58
$\begingroup$

Why are the windows of an airplane usually rounded at the edges?

enter image description here

$\endgroup$
  • 24
    $\begingroup$ <style>border-radius:50px;</style> That's about it for a programmer :P $\endgroup$ – Hanky Panky Jun 15 '15 at 7:06
  • 4
    $\begingroup$ That's HTML, probably not used for airplanes :) $\endgroup$ – anshabhi Jun 15 '15 at 7:55
  • 5
    $\begingroup$ @Iceman that CSS is lexically incorrect.. :P $\endgroup$ – QuarK Jun 15 '15 at 9:12
  • 4
    $\begingroup$ Airliner (Browser) Compatibility you know :) $\endgroup$ – Hanky Panky Jun 15 '15 at 9:17
  • 6
    $\begingroup$ Seems it's valid CSS: window{color:white;border-radius:15cm} $\endgroup$ – Gustavo Rodrigues Jun 15 '15 at 13:26
71
$\begingroup$

A pressurized fuselage needs rounded corners at all cut-outs to avoid the accumulation of stress in corners. If the corners were not rounded, the fuselage skin would need to be locally reinforced, or it would begin to crack there soon.

This was learned the hard way by de Havilland, when their Comet jet airliner began to plunge out of the sky for unknown reasons. In one of the first major crash investigations in aviation history, the wreck of one was lifted from the seafloor off Elba and re-assembled. Then it was learned that the pressurized cabin had burst from cracks emanating from the window corners.

de Havilland Comet

de Havilland Comet I. Note the square windows. The window which caused the destruction of G-ALYP was on top of the fuselage and cannot be seen in this picture. (Picture source)

Color coded stress plot in a fuselage panel with window cutout

Finite element simulation of corner stresses in a Comet fuselage skin. Red denotes the highest stresses. (Picture source)

Other aircraft companies had built pressurized cockpits twenty years earlier and did it right the first time …

$\endgroup$
  • 6
    $\begingroup$ A note on the Havilland: The cracks where detected not by reassembling the body but by building a full size container and using water to pressurize and depressurize: en.wikipedia.org/wiki/De_Havilland_Comet - gvf $\endgroup$ – anshabhi Jun 16 '15 at 9:30
  • 4
    $\begingroup$ If the corners were not rounded, the fuselage skin would need to be locally reinforced .. How much re-inforcing would the 90 degree angled corners require? And would re-inforcing really solve the problem? Did anyone try it again? $\endgroup$ – Firee Jan 22 '16 at 11:03
  • 3
    $\begingroup$ @Firee: In theory, a sharp corner needs an infinite reinforcement. In reality, the reinforcement is stretched out but will still need to be substantial, and would need to have rounded corners itself, so the cutout becomes bigger. Stiffness in the straight part of the window frame would then be much lower, so it would attract less stress. Rounding the corners is simply the better approach. $\endgroup$ – Peter Kämpf Jan 22 '16 at 11:57
  • 1
    $\begingroup$ Oh well, so much for triangular or hexagon or pentagon windows, in the name of design $\endgroup$ – securitydude5 Mar 29 at 13:45
  • 1
    $\begingroup$ @Sean. Right. Oval windows were also used with success, see here $\endgroup$ – Peter Kämpf May 12 at 6:50
45
$\begingroup$

Stress Distribution / Stress Concentration

The sharper you make a corner, the higher the stress will be. This is undesirable since it will increase the stress on the material and can promote fatigue which can lead to structural failure down the road.

You can avoid this problem by having a greater curve radius. This does not only apply to windows, but cabin/cargo doors as well on an aircraft.

pic

Source

When aircraft undergo conversions become freighters and add cargo doors, you will notice that they need to make the metal thicker around the door to help distribute stress, hence the area of unpainted metal:

pic
(source: transportarchive.org.uk)

Source

Calculation

Some material on this, although it make the simplification the stress is only in one direction rather than both as in real life. You can see how as $r$ becomes smaller the stress concentration $K_t$ goes up, raising the maximum stress ($\sigma_{max}$).

enter image description here


Footnote: There will have to be a little bit of 'play' between the window and fuselage for technical reasons, such as thermal and pressure expansion, making this type of arrangement necessary. If you could perfectly carry over the load in every direction to the window this would not be necessary.

$\endgroup$

protected by Jay Carr Jun 16 '15 at 8:38

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.