I was looking at some Shinkansen trains recently and one thing caught my eye. They are RECTANGULAR! Look at Shinkansen E5/H5, for just one example. It's absolutely square in cross section with slightly rounded corners. And some of those trains, at least in record runs, exceed 500kph, which isn't too far from some of the commercial jets in the sky. And the newest Japanese maglev exceeds 600kph which is actually FASTER than some bizjets. So I am thinking, WHY aircraft fuselages are round? They may be fully circular, or ovoidal, or 8-shaped, but they all have circled on top and bottom? Why not make them rectangular? We'd get a MUCH more spacious interior. And the benefits for cargo are ENORMOUS. No more tetris-shaped containers. A lot more cargo space. Increased intermodality - you don't need to transfer cargo from truck-friendly pallets into pyramid-shaped structures. MUCH easier building process. Why is it not done in today's every-penny-counts world?

Planes are pressurized. Trains are not. It takes much less weight to keep a round tube pressurized and intact than a square tube.

There have been square fuselages (see Shorts 330/360) but I don’t think they were pressurized.

I think because tubes are stronger when pressurized.

While I'm sure pressurization has something to do with it, I thought it was more because planes don't necessarily move through the air in the direction straight down the nose through the tail. A squarer cross section would be less aerodynamic to the crosswise winds. That's not a problem with slower vehicles like trains, or the Shorts 330/360.

Anyone remember the square windows in the Comet?

Polot wrote:

Planes are pressurized. Trains are not. It takes much less weight to keep a round tube pressurized and intact than a square tube.

There have been square fuselages (see Shorts 330/360) but I don’t think they were pressurized.

Solid answer by Polot.

Pressurization is the issue. In very simple terms, think of a balloon you'd inflate for a party - it get's it's shape principally from the pressure of the air inside it. A square pressurized fuselage could be built where you have corners, but those corners would have to be heavier than the sides. In other words, that square cross section fuselage would be less efficient than the round one.

Indeed it has to do with pressurisation. You can easily figure out what would happen if fuselages were square. Take a piece of cardboard and fold it into a square like that train. Then put your hands in it and push at the inside of the cardboard more or less equally everywhere like the pressurised air would do inside a cabin. That pushes the fuselage everywhere with the same pressure. You'll see that it changes shape into a round piece of cardboard, you'll push the corners out. Now if the cardboard was already round, it would already have the shape it would turn into anyway.

Polot wrote:

There have been square fuselages (see Shorts 330/360) but I don’t think they were pressurized.

They weren't.

Only airliner that isn't really round is the 747 which is more or less egg-shaped if you cut it through vertically. That's the next best thing, it's the reason eggs are shaped the way they are. Because they're more or less round they can take quite some pressure, a square egg would break much easier. Of course if you concentrate all power into one point the egg cracks, but since pressurisation is a force spread equally among the whole fuselage round or egg-shaped is the way to go.

Very interesting reading. Allow me to play Devil's Advocate. Pressurization may play a role and I would agree with this if I didn't see the shape of the A380 fuselage. It's 8-shaped. And most planes (with, I believe, the only exception being Boeing 777) aren't fully circular. So pressurization affects them and tries to make them into a circle, too. Besides, the "Comet factor" was relevant many years ago when airplanes were made from DurAl. These days it's all composites. As for crosswinds... 747 is shaped like a poorly streamlined combine harvester, yet it still flies in crosswinds, doesn't it?

BREECH wrote:

Very interesting reading. Allow me to play Devil's Advocate. Pressurization may play a role and I would agree with this if I didn't see the shape of the A380 fuselage. It's 8-shaped. And most planes (with, I believe, the only exception being Boeing 777) aren't fully circular. So pressurization affects them and tries to make them into a circle, too. Besides, the "Comet factor" was relevant many years ago when airplanes were made from DurAl. These days it's all composites. As for crosswinds... 747 is shaped like a poorly streamlined combine harvester, yet it still flies in crosswinds, doesn't it?

Yes, the A380 has a very fat 8-shaped cross section. But again, it doesn't have corners like a square would - does it? Further, corners - whether they're made from metal or composites - are still heavier than a curved panel.

And what do crosswinds have to do with your argument?

The A380 and many other aircraft have what is called double bubbles, a pair of circular cross sections truncated at the floor beams. The shell of each bubble have a constant radius except for some flaring at the bubbles intersect. The floor beams are in tension to retain the circles.

Flying at 35K with the interior at 6K elevation is about a 8.5 psi pressure difference, or 1,224 PSF. Tubes under internal pressure are totally in tension so are the gold standard in efficiency. An office building floor is designed for 150 PSF or less for the floor, the wall is designed for 30 PSF. To do it rectangular imagine the wall framing for 40 times the wall load or 10 x the floor load.

Being rectangular probably has over a 50% structural weight penalty.

FlyHossD wrote:

Yes, the A380 has a very fat 8-shaped cross section. But again, it doesn't have corners like a square would - does it? Further, corners - whether they're made from metal or composites - are still heavier than a curved panel.
And what do crosswinds have to do with your argument?

I'm not talking about sharp edges. Corners can be rounded. And some weight gain would be compensated by increase in internal volume, especially in the cargo hold. I meantioned crosswinds because someone above said rectangular shape is more susceptible to them.

BREECH wrote:

FlyHossD wrote:

Yes, the A380 has a very fat 8-shaped cross section. But again, it doesn't have corners like a square would - does it? Further, corners - whether they're made from metal or composites - are still heavier than a curved panel.
And what do crosswinds have to do with your argument?

I'm not talking about sharp edges. Corners can be rounded. And some weight gain would be compensated by increase in internal volume, especially in the cargo hold. I meantioned crosswinds because someone above said rectangular shape is more susceptible to them.

You're vastly underestimating the structure required to adequately strengthen it. The forces from pressurization are huge, and happen tens of thousands of times in an airliner's life.

A round/oval cross section is BY FAR the cheapest way to build an airliner. If you rectangular one, the structure would be so heavy, you wouldn't be able to put anything in it no matter what the interior shape.

The Douglas,DC-1 and the later DC-3 had round fuselages long before aircraft were pressurized. The round fuselage is more aerodynamic and and is structurally better than square ones which can be subject to failure due to the sharp corners. I remember the Comet and it's square windows which cracked at the window corners as they were pressurized. Several crashed and the Comet was was pretty much a failure even though oval windows were eventually used on the aircraft.
As an added note, the DC-1 was the first aircraft to have to have trailing edge flaps which helped the DC-3 to become the major success it was.

Making a square submarine would be more practical than a square airliner.

What you should be advocating for is an unpressurized aircraft.
But then, that introduces a whole host of efficiency issues. So basically, yes, you can build a more square airliner, make better use of volume for people and cargo, but it would cost far more to move anything than today's rounded, pressurized craft.

Small planes can be flat sided, as can military planes meant to haul vehicles, drop paratroopers, etc. These are horribly inefficient measured by any reasonable metric. This starts with their low altitude nature, not the shape of the fuselage.

Your train observations are irrelevant. Rectangular is optimal for trains.

JayinKitsap wrote:

The A380 and many other aircraft have what is called double bubbles, a pair of circular cross sections truncated at the floor beams. The shell of each bubble have a constant radius except for some flaring at the bubbles intersect. The floor beams are in tension to retain the circles.

Yes, there are plenty of double bubbles out there, 707, 717, 720, 727, 737, 757 to name a few. But not the A380.

A380 is in fact a triple bubble, the only triple bubble in the world. Both the lower and upper floor beams are in tension when the fuselage is pressurised.

One plane shows off better than any other what double bubble means, the Boeing 377. Have a look at
https://en.wikipedia.org/wiki/Boeing_37 ... ,_BOAC.jpg

Polot wrote:

Planes are pressurized. Trains are not.

Many high speed trains are actually pressurized

VSMUT wrote:

Polot wrote:

Planes are pressurized. Trains are not.

Many high speed trains are actually pressurized

You want to enlighten us on that?

I think pressurisation on high-speed trains is to reduce the effect of entering and exiting tunnels. I wager the pressure differential is nowhere near that on an aircraft.

Pressurization is a huge consideration for going round, but even on un-pressurized aircraft, the round fuselage cross section is the most structurally efficient, strength per weight.

Trains are just "air-sealed", even the Qinghai-Tibet Railway is not pressurized (as often falsely claimed on the news).

rbretas wrote:

Trains are just "air-sealed", even the Qinghai-Tibet Railway is not pressurized (as often falsely claimed on the news).

From an operator:

The train is pressurised and has an innovative exterior connection system for the air conditioning equipment, plus a hermetic seal using special gaskets for exterior doors and gangway doors between cars.

http://www.renfe.com/EN/viajeros/nuestr ... ficha.html

I can tell you that the DDG series of destroyers used by the US Navy are slightly pressurized and this differential can be increased to allow the ship to sail through an area contaminated by nuclear fallout or biological weapons


I spent a week in transit from HNL to SAN on the USS Milius

Max Q wrote:

VSMUT wrote:

Polot wrote:

Planes are pressurized. Trains are not.

Many high speed trains are actually pressurized

You want to enlighten us on that?

As Starlion said, some are slightly pressurized in order to prevent the sudden pressure surge when passing other trains or tunnels at high speed. The German ICE is one, you can hear the slight whoosh of rushing air when the doors open and the pressure is released.
The ICEs are also a lot more cylindrical than the newer Shinkansen's though, and the latter is a bit unique in being so square. Most train manufacturers don't even build square trains as by the OP.

BREECH wrote:

Pressurization may play a role

Let me just stop you there and state: pressurisation is the ONLY reason.

Think about this: which is easier to design and build: a box-shaped house or a spherical house? A rectangular smartphone or a round one? How many bridges are tubular-shaped, yet how many box-girder bridges have you seen? Life would be much easier for Airbus and Boeing if they could just bolt together a flying box-bridge.(*)

Gas cannisters, tunnels, tennis-balls... all these things have their shape due to pressure.

(*) metaphorically speaking, of course - in fact the structural efficiency of a thin-walled pressurised tube is very good

the shape of the A380 fuselage. It's 8-shaped. And most planes (with, I believe, the only exception being Boeing 777) aren't fully circular. So pressurization affects them and tries to make them into a circle, too.

The floors are part of the equation as well - there have been "double-bubble" designs which look just lilke when two soap-bubbles join - where the floor is the flat membrane in the middle. The A380 is similar to a triple-bubble but with a more continuous surface. Same story with the 747.

Besides, the "Comet factor" was relevant many years ago when airplanes were made from DurAl. These days it's all composites.

That changes nothing. Whatever you make the skin and joints from will still feel pressurisation stresses in the same directions.

Pressurization in buildings, trains, and ships is typically on the order of 0.1 PSI or less. At .1 psi doors to the exterior have a handle force of over 100 pounds, so airlocks or revolving doors are needed. Containments for asbestos cleanup require 0.02" water column. Building HVAC systems work usually at 0-2" water column, but high rises and other buildings can have the main trunk lines operating up to 6" water column. 12" water column = .43 PSI. At altitude a plane is seeing 8 PSI pressurization.

Most non-pressurised airplanes are also round. So the pressurisation must only be one of many reasons.

SomebodyInTLS wrote:

BREECH wrote:

Pressurization may play a role

Let me just stop you there and state: pressurisation is the ONLY reason.

Think about this: which is easier to design and build: a box-shaped house or a spherical house? A rectangular smartphone or a round one? How many bridges are tubular-shaped, yet how many box-girder bridges have you seen? Life would be much easier for Airbus and Boeing if they could just bolt together a flying box-bridge.(*)

Gas cannisters, tunnels, tennis-balls... all these things have their shape due to pressure.

(*) metaphorically speaking, of course - in fact the structural efficiency of a thin-walled pressurised tube is very good

the shape of the A380 fuselage. It's 8-shaped. And most planes (with, I believe, the only exception being Boeing 777) aren't fully circular. So pressurization affects them and tries to make them into a circle, too.

The floors are part of the equation as well - there have been "double-bubble" designs which look just lilke when two soap-bubbles join - where the floor is the flat membrane in the middle. The A380 is similar to a triple-bubble but with a more continuous surface. Same story with the 747.

Besides, the "Comet factor" was relevant many years ago when airplanes were made from DurAl. These days it's all composites.

That changes nothing. Whatever you make the skin and joints from will still feel pressurisation stresses in the same directions.

Disagree it’s the only reason



A flying box is a lot more draggy than
a cylinder

SomebodyInTLS wrote:

BREECH wrote:

Pressurization may play a role

Let me just stop you there and state: pressurisation is the ONLY reason.

Think about this: which is easier to design and build: a box-shaped house or a spherical house? A rectangular smartphone or a round one? How many bridges are tubular-shaped, yet how many box-girder bridges have you seen? Life would be much easier for Airbus and Boeing if they could just bolt together a flying box-bridge.(*)

Gas cannisters, tunnels, tennis-balls... all these things have their shape due to pressure.

Look at that circular cross section!

Boeing advanced blended wing body concept 2011 [Public domain], by NASA/The Boeing Company, from Wikimedia Commons

Yeah, I knew people would take my generalisation (which I even flag as such) and run with it

For trains, even high speed ones, a rectangular-ish cross section is undoubtedly more efficient.

Rail transport have ruling gages in both height and width. With these constraints, a rectangular body will have superior volume capability relative to a circular cross section.

Besides, in a high speed train, the relative wind is mostly straight down the track. The main drag difference between the two sections will be skin friction. On the basis of contained volume, the rectangular section probably has lower skin friction drag than a circular one.

In principle you could have a conventionally shaped pressurized cylinder inside a nonpressurized rectangular box. From the outside it would look like a rectangular fuselage. In the space "between the two fuselages" you could carry loads that do not require pressurization and that could be accommodated in what would be a rather oddly spaced compartment. Perhaps even fuel. Aerodynamically there would probably be massive flow separation at the corners, and the structural extra weight would be significant, so the issue would be whether the benefit of the extra cargo would offset these major disadvantages. My gut feeling is that it wouldn't but I haven't done even a back-of-the-envelope calculation.

D L X wrote:

Look at that circular cross section!

Boeing advanced blended wing body concept 2011 [Public domain], by NASA/The Boeing Company, from Wikimedia Commons

THAT is a great example! And my point exactly. Aluminium may have had a problem with cracking but with modern composite materials you can eliminate any problem and give the material any properties, including lack of cracks. Could it be that the aircraft manufacturers are just playing it safe and are concerned that the conservative airline executives would be scared of the new shape?

You can certainly make almost any shape, but physics is physics. Higher loads require higher strength which means higher structural weight.

The question is not "why don't designers have imagination?" but rather "is the weight penalty worth it?" At a certain point, materials will progress to where the weight penalty is small enough to make BWBs viable. Maybe we are at that point already but the industry has a certain level of inertia.

BREECH wrote:

D L X wrote:

Look at that circular cross section!

Boeing advanced blended wing body concept 2011 [Public domain], by NASA/The Boeing Company, from Wikimedia Commons

THAT is a great example! And my point exactly. Aluminium may have had a problem with cracking but with modern composite materials you can eliminate any problem and give the material any properties, including lack of cracks. Could it be that the aircraft manufacturers are just playing it safe and are concerned that the conservative airline executives would be scared of the new shape?

1. Where are the emergency exits?
2. How do the emergency exits work in the event of a belly landing?
3. How has the manufacturer overcome the problem of motion sickness for the outboard passengers whose seats rise or fall several feet every time the plane banks?

Just looking through pics of old aircraft, I'd say there is a transition during the 1930s from boxy designs to fully rounded designs, with an intermediate step being boxy-with-rounded-corners (à la Ju52, or many French designs of the era, like the De332)

I'd venture a guess and say this is likely due to the combination of several factors :
- aerodynamics "cleanup" begun in the late 20s, and accelerated by the ever more powerful engines, and thus speeds
- the introduction of metallic skins, in particular stressed ones (semi-monocoque/monocoque)
- and finally the introduction of pressurisation

Another possible factor may be the decreased influence of naval design and the move away from sea-planes towards pure "air-craft". Apart from the bow, ships are fairly flat faced (especially the superstructures)

BREECH wrote:

We'd get a MUCH more spacious interior. And the benefits for cargo are ENORMOUS. No more tetris-shaped containers. A lot more cargo space

While some of your points may be true, these ones are not so simple. The space between the round fuselage and the squared cabin/cargo area is usually used for systems : pipes, cables and all the various computers, pumps, valves, fans etc...There are even more of these on an airplane than on a train. And even on trains they have a hard time hiding all these away. Only the last generation have floors fully clear of systems, usually by tucking them away at the ends of the cars. I just read somewhere that the new Siemens HST concept would be among the first trains to have a fully end-to-end open flat space available for passenger facilities, without infringement from systems.

BREECH wrote:

Could it be that the aircraft manufacturers are just playing it safe and are concerned that the conservative airline executives would be scared of the new shape?

Flying wings/BWB have been investigated for 75 years at least, without success. The aerostructure aspects are more efficient if you want to carry unpressurised loads, like bombs
But for an airliner, you still need to stick a pressure tube inside, which cancels out most of the weight/drag advantage.
Getting pax in and out (especially in emergencies) is a pain in the backside. Same for fitting in all the systems. And for maintenance, getting access to the pressure tube is much more difficult than when the tube is directly apparent.
And as it is totally different, it is difficult to integrate into existing infrastructures :
- ATC, because its flight characteristics are different
- airports, because size, form factor, pilot visibility, support vehicle access etc... are different
- maintenance, because it may not fit in the usual hangars

And then there is the window/no window debate

To summarize : design issues, high costs of introduction, and operational costs at best similar to tube-with-wings (likely somewhat worse). It may be a bit quieter, though

BREECH wrote:

D L X wrote:

Look at that circular cross section!

Boeing advanced blended wing body concept 2011 [Public domain], by NASA/The Boeing Company, from Wikimedia Commons

THAT is a great example! And my point exactly. Aluminium may have had a problem with cracking but with modern composite materials you can eliminate any problem and give the material any properties, including lack of cracks. Could it be that the aircraft manufacturers are just playing it safe and are concerned that the conservative airline executives would be scared of the new shape?

One major reason why we don't all fly in BWB is that the need for pressurization and the non-optimized shape leads to extra structure and weight penalties, which for now offset the aerodynamic gains the shape offers. It isn't playing it safe but rather finding a compromise between your structures and your aero department. And while some technologies have advanced, the very basics of mechanical stress haven't changed for billions of years - you will find very few rectangular structures in nature.
All these groups have their own preferences how to design the perfect aircraft and rarely are they all identical:


For the shinkansen and other trains, it is obviously very important to make use of most of the rectangular crosssection the track offers. Weight is much less important than on aircraft, as are aerodynamics. When in doubt, you can always add some power. Doing so on aircraft means larger fuel tanks & wings, heavier engines and overall a major redesign. While some trains are pressurized, the pressure difference is much lower. On aircraft, pressurization is a key element of structural design, as are aerodynamics (which also benefit from a round fuselage).

Yes, some aircraft are not completely round. As always, there is a compromise. Sometimes, adding some weight to reduce drag or increase the available volume is a better solution and sometimes, it isn't. If you look at the proposed 797, this seems to be one of the important decisions delaying the decision: Oval or round fuselage?

Let me show you the relevance of different preferences on two examples:

The Eta is a glider optimized for aerodynamic efficiency, and only that. It doesn't have an engine* probably has the best aerodynamics of all aircraft (glide ratio ~70:1). It is capable of tranporting 2 people at 80 - 280 km/h.

The Cessna 152 is optimized for easy construction, easy ground handling and overall versatility. It has a - more or less - powerful engine and doesn't give a shit about aerodynamics (glide ratio ~10:1) It is capable of transporting 2 people at 80 - 200 km/h.

Similar mission differences influence the design differences of HSR and aircraft.

*yes it does but it is designed as a glider and spends most of the time without an operating engine

BREECH wrote:

Aluminium may have had a problem with cracking but with modern composite materials you can eliminate any problem and give the material any properties, including lack of cracks.

I can tell you don't design composite structures for aircraft...

SomebodyInTLS wrote:

BREECH wrote:

Aluminium may have had a problem with cracking but with modern composite materials you can eliminate any problem and give the material any properties, including lack of cracks.

I can tell you don't design composite structures for aircraft...

Someone should tell him about delamination of modern composite materials.........

IADFCO wrote:

In principle you could have a conventionally shaped pressurized cylinder inside a nonpressurized rectangular box. From the outside it would look like a rectangular fuselage. In the space "between the two fuselages" you could carry loads that do not require pressurization and that could be accommodated in what would be a rather oddly spaced compartment. Perhaps even fuel. Aerodynamically there would probably be massive flow separation at the corners, and the structural extra weight would be significant, so the issue would be whether the benefit of the extra cargo would offset these major disadvantages. My gut feeling is that it wouldn't but I haven't done even a back-of-the-envelope calculation.

A bit like a C130 !

airmagnac wrote:

Just looking through pics of old aircraft, I'd say there is a transition during the 1930s from boxy designs to fully rounded designs, with an intermediate step being boxy-with-rounded-corners (à la Ju52, or many French designs of the era, like the De332)

Thank you! That was very interesting and helpful!

mxaxai wrote:

That's a nice representation of the problems they need to solve. I see your point about different purposes and solutions. Very interesting. Thank you.

ZaphodHarkonnen wrote:

One very important difference that was lightly touched on is that trains have to deal with a structure gauge at all points along route of travel.

Well, so do airplanes. They don't have the actual structures they have to fit in, but they have to stay within certain limits dictated by the current infrastructure. I may be wrong, but didn't Lockheed Constellation have three tails because a single tall one wouldn't fit into the then existing hangars?

BREECH wrote:

...didn't Lockheed Constellation have three tails because a single tall one wouldn't fit into the then existing hangars?

Do you really think that is a helpful comparison?


Gruß, masi1157

D L X wrote:

While I'm sure pressurization has something to do with it, I thought it was more because planes don't necessarily move through the air in the direction straight down the nose through the tail. A squarer cross section would be less aerodynamic to the crosswise winds. That's not a problem with slower vehicles like trains, or the Shorts 330/360.

No, pressurization has everything to do with it.

The only things aerodynamic about the Shorts is...nothing. The box a refrigerator comes in is more aerodynamic, and they're known, when given a little freedom, for seeking out trailer parks for nesting purposes. The Irish concorde was never intended to break any barriers, but most repairs are conducted along the trailing edge to smooth out the bird strikes.

IIRC, the airfoil on the Shorts is aerodynamic.
https://www.airteamimages.com/short-360 ... 85113.html
... barely so. The 700 Series Shinkansen of the Air.
https://en.wikipedia.org/wiki/700_Serie ... en_700.jpg

Much in the same way that a barn door in a gale will eventually gain lift and fly away if pushed hard enough. There is some evidence that the Shorts flies only because the earth repels it.

planewasted wrote:

Most non-pressurised airplanes are also round. So the pressurisation must only be one of many reasons.

That is completely false. Most pressurized aircraft are essentially 2 circles that join at the aircraft main deck floor structure with the floor structure maintaining structural integrity were the 2 circles meet.