Missing windows discovered on U.S.-bound plane after departing London | Hacker News Reader

Missing windows discovered on U.S.-bound plane after departing London

npr.org

126 pointsasimpleusecase91 comments

Just to add some color on this from a lighting point of view. The film lights used look like Maxibrutes[1], which are 12,000 watts of old school, non-LED power per fixture. Looks like they used six lights, so 72,000 watts.

Now in a past life, I've stood a similar distance in front of a giant, super diffuse 72,000 watt light, while wearing full body welding protective gear/goggles, and it was cooking hot right through all the leather gear.

However, this looks even worse - it looks like they are using very tight beams on those fixtures. From the photos on the ground with the lights, it looks like only three and half windows are getting lit up. That's an insane amount of power being concentrated into a small space, and yeah, I'd believe it could wreck some damage.

(Hah, just read the AAIB report[2], yes, they are indeed Maxibrutes)

[1] https://www.filmgear.net/index.php?route=product/product&pro...

[2] https://assets.publishing.service.gov.uk/media/6544b3089e05f...

Not to diminish your valuable insight, but if they were simulating a sunrise, I doubt they had these at full throttle. My guess is they used old school lightning panels because of the orangey filament glow at lower wattages. WDYT?

Oftentimes a simulation will involve "overpowering the sun". You're not doing the lighting in the dark, you're adding light to an existing scene to simulate the sun being where you want it to be.

Day for night shooting is a similar idea - by changing color balance and adding more light to emulate the moon - it's possible to give the impression of darkness.

Even with the lights operating at half throttle, there is also a significant amount of energy in the non-visible spectrum from 6 of them focused on a small area for 5 hours. It’s amazing the acrylic windows didn’t melt, only the foam seal around them.

The window panes did melt, or at least deformed and shrunk. That's how they departed the aircraft.

The mentioned foam is on the inside of the cabin liner. The interior walls that line the actual pressure hull. The foam seals role is probably noise and vibration reduction, not pressurization.

These are almost always run full power, and if you want less light, you just flip some of the bulbs off. (You can actually see two bulbs off in the far middle fixture in the photo)

An yes, they are probably going for an orange glow inside.

Thanks for the context, that's really interesting: big numbers, wow.

Question: why not LED? Is there some quality of incandescent (I assume) bulbs which LEDs can't duplicate more cheaply? I guess so, but what would you say it is?

These lights date from the 60's, and were, and continue to be, the cheapest and most reliable possible way to throw boatloads of light at something. They are also flexible in that you can load them up with very narrow bulbs, as used here, or wide, or many choices in between.

It's hard to overemphasize just how much the film industry values reliability over everything else. These things are stupidly reliable. There's probably a big case somewhere on a truck nearby with a bunch of replacement bulbs of different zooms, and that's about the only thing that can go wrong here.

I've written the software stack for two different high-performance theatrical LEDs, and the hardware side is near rocket science to be ultra reliable and ultra performant. It's possible to get the same qualities of light out of LED's as from a tungsten bulb, but it's expensive and difficult, so most LED's lights are ether noticeably different, stupidly expensive, or both.

On the other hand, the maxi brute here costs 1/10th as much, and has been battle tested by generations of roadies.

Plus, the warm glow is nice.

> Question: why not LED? Is there some quality of incandescent (I assume) bulbs which LEDs can't duplicate more cheaply? I guess so, but what would you say it is?

I'd suspect the coverage of the light spectrum (CRI or equivalent) also plays a role.

Spectrum/shape of the spectral efficiency curve. Sure, you can get a vague color temperature match out of LED, but it's HARD to get a curve out of LED that looks at all like sunlight or like tungsten incandescents. And even the really good ones are noticeably deficient in reds. I can totally believe that if they're attempting to simulate a sunrise there's nothing like massive incandescents.

I mean insolation at sea level is on the order of several hundred watts/square meter depending on your latitude and the time of day. Sunlight is very bright compared to almost every artificial light source so I'm still surprised that a window designed to bake in direct sunlight for hours was damaged by these film lights.

The linked AAIB report is pretty interesting, and surprisingly easy reading.

https://assets.publishing.service.gov.uk/media/6544b3089e05f...

Also interesting from the article:

> An engineer and co-pilot went back to take a look at the window

Is that the same person? Or would they ordinarily be flying with 2/3 pilots and also an engineer? Or did they just happen to have an engineer among the nine passengers on board to call upon like they might a doctor?

Although, with 11 crew and 9 passengers, maybe they do just routinely fly with a doctor, engineer, dentist, therapist, ...?! (It's not actually stated, but seems like maybe it is the film studio or whatever's plane, not that it just happened to be loaned out for filming beforehand. So maybe a funny crew list is more likely than otherwise.)

I assume they mean “flight engineer” , the 3rd flight deck crew member on planes that require 3 person crew. Although those are rare these days.

The aircraft involved was an Airbus A321, so definitely not a "Flight Engineer". Likely a maintenance engineer (please correct me on the job title if I'm wrong here). Basically people who have to know the intricate details of a specific aircraft's systems inside-out.

In Europe they use Engineer interchangeably with Aviation Mechanic.

If this is a tour company, moving a plane, it's not unreasonable that they'd bring a mechanic along in case they needed an inspection on the way. It's cheaper than contracting out a lot of the time.

This was a repositioning flight of a smaller airline, and so everyone on board the airliner was in the industry / worked for the airline. Not a normal flight of passengers.

Well I thought that at first, but then why are so many designated 'crew' if not all? (When I said '9 passengers' I meant the 9 passengers stated in the report.)

Lights do deliver a lot of energy, even in the day and age of LED.

I worked on machine vision projects back in the pre-LED days and the heat on anything used for high-speed, especially with line scan cameras, where the photon density is higher, can be a real problem. I remember some fish inspection project that failed because the fish was cooked by the time it passed under the lights.

yup. and back in my rock'n'roll days, we used to toast bread for breakfast using two of the 1kW lights that are in the maxibrute

I’m really surprised lights used to simulate sunrise placed 20-30ft away could completely melt the foam holding the exterior windows in place.

Oh I don't know. The aircraft appears to be painted with a dark colored paint; that would naturally absorb any heat energy from the lights, and film set lights are really bright.

Yes, but what about the same place, sitting on the tarmac in hot latitude airport (Singapore, Doha, Las Vegas, etc.) ? It is a bit puzzling to know that heat from lights could damage windows of an airplane.

Stage lights can reach 90c.

Why don't they bake the actors then?

SPF 200. Also the heat will dissipate over distance. At a rock concert where the ambient temperature is 32c due to being packed with humans, the heat can’t dissipate at well and so you get 60c on stage. Melt cables. Input jacks. Doc Martin soles.

You’ll often see those giant fans blowing all over the stage. It’s not just for big hair stances, it’s to keep humans alive.

On sets, less humans, more space, so the lights heat will dissipate pretty well after 5-10 feet. However I suspect that a simulator that is simulating the sun using a physical lamp would be using higher power lights to mimic the suns bloom. Even keeping it within operating normals it will get hot.

the sun is very very hot and bright, but it is also very far away.

Inverse square law should apply to the lights too. The lights may be bright and very hot but each meter away from them would mean a significant drop in the transmitted energy.

Inverse square applies to spherically projected energy. It was directed in this case.

Care to elaborate?

Focusing a light, while still leaving the falloff ""spherical"", makes it so the center of the sphere is not at the light source, it's a virtual point significantly behind it. With a narrow beam, being 1 meter from the light and 10 meters from the light might only be 2x different in brightness, or even less.

And for lights with significant width, you drift away from spherical falloff the closer you are and the wider they are. As you get quite close, and the light fills a lot of your field of view, you approach the situation of an infinite wall of light and that has linear falloff.

These lights are huge and wide and can have tight beam angles.

Oh, I accidentally undersold it, infinite wall of light has no falloff.

Infinite line of light has linear falloff.

Your explanation suffices! It's not a point source, but if we half here and there we can kinda pretend it is, for some reason.

Cribbing off [another comment](https://news.ycombinator.com/item?id=38222198), there's as much as 72kW of lighting being focused on a few square meters of the plane's skin. Direct sunlight at noon at the equator is roughly 1kW/m²; the artificial lighting here is well over an order of magnitude brighter than the sun.

I was doing slomo capture of objects being dropped into a small fish tank. I only had a small lighting kit where my largest light was only 750w. To compensate, I placed the lamps mere inches away. By the end of the shoot, the heat melted the top frame of the tank that would normally hold the housing for the tank's light.

Some productions can bring out 10K lamps and larger. They definitely get H O T.

Only 750W?

That's as much heat as some electric stovetops produce.

Isn't that just half a hair dryer? Is it for keeping things warm?

Well, my biggest light might have only been 750w, but I had three of them running. So, a decent space heater. Only these are more useful, as the heat was generated but also made all of that wonderful light.

I guess I can imagine a grownup Easy Bake Oven with those.

I love you for this comment

Yeah, 1K, 2K, 5K, 10K are all available lighting options. Really, anything less than 1K, and people would look at you with that "ahh, isn't that cute" for bringing out your Fisher Price Baby's First Light Kit. At least that's what one gaffer told me he called the kits from Lowel that had 500W, 750W, and the big light was 1K. Before LEDs and fluorescent lighting, "hot" lights had that moniker for a reason. Having stuff ruined because of the heat of a light source was not a rare event, especially for noob gaffer/grip types.

That's about just under two RTX 4080 Tis, measured in techie units that you might be more familiar with.

Is it a common practice for commercial airliners to alternate between being used on a Hollywood set and for regular commercial air travel? Given how important failure analysis is in the airline industry, wouldn't this added complexity in the usage patterns of a commercial jet be a concern?

Not in my experience. It's typically an older body sitting on a back lot somewhere that they dress up as needed to make shooting easier. If you think lugging your carry-on luggage through the aisle is rough, just imagine trying to move cinema cameras through. They move the seats out of the way, so all of the support gear can be put in place.

Shots inside an aircraft usually use a set. If you watch any Netflix series at all, you start to see the same A320-based aircraft cabin set used over and over again in different shows. It looks just like an A320 cabin interior, but split in two lengthways and the aisle widened, with a flat panel added in the ceiling to fill in the gap from the widening. Once you know what to look for you realize it’s the same set every time.

It’s possible they used the actual aircraft in this case because they were filming an ad or some other promotional material for the tour company, perhaps.

The AV Herald article is also enlightening (hah)

https://avherald.com/h?article=50faa50b&opt=0

The studio lamps were placed much closer to the aircraft windows than the minimum safe distance of 10m, and there were a lot of them.

According to the spec sheet for the Maxibrute 12 lamp head [1], it has 12 1000W PAR64 halogen lamps and the surface external can reach 400 C.

The data sheet for an Osram PAR64 1000W lamp [2] has the following safety advice:

“To prevent personal injury or damage to property, halogen lamps may only be operated in suitable luminaires designed with suitable mechanisms (cover panel etc.) to ensure that in the event of a lamp bursting no parts/shards can escape and that during operation no ultraviolet radiation can escape. Because of the heat produced only heat-resistant lamp connections (holders) may be used. Do not operate near people, or any materials that are flammable, sensitive to heat or affected by drying out.

[1] https://www.filmgear.net/image/catalog/Spec/01Maxibrute/info...

[2] https://www.osram.com/appsj/pdc/pdf.do?cid=GPS01_1044353&vid...

Second lighting-related incident I've heard about in a week (after the Bored Apes UV incident). Seems proper lighting engineers are unsung heroes?

I’d love to read an AAIB report for that accident! It sounds like there were some significant eye injuries.

Probably won't be an AAIB report as it was at a crypto conference in Hong Kong I believe.

Indeed, the AAIB report I had in mind is purely hypothetical.

There are other organizations better-positioned to publish a visually appealing, well-written preliminary report of a non-aircraft accident. But I like the AAIB style of having concise descriptions, clear illustrations, and succinct tabular data. I prefer it to the NTSB’s standard format, and do believe I will read some more of these.

The headline seems like a great unit test for an NLP system (and, it is badly written). We all got the gist, but reading it literally it means the opposite of what it's trying to convey.

I think I failed the captcha, as I had a picture in my head of literal windows being found inside the cabin and was trying to guess why that would be so important. Stained glass? Stolen from an English country house?

Same for me. At first I thought maybe it meant historical versions of Windows that had been assumed lost.

Such word order shuffle German language can afford maybe but English cannot tolerate that much.

I really had to read the headline multiple times and assumed that some windows of historical value were being stolen from London and were probably heading to some museum.

I don’t understand how these two statements are both possible:

> "Several passengers recalled that after takeoff the aircraft cabin seemed noisier and colder than they were used to," investigators wrote

…

> the plane had remained "pressurized normally," investigators wrote.

Isn't that just because the pressurization system had enough capacity that it could keep up with the extra loss of air?

At up to 10,000 feet I imagine the pressurization system could keep pace with the differential from the missing windows. If they had gone much higher, it might not have been able to keep pace. If the cabin loses pressure at higher altitude, the procedure is to descend to 10,000 feet or less where pressurization is not needed for passenger survival. In this case, the problem was detected around the time they were crossing above 10,000 feet.

Shouldn’t that fire an alert of some sort that the system is having to work at higher capacity than normal to maintain pressure for the current altitude?

On a modern airliner, you'll get an alarm when the cabin pressure exceeds the pressure equivalent 10K feet. A few thousand feet of cabin altitude higher and the rubber jungle of masks appears. You won't get an alarm for just cabin pressure differential begin lower than usual.

So the air conditioning system essentially runs at a fixed capacity (although some very modern aircraft with electric compressors may vary). Pressure is regulated by controlling the amount of air exhausted by the cabin outflow valve near the back of the airplane.

Since they're are so many variables that can go into the inflow and outflow rates it's unlikely that anybody had characterized outflow valve position in all possible conditions. Nor is it's specific position of any concern, just that it be in the right position to maintain the selected cabin pressure.

So no, it's unlikely that they're would be an alert attached to its position, what is however likely is that they're is an alert to warn that cabin pressure has deviated from the commanded pressure. In this case though they never found themselves in a position where the aircraft was unable to regulate cabin pressure.

Planes that old are designed to actively pressurize to 6k/8k ft equivalent plus 13k feet is not that high, less high than mt Whitney.

Maybe the windows are double-paned, and the loss of the outer pane resulted in poorer thermal and sound insulation but not a drop in pressure.

Airplanes are not airtight, so the system is designed to continue to pressurise the cabin as the air escapes.

As they didn’t go above 14,000 feet, I’m guessing it could keep up with the pressure loss from the windows.

Aircraft windows are double paned. On the inner pane, there is actually a small hole that allows air to pass to the outer pane, and makes it so that the outer pane is holding in the pressure during normal operation.

If the outer pane is cracked (or lost entirely), then the inner pane will take on the pressure. The hole in the inner pane is small enough that the aircraft’s pressurization system can still maintain pressure even if that hole is leaking air.

However that hole likely makes a lot of noise if air is actively rushing through it, and also without the outer pane, there would be less insulation against the cold air and noise outside, and also the missing pane would create an aerodynamic change that I think would cause a lot more noise from the wind.

edit: actually per the AVHerald article, some of the windows were missing both panes (although from the pictures it looks like the inner pane is still intact?), so it must have just been that the pressurization system was just strong enough to keep up with the hole in the fuselage.

> and several window panes appeared to have been warped by the thermal heat

that's the worst kind of heat. They were lucky.

Strangely funny, that was the fate of the De Havilland Comet.

https://www.baesystems.com/en/heritage/de-havilland-comet-1-...

Not very similar, the design error of the Comet made the entire cabin fail.

> window panes appeared to have been warped by the thermal heat

Are there kinds of heat that are not “thermal”?

"Discovered" is an abuse of language IMO.

Sounds like somebody shipped a bug into production.

This was a hardware problem or, more specifically, a process problem. The outside inspection should have occurred or the indoor inspection should have caught it. There is no "bug". People are imperfect and failures of all sorts, happen due to that.

Sounds more like a Customer Use Error. Speculation is that the panes only fell off during the takeoff roll/climbout so inspection wouldn't have caught it.

> A U.S.-bound plane took off from London last month with four damaged window panes, including two that were completely missing,

Doesn't "sound" like that at all. Inspection could have caught it after the plane was modified...which should have never happened without an engineer present and a full post-filming inspection anyway. Their insurance is going to go up because they compromised a plane then let it fly with an immediate mechanical failure.

The crew are not customers nor did the customer passengers affect the failure. It's a process failure, not a bug or "customer" error.

Sounds like someone shipped a bug to me and QA didn't catch it

I'm curious, how does a flight crew not notice a window entirely missing?!? I get that most of the passengers were near the middle of the craft, but wouldn't it still have been fairly noticeable as the flight attendants were walking down the aisle? Or immediately upon takeoff? Why did it take them reaching 13,000 before it was obvious a window was not there? TFA doesn't really say.

My only guess is that the shades pulled down on most of the windows, not making the situation apparent. But even then, it one would imagine a substantial rattle from wind. Even cracking the window of my car at high speed is incredibly noisy, and my car can only go the fraction of the speed of an airplane. This whole thing is confusing.

Only the outside panes were missing. Hard to notice.

Ah, that makes much more sense. I must have missed that. Thanks for the clarification!

The window's windows, you could say.

The windows weren't missing during inspection. They were lost during takeoff and in flight. The article says the plane was used in a theatrical production that involved very hot lights that melted the the window seals. No one noticed the melted seals and when the plane took off the windows came out. They found the windows on the runway.

That's a very obscure test case...

Ah, thanks -- still disturbing but a little more understandable now.

I wonder if there is some kind of industry policy against things like this. One airline or plane owner tries to make a quick buck with a plane, that results in a new FAA restriction in perpetuity for all US flights, causing exponentially more cost to all carriers.

So unexpected to see actual “hot lights” used on location nowadays. Most have gone for LED. I am kind of glad and sad at the same time.

If you already have the gear, no need to not use it.

On set, sure, but on the move, LED can be battery powered. But I’m camp Tungsten. :)

way to move the goal posts.

I don't understand... I'm not arguing any position in particular here. :)

"on set...but on the move"

Being mobile and being setup on set are like comparing apples and oranges. The only thing similar is there's a camera and some actors. Run-n-gun shooters looking for battery operated lighting think totally differently than a crew shooting not "on the move". You're essentially comparing a cinematic setup to a newsie while essentially insulting both.

I am very sorry and didn’t intend to insult anyone! (This time.)

There is no way leds can give as natural lights as a glowing wire of Wolfram? I am still missing the glow of real light bulbs. The leds gotten better but they still look fake.

They found one shattered pane on the runway.

Reply on news.ycombinator.com