Many years ago I played with Microsoft Flight Simulator, turning off my engines and trying to land the plane. Coming down too steeply would crash, as would coming down too slowly. So I figured out the right attitude (angle of descent) to successfully stay in flight until reaching ground level.

Actually landing was another puzzle to solve. I had to "swoop" in—that is, just before landing, descend steeply for a short time to pick up velocity, then use that stored energy to glide in and land.

Was this realistic at all? Are there any changes that need to be made to the overall abstract process?

  • 3
    $\begingroup$ @GregHewgill -- Wish I could remember. That was 1987-89. Do you happen to know what the choices were? $\endgroup$ Jun 12, 2023 at 0:43
  • 8
    $\begingroup$ @GregHewgill: Someone using MSFS in '87–'89 would most likely have been flying a simulated Cessna 182. That was the default choice, though a Learjet 25G or a Sopwith Camel were also available. $\endgroup$ Jun 12, 2023 at 17:48
  • 1
    $\begingroup$ As a point of information, I was once called to dinner and left MSFS running in the '85-87 era, to find on return that the plane had landed itself. /However/, my recollection is that I was using a glider ("sailplane" etc.), and even if the software was prepared to waft that down gently it doesn't mean that it would have behaved the same for another type of aircraft. I'd add that in "real life" I believe there's been at least one case where an airman ejected from an uncontrollable fighter, which later recovered and grounded itself with only moderate damage. $\endgroup$ Jun 13, 2023 at 16:11
  • 6
    $\begingroup$ @MarkMorganLloyd: That would be the "Cornfield Bomber". $\endgroup$ Jun 13, 2023 at 17:04
  • 2
    $\begingroup$ Im surprised no one has yet posted real life examples - for example, the Gimli Glider, a 767 which lost both engines due to fuel exhaustion at cruising altitude, glided to a successful landing - en.wikipedia.org/wiki/Gimli_Glider Or Air Transat Flight 236, an A330 which again lost both engines at cruising altitude due to fuel exhaustion, glided for over 120KM to a successful landing en.wikipedia.org/wiki/Air_Transat_Flight_236 $\endgroup$
    – Moo
    Jun 15, 2023 at 1:17

4 Answers 4


Pilots of single-engine aircraft practice this kind of thing all the time, indeed it's a requirement to do it in most flight tests to actually get the licence/certificate.

There is a certain airspeed on every aircraft known as the best glide airspeed. This airspeed delivers the best lift to drag ratio, the end result being you will glide a further distance than you would at any faster or slower speed. You can maintain speed without engine power by pitching the nose down, and vice versa to slow down.

As for the actual landing itself, the swoop procedure you describe is not necessary as there is already enough energy at the best glide speed, so you can just flare as normal. For example, the Piper aircraft I fly has a best glide speed of 73kts (this can vary with weight and conditions), but the normal approach speed without flaps is 70kts anyway.

Twin engine jet pilots don't practice this as the likelihood of it happening is so remote, but the principles are largely the same.

  • 11
    $\begingroup$ And, just to be clear, when pilots of single-engine aircraft practice this they don't shut off the engine. They set it the throttle to idle; that way, if they come up short of the runway they can recover. $\endgroup$ Jun 12, 2023 at 15:22
  • 4
    $\begingroup$ You might want to be more specific about what you mean by "this kind of thing". I think you mean engine out scenarios, but given that the question is about the landing swoop, it initially seemed like you meant that. $\endgroup$
    – nasch
    Jun 12, 2023 at 15:29

This is not realistic for most scenarios. Every airplane has at least a published “best glide” speed that provides the best glide distance horizontally for the amount of altitude lost. My airplane has a glide ratio of just over 9:1 at this speed, so if I am 3,000 feet above the ground, I can under ideal conditions glide 27,000 feet, or just under 5 statute miles.

There’s a slightly different speed that maximizes time aloft. In other words, there’s a speed at which I won’t glide quite as far a distance, but I will be aloft longer. Most of our training focus is on maximizing distance (getting to a favorable spot) over maximizing time (having more time to diagnose.)

My airplane’s best glide ranges from 70 to 80kts indicated, depending on weight. On my typical mission it’s about 77 at takeoff and about 73 at landing weight. If it’s 77, I am going to trim the airplane up to fly right at 77 all the way to my desired landing point, and will maneuver around to arrive at a final approach position a bit high by gliding around at 77, not by altering my speed. I can use flaps and/or slip maneuvers to steepen my descent angle while sill maintaining a pretty consistent glide speed.

There should be no “zooming” or swooping, with rare exceptions. I want to bleed off the last speed—from 77 down to my touchdown target speed of about 55 kts—as I am on my last couple of hundred feet of descent. The goal is always to just do a very normal landing approach and flare, essentially the same as a “normal” landing, whenever possible.

There is no speed you can get down to that makes crashing upside-down at that slower speed better than crashing right-side-up at the higher speed. So hitting whatever you are going to hit at, or just above stall speed, is profoundly more survivable than stalling trying delay the inevitable forced landing another second or two.

  • $\begingroup$ What causes upside-down crashing? $\endgroup$ Jun 13, 2023 at 2:04
  • $\begingroup$ @JeremyFriesner Stall and a spin or spiral. $\endgroup$ Jun 13, 2023 at 8:48
  • 8
    $\begingroup$ I think to say this is "not realistic" is incorrect. Your answer provides the correct procedure to land in such a situation, but that doesnt mean Flight Simulator was unrealistic. What the OP described would also happen in real life if you chose to land that way. "Not recommended" or "not the correct procedure" would be better wording. $\endgroup$ Jun 13, 2023 at 13:04
  • 4
    $\begingroup$ @GavinCoates: I would think even if such an approach wouldn't generally be recommended, it might nonetheless represent the "least evil" among available options. If one has an engine failure and there's an open runway nearby, and one has too much energy to fly a stabilized approach, but not enough energy to fly a 360 degree circle to bleed off energy, an unstabilized approach to the runway might be better than a stabilized approach that lands off the runway. $\endgroup$
    – supercat
    Jun 13, 2023 at 18:40

I read this question as "Is the simulator accurately modeling the behavior of the aircraft" To which I think the answer would probably be "to a reasonable level of fidelity, yes".

There is a follow up question about your technique, which has been answered, and to which I'll add. The generally accepted technique for an engine failure scenario, as well as several other types of emergencies, are outlined in chapter 18 of the current version of the FAA's Airplane Flying Handbook (FAA-H-8083-3C). These procedures get further refinement, typically including airspeeds to use, in the operating handbook for individual aircraft.

It the case where you had to dive near the end of your descent, it may have been that your low airspeed corresponded to a high angle of attack and high descent angle. It is reasonable to expect this flight condition to be well modeled in your simulation. This condition would indeed result in a sub-optimal glide, and would also require increasing the airspeed in order to arrest the descent rate.

  • 1
    $\begingroup$ That is how I read the question too - I think this is the only answer that accurately answers what the user was asking. he wasnt asking what the correct procedure would be in the situation described, but whether what he experienced in FS would also be experienced in real life if you did the same thing. $\endgroup$ Jun 14, 2023 at 10:22

In 1987-9, computers of the age were more than capable of getting the math and physics right on flight simulation. Graphics fidelity was another matter :)

It is perfectly ordinary for pilots to be all over the map with airspeed and descent rate as you did. They may be trying to get enough airspeed across the prop to attempt a "windmill start". They may be tuning for best glide. They may be slewing or otherwise deliberately scrubbing off energy in order to arrive at the target landing site with enough energy but not too much.

In fixed-wing aircraft, energy is stored in two forms: airspeed (kinetic) and altitude (potential). Flight is all about trading one for another, and there's nothing magic about the trade - you can't create more energy than you had, nor destroy it merely by trading. Although... I hear some airplanes have an "engine" that can add new energy :)

This energy arbitrage was exactly what you were doing, and that's what it's all about.

If you arrive at the runway stripes with too much energy, you will either be too fast or too high in the sky. And you could find yourself in an "impossible turn" situation where your surplus is not enough to get lined up on a runway before you run out of energy.

So whatever you were doing to get on the ground, it was working :)

  • $\begingroup$ that said, MSFS back then, and AFAIK to this day, places the aircraft on rails rather than modelling actual aerodynamics. This leaves it with inadequate behaviour in certain corner cases, and this would likely be one such. X-Plane handles things better. $\endgroup$
    – jwenting
    Jun 15, 2023 at 8:26
  • $\begingroup$ While I have not audited available information thoroughly, MSFS 4.0, according to Wikipedia, was released in 1989 had an aircraft editor which allowed the user to alter paramaters such as frontal area, aspect ratio, etc. giving some clues about it's underlying aerodynamic model. Xplane appears to be more sophisticated; I believe it "bakes" a panelized model into easier to compute coefficients. On the other hand, I think saying that MSFS is/has been "on rails" is giving it short shrift. $\endgroup$ Jun 21, 2023 at 22:45
  • $\begingroup$ I just took another look at x-planes brag sheet on its model, which reminds me that they don't bake the panel model into a set of stability derivatives, but run the simulation on a somewhat simplified (10 elements per side) panel model. This is much more sophisticated than what is probably common, however, in the context of this question, this simpler models will be more than adequate. $\endgroup$ Jun 21, 2023 at 23:18

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .