Submersible missing while visiting Titanic wreck

[nctnico]

But my OCD brain says: nope, must have many more test hours on the bench.

Your brain needs some adjustment. You can't test quality into a design...  A design needs to be sound from the start.

[dietert1]

After whistle-blowing happened there were enough people who knew the design was marginal. Somewhere above i already wrote that in mechanical engineering safety margins of 10 or 20% are very unusual, especially if human life is at risk. Safety checks cause delays and cost money. I think that business wasn't viable as it was nonsense from the very start.

Regards, Dieter

[floobydust]

Well, the Coast Guard hearing 1st week revealed many malfunctions and quite the clown car that sub was.
I'd thought in the salvage pics, the sub had been disassembled. But from the deep-sea pic, I'm still shocked the titanium end bells blasted off, 17 bolts snapped. Wow. The viewport seems to have popped out too.

T-shirts and patches on Etsy have been around for a while, since last Christmas.
At least nobody made a bong out of it.

[johansen]

After whistle-blowing happened there were enough people who knew the design was marginal. Somewhere above i already wrote that in mechanical engineering safety margins of 10 or 20% are very unusual, especially if human life is at risk. Safety checks cause delays and cost money. I think that business wasn't viable as it was nonsense from the very start.

Regards, Dieter

They thought they had a safety factor of 2.25 on the hull, which is something they should have been able to achieve given that the demands on a 5" thick hull 66" outside diameter is only 35,600 psi in hoop stress compression. the us navy did a lot of work on this in the 80's and some of their E2 fiberglass hulls had 110K psi compression strength, even though the samples tested were curved sections cut from the hull and compacted in a press.

I can't seem to find reliable data on the viewport. Stockton actually claims he was 8% off the charts and had a safety factor of 4 on the viewport, in the released transcript of Lockridge's interview. But the manufacturer said we'd only certify it for 650 meters. if you take the usual factor of safety of 6 or 7 for viewports.. this is consistent with one expert's testimony that the veiwport would last "a few dives to 4000 meters" (650*7 is 4550)

a factor of safety 4 viewport would have been more like what you see on the subs that have gone down to 36,000 feet.

[tom66]

Stockton bragged about the viewport material (acrylic) being brilliant for this purpose because it tends to 'fog up' before it fails:

RUSH: So the viewport is seven-inch-thick plexiglass, acrylic, and that's another thing where I broke the rules. A lot of the submersible industry is run by Pressure Vessels for Human [Occupancy] standards, which acts like a standards body, but it's not a standards body; it's a volunteer group that has come up with some rules.

And there was a very well-known person, Clarence Statue (PH), was, like, the king of acrylic. And so he wrote a book—well, he had several books, but one of these is sort of the Bible. And even he admitted that he was super conservative. It has safety factors that—they were so high, he didn't call 'em safety factors, he called 'em "conversion factors."

You know, for most things, safety factors are one-and-a-half, two-and-a-half…and it's four to ten. And most of what he was looking at was lower-pressure applications. And so when you look at the charts, we're off the charts.

One of the things about acrylic that's really great is, before it fails, at one-third its failure pressure, it will start to "craze." So it'll often be distorted. So you know when that thing's gonna fail. And so when I was looking at this, that view port is—according to the rules, it is not allowed.

So there's these weird rules that are out there.

It will shrink. It's a semi-solid, the plexiglass; it'll come into the cabin by about three-quarters of an inch—all of the pressure that's there.

POGUE: Oh, man. And that's a good thing?

RUSH: Well, that's what it is. But the great thing about plexiglass that I love is, you can see every surface. And if you've overstressed it, or you've even come close, it starts to get this crazing effect.

POGUE: Okay. And if that happened underwater—

RUSH: You just stop and go to the surface.

POGUE: You would have time to get back up?

RUSH: Oh, yeah, yeah, yeah. It's way more warning than you need.

As we've seen from the Berlin Aquadrom failure, and other acrylic aquaria, it's certainly possible for acrylic under stress to catastrophically fail with little to no prior warning. But in those cases, human fatalities are rarely involved, so it might be acceptable to have a smaller safety factor.  Not so sure about this for a submersible viewport.

[floobydust]

I thought the viewport did not fail, it was not there mounted to the titanium end bell during the salvage lift.
I guess the explosive forces blew it out of the mounting ring. I'd guess those bolts snapped as well, just like the end-bell popping out.

Mission when the Titan tilted sideways, people fell down inside, it hit the landing platform and broke off the titanium dome (only 4) bolts is just outrageous and who knows could have even damaged the structure. Then there is the time it had one thruster working and could only go in circles.

[dietert1]

After whistle-blowing happened there were enough people who knew the design was marginal. Somewhere above i already wrote that in mechanical engineering safety margins of 10 or 20% are very unusual, especially if human life is at risk. Safety checks cause delays and cost money. I think that business wasn't viable as it was nonsense from the very start.

Regards, Dieter

They thought they had a safety factor of 2.25 on the hull, which is something they should have been able to achieve given that the demands on a 5" thick hull 66" outside diameter is only 35,600 psi in hoop stress compression. the us navy did a lot of work on this in the 80's and some of their E2 fiberglass hulls had 110K psi compression strength, even though the samples tested were curved sections cut from the hull and compacted in a press.

I can't seem to find reliable data on the viewport. Stockton actually claims he was 8% off the charts and had a safety factor of 4 on the viewport, in the released transcript of Lockridge's interview. But the manufacturer said we'd only certify it for 650 meters. if you take the usual factor of safety of 6 or 7 for viewports.. this is consistent with one expert's testimony that the veiwport would last "a few dives to 4000 meters" (650*7 is 4550)

a factor of safety 4 viewport would have been more like what you see on the subs that have gone down to 36,000 feet.

Verifying a safety factor of 2.25 or 4 would have meant pressure tests beyond 8 000 m. It never happened.

Regards, Dieter

[coppercone2]

hahaha
https://news.yahoo.com/news/hear-doomed-titanic-subs-sonar-170355060.html


I know that won't work too well I have seen spreadsheet blobs that could hang a super computer

it does not like going into the higher audio band, I imagine with ultrasonic there basically way no navigation system in excel

[floobydust]

Think of all the money saved! Just have to type fast. "We tried to do that every five minutes" lol.

[tom66]

Has it been confirmed or not if it was the hull or dome that failed? When the submersible was recovered the dome was missing, but this may have been because it was removed during recovery. 

[johansen]

Has it been confirmed or not if it was the hull or dome that failed? When the submersible was recovered the dome was missing, but this may have been because it was removed during recovery.

The released video of most of the hull neatly squished in the rear hemisphere as well as tym cattersons testimony of the inner lip of the front interface ring being sheared off all the way around show that the failure was a near uniform buckling failure of the front of the hull cascading to the rear.

Basically once water breaks through the glue joint and gets between the layers of the cf, it unloads the outer fibers.

Since the glue joint and stresses on the glue joint were not proved to be uniform nor ever tested for debonding, this was just a matter of time.

Had they checked for cracks at the glue joint (under the rhino liner truck bed liner)

Pulled a vacuum on the sub and bake it to get the water out of the crack, applied epoxy to the crack and then pressurized the chamber to drive the epoxy into the voids, they probably could have gotten hundreds of more dives out of it

[coppercone2]

Think of all the money saved! Just have to type fast. "We tried to do that every five minutes" lol.

it might be acceptable in a episode of stargate if they are stranded off world

[dietert1]

Yesterday i saw a US movie "Underwater" that was about a fictional accident in a fictional mining installation 7000 m below see level. One could recognize lots of unrealistic concepts about deep diving. Like they were showing damaged concrete structures with air for breathing inside. They were showing diving humans inside protective clothing who looked somewhat like astronauts. Those humans entered some installation from below a water surface into the air above it. Completely ridiculous.
I'm pretty sure Mr. Rush didn't understand what 400 bar means. It's 4x the typical pressure inside a gas cylinder as used by industry. The only structure proven to sustain 800 bar is a massive metal sphere. Mr. Rush was living and dying in a world of carbon fiber innovation.

Regards, Dieter

[tom66]

Yesterday i saw a US movie "Underwater" that was about a fictional accident in a fictional mining installation 7000 m below see level. One could recognize lots of unrealistic concepts about deep diving. Like they were showing damaged concrete structures with air for breathing inside. They were showing diving humans inside protective clothing who looked somewhat like astronauts. Those humans entered some installation from below a water surface into the air above it. Completely ridiculous.
I'm pretty sure Mr. Rush didn't understand what 400 bar means. It's 4x the typical pressure inside a gas cylinder as used by industry. The only structure proven to sustain 800 bar is a massive metal sphere. Mr. Rush was living and dying in a world of carbon fiber innovation.

Regards, Dieter

Whilst they aren't exactly mass production, hydrogen fuel cell cars store their hydrogen in carbon-fibre tanks rated to around 700 bar.  These tanks generally have walls around 3-4cm thick - surprisingly small for the pressure level required.

https://en.m.wikipedia.org/wiki/File:Hydrogen_tanks_for_Toyota_Mirai.png

Perhaps Rush was inspired by these.  Thing is, if a hydrogen fuel cell tank experiences a depressurisation, it's unlikely to lead to immediate fatalities.   Hydrogen cars contain leak sensors which can alert the occupants to the hazard and encourage prompt evacuation.  So the safety factors are likely far lower.

[Phil1977]

And please don't forget it´s a huge difference if you build a compression or vacuum tank. Compression can be handled by pure tensile forces, while the negative pressure as in a submarine is much more difficult to handle.

Just compare a water hose that is intended for high pressure to one that is meant for suction.

[iMo]

BTW. - if you look at the video carefully you may see the reminders of the bodies of the passengers still there in the hull.. Not easy to recognize, however..

[floobydust]

NTSB National Transportation Safety Board - Materials Investigation of the Experimental OceanGate Submersible Titan.pdf Sept. 25, 2024 CG-107

Pics of Ver.1 and Ver. 2 of the CF hull fab, cross sections etc. "waviness and wrinkles", "porosity", "voids in the epoxy" as well as the recovered broken pieces wreckage review.
Evidence of the glue having rubbing damage, so the joints were rubbing.
The sounds of the hull delaminating, the loud CRACK when surfacing in 2022... the change in the RTM strain monitoring after that...  all pretty damning. It was falling apart.
RTM logs of the fatal dive seem to be there as well, something bad at the one spot starting pg. 47 of 67.

Links for livestream, latest hearing-released documents and ROV footage etc. at Coast Guard: https://www.news.uscg.mil/News-by-Region/Headquarters/Titan-Submersible

[nctnico]

I'm pretty sure Mr. Rush didn't understand what 400 bar means. It's 4x the typical pressure inside a gas cylinder as used by industry.

100 bar is at the low side. Recreational divers use compressed air at 250 to 300 bar. Commonly used industrial gasses like oxygen, nitrogen, argon etc come in cylinders with pressures ranging between 200 to 300 bar.

[hans]

So from the reports published today, it looks like the large acoustic event heard onboard happened on dive 80 (after resurfacing). It was not the last completed dive at depth. Immediately following this acoustic event, some strain gauges showed a step response in their measurements. Subsequent dives also showed a non-linear response when the sub was strained initially at each dive. I guess this indicates a significant part had de-laminated, and perhaps had some flex to it before the CF material was fully compressed again resulting in a regular linear strain response onwards.

Everything that happened between dive 80 and the last fatal one was done on borrowed time. Probably the trashing around of the sub in storage and on seas (iirc the dive before it all went wrong?) didn't help. I didn't really look deep enough into those reports to see what root cause they actually identify.

What I find more remarkable is that OceanGate built a questionable design CF hull, installed a bunch of acoustic sensors (which they bragged about with patents), but didn't seem to act on half of those sensors never registering any acoustic event (it's a solid hull, so you'd think if you bang on 1 wall all sensors would pick it up something), and finally they had a full winter stop to look intensely in the data of dives 80 and beyond, but also didn't act on that.
They also only completed 14 or so dives at significant depth with that hull, so maybe they were hyper-fixated on some cycle number they needed to hit before replacing it?

It all suggests gross negligence to me.

[dietert1]

I'm pretty sure Mr. Rush didn't understand what 400 bar means. It's 4x the typical pressure inside a gas cylinder as used by industry.

100 bar is at the low side. Recreational divers use compressed air at 250 to 300 bar. Commonly used industrial gasses like oxygen, nitrogen, argon etc come in cylinders with pressures ranging between 200 to 300 bar.

So 100 bar would be an average pressure, i mean the average cylinder is half empty. Still if you would open a 300 bar gas cylinder 4000 m under the sea, it would get partially filled by water streaming inside..
As far as i remember a typical design pressure in hydraulic systems is 200 bar and now imagine a large industrial press with the dimensions of that sub. Many people in this world don't understand numbers without images.

Regards, Dieter

[johansen]

I'm pretty sure Mr. Rush didn't understand what 400 bar means. It's 4x the typical pressure inside a gas cylinder as used by industry.

100 bar is at the low side. Recreational divers use compressed air at 250 to 300 bar. Commonly used industrial gasses like oxygen, nitrogen, argon etc come in cylinders with pressures ranging between 200 to 300 bar.

So 100 bar would be an average pressure, i mean the average cylinder is half empty. Still if you would open a 300 bar gas cylinder 4000 m under the sea, it would get partially filled by water streaming inside..
As far as i remember a typical design pressure in hydraulic systems is 200 bar and now imagine a large industrial press with the dimensions of that sub. Many people in this world don't understand numbers without images.

Regards, Dieter

this forum is one of the worst, for people nit picking every perceived falsehood in someone's comments.

Yeah 3000 psi is pretty typical for hydraulics. they start getting expensive when you get beyond that.
When I get around to it, I have a 6" inside diameter steel pipe 5 feet long which should be good for 6000 psi tests if fully filled with water, its got .45" thick side wall.

A friend of mine just gave me a 3000 psi pressure washer so when I get around to it, I'm going to weld an end cap and threaded plug on the other end.

what a lot of the youtube videos don't capture properly is the intensity of the implosion, because they are using totally filled chambers. The energy displaced is limited to the energy stored in the pressure chamber's expansion under stress.. which ideally is zero.

if instead you fill the chamber with air to 1000 psi, then fill it with water to 3000 psi, you'll get a much more intense collapse.
i bought a block of acrylic i should be able to get a 1" diameter viewport out of.

[EEVblog]

this forum is one of the worst, for people nit picking every perceived falsehood in someone's comments.

It's an engineering forum full of engineers.

[EEVblog]

https://x.com/DJSnM/status/1838990454500839548
More photos in the thread that those here:

[PlainName]

this forum is one of the worst, for people nit picking every perceived falsehood in someone's comments.

It's an engineering forum full of engineers.

It's full of nit-picking pedants who like to fell trees instead of seeing the woods, point to the scratch on the display while missing the info it's displaying, argue the pronunciation of an ingredient instead of savouring the taste, etc.

(Some exceptions, though)

[Andy Chee]

this forum is one of the worst, for people nit picking every perceived falsehood in someone's comments.

It's an engineering forum full of engineers.

It's full of nit-picking pedants who like to fell trees instead of seeing the woods, point to the scratch on the display while missing the info it's displaying, argue the pronunciation of an ingredient instead of savouring the taste, etc.

(Some exceptions, though)

It's also full of people who are incapable of interpreting metaphors, and wondering why one would talk about chopping down forest trees in an electronics forum.