The Tiny Hole in Airplane Windows: What It Really Does

A Hole That Looks Like a Flaw

Lean toward the window on your next flight and look at the bottom edge of the plastic. There is a hole there, two or three millimetres across, drilled clean through. The first time you notice it, the reaction is a small cold drop in the stomach: you are at 11,000 metres with a hole in the window.

It is deliberate. It is called a bleed hole, or a breather hole, and it does not relieve pressure on the window. It decides which pane carries the load.

A cabin window is not one sheet. FAA Advisory Circular 25.775-1 describes the standard arrangement: two structural panes, a main load-bearing pane and a fail-safe pane, separated by an air gap, plus inner facing panes. So there are three sheets between you and the sky. The outer pane is structural, around 11 mm of acrylic; behind it, across a sealed gap, sits the 5 mm middle pane, the fail-safe; behind that is the sheet you can smudge with your forehead, a scratch pane belonging to the cabin trim, carrying no load and perforated by design.

The bleed hole is in the middle pane, near the bottom, and the confusion in popular accounts is purely about counting. Boeing patent US7267302 puts a breather hole on the inner pane of an assembly it defines as holding two structural panes; Airbus patent US5884865 puts a ventilation hole near the bottom of the inner pane of a double-pane window. Both mean the middle pane of the three you can see.

The Force on a Single Window

To see why this matters, put a number on the load. Air at altitude is thin, in the way the thinning atmosphere always is: pressure falls roughly exponentially with height. Run the International Standard Atmosphere barometric formula at 11,000 m and you get 22.6 kPa, a little over a fifth of sea level.

The cabin is not at sea level either. It sits at a pressure equivalent to between about 1,800 and 2,400 m, 6,000 to 8,000 feet. A 737 or A320 runs near 8,000 feet at cruise; the 787 and A350 hold closer to 6,000 feet, their composite fuselages being designed for a larger differential. At 2,000 m the cabin is at 79.5 kPa.

The difference, Δp, is 56.9 kPa — 8.25 psi. Then it is just pressure times area: F = Δp × A. Boeing does not publish window dimensions and the circulating figures are all secondary, but roughly 25 cm by 36 cm is the number usually quoted for a single-aisle window, an area near 0.09 m². That gives 5.1 kN, about 522 kilograms of force: more than half a tonne pushing outward on one rectangle of plastic, every hour in the air. Multiply by the hundred-odd windows on a narrowbody and the fuselage carries tens of tonnes through its cutouts alone — a different problem from the one that keeps the aircraft airborne.

Why Engineers Pick One Pane and Mean It

Because the middle pane is pierced, the gap behind the outer pane is permanently connected to the cabin and always sits at cabin pressure. So the outer pane has cabin pressure on one face and thin stratospheric air on the other: it sees the whole 56.9 kPa. The middle pane has cabin pressure on both faces. It sees essentially nothing.

Seal the hole and that collapses. Trapped air would expand and contract with temperature, so the gap would float between cabin and ambient and the two panes would share the differential in a ratio nobody could pin down. Worse, a slow leak through a crack in the outer pane would quietly transfer the load onto the middle pane with nobody aboard any the wiser.

The regulator has written the preference into law. AC 25.775-1 states that where there is an air gap, the fail-safe pane is not loaded unless the main pane has failed, and it requires manufacturers to prove it by suddenly failing the main ply at maximum cabin differential and showing the survivor holds. You cannot run that test without knowing which pane carried the load. A known load path is testable; a shared one is a guess.

Here is the caveat most accounts skip. The hole is far too small to equalise anything quickly. If the outer pane did fail, the gap would not calmly vent; the middle pane would simply become the load-bearer in a fraction of a second, which is exactly what it is certified to do. The hole is not an escape valve. It is a load-path switch, stuck in one position for the life of the window.

Stretched Acrylic, Not Glass

The panes are polymethyl methacrylate — acrylic — in the biaxially stretched grade. Glass has better optical quality, and windscreens do use glass laminates, but a cabin window must survive tens of thousands of pressure cycles, and glass fails catastrophically from a single crack because nothing blunts the crack tip.

Acrylic crazes instead: a web of microscopic voids spreads through the surface, scattering light into a milky haze long before anything structural happens. The FAA notes drily that stretched acrylic panels are generally replaced for loss of transparency from stress crazing before significant structural degradation occurs: the window goes cloudy, a mechanic swaps it, nobody is in danger.

Stretching is what makes that work. Biaxial stretching aligns the polymer chains into a laminar structure, raising crazing resistance, slowing crack propagation and cutting notch sensitivity — which matters enormously when you have just drilled a hole in the pane. Cast acrylic gets no such treatment, and the regulator prices the difference bluntly: in the fail-safe analysis stretched acrylic carries a load factor of 2.0, cast acrylic 4.0. Resistance to cracking rather than raw strength buys that factor of two.

Frost, Fog, and the Hole’s Second Job

The second function is real but routinely overstated. The gap holds cabin air, and the outer pane is pressed against air at roughly −56 °C. Moist air against a surface that cold in a sealed cavity gives condensation, then frost, then a milky window — the same dew-point physics that fogs your breath on a winter morning. The hole, near the bottom of the pane, lets vapour and meltwater out.

But it is secondary, and in some ways the cause rather than the cure. Both manufacturers call pressure equalisation the designed purpose. Airbus patent US5884865 exists precisely because the hole admits humid cabin air that frosts on the outer pane and puddles in the gap on descent; the patented fix is to plug it and feed the gap from a desiccant reservoir. A Boeing patent reports lab tests in which window vibration pumps air through the hole in pulses, driving moisture in. The hole vents moisture and also admits it.

Rounded Corners and the Comet Story Everyone Gets Wrong

Behind all of this sits the fuselage, a thin cylinder under internal pressure carrying hoop stress σ = pr/t. Cut a hole in it and stress crowds around the opening — more than three times the remote stress, the Comet inquiry found. Sharp corners concentrate it worse than curves, so windows get generous radii.

Here almost everyone reaches for the de Havilland Comet, and almost everyone gets it wrong. The story is that its square windows cracked at the corners. The 1955 Court of Inquiry under Lord Cohen found fatigue failure of a pressure cabin designed for 8.25 psi, and the crack origin identified on G-ALYP was a fastener hole near a corner of an automatic direction finder antenna aperture in the cabin roof — not a passenger window. The Court made no finding about passenger window shape. The RAE water-tank test of a second airframe failed elsewhere again, at a bolt hole near a forward escape hatch.

Paul Withey’s later analysis is sharper still. The Comet 1 windows were rectangular with rounded corners, nearly identical to the Douglas DC-6’s, which never broke up. The decisive factor was that window and cutout doublers were punch-riveted into 0.71 mm skin instead of bonded with de Havilland’s own Redux process, leaving sub-millimetre starter cracks in the fastener holes. On G-ALYP the crack began some 90 mm from the aperture edge, hidden under a bolt head and a doubler plate, and grew forward into rising stress until it passed critical length. Later Comet marks went to oval windows largely because ovals were easier to Redux-bond, removing rivets — not because corners were the problem. The escape hatch cutouts, where the tank test actually failed, stayed rectangular.

That is the bleed hole’s lesson too: the Comet was lost not to a shape but to an unknown load path with an undetectable flaw in it. The hole in your window is the opposite instinct, made small and permanent — decide where the force goes, write it down, then test what you wrote down.

Sources

FAA Advisory Circular 25.775-1, Windows and Windshields, 17 January 2003. Boeing patent US7267302B2. Airbus patents US5884865A and US9114870B2 — 11 mm and 5.25 mm pane thicknesses. Report of the Court of Inquiry into the accidents to Comet G-ALYP and G-ALYY, C.A.P. 127, 1955. P. A. Withey, Engineering Failure Analysis 4(2), 1997, with his 2019 revision. ISO 2533 International Standard Atmosphere.