It’s pretty bad, but when the door accessible from the inside implodes from the insane pressure of the water around it, that’s even worse.

The door can still be designed to open outward and also able to be unlatched from the inside: those aren’t mutually exclusive properties.

It needs to remain perfectly watertight at over 5000psi or everyone dies. That design constraint probably eliminates most clever latch designs.

I was imagining a round door in the middle of the hull, like a submarine hatch, but it looks like the entire front dome comes off and that’s the “hatch”. If so, can’t they just make the front a giant screw and add handles to both sides?

The dome shape is a structural necessity to withstand the pressure, like in a soda can. Else you would need more material, more material means more weight, more weight means more bouyancy required which again means more weight etc.

It's like designing a spacecraft but with different constraints.

It's pretty much the exact opposite of designing a spacecraft.

Then let me amend what I wrote: It's as difficult as designing a spacecraft, because you reach the limits of what materials that can be used can offer.

Well, once you are in space the forces are pretty tame until you get hit by something. But the ocean never rests from trying to kill you, even the smallest material defect can cause an implosion. Personally I think designing for a delta of 400 atmospheres in compression is much, much harder than for 1 atmosphere in tension.

But it's withstanding pressure(called external pressure, yesterday I learned), so can't they just have a clean, stepped flange to mate and to prevent slipping off, plus a simple lock to prevent unintended disengagement?

This is usually done with a plug hatch and it's not that complex. The Alvin submersible has one and it's from 1964.

Exactly. There are so many lessons learned already when designing a device like this there is no point in re-learning them the hard way, just adopt a tried and true design.

That's what they did? Alvin is an outlier; other bathyscaphes used the (tried-and-true) bolt-in method.

Assuming they're out there near the surface bouncing around in 6' waves and whitecaps, what they need isn't a door they can open and drown themselves with. Seems like what they need is a more effective location broadcast.

We don't actually know that they don't have good ways of being found. If they do, it makes it all the more likely that the sub is on the bottom :-(

I think in that situation there is absolutely no good outcome and you’ll take whatever ends it as quickly and as painlessly as possible.

Doesn’t seem like a good trade off against bobbing around the surface and slowly dying sealed inside a capsule to me.

It's traditional in deep submersibles design, dating back to bathysphere. You probably weren't hanging around when they were designing them

Tradition itself is probably the worst rationale for an engineering solution. Maybe it's the right way for other reasons?

I didn't interpret "traditional" as the rationale, but rather just pointing out that they're always done that way — I assume/hope for good reasons other than tradition.

Traditions emerge from "other reasons". There is probably not enough excess buoyancy to open the hatch without flooding, so without external support you're dead anyway.

I could see that as being the simplest/least failure prone way to design a closing mechanism that doesn’t need to resist the immense pressure to maintain integrity, but rather is held more forcefully to the hull as the pressure increases. Certainly you could design something like that that bolts from the inside instead, but perhaps the problem with that is that they cant exactly carry along with them various tools and equipment they might need in the event that there is difficulty reopening it.

If you pass out, how do they get you out?

Two hatches? One opens from the inside one opens from the outside? I'm sure there are other solutions.

But as somebody else already pointed out: even if you get out in the middle of the ocean you're in real trouble, which makes the hatch issue kind of moot.

And that's twice the sealing surface to potentially fail.

It could be the same hatch. Very common on submarines.

You know what's the worst? Someone barging in into an established engineering domain and making spot judgements. This is simply incredible hubris.

If some brilliant improvement occurred to you in the first 5 seconds of hearing about some technical solution chances are this thought visited others in the past 100 years working on the problem for a living.

It sounds like they could use some improvements, 5 people are probably dead now. Out of how many trips in that submersible? I don’t know, just wondering, but it can’t be that many. Certainly not comparable to commercial airlines. So that’s a pretty high failure rate. Being able to open the vessel from the inside in an emergency probably wouldn’t have saved them, it’s just the first thing that is too much of a red flag for me to personally ever step foot in that thing.

I also don’t even understand why that vessel needs to be manned at all. With today’s high def cameras and displays, including VR, they could outfit an unmanned vessel with high def cameras everywhere, a fiber optic tether to the surface to carry the high def signal (or even wireless to some degree if they could make that work), and the operator could sit in a control station either on the boat at the surface or streamed over the internet from home, controlling the vessel however he wants, with high def screens to simulate the viewing experience of being in the vessel. I’ll bet it would get you 99% of the way there without having to risk your life. It’s not like anyone is getting out of the thing to swim around the titanic in scuba gear anyway.

There is no hint that its probable demise had anything to do with hatch design so far.

This is a recreational dive, there's absolutely no need for it as for most recreational activities.

There are quite a few other manned submersibles designed for extreme depth that do allow for the hatch to be opened from the inside, in fact that's most of them. Alvin's design would be a nice starting point. It can be opened both from the outside and from the inside.

As for what the probable cause is: we don't know, we may never know. But with a device that was apparently designed by another party who declared it unfit for purpose anything goes.

Do passengers also need to undergo decompression before exiting the craft?

No, the crew remains at 1 atm and the hull bears the pressure difference. Ocean pressure at the Titanic’s depth is over 400 atm (6000 psi!) but the oxygen level in normal air becomes toxic under higher pressures. The record is 70 atm for a couple hours by breathing 0.5% oxygen and 99.5% hydrogen (this is so extreme that it can’t even explode).

That's what Gus Grissom said.

Lot of people probably won't get the subtext on this one.

(although he was recently-probably-exonerated in that case).

I think it’s more the Apollo 1 door opening to the inside and secured by nonexplosive bolts, rather than the spuriously blown door on Liberty Bell 7.

FWIW, while NASA may had thrown him under the bus, his comrades in the astronaut corps believed him. Wally Schirra on Sigma 7, refused to blow the door until the capsule was on the Kearsarge. When he climbed out, he showed the bruise on his hand — the same bruise every one got when they pushed the escape button. The same bruise Grissom did not have.

https://www.npr.org/2021/07/22/1019254674/gus-grissom-libert...