Tourist submersible visiting the Titanic is missing Part 2 68

[drwebb]

OSHA reports they have jurisdiction to investigate complaints of retaliation against employees under 25 different statutes.
To the extent Lochridge could be considered a "seaman", the SPA prohibits retaliation resulting from "the seaman has refused to perform duties ordered by the seaman's employer because the seaman has a reasonable apprehension or expectation that performing such duties would result in serious injury to the seaman, or other seamen, or the public."
Or it might be one of the other 2 dozen statutes.
For all the good it did in this case.

 

[LittleInch]

Definitive investigation under way.

https://www.bbc.co.uk/news/articles/cx2kk1g66n7o

Looks like they had some sort of notice as they said they had dropped two weights, but didn't get very far.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[IRstuff]

There's apparently a flurry of new USCG activity in the investigation

https://www.news.uscg.mil/News-by-Region/Headquarters/Titan-Submersible/

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[drwebb]

Pilot/ex-DoO/whistleblower Lochridge is telling his side publicly to the USCG:

https://www.nytimes.com/2024/09/17/us/titan-submersible-coast-guard-hearing.html

 

[jmec87]

There's a lot of interesting newly released info at the link from IRstuff

https://www.news.uscg.mil/News-by-Region/Headquarters/Titan-Submersible/

1. Preliminary design report for the carbon fibre sub completed by Boeing (2013)
- Suggests that it is feasible but acknowledges some of the risks inherent in CFRP and challenges due to differential thermal expansion, imperfections in the CFRP, and need for NDE. Suggests a need for further analysis.

2. Design report by the president of Spencer Composites (June 2015)
- Does not appear to give any consideration to the effects of imperfections on the design or any mention of a need for NDE.
- Claims that the design meets ASME BPVC Sec VIII Div 3 requirements.
- Suggests that either a titanium or composite dome could be used.

3. Stakeholder update by Stockton Rush (December 2015)
- The composite domes on the 1/3 scale model failed far below the predicted pressure.
- Claims that "Boeing with all their supercomputer capacity admitted they were unable to analyze such a structure" - Boeing's report didn't say that, but it did mention that "Compound curvatures such as spheres or ellipsoids generate additional discontinuities due to converging or diverging discrete material product forms which create gaps and/or overlaps between adjacent tows of slit tape composite material. These discontinuities can create local stress variations [which] reduce the capability of the structure. In general, composites are very efficient for in-plane loads, but complex loading near hardpoints or load transitions require cautious analysis for pristine and flawed structure. For this reason a hybrid concept was adopted..."
- OceanGate had apparently relied on the Spencer analysis for the composite domes. This analysis did not appear to include "cautious analysis for pristine and flawed structure" - it appeared to assume it was pristine with no gaps, delamination, or overlap.
- States "it is believed that the hemisphere buckled which was the reason our strain gauges did not give us warning of the failure. For the next test we will include acoustic monitoring which should give a better warning of this type of failure."
- Does not appear to recognize that they should not depend on analysis that significantly overestimates pressure at buckling failure (possibly because it did not consider imperfections).
- Also does not appear to recognize that buckling happens very quickly, so acoustic monitoring will likely result in an alarm sounding too late.

4. Quality Inspection Report by David Lochridge (2018)
- Very many significant problems, from those that should be easily avoidable (flammable flooring and internal materials that emit highly toxic gases upon ignition) to fundamental problems (visible significant delamination and porosity in the hull with no NDE to quantify the extent of the defects).

5. Risk Assessment for the ill-fated expedition (2023)
- Did not seem to consider any need for verification of the hull pressure integrity or inspection to find defects well before they led to failure. The only relevant item mentioned is "Monitor Acoustic Emissions system for extra ordinary acoustic hull events" as a "Control Measure" for the "Confined space" hazard during the dive.
- Note that there are items related to the "Pressure energy" hazard of the high-pressure air and low-pressure air systems, including inspecting for wear and damage, but they did not give the same consideration to the pressure vessel that would contain human occupants.
- Relies on the oft-criticized "severity x likelihood = risk" formula, with 5 levels with defined criteria for both of them
- Hazards do not seem to be well-defined
- Looks like somebody just assigned the severity and likelihood numbers based on intuition (and possibly avoiding any "High" risk ratings)
- Highest determined risk is 10 [Medium] - maximum possible on the scale is 25.

So far no design analysis performed after June 2015 has been released. I think the June 2023 failure was on a hull built by another manufacturer after there were some significant issues with the Spencer hull, but they reused the end rings, so they might have continued to rely on the same design analysis.

 

[SWComposites]

Thanks jmec87.

I quickly scanned the Boeing and Spencer reports.

As a former Boeing composites engineer, I'm not impressed by the Boeing report, even for a quick feasibility study.

The Spencer report is troubling. This statement "Defining the ultimate allowable for flexure as 85% of the mean as before results in an allowable strain of 12010. It is probably realistic to use this allowable where high strain is isolated and surrounded by lower strain material" is just IMO insane, as designing to that level of strain is just crazy. And it appears that most of the material design properties are based on assumptions and limited data. Such as:

"A compressive strain database is not available and a design allowable was derived based on available data in Mil-HDBK-17 (Ref. 2). This data is for ~34 msi carbon fiber/epoxy, for eleven fiber/resin systems. The compressive stress data for these eleven unidirectional laminate types were combined into a single data base consisting of 323 total test specimens with a mean compressive strength of 214 ksi (normalized to 60% fiber volume). The data in Mil-HDBK-17 is pre-1990. Fiber tensile strength for 37-800 has increased by more than 20% over the fibers used at that time, so it was assumed that the mean unidirectional compressive strength for 37-800 is at least 1.15x214 ksi = 246 ksi (60% fiber volume). The design allowable was defined as 85% of this mean value = 209 ksi. The compressive strain allowable corresponding to this stress in a unidirectional 37-800 laminate is 9400." This is just rubbish. Assuming compressive strength increases when fiber tensile strength increases is just wrong.

I'm not going to comment on the FEM analysis, other than to say the meshing looks terrible in some locations.

 

[SWComposites]

From the Quality Inspection report, showing the hull laminate "quality":

Just horrible.

 

[SnTMan]

LittleInch's post 17 Sep 24 07:37 woulda been a great time to start Part 3

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[Just Some Nerd]

From the Quality Inspection report, showing the hull laminate "quality":

Surely this picture is a joke???

----------------------------------------------------------------------

Why yes, I do in fact have no idea what I'm talking about

 

[rodface]

Surely this picture is a joke???

David Lochridge (OG's former Director of Operations and author of the notorious inspection report of Titan that got him sacked) discussed this photo as it pertains to his report during today's board hearing, it's at the 1:50:37 mark of the video:

https://www.youtube.com/live/SNrs8VA963w?si=PbwNbL1fMNiwh1ZE&t=6637

Yesterday's testimonies from Tony Nissen (former Director of Eng.) and Tym Catterson (contractor, experienced sub pilot, was topside on Polar Prince during the disaster), and today's from Lochridge are well worth listening to in full for anyone interested in this case. The sessions are 7-8 hours but I wait until after the live feed ends, speed up the video to 2x and skip through the breaks which reduces it to 2-3 hours.

 

[Reverse_Bias]

From this article, it appears that the weights could have been dropped to minimize the impact with the ocean floor, not because they were trying to return to the surface.

Tym Catterson, a former contractor for OceanGate, testified during Monday's hearing that there were "no red flags" on the day of the incident. He said he believes the intention of shedding the two 35-pound weights was to slow the vessel down as it approached the ocean floor. He thought the weight was dropped a little early than is typical -- not due to any emergency but to ensure a smooth landing, he said.

"Considering who was in the cab with Stockton at that particular time -- there are two billionaires in there with him -- I would absolutely guarantee that he's trying to make sure that this just goes as absolutely perfect and spot-on as he can," Catterson said.

https://abcnews.go.com/US/titan-sub...uard-hearing-last-messages/story?id=113729878

Also that pic of the hull material looks like the stuff they use to protect loading docks from truck impacts.

 

[TugboatEng]

The only part of the sub that we have seen post failure is one of the titanium domes. There was zero carbon fiber attached to it. I read that as the bond between the carbon tube and the titanium dome failing. Remember, this sub had made 13? dives to full depth. Is carbon fiber laminate so sensitive to fatigue? Meanwhile, metal to epoxy bonds are very sensitive to degradation from water exposure.

 

[drawoh]

TugboatEng,

One of the reasons for using titanium is that it is chemically very stable. How well do adhesives stick to it?

--
JHG

 

[LittleInch]

The issue that most people have noted and also brought up again recently is that the record of the vessel actually works against it.

So saying it's ok, it's been to depth 13 times already, is actually a bad thing. Any failure of a fibre is non repairing and hence every dive probably lobbed a few percent off its capacity to resist external pressure, which wasn't much above 1 anyway. So in this case sooner rather than later they just ran out of life. Not really fatigue, but more operating far too close to the margin and not having any way to NDT the vessel after each dive. Or indeed NDT it to start with.



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[SwinnyGG]

....titanium is that it is chemically very stable. How well do adhesives stick to it?

Bonding titanium components is not ground breaking - it's very common in certain parts of the aerospace world - but getting a high quality, high strength adhesive bond to titanium parts is still a non-trivial process. You can't just buy a product off the shelf and slather it on and get good results. Similar to adhesive bonding of stainless steels and to a lesser degree certain aluminum alloys, material types/surface cleanliness/surface preparation by chemical or mechanical means are very critical. To design a process for a bonded titanium component with a service depth of 4,000m, you'd need to invest serious time and capital into a true bonding engineering design. Which it doesn't seem oceangate ever did. And based on the videos of them setting the ring and hull joints, they don't appear to have been following any sort of meticulous or detailed process controls.

Just thinking out loud here, but I wonder if Rush got the thought in his head that because these adhesive joints were always in compression in service, investing the time to develop and fully understand the bonding process between those components was not necessary. That obviously isn't the case.

 

[JStephen]

"but I wonder if Rush got the thought in his head that because these adhesive joints were always in compression in service, investing the time to develop and fully understand the bonding process between those components was not necessary."
Well, that thought certainly entered my head, but then again, I'm not building my own submarine, either.
I'm not aware that you COULD bond carbon-fiber edgewise to a ring of metal and develop anything like the strength of the carbon fibers. If the glue was that strong, you wouldn't need the carbon fibers.
But one reason for thinking this joint was not so critical- if I recall correctly, on the old bathyspheres, they were actually forged in two or three pieces, then simply slapped together with no weld or glue. Of course, that was a spherical shell, which theoretically had no bending. Attaching metal hemispheres to a plastic cylinder, you'd get shear forces and bending moments at that joint, but it still seems very reasonable that the compressive stress and resulting shear friction could very well be higher than the bending or the developed shear at the joint. At any rate, I don't remember anyone claiming that was what failed, but rather just the cylinder itself.

 

[rodface]

Video of the carbon fiber wreckage and domes on the seafloor has been released:

https://www.youtube.com/watch?v=eEThkm0Dih8

 

[TugboatEng]

Drawoh, titanium, aluminum, and stainless steel all have a similar challenge when the bond is going to be exposed to a wet environment. All of the metals rapidly form a thin oxide layer on their surface in air. Adhering to this oxide layer is problematic because the electrons in the bond have a higher affinity for water than the epoxy. The metal oxides form hydrated and the bone is lost. We've all seen the way paint flakes off aluminum in large chunks.

In the advertising videos they show the joint being bonded. The joint is as machined. I assume there was some time elapsed between machining and assembly which means oxides are present. The joint should have been abrasive blasted immediately before bonding. If that is not possible there are primers available to extend the time between blasting and bonding. Silane based primers are the cutting edge.

 

[TugboatEng]

I've just watched the video and it's notable to see that nearly all of the craft is packed into the forward done and the hinged connection on the forward dome is blown open. This makes me think that the connection at the rear dome failed first which cause everything to get pushed forward into the forward dome.

 

[SwinnyGG]

if I recall correctly, on the old bathyspheres, they were actually forged in two or three pieces, then simply slapped together with no weld or glue

I have my doubts about that.. you would need to hold them together somehow. The first bathysphere ever constructed was a 1" thick steel casting, 5' diameter and cast in one piece.

Even if the 'bathyspheres were multiple pieces with no seals between' thing is true, there's one factor that would aid that approach, which didn't aid Titan; two halves of a sphere made from the same material would have the same stiffness in response to service pressure, and would have the same expansion and contraction rate when undergoing temperature changes.

Titan's hull was made from multiple materials and assembled at or near room temperature. For metal parts that big, 40 degree negative delta at the service depth would cause non-trivial dimensional changes, which would not be exactly matched by the CFRP tube. You're putting a fair amount of direct shear stress on that adhesive joint from the temperature change alone.