At the end of last week Aquadom, the world’s largest cylindrical aquarium, unexpectedly shattered and caused an emergency as it flooded both the Berlin hotel that housed it and the surrounding streets. From an engineering perspective it’s a fascinating story, because its construction was such that this shouldn’t have happened. We have an analysis of what might have gone wrong from [Luis Batalha] (Nitter), and from it we can learn a little about the properties of the plastic used.
The aquarium was made of an acrylic polymer which has an interesting property — at a certain temperature it transitions between a glass-like state and a rubber-like one. Even at room temperature the acrylic is well below the transition temperature, but as the temperature drops the acrylic becomes exponentially more brittle. When the outside temperature dropped to well below zero the temperature also dropped in the foyer, and the high water pressure became enough to shatter the acrylic.
Sadly few of the fish from the aquarium survived, but fortunately nobody was killed in the incident. News coverage shows how the force of the water destroyed the doors and brought wreckage into the street, and we’re guessing that it will be a while before any other hotel considers such a project as an attraction. Meanwhile we’ve gained a little bit of knowledge about the properties of acrylic, which might come in handy some day.
Header: Chrissie Sternschuppe, CC BY-SA 2.0.
Its sad I gto the news here first and not seeing any coverage from German news. Like they reported it as it happened live but thats it.
It happened LAST WEEK you know ?
And in the meanwhile was the corrupt soccer contest that took all the media space…
Corrupt or not I’m still in love with Messi 😍
You must be mexican mate. Enjoy the GOAT!
Messi is Argentinian.
Go sports ball! I hope they get a hole in one touchdown.
I’ll be happy if my guy just bowls a first down but if he can kick a 3pointer home run I’ll do a grand slam dunk in the field goal.
Of course nobody reported AFTER the event. Why would they? We don’t know how it happened yet. Anything we’ve heard until now, including Luis Batalha’s Twitter thread gives valid and plausible explanations of what MIGHT have happened. But speculation just isn’t NEWS.
Tell that to my local outlets. Speculation and fear sell.
Any glassblower knows that it most likely failed from thermal shock
If a piece of glass is going to change temperature…it had better do it slowly, and evenly
Coefficient of Expansion
You clearly didn’t even read the article. The tank was acrylic, NOT glass.
Victor is correct. Temperature and coeff of expansion applies to glass too just as it goes for acrylic
Ok. With a huge amount of water fast or uneven change is out of question.
Jurnalists live from number of words in articles, and curious people kliking on bombastic titles.
It was acrylic.
You do know it was acrylic
not glass
Clear polymeric materials are non-crystalline “glass” structures. Clear silicon dioxide (commonly called glass) is also a “glass.” What this article describes is a transition from random alignment “glass” to a semi- crystalline structure which is less flexible and more brittle. The sheets of semi-crystalline acrylic were more brittle and thus could not withstand the stress of the water pressure.
I’m not convinced. The enormous amount of water within would have moderated any temperature change in the acrylic to a very slow rate.
The high specific heat of the water may have even exacerbated the situation, as it would result in a larger temperature difference between the interior and exterior surface of the acrylic, but most likely any failure on the outer surface would propagate instantly towards the center as the exterior is in the greatest tension.
Should have gotten the plans from Scotty
@Marc: “Of course nobody reported AFTER the event. Why would they?”
Shame, indeed. News reports should cover events BEFORE they happen.
That was my thought. EVERY outlet reported AFTER the event
I believe you meant speculation isn’t news EXCEPT in the good ‘ol USA!
You mean coverage of a few random tweets from somebody on the internet who looked up the properties of acrylic?
Political news in the usa isn’t speculation–it’s fiction. Just like in a Communist country.
Now that transparent aluminum is truly a thing, time to rebuild using a better material!!!
Trabsparent Aluminum is $10 a square inch. So, very exoensive.
You have any idea how expensive 5 inch thick acrylic is.
How about polycarbonate? It can hold huge loads ( material aeroscreen and windows in racecars is made of) but it’d need some support structure as it flexes
I’m not from Germany so it’s no surprise that this is the first I’ve heard of it, but you? Yeah, that seems odd.
German outlets covered the incident the very hour it happened and then all day and all week. They didn’t tell you WHY it happened because that would be speculation.
I find it suspicious that there is not ccd cam footage leaked.
There is
Isn’t that what the news does speculation and fabrics stories to there truth
It was front-page news on the BBC website for a while.
As a German, I’m not surprised. 😂
I just hope the editorial office tried its best here, at least!
Poor fishes, though. 😔
Sure it was a coverage. In all prominent newspapers, online, you name it. At nauseam, but now it’s water under the bridge. US experts arrived (from the maker of the plexiglass and the aquarium) and we wait for the expert analysis.
I’m sure experts that hail from the manufacturers are going to “discover” it was not due to any flaw in their material or engineering, conveniently.
But of course. Kind of like everyone’s cable company.
It won’t be the first or last disaster caused by material failure due to an overlooked change in outside temperature. Bring back Richard Feynman…….
Well said.
He probably pongs a bit by now.
the fish required the water was kept at a very accurace and constant 26.0’C , how cold can the outside akrylic get while maintaining a 26.0’C water temperature?
If this is the cause, it would only have to get cold to a certain depth to weaken it to the failure point. But it would take time for the outer temperature to penetrate to a depth where this would be the case.
Coincidentally this cold snap in Berlin is thir first prolonged cold snap in nearly a decade. The last time the overnight lows were anywhere near this cold (and stayed there) was in 2014. So it’s certainly plausible that the acrylic has weakened over time.
Except the aquarium was not outside but in a hotel lobby. You know, a place where you typically don´t need to wear wintergear.
That they slightly tuned down the heating to accomodate for gas prices is a possibility, but if they did, it´s likely to 19°C or at most 18°C.
Not enough to crack the acrylic tank.
The air temperature isn’t the thing that matters here though, and the hotel is likely only really pushing hot air at the areas people should be anyway, so the air temperature everywhere else in the lobby could be rather colder. But what really may matter is if its connected to the outside of the building by something with good thermal conductivity, like a giant metal framework and the outside is some amount colder than cold enough. As then the frame around the tank can be cold enough to chill the plastic panels.
It does however seem unlikely to me – maybe the temperature aided the failure but it seems to me there must have been a flaw in the construction somewhere taking away the safety margins to which it would have been constructed for that to happen.
The thermal conductivity of steel ist not that good and warm air rises in the lobby. So even if the slightly reduced the heating it is unlikely to be the cause.
I think UV (daylight) induced degradation or damages during the renovation (scratches, stress) are more likely.
I agree Martin, it doesn’t seem likely and I said as much – the shear mass of water inside that needs to be warm is enough to make it that. But its not entirely implausible for the tank to in places be too cold.
And I have no idea what materials were used in its construction or how the hotel HVAC is set up – its quite possible they are using all sorts of fun forced airflow methods to keep the rising warm air from getting into the spaces people are not or very good thermal conductors in the frame (perhaps because they look nicer than steel?). Without the specifications on any of this it is all speculative.
It survived 15 German winters no worries so really what was any different about this one?
>It survived 15 German winters no worries so really what was any different about this one?
Who knows Andrew, but surviving the past winters is no certainty of surviving the present – if it was not built well enough it may have survived by the skin of its teeth all those previous winters and this one is just too much. I do doubt that though.
Perhaps some muppet with a power drill thought it would be a great idea to drill a few small holes in the tank walls to hang the fairy lights on this year…
Or perhaps it is simply some long standing flaw in its construction finally gave up and it being winter has nothing to do with it at all.
Usually a criteria for a material fail event involves a pre existing flaw….
Temperature gradients in solid materials that don’t conduct well can be equally damaging.
Acrylic will expand at a rate of 0.075 mm / m / K. So a 10m panel secrion would change about 1mm for every degree C it increases in temp.
If one side of the panel were 10C cooler than the other, then you would have a 1cm difference in the two surfaces of the same panel… That would cause a lot of internal stress.
I am sure they designed the tank to account for temperature variations, but perhaps they didn’t account for the gradient between the two sides of the panels.
Acrylic is an insulator with a 0.2 thermal conductivity, compared to 1.05 for glass, or 0.58 for water… So it could easily have a significant gradient across a 8-10cm thick panel.
Just spit balling, but my theory has as much credibility as any Twitter post.
Very much along my lines of thinking Jf – though I doubt it will be the cause in its own right still, but the expected air temp around it was probably in a range higher than it was currently experiencing, and with the extreme cold outside the water heater may even have been introducing water at rather higher temperatures than normal creating a little hot spot. (I assume all the mechanical are in the basement mechanical room, and that floor of the building is likely unheated – no guests down there, so it is easy for the water to get rather cold after leaving the tank and so get the heater working a touch harder than required…)
I’d still think the temperature alone should not have been enough however – it is a high end building so you would expect everything to rather more on the over engineered side – they can afford to double and more on the safety margins, and likely still afford to heat the place entirely normally if they wanted to…
Though as RW says bellow somewhere ‘chain of 1000 cuts’, and somebody else sounds like they know for a fact some of the wrong cleaners were used…
I’ve worked with the manufacturer of the tank, as they built one for my client.
At the time, mine was the largest free standing aquarium in the US.
Without doing forensics, I do not believe the acrylic the tank was a failure of the material or manufacture.
The renovation that was done earlier may have been the cause. Anything could have occurred… a small hit or disruption that facilitated a stress mark would spread over time and expand just enough to initiate a crack that spreads rapidly, just like a cracked windshield glass.
I’ve worked with material since 1978, and know the manufacturer very well, they are well-known wide and I’m pretty sure the tank didn’t spontaneously fail.
This article talks about the transition from a rubbery state to glassy and brittle as the temperature drops, but according to the very Twitter thread it links to, the transition temperature for acrylic is 108°C.
“ Even at room temperature the acrylic is well below the transition temperature, but as the temperature drops the acrylic becomes exponentially more brittle.”
I take that to mean that it was well away from its upper thermal limit, but the low temps got it.
It didn’t fail bc of becoming rubbery. The transition temp wasn’t even relevant. But being at a much lower temperature than that certainly makes the acrylic more brittle than if it was just below transition point.
Nobody has suggested creep of the panel. The units were in place for 15 years (as reported above) and during that time the panels would be subject to miniscule movement with thermal expansion and contraction. If there was insufficient packing to stop movement of the panel in the framework, the panel could slowly move until the edge failed to support the panel and failed. If the framework was out of tolerance the support provided on the tight side would help to push the panel out,
Just another thought
Also we have had cold days in recent years as well, like e.g. in Feburary 2021, where the weather station in Berlin Tempelhof recorded -12.5°C.
Actually, pretty cold. Acrylic is an insulator with a thermal conductivity of 0.2 vs 0.5 for water and 1.05 for glass. At great thicknesses like they would use here, there could be a significant gradient between the sides of the panel.
Because acrylic expands at almost 1mm/m/k (a 1m panel will get 1mm longer when heated by 1 degree C or K)… You can see how two sides of a panel with a large temperature gradient might introduce some stress.
I wondered about this too. Could it be a temperature difference between outer and inner faces causing a stress buildup?
Finally, someone asked a thoughtful question rather than blindly following a half baked speculative explanation.
fake project design, fake material technology
Do you remember the glass suspension bridge disaster in China ?
what amount of compensation was covered by the insurance policy ?>
The aquarium contained tropical fish. It was therefore certainly heated and the acrylic was kept at a constant temperature. The lobby of a luxury hotel is usually also heated. Outside low temperatures should not matter.
Seems fishy to me…
Best answer yet!!!
This hotel had no security cameras in the lobby? Or is the video being withheld by investigators?
sure, aliens with lasers …
🤣
I think Occam’s Razor suggests that sharks with lasers is a more likely scenario.
There would be the strong possibility that the security VCRs or DVRs were in a ground floor or basement room so might have been flooded.
That still elaves the strong possibility, that the storage media remained undamaged.
If the immediate cleanup and structure stabilisation efforts didn’t destroy or discard them, and if they didn’t happen to be wet and repowered, and if they aren’t ziplock bagged in an evidence room going mouldy and/or corroding as we speak instead of being carefully dried and preserved etc etc.
The hotel as absolutely no obligation to share any video from their own CCTV with the internet.
ask for the hammerhead shark
Why you should never tap on an aquarium….
hmmm… interesting stuff, but I have trouble with:
“At the bottom of the tank the pressure was P=ρgh= 10³×10×16 = 160kPa = 2.5 atm”
How do we get from 1.6 to 2.5 ?!?!
What am I missing here? I can’t be the effect of the ambient air pressure, since that is present on both sides of the glass.
after someone on twitter pointed that out, he claims that he ‘forgot to add in atmospheric pressure’: https://nitter.net/luismbat/status/1604766252823728128
(although, as other then pointed out – what, was it operating in a vacuum?)
The link in the article above doesn’t show me the same as the link you mentioned here, so thanks for showing me the better link here.
I wonder, auto centre punch?
Taking clues from other disasters of this type…
Maybe the molasses fermented and pressurised the vessel excessively? ;-)
Well, there was that Chinese kid a few years back who scratched his name into one of the ancient megaliths in Egypt… Maybe he was in Berlin…
He is joining a long and proud tradition going back thousands of years
We didnt learn anything about acrylic we learned that they under engineered the project and whoever designed it was a moron.
I’d almost put money on it being one of those chain of a thousand cuts things, Designer specified W, Architect said it was okay to do X, Engineer okayed Y and Contractor did Z
It was installed in 2003 so it nearly made it to 20 years old.
Would have preferred a Twitter warning. I would not have clicked on the link.
Twitter/FoxNews: Same thing.
Not really. Twitter content is generated by users, not a corporation. Even the NY Times uses twitter constantly.
That said, I’m very happy to see HaD include a Nitter alternative link. Nitter is a Twitter front-end with less cruft and tracking. Always better to link there than directly to twitter.
Would be interesting to check how Embassy Garden hotel in London will manage this with their Sky Pool.
“The structure that forms the pool was fabricated in the United States, in Colorado, by acrylic engineers Reynolds Polymer, Inc.”
https://en.wikipedia.org/wiki/Sky_Pool,_London
this reminds me of something i just read about ‘supertall’/’superthin’ sky scrapers. they’re one-offs so they are their own prototypes in more ways than a regular building is.
so now we’re learning about the aging / maintenance / failure modes of large acrylic structures at the same time as people are swimming in them or walking underneath them. that’s what makes me so uncomfortable about this. i don’t think anything really novel will be learned about acrylic but the practical day-to-day of it is gonna be a voyage of discovery. there’s not a huge experience to draw upon, though i assume there are massive aquariums at museums and so on.
the other thing that makes me real uncomfortable about these kind of structures is that it’s a single piece so it seems like a nick / scratch / deformity will tend to compromise the whole thing, perhaps all-at-once. i don’t know what meaningful structural alternatives there are, but it’s just hard to imagine that unavoidable surface scratches wouldn’t alter stresses throughout the whole thickness of it.
i remember there were some glass-bottomed structures that are made of layers of glass, with even the explicit expectation that the outer layers will eventually be destroyed as wear items. that sort of stuff makes a lot more sense to me, seems a lot easier to reason about as an engineer.
I doubt anything will change any time soon – We will have to wait and see what the results of failure analysis is – I wouldn’t be at all surprised to find the failure wasn’t anything about the age or maintenance of the tank itself but something in the surrounding building being changed – perhaps even just a very tiny but uneven subsidence of the foundations putting a tiny unexpected side load on the tank.
We have so many examples of the same material in use in so many ways over a very long period of time. Many of them similar enough to this for ages before this build, so the material shouldn’t really be throwing up any shockers if the tank builders made it properly and the owners didn’t treat it carelessly somehow.
Also I find it hard to accept that the wall thickness was not so over specified that you could have cut it half and still have a perfectly sound tank for the expected conditions – It is a building with effectively a blank check to build not a social housing project – there isn’t a major need to economize… I suppose the designer could have gotten greedy and insisted on a tank so large it is right on the ragged edge of plausible to create, but its such a modern building that it is from the age of litigation when everyone is going to play it safe you would think…
I believe that you are correct that very high margin for static failures were designed in. The actual designing load is likely an impact abuse load to account for accidents. In my opinion, the failure mode is probably creep rupture. Creep rupture can occur at stressed far below the static load allowables. Polymers are notorious for having poor creep characteristics at slightly elevated temperatures. (Aquarium water temp). Furthermore, it is difficult to obtain design data for material creep characteristics over extremely long loadings. The situation might be exasperated if they used a thin pressure shell assumption when a thick pressure shell analysis was required. For more information about creep failures, search for jet turbine blade creep ruptures. The assertions in this article are nonsensical.
For reasons already mentioned in the comments the “explanation” from that twitter guy makes no sense:
– water was heated for the (saltwater) fish to live
– hotel lobby != outdoors
On the other hand there’s been work done on the aquarium in 2019/2020 which is now being investigated as well…
There’s a whole (German) documentary on the building process of that thing on youtube: https://www.youtube.com/watch?v=E3KjGJ8AZzI
If only we had transparent aluminum… https://memory-alpha.fandom.com/wiki/Transparent_aluminum
Sapphire? More brittle I would think
1/2″ thick Aluminum OxyNitride (aka ALON) can shrug off .50 BMG bullets with just a little spalling. Those same bullets will go through 3″ of “bulletproof” glass.
An ALON tank the size of the Belin one would be super expensive to make. It would take a massive press to compress and melt the powder. Then the smoothing and polishing to make it transparent would take a long time.
But make it 2″ thick and it would literally be bulletproof, and it should easily be able to hold that amount of water. Might even be able to do the job at only 1″ thick.
Making such a massive ALON cylinder could be monetarily feasible if other uses could be found for cylinders of the same diameter and thickness. Imagine turning them horizontal for the main hull parts of tourist submarines. Or sink them vertically into shallow water and put several platforms in with a lift up the center so people can go down underwater to view the sealife.
Line up 100 or so such uses and I bet the funding to make the tubes could be collected.
“1/2″ thick Aluminum OxyNitride (aka ALON) can shrug off .50 BMG bullets with just a little spalling. Those same bullets will go through 3″ of “bulletproof” glass.”
Must be what the bulletproof windows in tom clancy’s ghost recon: breakpoint are made of.
So it will be interesting to see the report on where the failure occured. Because the inner section around the elevator is in compression, and Acrylic should be quite strong in that way. But the outer rim will be in tension, and the highest tension would be at the bottom.
They guy saying that the cold would have done the work of weakening it doesn’t make much sense to me because the 100 tons of water (to pull a number out of the article, haven’t done the math on the actual volume) would have been a huge heat sink. So unless the acrylic has horrible heat transmission properties and the outer layer got cold enough that maybe a shear stress was setup to cause the failure, it just doesn’t make sense.
Seems to me that there was a small crack which grew overtime. Maybe someone rammed a cart into it down load, and it just took time for the crack to propogate and them finally break free.
The number out of the article is one magnitude off. I read it have been 1000 tons of water.
Jup, a million liters.
If the tank was truly 25m tall, there would be a pressure at the base of 35.6 psi. Given approximately 1550 sq in/sq m, that leads to 35.6 * 1550 = 50,530 pounds/sq meter, or nearly 23,000 Kg / sq m. No one is reporting the thickness of the acrylic, nor how strong acrylic is in tension vs. compression, which is also a critical factor. The submersibles using acrylic spheres are incompression, not tension, as is the case with a tank. As to hydrostatic pressure:
https://www.engineeringtoolbox.com/hydrostatic-pressure-water-d_1632.html
Long retired SCUBA instructor…
In the documentary (see my comment above) the American guy somewhere states it to be around 200mm thick.
The tube is made of several pieces which are then glued and polished…
According to the Wikis, the *tank itself* is 25m high, but the water column is – er, I mean was – only 14m.
14m == 46′ == 552″ so for cool salt water (appx 0.038 lb/in^3), the pressure at the bottom was right at about 21psi.
The inner diameter was about 35′ or 420″ (nice) and the walls were about 8.5″ thick, so the lowest part of the acrylic was resisting a hoop stress of about 518 psi ( 8820 lb / 17 in^2)
The lowest rim was also under a right-angle compressive load of about 26 psi from the weight of the 600″ or so of acrylic above it, but this seems like a small factor compared the tensile load of 518 psi ( 3.27Mpa ).
Is 518 psi of tension a lot for acrylic? I have no idea. But without a background in plastics engineering, and given a general feeling that cool acrylic is a brittle-ish material more likely to fail in tension, it sort of feels like it’s getting up there into range where a small stress crack could propagate over time.
The only remotely comparable aquarium I’ve been to, which was huge, was Monterrey bay. Thickness of “glass” (acrylic) is 15″ for a panel 54feet long and 15 ft tall. Sorry for goofy units that’s what press release says. That is still twice as thick for a tank not nearly as deep. Unclear if that is any help for comparison but there it is.
MBA has flat panels and thus the bending stress is much higher than a cylindrical tank.
I wish they would tell us how many fractions of the Earth’s circumference it is instead of how many feet it is. I mean who even still has feet? They are so 19th century!
” News coverage shows how the force of the water destroyed the doors and brought wreckage into the street, and we’re guessing that it will be a while before any other hotel considers such a project as an attraction. ”
And then there are zoos.
My tin foil hat mentality finds the complete and entire lack of actual news and information about the event, other than it happened, suspicious. No video. No witness reports. No engineering description of the tank and no reported analysis at all. No photos of the broken sections with any sort of damages, cracks or anything that could provide scientific objective data as to what happened. Nothing at all. This smacks of some kind of deliberate attempt to suppress information. One dude speculating on twitter is not newsworthy. This comments section is the best source of educated speculation with engineers. It’s still speculation but comments section on a hacker website being “best” source of objective analysis is super awesome and super sad at same time.
I think we would all rather have some actual details to work from, and it does seem bonkers to me that so much effort is spent in the media, despite them running 24hour news channels, in just repeating the same tiny handful of stories with some new talking heads to cover them, and often without once quoting a source or deliberately quoting that source out of order and with all frame of reference removed…
I’m not sure your tin foil hat is actually required here though – there is probably nothing to report on yet anyway. The onsite cleanup and analysis will take time, and there are other things that really should take precedence over some dead tropical fish in a luxury hotel that are not getting widely reported either.
If somebody can go to the local planning office and find the building specs when this tank was put in that would be good though – and being in Germany you can’t tell me there is not an enormous body of paperwork, filled out in triplicate and filed carefully away in the records to dig up so we know how it was built etc.
https://www.youtube.com/watch?v=E3KjGJ8AZzI Building documentary Starts at 3:30. Auto generated subtitles should work just fine.
Outer tube is made of 12 parts, inner tube of 4.
And of course there’s everything documented they need to get to the root cause. It’s Germany lol, the worlds capital of bureaucracy haha
Yes, and the world’s capitol of “you need authorization to see that” along with its own bureaucracy. The documentation is there, it’s just that getting a copy of it might be a pain.
“Your papers are not in order!”
German guy here. News was full of it. Several articles and interviews of eye witnesses and the local fire brigade. The builders of that tank (Americans) reportedly want to come visit and help inspecting what is left and what might’ve happened. It happened in the late night/early moning when most people are still sleeping. It’s a hotel…
Recent news says the completely 3D-scanned the place and did HQ photos and videos of everything.
Instead of spreading rumors they want to get to the cause. Even one or two “experts” asked on tv said: We just don’t know yet. We’ll see what happened soon.
Get out of your conspiracy bubble just because stuff is under investigation. Jeeeeez.
Yes. Thank you. We don’t need talking heads speculating, we need experts who explain it after analysing it.
Yes. It was in the news here in Germany. I saw it in the newspaper the day it happened.
Thank you. Conspiracy theories are the refuge of simple minds.
While I’m pretty sure the eventual explanation will be a nice detailed and (to public) boring engineering description of the failure, there are a bunch of other plausible explanations or at least possibilities ranging from benign (worker backed a forklift into it) to much less so. For insurance and liability reasons there would me many good reasons to not release self-damning evidence on pet or the hotel, the engineering firm, the insurance carriers etc. if even considering those possibilities makes me simple minded, so be it.
Polycarbonate is vastly superior to acrylic. I would have never used acrylic for something like this at all… I feel like someone cut corners in the bid. Wow.
No one uses polycarbonate for aquariums because it yellows over time.
You know what, could have been the cleaning lady diligently sanitising the ground floor surface with alcohol these past 20 months. Or other sanitising chemicals whose use was not envisioned pre Covid.
Maybe they had mantis shrimp.
Best comment yet.
I live in Grand Junction, Colorado where the acrylic for this tank was manufactured by Reynolds Polymer and when I heard of the collapse, I went directly to the source. Apparently, the hotel bartenders were cleaning outside of the tank with Windex and so Reynolds recently repaired the tank for the chemical damage. My friends at Reynolds informed me long ago that chemicals like acetone actually reverse the chemical process of acrylic and so I wasn’t surprised to hear that Windex containing ammonia was the culprit to the recent chemical damage.
I own an acrylic cylinder tank and I only use acrylic cleaner made for aircraft windshields and Meguiar’s clear plastic polish on my multiple acrylic tanks because I know enough about organic chemistry that I know better. Furthermore, I prefer my glass fish tanks don’t scratch like acrylic but they are very heavy in comparison. Acrylic is both strong and light but it’s chemically reactive to cleaning agents. Tragically, alcohol is also an acceptable cleaner on acrylic and so I’ve used isopropyl many times when building acrylic projects in the past because it easily removes Sharpie marker ink. I even used Everclear grain alcohol during the pandemic when isopropyl alcohol was unavailable and so sadly, the bartenders had an acceptable cleaning agent at their fingertips and never knew it.
Now that you mention it, in my biochemistry days we used acrylic (presumably) small tanks to run gels and stuff. If you put the wrong solvents they would soak it up like a sponge (the material noticeably swelled and got all floppy) and then leak at the joints too . I think we repaired them with carbon tet or chloroform or something that basically re-melted it back together like how model airplane cement works. I also built some acrylic boxes to display model airplanes and used some solvent or another to melt the edges together. So a very plausible explanation could be repeated exposure to solvents while cleaning, the material soaks it up and doesn’t have enough time to off-gas it or whatever. Over time with small repeated applications it could significantly penetrate and greatly alter mechanical properties of the material. Especially since ground level cleaning is where the highest pressure exists in the structure of the tank and presumably the most application of cleaners existed.
Chemical damage would be my best guess too. I’m actually thinking it was possible/likely a flushing/cleaning chemical used after the recent maintenance that damaged something out of sight and unnoticed.
Interesting and very plausible as the cause, though I’d still be rather surprised if such surface mistakes could actually cause the tank to fail structurally – using a normal cleaning amount of the wrong volatile organic/chemical cleaner really should ruin the surface finish to warn you off marinading the thing in more of the wrong stuff, and not much else – as it really should be rather more overspec than that structurally, giving lots of time for proper remedial work whatever that may be for the chemistry in play…
However as yet another little mistake adding up to a big failure… and if the operator didn’t have the big instruction manual with all the info required – clean only with x,y and z its not exactly confidence inspiring…
Lattest news :
“A hotel aquarium containing as much water as several Olympic swimming pools and whose walls were thick as several bananas eventually flood an area the size of several football fields”
I’m quite sure, that for the most part thickness isn’t as relevant as evenness
…at least not with glass, as it pertains to strength-durabilty
Perhaps monitoring of such structures could be automated using polarized light / birefringence?
One thing I haven’t seen anyone else mention is this theory I had: What’s changed this year to previous cold years? – The cost of energy. Considering the state of the material changes at particular temp ranges to make it more brittle, I think one possible theory to the failure is that the hotel reduced the temperature of the atrium to save money on their bills, which unfortunately had more of an effect than they realised on the aquarium during the recent cold snap?
+1
Acrylic will stress-crack when exposed to chlorinated solvents.
An improper or unapproved cleaning process/procedure can damage the plastic.
Sometimes a protective film/surface is applied to acrylic, but that can wear out over time if not reaplied.
I would look also at differential thermal expansion (or contraction in this case) As the temperature in the atrium drops the outer surface temperature also drops. This causes the outer surface material to try to contract. The inner surface has its temperature maintained by the heated water, so there is a difference in strain – which causes stress.
The linear elastic modulus is about 0.4 psi/ppm; the coefficient of thermal expansion is about 80 ppm/C, so the combination is 32 psi/C. The fish need about 25C to be happy and the atrium likely reached 5C so the difference is 20C => about 600 psi additional stress at the outer surface of the aquarium.
The linked estimate is that the water pressure was about 2.5 atm -> 37 psi , but that’s in the water. The tank wall will see a pressure as an inverse ratio of the diameter of the tank to the wall thickness – 432 inch diameter/(2*8.7 inch wall) => 920 psi. (Dims from WIkipedia)
Note that bulk tensile strength of acrylic is around 9000 psi at the low end, I don’t know what the bond strength is for glued panels.
The temp difference would be especially bad at the bottom of the tank as colder air would pool at the lower end and warm air would rise. There might have been circulation as air cooled at the upper windows would cascade down the walls, more rapidly removing heat at the lower end.
Add thermal expansion to changes in material properties and to damage from inappropriate cleaners and there is plenty to examine.
As AvE says, put the wrong answer out there and someone will provide the correct answer! Not too deep in the twitter thread someone posted the following link https://www.plasticstoday.com/materials/when-acrylic-aquariums-fail
Creep is the likely culprit. (That or unseen notches in the acrylic, or poorly manufactured panels with too much internal stress). Creep is such an interesting mechanical phenomenon. Material moves at stresses *well below* yield strength. Yup. Well ***below***. Heat makes it worse in most materials. I’ve seen numerous mechanical engineering articles about it…many calling it “creepy” Ha! (And it is, but that’s not what it’s named for). As someone not responsible for the design of things likely to experience creep, it’s a fun read. (If you’re on the other side, it’s a terrifying read. Temperature may have accelerated the failure, but it wouldn’t have caused it.
With the folks involved in building this and its sheer scale its not likely to be creep (on its own anyway), it is a rather slow process generally, but not so slow as to take this long to happen – if creep alone was going to be an issue it should have been noticed years ago that the tank was slowly failing, or it would have failed long ago.
It would be rather visible on such a large structure if you are paying any attention to it at all as well, and you would think they would… Plus its so well known and was long long before this tank was built they would have to massively screw up at the design and/or build phase for it to be an issue – not likely for the sort of building where money is no object…
Its not impossible, but as a cause in its own right it seems rather unlikely…
+1 great linked article
I once made the mistake of cleaning acrylic with acetone. The acrylic very rapidly split along internal stress lines that seemed to attract more acetone into the cracks. The cracks propagated very rapidly and essentially shattered the acrylic without the acrylic moving at all.
A possible explanation in the failure of this Acrylic tank could be that someone had stuck an Xmas decoration onto the Tank wall earlier.
If the adhesive used contained Superglue ie ( Cyanoacrylate) liquid this would have created a thermal stress point reaction raising the acrylic envelope temperature to 275°F at that single point.
If an excess of Superglue had been applied this could also have run down the tank wall thereby inducing a thermal crack in this structure.
A sharp thermal shock of 250+°F even if patially penetrating the tank wall within a small linear area, could cause a polymer chain force stress reaction.
As an air transport engineer I have gained some prior knowledge of modern materials and their ‘reactive’ restrictions.
They should have used AlON (i.e. Aluminum Oxynitride) instead…
As we all know these small backcountry Insurance Company as in Lloyd’s of London Etc never do their homework on mediocre small projects that could alter or destroy lives and property much less cost millions and millions in reparations. But I also have to admit the latter thirty or 40% of the comments on this are pretty close. Seldom do we find under engineered projects of this scale. And I doubt this was under engineered. Good Times Merry Christmas and don’t screw with the aquarium.
Any idea what kind of fish they were that did not make it? Just wondering as I simply LOVE fish and would have taken some home to cook up and eat in celebration of free food!
Just say’in.
One Of the photos appeared to have a ray lying in the debris.
Clearly, an unfortunate coincidence of circumstance coupled with a technical error of judgement to do with ignorance of material property and strength. However, there are at least two indicators that should have been recognised as possible dangers:
1.
The high pressure created within the tank due to the sheer height of the column of water,
2.
The antithetical strength of cylindrical structures experiencing pressure from the interior curvature, where the arch strength principle simply does not work when the direction of force is reversed – in fact, it is considerably weakened so that the curved container was weaker than a square, flat sided container and was vulnerable because of this.
I designed, engineered, fabricated, installed, and inspected this aquarium and can set the record straight by saying it was caused by a bad kid tapping on the glass and causing a shark to smash into it in attempts to eat said bad kid. That kid will be getting coal in his socks this yule time. And it serves him right that the shark bit of half his ear before he suffocated in the cold air. Maybe the kid will use the other half of his ear and listen to adults when they tell him not to tap on tanks any more.
Outside temperature as the cause doesn’t seem likely to me. Shouldn’t the aquarium be designed to not fail, just because the heating in the hotel failed for a day or was shut off for maintenance?
This wasn’t reported on any new station I listen to.