 11, 2021 at 20:53
  • 1
    $\begingroup$ Note that the Spit has an unusually thin wing for the day, at 13.5% T/C, and a tail a little above the wing. Washout was the key. $\endgroup$ Jul 13, 2021 at 9:20

The Spitfire wing had this behaviour designed in. Mitchell's aerodynamicist Beverley Shenstone had a good deal to do with it. I can't remember who was the main design lead on the wing, but it was not Mitchell himself.

A constant-chord wing will experience some burble before the stall, due to buffeting as the airflow begins to separate near the trailing edge. It was well known that on a tapered wing with no tip washout (reduced AoA), the tips stall first and very suddenly. Worse, the stall then made the ailerons ineffective. In order to avoid this, the Spitfire was given 2.5 degrees of physical twist or washout. The tip aerofoil camber was also reduced (sometimes referred to as aerodynamic washout), to further reduce loading and lower its stall speed.

As a result, the inboard section stalled first, complete with warning burble, while the ailerons remained effective.

The Spit also had a much longer mean chord and hence lower wing loading than the Bf 109, which also lowered its stalling speed.

Although the Bf 109 wing had leading-edge slats to increase tip lift and delay the stall, it had a shorter chord and higher loading; the combination of features on the Spit still allowed it to out-turn the Messerschmitt.

  • 1
    $\begingroup$ Very interesting. Earlier in the book, it is mentioned that the Harvard training aircraft had very sudden stalls which needed some care to recover from, was that a result of the wing design? $\endgroup$
    – DrMcCleod
    Jul 11, 2021 at 9:37
  • 2
    $\begingroup$ @DrMcCleod Don't know much about the Texan/Harvard, but a nasty stall was still common in the mid-thirties and quite likely expected, even if not an actual design requirement for an advanced trainer. $\endgroup$ Jul 11, 2021 at 11:14
  • 4
    $\begingroup$ The 109 had outer slats which prevented outer stall at any reasonable AOA and so could do pretty much the same thing. Your last sentence is the real answer. The difference was the wing loading. $\endgroup$
    – John K
    Jul 11, 2021 at 12:56
  • 4
    $\begingroup$ @JohnK Really, it was the combination of the two. A Spit lacking washout would also have stalled at a higher speed. I updated my answer accordingly $\endgroup$ Jul 11, 2021 at 13:52

The reason for this is a phenomenon called buffeting. The initial separation of the boundary layer induces vibrations in the wing (which at certain speeds can cause flutter). In most aircraft the stall warning in the cockpit is set off by buffeting.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.