totm october 2020 Airplane Design Q&A

[mrfox]

Dihedral and anhedral had always seems to me to be such an inefficient way to generate stabilising moments. In this age of computers and fly by wire, can’t the (de)stabilising effects be generated artificially via control systems design? IE larger control surfaces and gains and feedback loops actively modified on the go to obtain the desired response? Is it a certification issue or is it a basic aerodynamics issue?
On the same note - why so much dihedral from wing flex on the 787 compared to other current designs? 
 

[mikegarrison]

Keeping an airplane stable by computer is fine for a fighter jet. If anything goes wrong -- out goes the pilot! But it's a bad idea for an airplane that doesn't have ejection seats.

[mikegarrison]

2 hours ago, mrfox said:

why so much dihedral from wing flex on the 787 compared to other current designs? 

The 787 wing is simply the most beautiful wing I've ever seen. From the first moment I saw it, I loved it.

I attended the roll-out ceremony. I could invite one person, so I invited my grandfather. He started by bucking rivets on B-17s and finished his career working on 747s.

I know that doesn't answer your question, but even if I did know the recipe for the secret sauce I wouldn't be able to say it here.

[JoeSchmuckatelli]

4 hours ago, mikegarrison said:

The 787 wing is simply the most beautiful wing I've ever seen. From the first moment I saw it, I loved it.

I attended the roll-out ceremony. I could invite one person, so I invited my grandfather. He started by bucking rivets on B-17s and finished his career working on 747s.

I know that doesn't answer your question, but even if I did know the recipe for the secret sauce I wouldn't be able to say it here.

After reading this I had to go check it out. 

 

Someone spent an evening with a bottle of whisky and stared at pictures of albatross to gain that inspiration. 

 

Engineering taking notes from nature if I'm not mistaken 

[wizzlebippi]

10 hours ago, mrfox said:

Dihedral and anhedral had always seems to me to be such an inefficient way to generate stabilising moments. In this age of computers and fly by wire, can’t the (de)stabilising effects be generated artificially via control systems design? IE larger control surfaces and gains and feedback loops actively modified on the go to obtain the desired response? Is it a certification issue or is it a basic aerodynamics issue?
On the same note - why so much dihedral from wing flex on the 787 compared to other current designs? 
 

It's not that you couldn't fix it in fly by wire control logic, just that you would be signing up for a lot more work. Dihedral and anhedral are practically free for small angles anyway and I'm not aware of any benefits to loose or negative lateral-directional stability.

The 787's wing flexes more because it's mostly carbon fiber and designed to do so to keep weight down. Keep in mind that a large aluminium wing also flexes significantly when loaded. Look up a video of a B52 takeoff and watch it's wings lift before the fuselage.

[mrfox]

11 hours ago, wizzlebippi said:

The 787's wing flexes more because it's mostly carbon fiber and designed to do so to keep weight down. Keep in mind that a large aluminium wing also flexes significantly when loaded. Look up a video of a B52 takeoff and watch it's wings lift before the fuselage.

The B52 was exactly what came to mind when I thought of this subject. But in the BUFF’s case - wing flex reduces anhedral - which to me seems be be such an elegant design outcome.

This might be anecdotal, but Boeing aircraft had always looked and felt floppier to fly compared to an equivalent Airbus. The A350s wing  - also carbon - does not flex nearly as much as the 787s.

 

[JoeSchmuckatelli]

1 hour ago, mrfox said:

The B52 was exactly what came to mind when I thought of this subject. But in the BUFF’s case - wing flex reduces anhedral - which to me seems be be such an elegant design outcome.

This might be anecdotal, but Boeing aircraft had always looked and felt floppier to fly compared to an equivalent Airbus. The A350s wing  - also carbon - does not flex nearly as much as the 787s.

 

Do you fly these?  How does it feel to have a 'floppy' wing when you're holding the 'stick'?

[mikegarrison]

18 hours ago, wizzlebippi said:

The 787's wing flexes more because it's mostly carbon fiber and designed to do so to keep weight down. Keep in mind that a large aluminium wing also flexes significantly when loaded. Look up a video of a B52 takeoff and watch it's wings lift before the fuselage.

There are a lot of misconceptions about CFRP (carbon fiber reinforced plastic). You can make it more stiff than aluminum if you want. You can make it lighter than aluminum for the same strength, too.

But the really interesting thing is that it is non-isotropic. Metals like aluminum are isotropic -- they have the same strength in every direction. But CFRP is a designed material, and the way you lay the fibers can make it stronger in one direction than another.

I don't do structures, but I know that it's generally considered wasteful to make "black aluminum" -- CFRP that behaves just like aluminum would. Instead, the goal is to lay the fibers in such a way that you can tailor the stress/strain relationship to do your bidding.

[mrfox]

19 hours ago, JoeSchmuckatelli said:

Do you fly these?  How does it feel to have a 'floppy' wing when you're holding the 'stick'?

Well, one place it felt distinctly different between the two is during crosswind landings. One of the biggest concerns during the touchdown, derotation and rollout in high crosswinds is the possibility of engine pod strikes. A floppy wing makes engine pod clearance very tricky to judge. Although the Airbus control laws makes flying with crossed controls somewhat unintuitive, once I got my head wrapped around the logic, I ended up feeling more comfortable in on the bus compared with their Boeing counterparts.
 

[mikegarrison]

47 minutes ago, mrfox said:

Well, one place it felt distinctly different between the two is during crosswind landings. One of the biggest concerns during the touchdown, derotation and rollout in high crosswinds is the possibility of engine pod strikes. A floppy wing makes engine pod clearance very tricky to judge. Although the Airbus control laws makes flying with crossed controls somewhat unintuitive, once I got my head wrapped around the logic, I ended up feeling more comfortable in on the bus compared with their Boeing counterparts.
 

That's interesting. The control systems of the two manufacturers are so different. I didn't think a lot of pilots fly both.

[mrfox]

Not both at the same time, although some in our test and engineering dept do. It’s a nightmare to keep current in multiple types, legally and practically.

The Airbus control system would be another interesting discussion topic on this thread. In essence, it’s an attitude stabilisation system - like having the KSP SAS on full time. Its origins can be traced back to both the space shuttle FBW control laws, the CWS mode on older analog(and a few modern digital)autopilots, and all the way back to WWII formation  bombing autopilots.

https://ntrs.nasa.gov/citations/19760024058
 

https://www.ima-usa.com/products/original-wwii-u-s-b-17-flying-fortress-c-1-autopilot-formation-control-stick-assembly?variant=12460921192517

[benzman]

As I recall, bendy wings allow an aircraft to more comfortably ride out turbulence by allowing the wings to flex without jolting the fuselage and hence the passengers. Incidentally, in the 1980s I did quite a bit of flying in sailplanes. The modern, fiberglass open class single-seaters have very long, narrow wings and they can bend their wings like you would not believe!

[wumpus]

On 10/21/2020 at 6:50 PM, benzman said:

As I recall, bendy wings allow an aircraft to more comfortably ride out turbulence by allowing the wings to flex without jolting the fuselage and hence the passengers. Incidentally, in the 1980s I did quite a bit of flying in sailplanes. The modern, fiberglass open class single-seaters have very long, narrow wings and they can bend their wings like you would not believe!

Back in the 1990s I briefly worked at a tiny startup that took this to extreme levels.  Called Freewing (then Freewing Aerial Robotics, they were getting into drones), the whole point of the company was to build airplanes whose wings would rotate freely around a shaft (the angle of attack would be near zero, the wings would simply rotate to be parallel with any airstream).  The only  way to stall the plane would be to drop down to stall speed, but in practice this wasn't true as the only way to use flaps would be to lock the wings.

The aircraft (drone) I was working on (just the telemetry, I knew even less about aircraft then than I do now) was designed (more like architected, we had a grad student doing the detailed engineering) by Burt Rutan and the basics was that if the wings could rotate freely, and the propeller was fixed to the fuselage, you could angle the propeller (and thrust) by an actuator in the tail.  I'm not sure if they were still around by the Iraq War, or partnering with the French closed that market, but they didn't last much longer.  It was pretty much stuck with being a drone unless you want to takeoff/land while sitting at a 60 degree angle pointed upwards.

But one of the claimed advantages was that it was much less effected by turbulent air, especially in smaller planes.   Since the wings could slough off any sudden change of direction of air, that didn't leave much else to be pushed around.

[mikegarrison]

14 hours ago, wumpus said:

Back in the 1990s I briefly worked at a tiny startup that took this to extreme levels.  Called Freewing (then Freewing Aerial Robotics, they were getting into drones), the whole point of the company was to build airplanes whose wings would rotate freely around a shaft (the angle of attack would be near zero, the wings would simply rotate to be parallel with any airstream).  The only  way to stall the plane would be to drop down to stall speed, but in practice this wasn't true as the only way to use flaps would be to lock the wings.

Hmm.

Despite the occasional instance where people want to build an airplane with forward-swept wings or swiveling wings or flapping wings or whatever, in practice airplanes are usually designed in the following way:

1. A new engine becomes available
2. Airplanes are designed to make use of it.

There is some feedback between 1 and 2, because engine manufacturers need someone to buy their engine, so it often ends up working like this:

1. A new engine capability becomes available.
2. An airplane concept is developed to make use of it.
3. The airplane manufacturer requests a commitment from one or more engine manufacturers to provide an engine to meet a given specification.
4. The engine and airplane detail design are both done and the two get mated together.
5. Everyone argues about whether any performance shortfall is the result of the airplane or the engine.

(Step 5 is nearly guaranteed.)

[SOXBLOX]

6 hours ago, mikegarrison said:

Everyone argues about whether any performance shortfall is the result of the airplane or the engine.

This part sounds like the most fun. 

[wumpus]

On 10/26/2020 at 6:05 AM, mikegarrison said:

Hmm.

Despite the occasional instance where people want to build an airplane with forward-swept wings or swiveling wings or flapping wings or whatever, in practice airplanes are usually designed in the following way:

1. A new engine becomes available
2. Airplanes are designed to make use of it.

There is some feedback between 1 and 2, because engine manufacturers need someone to buy their engine, so it often ends up working like this:

1. A new engine capability becomes available.
2. An airplane concept is developed to make use of it.
3. The airplane manufacturer requests a commitment from one or more engine manufacturers to provide an engine to meet a given specification.
4. The engine and airplane detail design are both done and the two get mated together.
5. Everyone argues about whether any performance shortfall is the result of the airplane or the engine.

(Step 5 is nearly guaranteed.)

Somehow I suspect this is the reason the company is no longer in business.

[Silavite]

On 10/26/2020 at 5:05 AM, mikegarrison said:

Everyone argues about whether any performance shortfall is the result of the airplane or the engine.

In the case of my current uni design team, we argue about why the powertrain efficiency is so bad before realizing that different sources used different definitions for advance ratio, thrust coefficient, and power coefficient. Then we get the conversions wrong. Hopefully the 3rd time's the charm!

[mikegarrison]

What happens in flight testing is that you can get a reasonably good measure of fuel flow and you can get a reasonably good measure of speed.

You can work out a reasonably good measure of weight and therefore lift by weighing the airplane before and after the flight and integrating the fuel flow to get a weight at any time in flight.

What you can't directly measure is either drag or thrust. So while you know that drag = thrust, you don't know either one.

This means that if the airplane has higher than expected fuel flow, either it was draggier than expected or the thrust SFC was not as good as expected, but you can't directly measure which of these is the actual reason. So there is a lot of analytical modeling done to try to piece together all the available data and figure out what the answer is. Needless to say, the engine company and the airframe manufacturer often have different points of view in this process.

[mrfox]

On 10/26/2020 at 3:05 AM, mikegarrison said:

Hmm.

Despite the occasional instance where people want to build an airplane with forward-swept wings or swiveling wings or flapping wings or whatever, in practice airplanes are usually designed in the following way:

1. A new engine becomes available
2. Airplanes are designed to make use of it.

Prior to the 2000s, re-engining older airframes was a common practice to extend their useful economic life. Any thoughts as to why this practice has fallen out of favour?

[SOXBLOX]

I guess it's partly because the frames which would benefit from this are nearing the end of their lives, and we really need to replace them, or that other frames durable enough to last were designed specifically around one engine's structural loads. But I don't think it's fallen out of favor; the B-52 is scheduled to get a reengine soon, to serve till 2050ish, and there was a proposal to reengine the F-35. IDK.

[mikegarrison]

3 hours ago, mrfox said:

Prior to the 2000s, re-engining older airframes was a common practice to extend their useful economic life. Any thoughts as to why this practice has fallen out of favour?

I think what you are remembering was actually a reaction to the Stage 2 noise phaseout. In 1990, the US made a law that all civil airplanes >75,000lb had to meet Stage 3 noise levels by Jan 1, 2000. Otherwise they would no longer be allowed to operate in the US. Europe then followed suit, with a phaseout of Chapter 2 airplanes that went into effect April 1, 2002. (US laws had "Stages", while Europe used ICAO "Chapters". They were nearly exactly the same.)

This drove a market for "hushkits" and re-engining programs that would meet Stage/Chapter 3. Once the phaseouts were complete, this market no longer existed.

Edited November 2, 2020 by mikegarrison

[mrfox]

Good point regarding the noise deadline. I imagine the incremental benefits of going from a 1950s gen to a 1990s gen engine is far greater than from the 90s to today - both with regards to noise and operating cost.

The B52 re-engine brings up an interesting question - what exactly is a B52 capable of doing that couldn’t be done by retrofitting equipment to something more modern like a C17?

 

[magnemoe]

2 hours ago, mrfox said:

Good point regarding the noise deadline. I imagine the incremental benefits of going from a 1950s gen to a 1990s gen engine is far greater than from the 90s to today - both with regards to noise and operating cost.

The B52 re-engine brings up an interesting question - what exactly is a B52 capable of doing that couldn’t be done by retrofitting equipment to something more modern like a C17?

 

Well its the bombing thing. yes you can drops bombs out the back but this get into issues if you try to drop heavy bombs, an high number of bombs or large fragile stuff like cruise missiles. 
 

[mrfox]

8 hours ago, magnemoe said:

Well its the bombing thing. yes you can drops bombs out the back but this get into issues if you try to drop heavy bombs, an high number of bombs or large fragile stuff like cruise missiles. 
 

Agreed that an aircraft designed to drop bombs will be better suited than something modified. I should rephrase my question as one pertaining more to the actual physical capability required of an aircraft of a C17s vintage to fly a B52s mission profile... Or perhaps better phrased - how would a B52 be designed today using current tech?

[mikegarrison]

3 hours ago, mrfox said:

How would a B52 be designed today using current tech?

High-bypass turbofans. It would probably look more like a 777 than a B-52.

The B-52 replacement was first supposed to be faster planes, then stealthier planes. But the current plan is to keep B-52s flying for 100 years.