Learning to fly is worth it just for the stories you can tell 50 years later.

I never thought of it that way, but you're right.

 

It is not a special jargon term. I am just saying that I was taught how to get out of a spin in ground school, but not done it in practice. Not until a spin happened by accident, flying solo.

I was drilled that you push the rudder pedal opposite the spin direction to the floor. In that moment of panic, I actually pushed the wrong pedal. When the spin tightened, I quickly shifted to the correct one, fortunately, because otherwise, I would not have lived to tell the story. A spin is a stable state until you hit the ground nose down.

At the time I was training, a spin was considered an aerobatic feat.

Hi,

So in real life, what do you think causes a spin in the first place during otherwise normal operation?

 

I'd be happy if we could have a self-driving car that can park itself. A virtual Valet

Drop me off at the door, go find a parking space. When I come out, come and pick me up.
Last but not least: turn OFF the self-driving mechanism (ha ha).

 

Hi,

So in real life, what do you think causes a spin in the first place during otherwise normal operation?

A spin is induced by a stall while in uncoordinated flight. This results in one wing stalling before the other and, as a result, the lift produced by the stalled wing is reduced dramatically while the other wing is still producing full lift.

 

Hi,

So in real life, what do you think causes a spin in the first place during otherwise normal operation?

It was not during normal operations. I was practicing stall recovery. After lowering the nose to gain airspeed, which is how you recover, I pulled up a little too hard to get back to level flight, and a secondary stall, which I was not expected occurred. That is when the right wing dropped suddenly inducing the spin. Fortunately, that kind of practice is always done at a minimum of 1200 feet above ground level, so I had plenty of time to get out of the spin.

 

Hi,

So in real life, what do you think causes a spin in the first place during otherwise normal operation?

In my earlier response, I addressed the technical aspects of what causes a spin. But I neglected to address the "during otherwise normal operations" part. So let me take a shot at that.

Probably the most common scenario is when coming in for a landing and making the final turn from the base leg to final approach. At this point you are likely to be about five hundred feet above the ground and your airspeed is down to a little bit above your touchdown speed and you have minimal power being produced by the engine -- it may even be all the way at idle, but probably just a bit above it. In other words, you are "low and slow". As you make the turn to final, you realize you are going to overshoot the centerline. If there is traffic landing on a parallel runway, you don't want to do this. So you tighten up your turn. But, in doing so, your are in what is known as an "accelerated maneuver" and your stall speed increases because of the additional lift you need to support the turn and an increased bank. Because you are slow, your lower wing starts to drop and so you turn the ailerons to lift it, but this puts you in an uncoordinated turn because you didn't release rudder pressure at the same time because of not wanting to overshoot the centerline. Turns increase drag and uncoordinated flight really increases drag (it's one trick we use to slow down and/or loose altitude quickly), and because you have minimal thrust from the engine, you lose energy rapidly and the nose starts sinking. You pull back on the yoke to counter that, and that initiates the stall. Since you are uncoordinated, that induces the spin. At a few hundred feet above the ground. With minimal airspeed. With an engine near idle. A really good way to end up having a really bad day.

Another scenario is shortly after takeoff. You are climbing under full power and now need to turn to either fly the crosswind leg of the pattern or to exit the pattern entirely. You might be at a steeper than normal angle of climb in order to clear obstacles (tree, power lines, buildings, terrain, whatever). Any time you turn, you need more lift to sustain it, which means that the speed at which you will stall goes down. So, if you were already pushing things hard and climbing near the limits, you aren't much above stall speed to begin with. Now you turn and the stall speed drops to less than you are going and the aircraft stalls. If you weren't making sure that your turn was coordinated, the aircraft will try to enter a spin.

Before DIA there was Stapleton International. There was (I think it's still there) a small airport just south of I-70 east of Denver. The surface area of Stapleton's Class B airspace used I-70 as it's border. Normally when a ground feature is used, it's not specified any tighter than that. Officially, if push comes to shove, it's the centerline of the road. But Denver very explicitly specified it as the northern-most edge of the northern-most lane. The reason is that there was very little distance from the end of the runway and I-70, so pilots needed to turn very soon after takeoff and the FAA wanted them to basically turn out over the center of I-70, so that they had a solid visual reference available. Similar for landing coming the other way. That created a situation in which a pilot could get distracted trying not to bust the Class B and find themselves in a world of hurt because they forget the first rule of aviation -- fly the damn airplane! Everything else is negotiable. Fortunately, and contrary to what some pilots believe, ATC is not out to get you. Controllers at Stapleton were fully aware of this situation and simply ignored you if you busted the Class B while operating out of Arapaho as long as it was promptly corrected. Neither you nor they said anything over the radio about it, because if it was brought to their attention, their ability to pretend they hadn't noticed goes away.

 

In my earlier response, I addressed the technical aspects of what causes a spin. But I neglected to address the "during otherwise normal operations" part. So let me take a shot at that.

Probably the most common scenario is when coming in for a landing and making the final turn from the base leg to final approach. At this point you are likely to be about five hundred feet above the ground and your airspeed is down to a little bit above your touchdown speed and you have minimal power being produced by the engine -- it may even be all the way at idle, but probably just a bit above it. In other words, you are "low and slow". As you make the turn to final, you realize you are going to overshoot the centerline. If there is traffic landing on a parallel runway, you don't want to do this. So you tighten up your turn. But, in doing so, your are in what is known as an "accelerated maneuver" and your stall speed increases because of the additional lift you need to support the turn and an increased bank. Because you are slow, your lower wing starts to drop and so you turn the ailerons to lift it, but this puts you in an uncoordinated turn because you didn't release rudder pressure at the same time because of not wanting to overshoot the centerline. Turns increase drag and uncoordinated flight really increases drag (it's one trick we use to slow down and/or loose altitude quickly), and because you have minimal thrust from the engine, you lose energy rapidly and the nose starts sinking. You pull back on the yoke to counter that, and that initiates the stall. Since you are uncoordinated, that induces the spin. At a few hundred feet above the ground. With minimal airspeed. With an engine near idle. A really good way to end up having a really bad day.

Another scenario is shortly after takeoff. You are climbing under full power and now need to turn to either fly the crosswind leg of the pattern or to exit the pattern entirely. You might be at a steeper than normal angle of climb in order to clear obstacles (tree, power lines, buildings, terrain, whatever). Any time you turn, you need more lift to sustain it, which means that the speed at which you will stall goes down. So, if you were already pushing things hard and climbing near the limits, you aren't much above stall speed to begin with. Now you turn and the stall speed drops to less than you are going and the aircraft stalls. If you weren't making sure that your turn was coordinated, the aircraft will try to enter a spin.

Before DIA there was Stapleton International. There was (I think it's still there) a small airport just south of I-70 east of Denver. The surface area of Stapleton's Class B airspace used I-70 as it's border. Normally when a ground feature is used, it's not specified any tighter than that. Officially, if push comes to shove, it's the centerline of the road. But Denver very explicitly specified it as the northern-most edge of the northern-most lane. The reason is that there was very little distance from the end of the runway and I-70, so pilots needed to turn very soon after takeoff and the FAA wanted them to basically turn out over the center of I-70, so that they had a solid visual reference available. Similar for landing coming the other way. That created a situation in which a pilot could get distracted trying not to bust the Class B and find themselves in a world of hurt because they forget the first rule of aviation -- fly the damn airplane! Everything else is negotiable. Fortunately, and contrary to what some pilots believe, ATC is not out to get you. Controllers at Stapleton were fully aware of this situation and simply ignored you if you busted the Class B while operating out of Arapaho as long as it was promptly corrected. Neither you nor they said anything over the radio about it, because if it was brought to their attention, their ability to pretend they hadn't noticed goes away.

Hi,

Oh that sounds nasty.

So you are talking two engines then?
If so, are they the type where if you do a deep dive you can restart the engines, or is there another way to restart one or both of them?

I was going to ask about doing a deep dive after shutting off the other engine, but if none of the above works that won't work either. Also, that was assuming you were high enough to do the dive, which it sounds like during the landing phase you would be nowhere near high enough. Too bad. I guess the only way then is to be certain this never happens. Circle around again if you have to, is that possible?

 

It was not during normal operations. I was practicing stall recovery. After lowering the nose to gain airspeed, which is how you recover, I pulled up a little too hard to get back to level flight, and a secondary stall, which I was not expected occurred. That is when the right wing dropped suddenly inducing the spin. Fortunately, that kind of practice is always done at a minimum of 1200 feet above ground level, so I had plenty of time to get out of the spin.

Hi,

Oh ok yeah, I had thought that a deep dive would recover, but near landing that won't be possible right?

 

Hi,

Oh that sounds nasty.

So you are talking two engines then?
If so, are they the type where if you do a deep dive you can restart the engines, or is there another way to restart one or both of them?

I was going to ask about doing a deep dive after shutting off the other engine, but if none of the above works that won't work either. Also, that was assuming you were high enough to do the dive, which it sounds like during the landing phase you would be nowhere near high enough. Too bad. I guess the only way then is to be certain this never happens. Circle around again if you have to, is that possible?

One engine, two engines, same thing. I only flew single-engine. Multi-engine is a separate rating and a lot more expensive to get and a lot more expensive to use (since you are paying for an aircraft that has twice as many engines).

I think that you are confusing what is meant by "stall". It's not the engine that stalls, but the wings. The amount of lift produced by a wing depends on a number of factors, such as air density and speed through the air. But a big factor is the angle of attack, which is the angle between the chord of the wing (roughly line from the middle of the leading end to the trailing edge) and the air. As the angle of attack increases, the amount of lift increases, all else being held constant. But at some point, called the critical angle of attack, the airflow over the top of the wing breaks up and becomes very turbulent. At that point, the amount of lift drops enormously. This is what is known as a stalled wing. Aircraft stalls always refer to wing stalls. An engine stall falls under the category of engine failure.

To address about three things you touched on.

Engine stalls: If the engine stalls in most aircraft, it will keep turning because the airflow past the propellor (or the compressor blades in a jet engine) will drive the engine much like a stick rolling down a steep hill might do. This causes a lot of drag, so on some multi-engine aircraft the propellor will turn it's blades straight forward and back, known as feathering, to reduce drag. On a fixed-pitch prop, it is what it is and the engine will stop turning once the airspeed gets low enough, which might be well below stall speed. I've done true power off stalls in which the fuel flow was cutoff completely and I had to hold the aircraft in level flight as long as I could, which means until it stalled. Then get the nose down, pick up enough airspeed to recover from the stall, establish best glide speed, then get fuel flowing again to the engine and go through the restart procedure. I never had to go beyond getting the fuel flowing again because the engine almost immediately lit off every time

Multi-engine: Engine failure on takeoff of a multi-engine aircraft is one very serious situation. You are usually better off in a single-engine aircraft under those conditions. There's a saying that when one engine fails, the other takes you straight to the crash site. It's a situation that can be dealt with, but you've got to be Johnny on the spot. Many of the pilots I knew that were multi-engine rated said that any time they took off in a twin, they always consciously prepared for and assumed that one engine WAS going to fail on takeoff so that, when it eventually happened, they would be able to react immediately. If you lose the engine on a single, it is pretty much a non-event from a control stand point. Depending on where you are in the takeoff, you may or may not have time to attempt a restart, but if not, you just fly straight ahead (which the plane pretty much wants to do anyway) making only shallow turns to avoid obstacles and set it down as gently as you can. In a twin, the loss of an engine on takeoff, especially once you have left the ground, results in the other engine trying desperately to take the aircraft out of control and the pilot has to respond pretty much instantly with the correct actions to maintain control.

Altitude for recovery: The only way to recover from a stall is to get the airspeed up and unless you have serious power to spare, as might be the case with some over-powered aerobatic aircraft, but not anything a typical pilot is going to be flying, the only way to do that is to get the nose down and let gravity exchange altitude for airspeed. This applies no matter how low you are and you have to keep that nose down attitude until you have flight speed and then you have to bring the nose back up slowly enough that you don't induce a second stall. When the ground is racing up at you, that can be hard to remember, but it is even more critical to survival the closer you are to the ground, because you don't have enough altitude to recover from that second stall if you are already that low. If you simply don't have enough altitude to recover, then you are screwed. This is why stall awareness and avoidance is so important, because almost all stalls occur when you are low and slow while doing either a takeoff or a landing.

 

 

I've long contended that the shift to autonomous vehicles will be driven by the insurance companies. If you - a human that's prone to being drunk, preoccupied or asleep - want to operate dangerous machinery around the rest of us, you're going to pay and pay dearly. Once the robots take over, human drivers will be viewed the way drunk drivers are today. We don't shoot them on sight but we want them off the road.

https://www.roadandtrack.com/news/a70074757/autonomous-car-insurance-cut-tesla-fsd-rates/

https://techcrunch.com/2026/01/22/t...-rides-in-austin-with-no-human-safety-driver/

 

I've long contended that the shift to autonomous vehicles will be driven by the insurance companies. If you - a human that's prone to being drunk, preoccupied or asleep - want to operate dangerous machinery around the rest of us, you're going to pay and pay dearly. Once the robots take over, human drivers will be viewed the way drunk drivers are today. We don't shoot them on sight but we want them off the road.

https://www.roadandtrack.com/news/a70074757/autonomous-car-insurance-cut-tesla-fsd-rates/

https://techcrunch.com/2026/01/22/t...-rides-in-austin-with-no-human-safety-driver/

That's an interesting viewpoint.
If the insurance companies deem the self-driving stuff to be better than humans, then they might charge more for insurance for a human driver. It could end up being the other way around though.

This happens a lot where we see different driver factors in industry, and some of them come from places we would not expect at first.

 

I've long contended that the shift to autonomous vehicles will be driven by the insurance companies. If you - a human that's prone to being drunk, preoccupied or asleep - want to operate dangerous machinery around the rest of us, you're going to pay and pay dearly. Once the robots take over, human drivers will be viewed the way drunk drivers are today. We don't shoot them on sight but we want them off the road.

https://www.roadandtrack.com/news/a70074757/autonomous-car-insurance-cut-tesla-fsd-rates/

https://techcrunch.com/2026/01/22/t...-rides-in-austin-with-no-human-safety-driver/

I didn't see how that second article had any relevance to the assertion about insurance, but maybe I missed it.

The first one seemed more like wishful thinking and also is clearly relying on an apples-to-oranges comparison of risk.

When considering apples-to-apples (which is not the easiest thing to do, admittedly), namely human drivers in the same areas under the same conditions, full-AV vehicles have about half the police-reportable accidents but two to three times as many minor fender-benders (on a per mile basis) as human drivers. While Tesla Level 2 semi-autonomous has about twice the crash rate of human drivers under similar conditions.

And we are still a LONG way from having fully-autonomous vehicles that can operate on more than a couple percent of roads in the U.S. (even if regulatory constraints were simply removed).

 

I didn't see how that second article had any relevance to the assertion about insurance, but maybe I missed it.

No, it's just related news.

The first one seemed more like wishful thinking and also is clearly relying on an apples-to-oranges comparison of risk.

I don't follow. The insurance industry is seeing the lower accident rates provided by assisted driving and adjusting premiums down accordingly. Even my Subaru, which has adaptive (radar) cruise control, gets a break.

And we are still a LONG way from having fully-autonomous vehicles that can operate on more than a couple percent of roads in the U.S. (even if regulatory constraints were simply removed).

Agreed. The dystopian future I allude to won't happen in my lifetime. My next car purchase may well be my last, and it's not going to be an EV. Real cars have to operate reliably
on country roads during snow/ice/dust/mud storms. I know they're putting heaters in the sensors to help with ice and snow but these will surely be overwhelmed at times. Dust and mud could render them useless in minutes. Are they going to add wipers?

On the other hand, in dry and southern cities, I could see robotaxis becoming the norm in just a few years. It's a niche application they're well suited to. They'll grow into other niches over decades.

 

How do robotaxis actually make money?
https://electrek.co/2025/04/16/elon...alysis-that-showed-robotaxi-would-lose-money/



If the vehicle really is autonomous.

Being competitive and displacing competitors means significantly under-pricing existing taxi and ride-share networks.
What's the resale value of a car with no steering wheel, very high miles and likely needing a expensive battery replacement for normal driving. A fleet of robotaxi will be depreciating to 100%.

 

What's the resale value of a car with no steering wheel, very high miles and likely needing a expensive battery replacement for normal driving.

Tesla designs for 1 million miles life expectancy, and the battery life is part of that. There's no such thing as replacement. Does anyone really care what it's worth at that point?

But yeah, an interesting analysis.

 

I don't follow. The insurance industry is seeing the lower accident rates provided by assisted driving and adjusting premiums down accordingly. Even my Subaru, which has adaptive (radar) cruise control, gets a break.

I don't think they are -- we have to be careful not to confuse autonomous driving capabilities with other features such as radar-assisted cruise control. The latter are frequently given discounts very much like discounts for other safety and security features have, such as side airbags or antitheft devices and tracking/recovery capabilities.

This particular article is talking about a company, Lemonade, that is trying to enter the auto insurance market and essentially using a discount for autonomous driving capabilities as a marketing gimmick. Notice that they themselves are claiming that this is a first-of-its-kind offering, indicating that the insurance industry is NOT adjusting premiums down based on autonomous capabilities. Lemonade's rationale is based on a fanciful premise, as indicated by their own president's statement that, "a car that sees 360 degrees, never gets drowsy, and reacts in milliseconds can’t be compared to a human."

There are also a number of confounding factors when it comes to how this will all affect insurance rates. The costs to repair cars with autonomous driving capability are significantly higher because of all of the sensors and systems involved. A much bigger one, and one that will probably have to be hashed out in courts, is who is it that eventually needs the liability protection. Traditionally, fault for accidents is usually assigned to the driver, with fault occasionally being assigned to the owner in the case of a defective vehicle due to maintenance issues. Manufacturers are seldom assigned blame for specific accidents, though they can face class action lawsuits for systemic issues with a vehicle. But when the driver of a car is merely a passenger and they get in a wreck, how can the driver be held liable? This would be like blaming the passengers on a bus for an accident and requiring them to carry liability insurance that pays for the damages in order to ride on a bus. So who should pay? The vehicle manufacturer or the AI software developer? If so, they will have to factor the costs of all of those accidents and lawsuits over the entire life of the vehicle into the price of the vehicle.

This is actually something that we've seen before with the small-aircraft industry. Manufacturers were getting hit with lawsuits on fifty-year old airplanes being operated in irresponsible ways resulting in huge settlements, since juries are notorious for giving out large awards against a big company that they would never assign if the defendant were some soccer mom. As a result, Cessna ceased making small aircraft in 1986 solely and specifically because of this issue and, overall, production of single and light twin piston aircraft dropped more than 90%. Insurance carried by the manufacturer went from about $50/yr in the early 60s to over $70k/yr in the late 80s. It took tort reform that capped liability at 18 years post-sale before production started up, but even now it is a tiny fraction of what it used to be. In the 1970's there were about 17,000 small aircraft made in the U.S. and today it is well under 1,000/yr. Also, the cost of an aircraft when up by about a factor of ten after production started again because of the remaining liability picture.

And the "solution" that allowed aircraft production to start again isn't one that would be available for autonomous vehicles. It's one thing to say that liability ends for an aircraft once it has been out there for nearly two decades, since the manufacturer isn't involved in the day-to-day operation of the aircraft and that liability shifts to the owner/operator. But the manufacturer of an autonomous vehicle is inextricably involved in every driving decision that vehicle makes for its entire life and it is unreasonable to ever shift that liability to the owner/operator for those decisions. So what's going to happen to the costs of those vehicles as the liability landscape establishes itself? Another thing to consider -- what happens to cars that are already on the road when the manufacturer goes out of business because they can't afford the liability costs any more?

Agreed. The dystopian future I allude to won't happen in my lifetime. My next car purchase may well be my last, and it's not going to be an EV. Real cars have to operate reliably
on country roads during snow/ice/dust/mud storms. I know they're putting heaters in the sensors to help with ice and snow but these will surely be overwhelmed at times. Dust and mud could render them useless in minutes. Are they going to add wipers?

On the other hand, in dry and southern cities, I could see robotaxis becoming the norm in just a few years. It's a niche application they're well suited to. They'll grow into other niches over decades.

I could easily imagine islands, such as downtown areas or large university campuses, that move toward eliminating human-driven vehicles and relying primarily on autonomous vehicles. But these face all kinds of challenges. First, it is very difficult to re-engineer existing areas to make them car-free friendly. Second, there is a huge resistance on the part of the public to accept them because of the added inconvenience (some of which is an illusion). It can work if the area is designed from the ground up with that goal, consider large amusement parks like Disney World. But those are fundamentally different environments. First, people are paying a lot of money to go there, which provides the funds to establish, maintain, and operate the infrastructure. Second, people aren't going there every day or even every month in order to do short-term business. Most people are going there once every many years and are going to spend the entire day there. Could they operate if, instead, they had to rely on people that wanted to come there once a month to spend an hour shopping?

 

 

I have to question whether that video was staged. None of the people involved acted as I would expect them to. I cannot believe the construction workers just went about their business instead of trying get the car to turn around. The passenger seemed all too calm. And the control center rep said nothing about what she was doing to fix the problem. It just seems very off to me, not the way people would react to an unexpected situation.

 

I looked at other videos of that user. Anything is possible today but it's pretty dull if staged. Construction guys see lots of confusion from random drivers, a slow moving car moving though a street work site is not totally unexpected, you stay alert, do your job and let idiots pass.