Stalling - why is it so dangerous?

[John Cole]

Well, obviously, loss of lift is always bad for a aeroplane, but I don't think Martin Bedding's article (August RCM&E p. 88: Stall School) fully explained why it makes planes crash even though he gives excellent recommendations on how to avoid stalling - particularly on the approach.

It's a problem if the plane's wings both stall, as the plane goes nose-down until (if you release the up-elevator) it reaches flying speed again.  Hopefully, this is before the plane "lands".

 

It's a BIG problem if ONE wing stalls, as recovery is then a bit more difficult.

 

Let's get some terminology straight.  A wing's Angle of Attack is the angle between the (local) airflow and the chord line of the aerofoil; the line drawn between the leading edge and the trailing edge.  What makes a wing (or any other flying surface) stall is if the AoA exceeds the stalling angle.  I say "local" as the AoA can vary down the wing, and vary between the two wings: see below.  When this happens, the (local) airflow over the top surface of the wing goes highly turbulent and fully breaks away from the wing, causing a dramatic loss of lift.  What makes it exceed the stalling AoA?  Excessive use of the elevator.  That's it.

 

Note I say LOCAL airflow.  Here are two key examples of how you get local stalling.  Take a parallel-chord wing,  the plane flying straight - with up-elevator being gradually fed in (at constant lowish power setting).  The plane will slow, nose-up and eventually the wing will stall and nose-down.  Now the air flowing over the wing near the root goes more-or-less straight back.  The air on top of the wing is at lower pressure (that's what gives the lift); that underneath is at higher pressure.  Near the wingtip the air on the upper surface does NOT go straight back but veers inwards towards the root, "pushed" there by air flowing round the wingtip (from bottom to top) because of the pressure difference.  So what?  Well, as it goes sideways-and-backwards, the air travels further going from the LE to the TE near the tip than it does near the root.  What this does is to reduce the local AoA near the tip.  So a parallel-chord wing will tend to stall near the root first, and at both roots simultaneously if the plane's flying dead straight.  That's good news, as the plane is more likely to stall straight-ahead, with simple recovery.

 

But only MORE likely.  Consider the second example.  If the plane is turning - say a level turn to the left - then you can see that by the time it's done a full circle the right wingtip has travelled further through the air than the left wingtip.  The difference is good old Pi times twice the wingspan.  It's easier to understand the consequences of this if we think of the plane gliding, as during this full turn both wingtips will have descended through the air.  The average AoA is then given by the amount of descent and the circumference of the turn.  What this means is that the AoA at the left wingtip is greater than at the right wingtip (because it descends the same amount, but over a shorter distance - it's flying at a lower airspeed), and the same applies for any two corresponding points on the two wings.  So the left wing stalls first.  Lots of lift on the right, almost none on the left, quick quarter-roll left and dive towards the ground.  Crunch!  So that's why planes crash when you do a turn at low speed, using enough up-elevator to provoke this misbehaviour (termed an Incipient Spin).  I explained using the example of a plane gliding; it's essentially the same for under power, but more complicated to explain.

 

Notice I said nothing specific about tip-stalling.  Whether this happens depends on lots of things (spelled out by MB).  But if one wing stalls before the other, you're in trouble if you're close to the ground - whether it's tip-stalling or not!

 

So the recommendations for the final turn on the approach are: don't fly too slowly; keep the power on and don't bank too much - and don't do a tight turn.  Or it really will be a final turn!  Similar advice applies immediately after take-off.

[Martin Harris - Moderator]

On your last point, although not a problem for the average model flyer, it's important to avoid any temptation to over-rudder the final turn.  While I agree that a tight turn leads to an increase in stalling speed, in a typical deadstick approach getting low and trying to stretch the glide, then using rudder to assist in  "avoiding a tight turn" is a great way to spin in.

 

This is definitely more of a problem in the real world where use of the rudder is more instinctive but it's important not to get people to equate a tighter but balanced turn with sufficient airspeed (derived from AoA of course) with being unsafe as opposed to a skidding turn with a lesser bank angle and far more chance of a spin resulting.

[Gemma Jane]

Timbo any chance John's thread and mine on this subject can be merged? I think this is important and could save a lot of models.

 

John, though your description of 'why aeroplanes stall in the turn' is not incorrect there is also the factor of side-slip.

 

Generally I now believe this is actually the biggest factor for stall/spin on the final turn due to breakaway of the airflow over one wing. Regarding full size this was often because of pitot tube ' positioning error' and killed many pilots, they were flying slower than they thought they were, modern aircraft largely avoid the problem with better positioning and pitot design .

 

However I also believe it is a major factor with models. The cure is co-ordinated use of rudder/aileron to prevent side slip and keeping bank angle low. Really for a model with a high wing loading or a tendency to 'tip stall' and I use the term in the modelling sense of 'dropping a wing at the stall' rather than the true sense, bank angles as low as 20 - 30 degrees will give a reasonable margin of safety.

 

If your wondering why side slip is important, consider if you add to your description above what has happend to the relative airflow (in my terminology free stream velocity) if the aircraft 'slips' into the turn.

 

 

Edited By Gemma Fairchild on 04/07/2009 01:17:06

[Martin Harris - Moderator]

Gemma,

 

Isn't skid more of a factor in an incipient spin than sideslip?  I remember reading an article by Brian Lecomber many years ago, where he advocated the use of the slipping turn as a method of turning back after engine failure with relative safety when faced with no viable alternative within the conventional 30 degrees either side of the runway heading - the example given being a runway that he trained students from that projected into shark infested waters in the West Indies - where he claimed to have actually taught this technique.

 

The reasoning was that reduced elevator effectiveness due to blanking in the sideslip meant that it was unlikely that a stall would be induced and a turn could be completed faster with less height loss than a balanced turn due to the lack of the G build up that occurs in a tight balanced turn.  I don't recall all his reasoning but he did point out that it was a strange aircraft that would spin against opposite rudder. Possibly a flick over the top might occur in extreme circumstances?

 

I would say from experience in gliders that this has some credibility.  Certainly, my 1935 Rhonbussard which had generous side area and sideslipped most effectively would require pretty much full elevator to hold the nose up in a fully developed sideslip.

 

A major part of training students in the circuit was avoidance of the over ruddered final turn which I was always given to understand has killed many  a full size pilot. I certainly agree with you that in normal circumstances a balanced turn is the ideal condition to aim for and in the circuit the norm should be well planned turns at reasonable bank angles but given adequate airspeed there is nothing inherently dangerous in a tighter turn or sideslipping turning approach which might be appropriate when flying an aerobatic model in a scale display manner, for example.

Edited By Martin Harris on 04/07/2009 02:38:29

[Tim Mackey]

Gemma says..

"Timbo any chance John's thread and mine on this subject can be merged?"

Sorry Gemma...which of John's threads are you meaning?

[ken anderson.]

hello all-i'm not in to the dynamic's of it all as some of you are-very educated on the subject--i take my hat off to you's.........but surely we/you can't equate what you are talking about 100% wholeheartly to a small model aircraft--i was allway's told that a model wing wouldn't behave anywhere near as effective as a full size one---and over the year's it amaze's me when i hear people comparing their model's to the fullsize counter part's!...."the fullsize spit had the same fault's etc"......i,ve been watching for year's at all the show's etc for a full size 'junior 60'..........

 

        ken anderson........

[Martin Harris - Moderator]

Ken,

 

There's no denying Reynolds numbers (which in simple terms is a correction factor aerodynamicists use to take into account the relative thickness of the air - Gemma would no doubt explain this much better) and the larger a model gets, the more it takes on the characteristics of the full size, however, the principles are much the same for any flying object (the bumblebee excepted!)

 

Certain factors, such as tip stalling being accentuated by wing planform, the propensity for Dutch rolling from swept wings, pitch stability from a long tail moment etc. etc. are directly related to the shape of the model / full size so some of these people may have had a point although I agree that much of it may have been pure speculation!

 

 

 

 

Edited By Martin Harris on 04/07/2009 10:07:07

[Gemma Jane]

Martin I think essentially we shouldn't get too hung-up on whether it is slip or skid.

 

Yes I think in many ways a skid is worse, not least because of the additional factor of fuselage blanking of the airflow over the lower wing.

 

Is it skid or is it slip? It is either, it is the lack of co-ordinated turn that is the issue and the tendency to skid or slip depends on a number of factors including momentum entering the turn, the secondary yaw characteristics of the aircraft (model) in question, fuselage side area etc etc. Slip still modifies the airflow over the low wing as does skid. I also fall over my pilot/engineer terminology blocks on this as to me a slip is either way to a pilot it is a slip in and skid out!

 

Ken, this isn't a direct comparison of a particular scale model to its full-size counterpart. Models do exhibit the same physics as larger aircraft, smaller models may be at lower Re numbers but that has nothing to do with flight dynamics. Unfortunately it is part of modelling 'folk law' that they are somehow different which appears to fuel much 'club expert' talk - lets face it it allows an club member to be an 'expert' . The fundamentals are exactly the same whether it be a junior 60 a full-size Piper Cub or a 747. A classic example is club use of 'tip stall' for any stall where the wing happens to drop

 

Tim, I meant merging this thread with the one I started regarding the technical inaccuracies in the article. Both threads raise many of the same points.

 

I think at some point we will need a summing up 'what you actually need to do' rather than the technical talk but I think it is good to air the issues involved in the first place.

[Gemma Jane]

A cross post there Martin, essentially the same points made though

[John Cole]

Martin: I said "avoid a tight turn".  This was not meant as a technical term, and I absolutely agree with you that excessive rudder on the turn onto finals is likely to provoke a one-sided stall.  I said "tight turn" rather than "steep turn" with exactly that point in mind, but it's good you clarified it.

 

Gemma: yes, spot on.  But before I set out my thoughts on sideslip (or indeed skidding) let me add something to my initial comments - which I really should have put in to start with.

 

What I was trying to do is explain what (in my opinion) it about stalling that makes it so dangerous.  And I missed out something.  First let me expand on "tip-stalling".

 

When a model starts to stall and falls off sideways, we often call this a tip-stall.  As I explained in the example above, for a parallel-chord wing, one wing will stall before the other if the plane is turning at the point of stalling.  What I left out is what happens next.  Partial loss of lift from (in my example) the left wing means that the plane rolls left and the left wing goes down.  This action INCREASES the AoA on the left wing (with more increase as go out along the wing) and straightaway the whole left wing is stalled. Conversely, this roll reduces the AoA of the right wing so it remains unstalled.  That's why an initial one-sided stall turns into a flick.  Of course, all this changes when a quarter-roll is complete, but on the final turn it may not get that far!

 

Sideslip and skidding.  The links to one-sided stalling are I think similar.  In both cases the plane is moving sideways through the air, and I'm going to call it "sliding" and not try to distinguish between the two.  The link to stalling is I think different depending on whether the wing has dihedral or not.  Consider the case where (as in my example) the plane is turning left (and banked left) , and is also "sliding" to the left - so the airflow over the entire plane is coming from a direction slightly left of straightahead.

 

Now, consider a plane (with dihedral) near the stall, turning and banked left and sliding left: airflow over the left wing will be slightly sideways going in towards the root, increasing the distance travelled from LE to TE and thereby decreasing the local AoA (the "simple sliding effect").  But the dihedral means that the angle of incidence of the left wing is in effect increased (that's what makes rudder/elevator models bank when rudder is applied).  So these two effects are in opposition.  For the right wing the sliding also decreases the AoA (with the air moving OUT away from the root as it goes back) but so does the dihedral effect (effectively reduced incidence).  So a plane with dihedral sounds to be more prone to one-sided stalling in a sliding turn.  But the two opposed factors affecting the left wing: which is the bigger effect?  The more dihedral, the bigger the "increased incidence" effect, so for any degree of "sliding" there's a dihedral angle beyond which stalling is made MORE likely by "sliding" than not.  And that angle is VERY small.  Take an arbitrary example: with 10 degrees average AoA and 5 degrees of sliding and 5 degrees of dihedral (that is, each wing points up 5 degrees from the horizontal) the effect of "simple sliding" on AoA is less than 0.1 degree but the dihedral effect on AoA  is nearly 1 degree. But it's also affected by the issues that affect a plane with no dihedral

 

What about a plane with no dihedral?  Well, it then all gets a bit messy as the primary effect of sliding is to put the inner part of the right wing into dirty air which has flown round the fuselage.  And it's different for a high- or low-wing plane; for a low-wing plane this disturbance primarily affects the flow over the upper surface (where the stall might happen), and for a high-wing plane it affects primarily the flow underneath.

 

Gemma's right: what matters most is what we need to do to avoid a stall-provoked crash.  And it's pretty much down to avoiding excessive elevator (as this is the principal control that stalls the wing), and being particularly cautious about this in a turn.  My personal opinion is that it's safer to add power in a turn rather than increase elevator.  And of course, keep the CoG well forward! With i/c powered planes landing with a near-empty tank the CoG will be at its rearmost point during the flight, and hence it will be most stall-prone AND elevator-sensitive.

[Tim Mackey]

Sorry Gemma, I dont know where that thread of yours is... I am up to my eyes in it at the moment, so please post a link to it here, and I will try and merge...although believe it or not..... its hard to do actually

Edited By Timbo - Moderator on 04/07/2009 12:09:44

[Gemma Jane]

Tim, things are progressing on, it might just confuse a lot of issues now! Let it go but thanks for considering it.

 

I agree with you points John entirely.

 

I'm not sure if it is excessive elevator that is the primary cause. But I do see your point and also how one can perceive it as that.

 

I think you'll agree 'excessive use of the elevator' is actually a symptom of 'too low an airspeed' This is where things get very confusing to explain regarding models, as our job is a little harder on the ground to fully appreciate the airspeed the model is experiencing without the aid of an ASI that a pilot would have. We have to picture in our  minds the affects the wind is having on our perceived airspeed of the model.

 

Of course the other big factor with the an excessivley steep turn which we haven't mentioned in the discussion is G loading or increased wing loading. We also now  have the complication that as the bank angle increases to the wing the model appears to be getting heavier and heavier, in fact the stall speed is increasing ready to bite us with a nice break away of airflow from the lower wing compounded by all the other factors raised.

 

Edited By Gemma Fairchild on 04/07/2009 12:52:38

[Terry Whiting]

I just can't believe what I'm reading here.

I feel sorry for any potential flyer  reader this jargon who was  thinking taking up this hobby may well have been put off for life. 

 

How did I manage all these years? 

 

 

 

[John Cole]

Gemma: you are of course correct about high-speed stalls (incluuding in steep turns) but that was pretty much covered in the original article.

 

Ken (Anderson): I think there's another point about the difference between models and full-size, and that's the wind - and in particular gustiness.  As models are smaller and fly slower they are potentially more affected by any particular speed of wind or gust - as it's a greater proportion of the total airspeed. And I think it's true that we will fly our models in gust-speeds well in excess of the proportionalte gust-speed which would stop full-size flying.

 

The particular relevance of this ot stalling is that if the plane is flying just above the stall then a change in windspeed can provoke the stall.  As I set out above, once a wing starts to stall and drops, the AoA increases further, deepening the stall.  So if the change in windspeed reverts (even very quickly) the plane then remains stalled.

 

Terry: how did you react to the original article?  I agree some people like to know how things work, and that others don't find it interesting (and indeed boring).

[Gemma Jane]

Well Terry, you know what, I use to think that about Timbo's posts on electrics. How wrong was I about that.

 

Those who can follow it may benefit.

 

 

John as both full size pilot and model flyer, no sorry the proportional 'gust' speed is not an issue. Trust me a full size light aircraft feels gusts too, remember though you are thinking in terms of the 'gust' being proportional to the scale, sorry but a light aircraft has a much lager suface area and beleive me on finals you know all about gusts.

 

It has been talked about before on here, a gust does not and cannot cause a section to stall. What a gust can do is confuse the model pilot, who then stalls the section themselves

[Stephen Grigg]

Isnt stalling not flying the model fast enough,therefore if you put the model high up,youcan then see when it stalls and ensure you dont fly that slow in future

[Gemma Jane]

If you look at it just like that Stephen, which is probably as far as instruction for the A test goes, don't be too surprised when you break a model when it stalls when you didn't expect it. We could essentially all fly so fast on approch we don't stall, then you hear everyone moans because the undercarriage isnt' strong enough on their models

[Stephen Grigg]

Thats what I get when approaching a landing,helpfull people telling me not to stall.I do find that approaching a little faster than I want tends to make the model more stable and controllable when landing.If I get to find my Super Air and its any good that will be the test for me.The first time I was ready to land I ran out of fuel,and then I lost signal for some reason and lost site of where it went,so havent tried to land it.I hope it landed itself

[Gemma Jane]

The greater airspeed gives you more airflow over the controls Stephen, that is why it feels more positive. But get away from tough as old boots trainers and you will find those sort of fast approaches will end up ripping the undercarriage clean out of the model.

 

Try flying a few high circuits but throttling back and getting the feel of 'sloppy' controls. It's all practice really and any amount of jargon doesn't change that. Though for some people understanding the physics can help them interpret what they are seeing and doing.

 

[Tim Mackey]

I darent really enter this mysterious and technically challenging room, as I am way out of my aerodynamic depth here, but one thing which I think is fair to say to the likes of Stepehen and others who are still fairly early on in their flying and model types is that many models will not bite really anyway.

Take for instance that classic WOT4.

A large symemetrical surface area, and very thick section wing produces oodles of lift, and of the many I have owned and flown with small and large engines, not once have I ever stalled one accidentally.... landings can be really slow, it can be put into silly steep banking turns when it goes deadstick etc, and it it likely isnt gonna tip stall unless you do something to deliberately provoke it...even a deliberate spin or stall turn often requires a fairly rearward COG to get it going.  Rambling here...so what I really mean is, dont be getting too worried about all this stuff Stephen and Co... the type of model you are generally flying around with at the moment should not bite.

I shall get out quickly now before the aerodynaboffins shoot me out the sky

Edited By Timbo - Moderator on 05/07/2009 13:16:24

[Gemma Jane]

It's perfectly true, really most of this is going to apply to models with higher wing loadings and tapered plan wings, so most likely the more scale end rather than sports models or trainers. Even the full-size I trained on has had most nasty traits removed from it in the design process, stalling lessons were more a case of, 'can I actually get the thing to stall'

 

I kind of look at it all in the same way at times, what does any of it matter, yet on the other hand for some with engineering backgrounds etc it is interesting to discuss - it shouldn't put people off as nobody needs to know it.

 

 

I once got into a massive debate on a full-size forum regarding aerodynamics that seemed to scare a lot of people.

 

Fact is, whatever we do or do not know, birds still fly just fine and no amount of jargon makes them fall out of the sky

[Stephen Grigg]

Doed that mean we should build birds Gemma.I do notice a lot of the Limbo Dancer type models do unbelievable acrobatics then land at almost no speed at all.Is there a direction there for us who are moving from  high wing trainers.

[Gemma Jane]

I think the issue underlying it Stephen is some of us are interested in this stuff and that means others do not have to be.

 

It's all very well Terry saying 'I got by fine without it' but the truth is that is because other people understood the issues involved and saved him the trouble.

 

If I'm stupid enough to design bad traits into the BV 138, rest assured I'm smart enough to remove them also 

[Tim Mackey]

Posted by Gemma Fairchild on 05/07/2009 13:43:02:

It's perfectly true, really most of this is going to apply to models with higher wing loadings and tapered plan wings, so most likely the more scale end rather than sports models or trainers. Even the full-size I trained on has had most nasty traits removed from it in the design process, stalling lessons were more a case of, 'can I actually get the thing to stall'

 

I kind of look at it all in the same way at times, what does any of it matter, yet on the other hand for some with engineering backgrounds etc it is interesting to discuss - it shouldn't put people off as nobody needs to know it.

 

 

I once got into a massive debate on a full-size forum regarding aerodynamics that seemed to scare a lot of people.

 

Fact is, whatever we do or do not know, birds still fly just fine and no amount of jargon makes them fall out of the sky

 

LOL...well said Gem !

 

[Martin Harris - Moderator]

Getting back to Stephen's questions on the practical application of all this and getting used to the model's stall, the one way to judge a model's airspeed is by its attitude.

 

Although with sufficient elevator power it is perfectly possible to stall a model (or of course, full size) with it pointing directly at the ground, in normal flight conditions with approach  power or in the glide, the elevator controls the airspeed to the major degree.

 

The airspeed is directly related to the aircraft's pitch attitude in steady flight.  With sufficient power to maintain height, lowering the nose will increase the airspeed, raising it will reduce it.  Yes, the height will decrease or increase (at least initially) but we are adding or subtracting total energy by lowering or raising the nose. Neither of these is of much use when thinking about landing so the way to alter height is by the subtraction or adding of power.

 

Moving on, we can establish a safe attitude by learning where the stall occurs (which can't be too high otherwise we can't see it properly) by gradually raising the nose at typical approach power until it all goes mushy (most trainers won't give a sudden nose or wing drop but be aware of the possibility by trying it with plenty of height first) and you'll then know what the model will look and feel like at the stall.  You don't want to go anywhere near there until your an inch or two off the ground when you're in the circuit.

 

What we need is sufficient airspeed to ensure an adequate margin for inattention or more particularly, gusts.  So in light wind and no turbulence we don't want too much airspeed - especially if we have a restricted landing length.  In a higher wind and particularly if we're expecting turbulence we need a margin of airspeed in hand and so we need a more nose down attitude.  If this means the model is coming down too quickly we need some more energy to hold it up in the air - and we've got some of this on tap via our left hand (mode 2) i.e. add a little throttle.  What we don't do is try to stay away from the ground with the elevator which will simply raise the nose and reduce the airspeed while giving a temporary increase in height.  Of course, these are the primary effects of the controls and they always need to be balanced against each other but this is something that becomes instinctive with a little practice.

 

So get used to how the model sits in the air i.e. how much nose down it looks at a safe and suitable approach speed and aim to keep it looking like that from the turn onto base leg until you are close to the ground. Full size practice is to trim for this attitude but it's not very practical with digital trims although perfectly possible.

 

The landing phase is simply an approach to the stall carried out at zero altitude i.e. reduction of power to minimum (removing the energy to stay airborne) coupled with a gradual increase in pitch attitude to compensate and reduce airspeed to the point where the aircraft ceases to fly.  Achieving this 2 feet high or potentially 2 feet below ground level results in a long face, within an inch or two of the ground gives a warm glow of satisfaction!

 

Hopefully none of this involves anything very technical and will be useful to those who don't feel the need to understand the aerodynamics.  As Gemma has eluded to, birds know nothing of any of this but they are in a position to feel every nuance of the air and have several million years of evolution to tune their instincts to the task in hand compared to a couple of hundred years worth of theory and only hours of practical experience for a new pilot.

Edited By Martin Harris on 05/07/2009 15:36:41

Edited By Martin Harris on 05/07/2009 15:42:50