Can any plane glide unpowered? (unless it is a brick of course)

[mrfox]

Glide ratio is a function of a Reynolds number.

Negative. Reynolds number in the case of airfoils, is a property of the environment - it is useful if you want to compare the behavior of vehicles of vastly different scales, or how a 747 would fly in a vastly different atmospheric environment, such as on mars. It does not vary in a meaningful manner given the flight regime of a subsonic, atmospheric vehicle.

Here is an analysis of a typical foil used by Boeing. 747 uses several variations of this foil along the length of the wing. Take a look at the Cl/Cd polar. The slope of the tangent will give you best glide ratio. 747 has MAC of 327.8" and V_Y of around 180KIAS at maximum landing weight. That gives me R = 50x10⁶ at sea level. At cruise altitude, air density is about a third, so under ideal circumstance, the velocity needs to increase by sqrt(3) to maintain lift at the same glide slope. That puts Reynolds number at less than 30x10⁶. I don't have polars for Reynolds numbers that high, but if the trend holds, this would actually reduce glide ratios.

A 747 consists of more then just a wing. You are off by a factor of 50 in terms of the reynolds number for the entire aircraft - its closer to 2 billion.

And the trend does not hold as you assume. Once the scale of the problem results in very large reynolds numbers, viscous forces become inconsequential and the flow is basically inertial.

However, real wings aren't constant 2D cross-sections of infinite span. The above is only useful to demonstrate the fact that the statement that glide ratio does not depend on altitude is wrong at a very fundamental level. For a real airplane, things are far, far more complicated. I'm going to skip all the factors related to the way the wing profile changes along its length, and the fact that body contribution is very different at different air speeds. What's critical for a 747 is actually the wing loading and finite wing span. That's what kills the 747 performance at low air speeds and, consequently, high air densities. Without getting into all complexities, lift depends on transverse circulation induced by an airfoil. Unfortunately, due to finite span, that tends to induce longitudinal circulation. Id est, tip vortices. While circulation is actually higher at cruise altitude, the low density and high air speed means that it is distributed over much larger volume of air. That results in much gentler wing tip vortices as well. At low speeds, the wing tip vortex is much more compact. It has a smaller radius, and the air flow past the wing is much slower, all while the impulse transfered to the air remains the same. This results in much higher air speed in the wing tip vortex, so more energy has to be transferred to air. More power lost to wing tip vortices at lower air speed means dramatically higher drag.

You have managed to describe the concept of induced drag in a very complicated and round-about way.

This also brings up another interesting aspect. For a 2D air foil, wing loading isn't even a factor in the performance. You have a fixed glide ratio, and it depends on qualities of the wing only. This is absolutely not the case for a real airplane. A heavily loaded airplane does not glide as well as an empty one. This is pretty intuitive, but another thing you don't seem to account for in your statement.

It may be intuitive, but it is absolutely wrong. A heavily loaded airplane has the same glide ratio as an empty one. Ask any sailplane pilot why they carry ballast.

But, I believe, people will still ask me for references. Here is the manual. On pages 42-43 you can find the climb rate for maximum takeoff weight to be 2000FPM at 210KIAS with 880,000 lbs of weight and about 250,000lbs of thrust. That's a climb at a slope of +9%. That indicates an excess thrust of 9% of weight, meaning drag of 160,000lbs, indicating a glide ratio of 5.5. That's a bit higher than ratio of 4 I've mentioned earlier, but this is clean configuration. Add a bit of flaps to try to bring the landing speed down, and you're right back to 4 or bellow.

You are still stuck on the fact that a pilot would approach at normal landing speed. With no power available, the pilot will maintain best glide speed (for the 747, Vref+80, appox 5.5 degs AOA - From actual Boeing 747-400 manuals I have access to at work.) until he/she is assured that the landing site is made, and only then will start slowing down moments before touchdown, bleeding off energy to ensure a slower, more survivable landing - just like a space shuttle or a X-15 landing.

Like the long-retired 777 Captain I mentioned in an earlier post, they never pitched for best glide. According to the guys at the flight school, the extra speed allowed them to restart the prop.

Why didn't they use the starter?

[K^2]

Negative. Reynolds number in the case of airfoils, is a property of the environment - it is useful if you want to compare the behavior of vehicles of vastly different scales, or how a 747 would fly in a vastly different atmospheric environment, such as on mars. It does not vary in a meaningful manner given the flight regime of a subsonic, atmospheric vehicle.

Maybe you'd like to remind us all what the definition of Reynolds number for an airfoil is? Or for a general flow? Or precisely in what ways does atmosphere of Mars differ from that on Earth, other than density? Dynamic viscosity of both is essentially the same.

The fact that actual ascent profile of an airliner happens to follow more or less constant Reynolds number is besides the point here. Cruise speeds are nowhere near best glide.

A 747 consists of more then just a wing.

I'm pretty sure I covered that.

viscous forces become inconsequential and the flow is basically inertial.

Inertial flow cannot generate lift, but I'm sure that's not important at all for an airplane.

You have managed to describe the concept of induced drag in a very complicated and round-about way.

I've managed to describe its physics, which tends to be more important than whatever hand-waving explanations that lead you to that level of confusion on Reynolds numbers and importance of viscous forces in aerodynamics.

It may be intuitive, but it is absolutely wrong. A heavily loaded airplane has the same glide ratio as an empty one. Ask any sailplane pilot why they carry ballast.

And this kind of confusion, too. Which aircraft has more drag, in proportion to weight, due to wingtip vortices, 747 or a glider? And why do you think that is? And as the final question, based on that, for which is extra weight going to make a larger difference?

Yes, I would not doubt that advantages of bringing ballast on a plane with glide ratio over 60 significantly outweigh (sorry) the losses to wingtip vortices. But if you think that applies across the board, you understand nothing about the induced drag.

You are still stuck on the fact that a pilot would approach at normal landing speed. With no power available, the pilot will maintain best glide speed (for the 747, Vref+80, appox 5.5 degs AOA - From actual Boeing 747-400 manuals I have access to at work.) until he/she is assured that the landing site is made, and only then will start slowing down moments before touchdown, bleeding off energy to ensure a slower, more survivable landing - just like a space shuttle or a X-15 landing.

Even the most casual glance at the polars, both the foils and the aircraft, show that glide ratio is very soft near best glide. The difference between ref+20 and ref+80 is not going to be dramatic. But hey, if you have specific references that show a completely different glide ratio at sea level, go ahead and share. If you can show me data that gives glide at sea level at anything like 10+ you get in cruise^*, I'll be genuinely surprised and admit my mistakes.

^* I was actually surprised to find that glide ratio in nominal cruise of 747-400 is 12, after I found some polars. I was expecting a slightly lower number, but it's in the ballpark.

[mrfox]

Maybe you'd like to remind us all what the definition of Reynolds number for an airfoil is? Or for a general flow? Or precisely in what ways does atmosphere of Mars differ from that on Earth, other than density? Dynamic viscosity of both is essentially the same.

The fact that actual ascent profile of an airliner happens to follow more or less constant Reynolds number is besides the point here. Cruise speeds are nowhere near best glide.

I'm pretty sure I covered that.

Inertial flow cannot generate lift, but I'm sure that's not important at all for an airplane.

I've managed to describe its physics, which tends to be more important than whatever hand-waving explanations that lead you to that level of confusion on Reynolds numbers and importance of viscous forces in aerodynamics.

And this kind of confusion, too. Which aircraft has more drag, in proportion to weight, due to wingtip vortices, 747 or a glider? And why do you think that is? And as the final question, based on that, for which is extra weight going to make a larger difference?

Yes, I would not doubt that advantages of bringing ballast on a plane with glide ratio over 60 significantly outweigh (sorry) the losses to wingtip vortices. But if you think that applies across the board, you understand nothing about the induced drag.

Even the most casual glance at the polars, both the foils and the aircraft, show that glide ratio is very soft near best glide. The difference between ref+20 and ref+80 is not going to be dramatic. But hey, if you have specific references that show a completely different glide ratio at sea level, go ahead and share. If you can show me data that gives glide at sea level at anything like 10+ you get in cruise^*, I'll be genuinely surprised and admit my mistakes.

^* I was actually surprised to find that glide ratio in nominal cruise of 747-400 is 12, after I found some polars. I was expecting a slightly lower number, but it's in the ballpark.

http://www.faa.gov/regulations_policies/handbooks_manuals/aviation/media/00-80t-80.pdf

Page 371-72

The effect of gross weight on glide performance

may be difficult to appreciate. Since L/D(max)

of a given airplane configuration will occur at a

specific value of C(L), the gross weight of the airplane

will not affect the glide ratio if the airplane

is operated at the optimum C(L),. Thus,

two airplanes of identical aerodynamic configuration

but different gross weight could

glide the same distance from the same altitude.

Of course, this fact would be true only if both

airplanes are flown at the specific C(L), to produce

L/D(max). The principal difference would be

that the heavier airplane must fly at a higher

airspeed to support the greater weight at the

optimum C(L),. In addition, the heavier airplane

flying at the greater speed along the same flight

path would develop a greater rate of descent.

The effect altitude on glide performance is

insignificant if there is no change in L/D(max).

Generally, the glide performance of the majority

of airplanes is subsonic and there is no

noticeable variation of L/D(max), with altitude.

Any specific airplane configuration at a particular

gross weight will require a specific value

of dynamic pressure to sustain flight at the

C(L), for L/D(max). Thus, the airplane will have

a best glide speed which is a specific value of

equivalent airspeed (EAS) independent of

altitude. For convenience and simplicity, this

best glide speed is specified as a specific value

of indicated airspeed (IAS) and compressibility

and position errors are neglected. The principal

effect of altitude is that at high altitude

the true airspeed (TAS) and rate of descent

along the optimum glide path are increased

above the low altitude conditions. However,

if L/D(max) is maintained, the glide angle and

glide ratio are identical to the low altitude

conditions.

[K^2]

Have you tried to actually understand the physics of flight, rather than quote simplified rules of thumb written for pilots? Especially a book for naval pilots, when we are discussing airliners?

Thus, two airplanes of identical aerodynamic configuration

but different gross weight could

glide the same distance from the same altitude.

Of course, this fact would be true only if both

airplanes are flown at the specific C(L), to produce

L/D(max). The principal difference would be

that the heavier airplane must fly at a higher

airspeed to support the greater weight at the

optimum C(L),.

Which is perfectly true, so long as L/D(max) occurs at the same place for any air speed. Which, of course, doesn't happen. I don't have 747 polars across a sufficient range, but lets look at 787 for sake of example.

Even the text you quote is in agreement that if I load the plane heavier, it will have to glide faster to do best glide. So suppose, I load a 787 to have best glide at mach .87. Looking at the nice blue curve for that speed we see that L/D(max) is at CL .45 and the L/D is about 19.5. Now I load the plane more, so that at the same CL it has to glide at mach .88. Suddenly, we're on the orange curve, and the L/D is now just over 17. And isn't L/D(max). So the planes with different amount of load will have to glide at different CL to achieve L/D(max).

But again, any time you'd like to bring out charts that show different, feel free.

The effect altitude on glide performance is

insignificant if there is no change in L/D(max).

"If there is no change in L/D(max)." Bam. Right on the first sentence. There is a huge change in L/D(max) with altitude. See that 37k' on the polars? It's there for a reason. Polars for sea level are completely different. Now, I've been trying to explain why there is a change across several posts. I've shown via computation that 747 has L/D(max) a little over 5 at sea level, and well over 10 in cruise. Yet you keep insisting that L/D(max) doesn't depend on altitude with absolutely no argument, reference, or drag polars to back it up. So again, if you claim you have actual charts for 747, go ahead and post them.

[mrfox]

http://www.lissys.demon.co.uk/samp1/polarb.png

So suppose, I load a 787 to have best glide at mach .87. Looking at the nice blue curve for that speed we see that L/D(max) is at CL .45 and the L/D is about 19.5. Now I load the plane more, so that at the same CL it has to glide at mach .88. Suddenly, we're on the orange curve, and the L/D is now just over 17. And isn't L/D(max). So the planes with different amount of load will have to glide at different CL to achieve L/D(max).

Interesting that the two data points that you happened to have picked is way above best glide speed, not to mention beyond the drag divergent mach of the aircraft, where transonic drag effects dominate. Do you think that is a good example to demonstrate gliding performance?

"If there is no change in L/D(max)." Bam. Right on the first sentence. There is a huge change in L/D(max) with altitude. See that 37k' on the polars? It's there for a reason. Polars for sea level are completely different.

So the graph states the data is from 37000' in its title. How does that mean that it is completely different for sea level conditions? Please Elaborate.

Have you tried to actually understand the physics of flight, rather than quote simplified rules of thumb written for pilots? Especially a book for naval pilots, when we are discussing airliners?

So you are saying these "rules of thumb" are not only inaccurate, but completely wrong, and reality is the complete opposite? Please elaborate.

[PakledHostage]

^* I was actually surprised to find that glide ratio in nominal cruise of 747-400 is 12, after I found some polars. I was expecting a slightly lower number, but it's in the ballpark.

I mean no disrespect, but can I ask you to elaborate on an earlier "back of the envelope" calculation that you did in light of the above?

You calculated the L/D performance of a B747 based on climb performance information that you found in a manual for a flight simulator program. You determined that the aircraft, when loaded to maximum takeoff weight and in the early part of its climb, has an L/D of 5.5. You extrapolated this number to predict that the aircraft's L/D in landing configuration was about 4, while your new data above suggests that the L/D may be as high as 12 in normal cruise.

Fair enough, but could you please speak to how calculations similar to your earlier calculation (but using early climb performance for a fully loaded B767-200), would predict the likelihood of pilots being able to successfully land a B767 after both engines flame out? The question is relevant because The Gimli Glider was a B767-233 (where the "33" means that it was a variant built to Air Canada's specifications) powered by two P&W JT9D-7R4 engines and it was landed successfully in just such a situation.

We started this discussion with you arguing that it is not possible to land a B747 deadstick. Notwithstanding your efforts to justify your position, I think your analysis is oversimplified and it is pointless to draw such conclusions from it. I suspect that the same oversimplified analysis that you applied to the B747 would predict equally bad glide performance for the B767. But clearly it is possible to glide a B767 over 100 NM and land safely (if you start at 41000 feet) because it has been done in real life. If your numbers work out the way I anticipate they will, then I would argue that the empirical data invalidates your analysis.

[rdfox]

To boil it down: All objects have a certain ability to glide; the lift-to-drag ratio defines how good they are at it, and the glide ratio tells you how far you can glide from a given height. The Space Shuttle *was* a pretty lousy glider (one engineer claimed in 1980 that it had "the glide ratio of a pair of pliers"), but even it landed. For the record, the Shuttle's glide ratio was comparable to that of an autorotating helicopter (which, to the uninitiated, feels more like an out-of-control plummet than a controlled glide). Also, in general, the faster an airplane is designed to fly, the worse its glide ratio at best-glide speed will be, as it's designed to have limited low-speed lift to reduce high-speed drag.

Also, the B-2 does NOT need its engines for yaw control; it uses "drag rudders" for that, split-surface ailerons akin to the Shuttle's split-surface rudder; they open as needed for yaw control and general drag-increase when you're trying to shed speed quickly; asymmetric deployment for yaw control and symmetrical deployment for general drag.

[Rickenbacker]

some fighter jets cannot glide very well... like that F-117 one...

The published glide ratio for the F-117 is 15.3/1, which is about the same as your average GA aircraft, so it'll glide just fine. Not sure if it would have any hydraulics or how unstable it might be if the fly-by-wire system isn't powered, though.

The Cessna incident in another post sounds odd to me, though. Most pilots I know would instinctively pitch for best glide as soon as the engine quit, and as long as the starter and battery worked, they'd then be able to attempt a restart AFTER making sure the plane was still flying and they could reach a landable spot. That's standard procedure. But they might of course have lost electrical power, in which case a dive might be able to increase RPM to the point where the ignition worked... The only time I experienced something like this in flight was in a Grob-109 motor glider: The battery died in flight, and I didn't even notice until I landed, shut down and let my passenger out. When I tried to restart the engine, nothing happened as there was no power to the starter or fuel pump. But since I was parked by the hangar, we just got another battery from the shelf .

I might be biased as a glider pilot. Even touring motor gliders, which are really meant to fly under power, tend to land like a glider. You leave the engine at idle and adjust your pattern so that you don't need it. Just remember to turn on carb heat, or you'll kill the engine when throttling up to taxi to the parking, since you'll have some carb ice by then. Not to mention go-arounds, where you QUICKLY turn off carb heat and push the throttle to full. If you forgot to turn on the heater before landing, you'll then be greeted with a deafening silence...

[mrfox]

Some other examples of airplanes with extremely poor glide ratios, yet are more than capable of landing with no engine power:

F-104:

Glide ratio - appox 5 :1

X-15:

Glide ratio - appox 4:1

X-24B:

Glide ratio - appox 2.5:1

[shynung]

X-24B:

http://www.unrealaircraft.com/wings/images/mariet2.jpeg

Glide ratio - appox 2.5:1

Does that thing even fly, for that matter? Looks really weird.

[mrfox]

Does that thing even fly, for that matter? Looks really weird.

Here she is, engine off, in formation with a F-104, gliding in for a landing

Kinda brings the topic full circle - even a brick can glide!

Edited April 27, 2014 by mrfox

[rdfox]

The X-24 was one of a series of "lifting body" prototypes of the late 60s and early 70s, which were expected to be the basis for a future reusable spaceplane, being able to fly on lift generated entirely by the body shape rather than by wings. While frequently credited as the basis for the Space Shuttle, in reality, pretty much the only thing they contributed to it was demonstration of the ultra-steep gliding approach and landing used by the Shuttle could be done with more than just the stubby-winged X-15/F-104 configuration, which allowed NASA to eliminate the planned pop-out jet engines for approach and landing, saving much weight and interior volume for cargo. That said, lifting body designs *are* the basis for most small reusable spaceplanes that are intended for carrying only people instead of cargo.

As for the F-104, Mike Collins had comments on it, specifically regarding its glide ratio, actually: "The plane was designed ('optimized') for all-out speed, and it did go like hell, preferably in a straight line. When slowed down, it was simply too heavy for the amount of wing area (too high a 'wing loading'), and if the engine quit, you'd better be over the Edwards dry lake, because it dropped like a stone."