Hacker News Reader

Comment by jcalvinowens | Hacker News Reader

It is not necessary to empirically determine a specific airplane's stall speed in order to operate it safely. It's not required in the US, we just use the number the manufacturer publishes.

It's normal for airplanes of the same model to fly differently: I fly a little fleet of six Citabrias, and their stall characteristics are radically different. You'd expect more uniformity from a modern aluminum airplane, but still: nobody should be flying an airplane like this so close to the edge the exact stall speed needs to be known numerically within one knot.

The 40lbs of gas I burn flying for an hour decreases the stall speed by more than 1mph on those Citabrias I fly.

EDIT: I was mistaken, this isn't a requirement in Europe either.

View on HN

> nobody should be flying an airplane like this so close to the edge the exact stall speed needs to be known numerically within one knot.

An experienced pilot can feel the stall coming on with a bit of a "burble" in the stick.

My dad (fighter pilot) told me that knowing exactly where the stall point is is life and death. When you're in a dogfight, the winner often is the one that can turn inside the other. Turning as tight as you can requires getting exactly on that edge of the burble.

It's the same thing as in automobile and motorcycle racing. How close you can get to a slide without sliding is the difference between victory and ignominy.

Imagine dogfighting in a Cessna! I imagine the arms to be pilot wielded colt 45s and first officer to be a jug of whiskey.

Love this description lol

One of the last aircraft shot down in the European theater of WWII was a Piper Cub that shot down a Fieseler Storch with a .45 pistol. They landed next to the crashed Storch, captured the pilot, and treated his injuries. Then handed him over to the Russians, who I am sure also treated him in accordance with the provisions of the Geneva convention.

https://theaviationgeekclub.com/that-time-a-usaaf-piper-l-4h...

Perhaps in accordance with the spirit, but the Geneva convention was in 1949, so those provisions didn't exist yet.

There were several Geneva Conventions prior to the 1949 one; the "Convention relative to the Treatment of Prisoners of War" was signed in 1929.

https://ihl-databases.icrc.org/en/ihl-treaties/gc-pow-1929

The current Geneva Convention was in 1949, but the original Geneva Convention dates back to 1864.

Of course, the previous poster is noting that the USSR blithely disregarded the conventions for others' POWs as well as its own: most returning soldiers were considered traitors to the Motherland and summarily executed or sent to the Gulag to die more slowly.

if you like that idea and science fiction, see the second book in the Greatwinter trilogy by Sean McMullen (read them in order, though)

https://en.wikipedia.org/wiki/Cessna_A-37_Dragonfly

https://en.wikipedia.org/wiki/Cessna_O-2_Skymaster

https://www.airforce-technology.com/projects/ac-208-combat-c...

Sure, that's why I said "airplane like this": in car terms, the plane we're talking about here is a minivan.

> An experienced pilot can feel the stall coming on with a bit of a "burble" in the stick

Stall recovery has also been an essential part of my flight training.

Although can you recover at a height 15m above terrain?

. . . probably not, but does that mean you never practice? Most stall/spin incidents occur at low altitude, but not THAT low. It's often pilot inattention in the landing pattern.

I agree! IANAP though.

Also in this case I suspect they needed to learn proper takeoff procedure. Since you can’t practice dying more than once the mitigation has to be avoidance (in addition to the stall practice you mentioned)

Technically, power-on stalls are supposed to teach how to recover from bad technique on takeoff/go-around and power-off stalls are supposed to teach how to recover from bad technique in the landing pattern.

In reality, they're both training-wheels versions of how to deal with any departure from controlled flight, which is why any professional pilot worth their salt should have a decent grounding in aerobatics and some spin/departure/out-of-control flight experience.

-Am (or at least was) a pilot . . . I could jump in and fly a bugsmasher today if I really needed to, just not legally without a doc visit and a checkride.

The most dangerous situations are when you're flying low and slow like before a landing though, when there's often no time for recovery (especially when you end up in a spin due to asymmetric stall). When I was flying I always worried about this too the point of approaching too fast. But I never finished my PPL anyway

> An experienced pilot can feel the stall coming on with a bit of a "burble" in the stick.

Without any disrespect to the flying abilities of your Dad, as an aerospace engineer, the short answer to that is "no".

You will have the luck to feel a "burble" if the airplane has been been built to be give you that warning. You have absolutely no guarantee that once the airplane has been modified beyond its original configuration, you will get any form of warning. This extreme behavior is actually partly present in the current case, with the stalling point jumping from the wing tip immediately to the wing root. That's no good.

It can be even worse if you're flying a plane that is licensed under the "experimental" category by the FAA. One of the preferred airfoil for some time was the NACA 5-series, as it can have very little lift-induced pitching moment. It has also criminal stalling characteristics with absolutely no warning. Something beautifully illustrated by the following lift polar:

https://media.cheggcdn.com/media/76b/76b33250-7f26-41f6-8f2c...

Is it possible there's a mechanical/design cutoff where most planes designed to be fully mechanical would feel the "burble" while a more modern plane would not, similar to ABS vs non-ABS brakes or powersteering vs none? Are there hydraulic assists for flight controls that might smooth out those mechanical tells?

Not necessarily, no.

It's a design feature to give pilots an advanced, physical warning that they are about to stall, and fall out of the sky. Electronic systems are all nice an well, but they can fail. So why not simply make it a physical feature of your airplane design, that triggers reliably not matter what?

In some very specific edge cases a designer might decided that the planform trade-offs needed to bake the feature into the wing is not worth the performance loss in other metrics, and might forgo a nice, buffeting pre-stall wing. But plenty of pilots have died because they unexpectedly stalled, so it is not a design decision that should be taken lightly.

> You will have the luck to feel a "burble" if the airplane has been been built to be give you that warning.

Airplanes which are designed to do this (most airplanes) do it very reliably, it's not luck. Several of the Citabrias I fly don't have stall horns, all I have is the buffet.

Complexity is always a tradeoff: it's harder to fuck up building and/or repairing a simpler airfoil.

> Several of the Citabrias I fly don't have stall horns, all I have is the buffet.

That's fine, as long as the airfoil and wing planform you chose have a forgiving-enough lift vs angle-of-attack curve that you get plenty of buffeting before you start loosing significant lift and fall out of the sky.

> Airplanes which are designed to do this (most airplanes) do it very reliably

Yes, this is by design. Enough pilots were killed because their planes were not doing this reliably, and they literally fell out of the sky.

Not really, as to your final point; the race driver’s skill is not in avoiding the slide, but — with the mechanic’s help — finding it, and using it. Recall: loose is fast, and on the edge of out of control.

I'll admit to complete and total ignorance, but:

> ...we just use the number the manufacturer publishes.

From the article it sounds like this plane was radically modified, to the point where the manufacturer's stall speed would be irrelevant.

Why wouldn't you want to confirm for yourself where the speed is after so many changes?

The plane had a cargo pack and a Robertson STOL. Cargo packs are essentially for bush planes and as an example, the 1975 Skywagon’s owners manual even had one diagrammed. Robertson STOL’s are extremely common in Northern Canada to the point that even as a passenger I know about them.

It’s nothing too radical.

Edit - Here’s a copy of the 1975 owners manual:

https://www.seaplanescenics.com/documents/1975-cessna-185f-p...

My understanding is that in the US, part of the modification would be updating the plane's official operating limitations, and there could be a new stall number. Still a number from the manufacturer, not a number empirically determined by testing that specific airplane. For special one-off modifications, I don't know: I've always been told that's almost impossible with certified airplanes.

One Citabria I've flown had vortex generators installed on the leading edge of the wings, but the club sold it over a year ago and I don't remember if it listed a modified stall speed. I do remember it said "not airworthy if more than N of the VGs are broken off", I think it was three?

Stall speed depends on so many factors that it can change significantly in a single flight.

Weight, altitude, density altitude, angle of attack etc. are all going to have an effect.

In other words, sure, you might want to confirm it, but you should also give yourself some margin since you don’t ever really know what the stall speed is until you stall.

How do you determine a margin without some form of a baseline?

You read the operator’s handbook which will give you that information.

In a certified aircraft the manufacturer has already done the test flying

Your airplane can also be in the "experimental" category, which generally means "homebuilt" so the stall speeds and characteristics can only be determined through test flight. I made the decision to empirically demonstrate stall speeds on the very first flight of my homebuilt, before attempting to land, so I knew for sure what they were.

I also have a homebuilt, and went and found out my stall speed.

But most homebuilts will still have a stall speed for planes built to plan.

In any case, this is an article about a Cessna. Where the stall speed is in the POH and also marked on the ASI.

Don't forget ice!

And temperature...

> angle of attack etc

This is wrong.

The angle of attack is not a parameter of the stall speed, it is the cause of the stall (for a given configuration, assuming well below transonic speed). This is why for example, for precise handling you should use an angle-of-attack indicator (a large majority of fighter jets and more generally military aicraft have it).

> you should also give yourself some margin since you don’t ever really know what the stall speed is until you stall

The manufacter speed already take into account the afore-mentioned parameters into account, and the resulting speed is the worst case scenario, if not told otherwise (usually max weight, max forward CG).

One should not fly with an arbitrary speed margin, but instead use well-known speedq (1.3Vs, 1.45Vs, where Vs=stall speed in the given config) depending on the flight phase, and remember well the bank angle limit associated with them.

> The angle of attack is not a parameter of the stall speed, it is the cause of the stall

You're right of course, based on the real definition of angle of attack which is based on relative wind. When people say angle of attack is a parameter of stall speed, they're separating speed from pitch.

This isn't really useful but people seem to struggle with the concept of relative wind, so it's a kind of shorthand.

This is what I was getting at. Thanks.

Also there’s planes that have slats and other devices that effectively change the angle of attack of the wing.

And then there’s planes that direct prop wash over the wing so that the power on stall speed is much lower than power off despite the angle of attack being different.

As Juan Browne (blancolirio) likes to say, you can stall a wing at any airspeed and any attiude but only one critical angle of attack.

Radical has all the wrong implications. It's a "major alteration" in a regulatory sense, done from an approved kit of parts, with a very well documented installation and post-installation operation and maintenance procedures.

The aircraft was modified with a Robertson STOL kit. A common type of modification to make to a "bush aircraft". In the USA the modification is covered by an STC (Supplemental Type Certificate), the installation needs to be supervised and approved by an A&P technician with IA (Inspection Authority). The STC will modify the airspeed indicator markings, including the stall speed markings (bottom of green and white arcs), and modify the approved flight manual/pilot operating handbook and maintenance documentation for the aircraft. Since this is a major alteration (in a 14CFR regulatory sense) that modifies the flight characteristics of the aircraft it needs to be test flown after the work, and the STC will also separately requires this. I expect/hope the STC includes instructions for checking stall characteristics including airspeed. In European countries a similar level of regulation/documentation is followed based on the USA STC.

Give the description of the pilot's sad lack of understanding of basic operation of the aircraft I am doubting they even read the pilot operating handbook.

As noted in the article, the plane had been modified with far more than just the STOL kit.

> A further issue was that his Cessna 185 had been extensively modified. The addition of floats, cargo pack, short take-off and landing kit and a three-blade propeller had never had their combined effect documented.

That’s a lot of things that modify the flight characteristics of a plane, all interacting together in what seems to be a previously untested configuration.

I can completely see how each individual modification might modify the planes flight characteristics in a well know manner. But I struggle to see how anyone could realistically predict the result of all the modifications without some basic empirical testing.

A three bladed prop modification along with floats is very common. Wipaire, the leading float manufacturer has its own STC for that. In combination with STOL kits is not uncommon either. I expect the Robinson STOL STC explicitly accommodates floats (Wipaire’ STOL STC sure do) but not sure since I don’t have the paperwork in front of me.

You are over inflating issues here. Are you a pilot? The issue here is straight up incompetent operation of an aircraft.

Operating a floatplane and separately operating with a STOL kits can increase complexity and risks especially when aircraft are mishandled/abused like here with an incompetent pilot who does not know what they are doing. You don’t need to dream up issues from combination of stuff here, the simple linear addition of issues was more than enough for an incompetent pilot to get into trouble, and seems they are lucky they did not get into more trouble before, and lucky they were not killed.

The C185 is a beautiful well behaved workhorse used extensively in bush and floatplane flying, often with multiple STCs to upgrade these aircraft.

I get what you're saying, and I agree this sounds like operator negligence to me.

But the article does say that investigators had to get another aircraft set up with the same modifications and fly it around with bits of wool all over it in order to understand the stall characteristics.

So I assume this plane setup isn't so widespread that its stall behaviour is common knowledge.

European pilot here :) it's also not a European thing. It works the same here as it does in the US, you use the published number in the flight manual.

But this plane had significant modifications done. And if you do things that significantly change performance, you'll need to get updated performance data. Or the provider of the supplemental type certificate that allows the modification has to provide an updated flight manual with that data.

I'm not a pilot or an aviation expert, but I think you're arguing against a strawman here. The investigation report said nothing about determining the precise stall speed, it just recommended "that pilots be informed about the stall behaviour of the Robertson short take-off and landing kit on the Cessna 185 aircraft, both by listing the issue in the aircraft flight manual supplement and through education through the aviation authority". The fact that the stall speeds were not listed on the test flight report was just further proof that the pilot didn't actually test the stall behaviour of the aircraft.

What the investigation report also didn't mention is that pilots should actually practice stall recovery and not just think "if I believe strongly enough that my plane is unstallable, I can ignore stall recovery". But that probably goes without saying...

> It's normal for airplanes of the same model to fly differently

In my experience any equipment this large is effectively handmade, with all the variability implied by this. I'd hope aircraft are made to tighter tolerances than the stuff I work with but a couple of millimeters is often enough to have a noticeable effect on things like actuator travel.

> nobody should be flying an airplane like this so close to the edge the exact stall speed needs to be known numerically within one knot.

This is exactly my takeaway as well. Rather than trying to determine the characteristics of the machine to the n'th degree, assume a realistic degree of variation and design for it accordingly.

> It is not necessary to empirically determine a specific airplane's stall speed in order to operate it safely. It's not required in the US, we just use the number the manufacturer publishes.

It is not necessary as in "there is no legal obligation", true.

But if you want to live, it is absolutely necessary. The numbers provided by the manufacturer will tell you absolutely nothing about the stalling characteristics of the airplane as soon as you start modifying it.

The non-linearity of the phenomena involved in stalling also mean that "intuition" and "small changes" will be of absolutely not help to determine by how much you changed the characteristics of your airplane.

> nobody should be flying an airplane like this so close to the edge the exact stall speed needs to be known numerically within one knot

Citabria's are often flown in aerobatics (Citabria backwards is airbatic and for a while, they were the only aerobatic aircraft being commercially manufactured in the US) and a lot of aerobatic maneuvers involve stalling the wing.

Yes, but anyone doing aerobatics so close to the ground that they can't recover from a stall is understood to be doing that at their own risk.

Anyone doing the above with unfavorable wind has a death wish.

You don't really look at the airspeed indicator for that, you feel the stall in the reversible controls. That feel is incredibly precise.

Decreases the stall speed? How does that work?

Things that can influence stall speed include weight, power, center of gravity, flaps/landing gear configuration, and more.

Why? Well, stall speed isn't a real thing. There isn't a speed at which you stall, that's not how it works. It's a convenient short-hand that we use for the more complicated reality. The physical reality is that stalls happen at a particular angle of attack (AOA) into the apparent wind. That is, the angle of your wings relative to the airflow. Up to the critical angle a higher AOA means more lift to counteract gravity. As you slow down you generate less lift because there's less airflow over the wings. So as you slow down, in order to generate a similar amount of lift you have to increase your AOA. If you keep slowing down and adjusting your AOA to compensate, you'll reach a speed that's low enough and therefore AOA high enough that adding more AOA no longer adds more lift (the air no longer flows smoothly over the wing). That's the stall speed, the speed at which more AOA no longer generates more lift. But it's the AOA that's the problem, not the airspeed.

In addition to lower speeds needing more AOA, you also need a higher AOA if you weigh more. A wrong but illustrative way to think about it might be that you need the engine's thrust pointed more towards the ground the more you weigh. That means that as you burn fuel (lose weight) the AOA that will stall you doesn't change, but the excess AOA available due to your weight-change does so in effect the air speed at which you would be near the critical AOA to stay airborn does change.

Stall speed is still a useful concept especially while landing but it's misleading outside of landing and when anything else is remotely unusual like weight or modifications to the plane. For this reason the FAA has been trying to get AOA indicators installed in planes and to train pilots to look at those instead of thinking about stall speeds https://www.faa.gov/sites/faa.gov/files/2022-01/Angle%20of%2...

The published speed is at maximum takeoff weight, with the most unfavourable center of gravity (usually most forward) and idle power

If the conditions are better (not at max weight, rear center of gravity, engine power adding more airflow over the wings) you can fly below the published stall speed number.

Yeah I was thinking backward. Lower stall speed is better, not worse behavior.

This is a good resource:

https://ciechanow.ski/airfoil/

Useful to think of the airplane as standing still while the engine accelerates the air around it. To fly, you need the air to move over the wings quickly and in the right direction.

You can sort of trade how fast you need the air to go for how ideally the air is flowing over the wings. If you angle the wings just right against the air flow, and/or you bend them out of shape just right with flaps, you can slow down a lot while relying on the air itself to carry your plane. If you're flying against the airflow, you need to go faster.

This is usually done during take off and landing. The pilot lowers the flaps when approaching to land and flares the aircraft before touchdown, all to make the air flow efficiently into the wings, thereby allowing the aircraft to slow down without falling straight down like a stone.

That's why weather is so important for flights. Pilots need to be ready to call TOGA and go maximum thrust at a moment's notice just in case some crosswind or heat wave or something screws up the direction of the air flow just as they're about to land. Many an admiral cloudberg article has been written due to that sort of thing. You angle the plane just right, slow it down just to the edge of stalling, then some phenomenon happens and increases your stall speed past your current speed...

The Robertson STOL mod droops the ailerons with flaps, changing the effective angle of incidence of the wing. A friend had a Robinson-equipped 182 and we could quite comfortably operate in/out of Marlboro Airport (1650' paved with trees at one end and a fence at the other).

That's not TERRIBLY impressive, considering a stock 182P has a 50' obstacle take off distance of 1350' of at max gross on an ISA day

Reply on news.ycombinator.com