SpaceX Starship Mk1 pressurisation test failure

[trurle]

The SpaceX Starship Mk1 prototype has been damaged beyond repair 20 November 2019 during a cryogenic pressure test. Entire top been torn off.
https://www.space.com/spacex-starship-prototype-anomaly-pressure-test.html

 

[jackwhirl]

I'm excited to see Mk3. I've been trying to picture the plan for the new construction method and I'm not sure if they plan to continue stacking rings (each ring with a single weld, as it comes off the roll) or if they're going paper towel tube style with one long helix weld from top to bottom. Did we get any clarification on that?

 

[mfb]

They will keep stacking rings, but the rings will be single steel sheets.

Looks like Mk 2 in Florida won't fly either, so we have to wait for Mk 3 to be built before there will be test flights. That one could reach orbit later, or maybe Mk 4 will do so from Florida. Super Heavy construction should start somewhat soon after Mk 3 or 4.

 

[trurle]

They will keep stacking rings, but the rings will be single steel sheets.

Strange decision. Driven by desire to reduce assembly and test costs may be. From strength perspective, the failure looked like methane tank broke along weld line. The staggered pattern of sheets would at least make failure slower...

 

[mfb]

The steel mills don't produce endless sheets and you can't transport sections well if they become too long.
What broke was the connection of the bulkhead to the outer structure, that is not the welding of the outer structure.

I split this into a separate thread.

 

[trurle]

The steel mills don't produce endless sheets and you can't transport sections well if they become too long.
What broke was the connection of the bulkhead to the outer structure, that is not the welding of the outer structure.

I split this into a separate thread.

Seems the Starship is using stressed outer skin architecture if not integral bulkhead.

The outcome of incident (cited from https://www.teslarati.com/spacex-starship-hardware-mystery-solved/ )

Starship Mk2 and Mk4 are now effectively dead before arrival as a result of several challenging and reoccurring technical issues. Starship Mk2 likely shares some significant heritage with Starship Mk1, which lost its top during a pressure test. Roughly two-dozen steel Starship Mk4 rings may also be scrapped after SpaceX’s Florida team could not overcome a technical hurdle. Per the source, many of those single-weld steel rings were slightly different diameters, making it next to impossible to build a sound pressure vessel (i.e. Starship Mk4) with them.

 

[mfb]

SN1 (old name Mk3) faced the same fate a few hours ago. It did "launch" in the process!
SpaceX planned to install a Raptor engine later, so this was certainly not planned (but also not completely unexpected - they are still working on the welding process to make the tanks tolerate higher pressure).

 

[jackwhirl]

Article about the same.

 

[mfb]

Non-AMP link
Why it matters

 

[etudiant]

No expert, but it seemed to be a failure at the bottom of the tank, presumably the place where the valve fittings are mounted. The actual tank structure appeared then to implode, presumably because the pressurized LN2 burst out of the breach leaving a near vacuum behind.
If correct, this should be fixable fairly quickly.

 

[mfb]

So far the connection between bulkhead (upper/lower end of a tank) and outer walls seemed to be the weak spot, at least that's what people who know more about rockets said.

 

[trurle]

No expert, but it seemed to be a failure at the bottom of the tank, presumably the place where the valve fittings are mounted. The actual tank structure appeared then to implode, presumably because the pressurized LN2 burst out of the breach leaving a near vacuum behind.
If correct, this should be fixable fairly quickly.

I agree with sequence of events, but not with conclusion. The tank have imploded under differential pressure of no more than 70 kPa (likely about 20 kPa). This is not the way a well designed tank intended to withstand the stress of hypersonic ascent should behave. For example, Space Shuttle Max-Q dynamic pressure during ascent was 33 kPa. Therefore, i.m.h.o., the implosion failure mode do indicate a too flimsy construction of Starship fuel tank.

 

[mfb]

The dynamic pressure on ascent is along the direction of the rocket, not orthogonal to it. It's also nowhere near the tank section. The pressure on descent is orthogonal to the rocket axis but the tank will be pressurized to keep the interior pressure higher.

 

[trurle]

The dynamic pressure on ascent is along the direction of the rocket, not orthogonal to it. It's also nowhere near the tank section. The pressure on descent is orthogonal to the rocket axis but the tank will be pressurized to keep the interior pressure higher.

That is what i call "flimsy construction". Baloon tank which do not tolerate even small damage. Initially SpaceX boasted construction to be much more stable and rigid than Atlas III at same fuel fraction, but after all Starship fuel tank seems to behave similar to Atlas III.

 

[etudiant]

SpaceX has not made any statements afaik about the Starship structure.
It may well be a balloon tank design, although that makes the reentry pretty interesting.
The Atlas needed to be pressurized at all times, else the tank collapsed. I remember seeing pictures of one unfortunate vehicle, it truly looked deflated.

 

[mfb]

Starship doesn't need to be pressurized at all times, so it's more robust than Atlas III. It just needs to avoid pressure losses so rapid that the whole structure is shot into the air.

 

[mfb]

A comparison how quality improved over time:

https://www.usnews.com/news/national-news/articles/2019-08-26/big-test-for-elon-musks-spacex-mars-colonizing-rocket-prototype-starhopper - thick metal plates, visibly deformed and with many, many welds to connect all the plates. Good enough to make a 150 meter hop.
Mk1 - rings segments instead of plates, less steel overall because the material is thinner, fewer welds, still many visible dents.
SN1 - full rings, they look more round as well (but it's not shiny, this hides dents). Welding connections still very visible.
SN2 - it's reflective enough to spot dents, if it would have them. The welding lines look smoother, too. Apparently they improved settings for the welding.

SpaceX already has the key segments of the tank section (the parts where bulkheads need to be attached), so we can expect more tank pressurization tests soon.

 

[etudiant]

A comparison how quality improved over time:

SpaceX already has the key segments of the tank section (the parts where bulkheads need to be attached), so we can expect more tank pressurization tests soon.

The failure was reportedly at the engine mounting, which was not adequately designed. The tank itself seemed just fine in both cases, just imploded after the failure at the fittings.
Apparently SpaceX has designed a novel welding assembly for the rings that will automate that aspect.

 

[gleem]

SN7 was pressure tested yesterday. Began leaking at 7.6 Bar but did not burst. They used 304 SS but Musks wants better and plans to develop their own alloy. Follow the testing at https://twitter.com/spacepadreisle?lang=en

 

[mfb]

The nominal pressure is 6 bar, for uncrewed flights 25% safety margin is typical, so it needs to hold 7.5 bar. It did - barely, but the safety factor is already covered by requiring 7.5.