Starship tests: SN4 passed cryogenic test + static fires

[mfb]

TL;DR Summary

The most recent prototype of Starship is expected to get tested in early April

SpaceX loves confusing numbering schemes, but so far we have seen five relevant test articles: Starhopper ("water tank") made a 150 m flight, Mk1 was the first full-scale prototype but got destroyed in a pressure test, Mk3=SN1 got destroyed in a pressure test, SN2 was a tank test only (successful), now SN3 is being prepared as full-scale prototype and it is expected to make shorter flights later.

Yesterday the tank section was finished (image). They got a one-hour road closure on Friday to transport SN3 to the test/launch site, and a full-day road closure Wednesday April 1 (with backup dates the following days). It is expected that they will do pressure tests with the tanks, with a static fire planned for later. The payload section and nose cone are expected to join the tank section then, followed by a flight test.
In parallel to these activities SpaceX is preparing parts for the successor, SN4. It will be built for longer, high-altitude flights. It will also need the external fins for that.

Edit: They are faster than I expected. The road closures are for a static fire on April 1 and a 150 meter hop on April 6 (source).

 

[mfb]

Failed during cryogenic testing

They keep building new ones, we'll see how far SN4 gets (already under construction).

 

[berkeman]

Bad weld?

 

[etudiant]

Peculiar, again it seemed that the shell imploded.
The upper part of the tank looked to be filled with a cryogenic, but the lower part appeared to be empty.
Possibly chilled the lower part and dropped the internal pressure because of a valve failure?

 

[mfb]

An earlier test was stopped because of valve issues, but that was not the one where it collapsed. Doesn't look like a weld failure to me, more like something that created underpressure inside.

"We will see what data review says in the morning, but this may have been a test configuration mistake"

Better quality video

 

[mfb]

"Pretty much. Good news is that this was a test configuration error, rather than a design or build mistake. Not enough pressure in the LOX tank ullage to maintain stability with a heavy load in the CH4 tank. This was done with N2."

On the positive side: Not a design flaw. Build the next one and proceed. On the negative side: How could such a simple mistake happen?

 

[Tom.G]

How could such a simple mistake happen?

How many people do you know (other than politicians) that have never made a mistake?

 

[trurle]

"Pretty much. Good news is that this was a test configuration error, rather than a design or build mistake. Not enough pressure in the LOX tank ullage to maintain stability with a heavy load in the CH4 tank. This was done with N2."

On the positive side: Not a design flaw. Build the next one and proceed. On the negative side: How could such a simple mistake happen?

ITS/BFR/Starship was initially slated for self-pressurization (VAPAK) configuration. And now SpaceX is using gaseous N2 pressurant - a more conservative approach. I suspect this lack of architecture persistence is behind the repeated implosion/buckling failures. Pressurization system is simply not mature yet.

 

[mfb]

Starship is still using autogenous pressurization. Nitrogen is only used for tests. Less dangerous if things go wrong but very similar to the real conditions if there is no accident.

 

[trurle]

Starship is still using autogenous pressurization. Nitrogen is only used for tests. Less dangerous if things go wrong but very similar to the real conditions if there is no accident.

Then i am even less surprised the accidents happened. Adding "temporary" pipes and valves in addition to "permanent" ones is a reliability problem.

 

[mfb]

I don't think they modify anything about Starship for the tests. They just connect a different tank outside. Or just open a different valve.

 

[trurle]

I don't think they modify anything about Starship for the tests. They just connect a different tank outside. Or just open a different valve.

90% of R&D failures are starting from word "just"...

 

[mfb]

So do most successes.

Luckily SpaceX can iterate fast. About half of the next prototype is assembled already: Overview (each color is one piece).
SN3 didn't follow the same order but it had a similar status March 10. From there to the cryogenic test it was three weeks. We might get cryogenic tests of SN4 by the end of this month.

 

[mfb]

Updated overview
One stacking event left for the tank structure. It has been speculated that they might start to fly with only the tank structure. SN4 won't get the flaps so it can only make shorter hops (similar to Starhopper), that doesn't need a finished nose cone. Flaps should come back with SN5 or SN6.
Cryogenic tests later this months are certainly realistic.

Edit: The two parts of the tank structure have been combined. Beach closures suggest a cryogenic test April 26.

 

[mfb]

Edit: The two parts of the tank structure have been combined. Beach closures suggest a cryogenic test April 26.

Spot on!
SN4 passed cryogenic test. It also passed the test at room temperature before.
Static fire hopefully later this week and a short flight (150 m) in a few weeks.
After months of tank and welding development it looks like SpaceX is on the way to a flight again.

(months is still an extremely short time in the world of rocket development)

 

[mfb]

Static fires complete - one from the main tank, one from the smaller tank for landing fuel.
Next: A short flight!

SpaceX is serious about ramping up the production rate. While SN4 is on the launch pad for tests, all the segments for the next one have been completed, and they started stacking the different segments. They even have some segments of SN6 already.

A single reusable vehicle is great, but imagine what you can do with one additional vehicle every month...

 

[gleem]

It seems that SpaceX has been given permission to carry out suborbital testing of the Starship prototypes for the next two years.

https://www.faa.gov/about/office_or...aceX Starship Prototype LRLO 20-119)lliu1.pdf

 

[mfb]

They attached a large dummy mass to the top of SN4, clearly in preparation of a hop/flight.

The core of SN5 is essentially complete (status a week ago), all elements for the tank section of SN6 are around (status a week ago), and a ring for SN7 has been spotted.
They produce them faster than they can destructively test them.

 

[mfb]

No SN4 hop. Unless you could that as a hop. They tested static fires when this happened. Good that SN5 will be ready soon!

 

[trurle]

No SN4 hop. Unless you could that as a hop. They tested static fires when this happened. Good that SN5 will be ready soon!

According to video, service tower at launchpad has collapsed too. This is definitely a major setback for SpaceX.
The string of failures with Starship tends to happen at earlier testing stages and are more numerous than failures of Falcon 1. For Falcon 1, it took two and half years for SpaceX to make things right. For Starship, my estimate is at least five years. Of course, if the project is not cancelled.

 

[mfb]

Falcon 1 needed to go to orbit within the first 4 flights or the company would have run out of money. Starship doesn't have that constraint and it is much more ambitious. If it doesn't work as planned they'll make the walls a millimeter thicker and fly with that. Plenty of safety margin, but if you don't push the limits you end up with a worse product than you can. With one prototype per month or so they can move much faster, too. SpaceX is not just developing the spacecraft . They are also developing a factory to build 2 per month or so.

 

[jackwhirl]

Rapid unscheduled sadness. D:

 

[sandy stone]

That was one violent explosion! Looked sort of like the ignition of a large vapor cloud. I'll be interested when they reveal the results of their investigation of the failure.

 

[gleem]

Here is the explanation of the explosion:

https://www.teslarati.com/spacex-starship-explosion-explained-elon-musk/