3 ms·

All of this work is designed to be highly iterative, (relatively) low cost prototyping. Cutting corners is exactly what they are trying to do. If you hadn't see

All of this work is designed to be highly iterative, (relatively) low cost prototyping. Cutting corners is exactly what they are trying to do. If you hadn't seen before, until SN8 they were flying this thing without a nose cone. Cutting corners here is OK. The point is to find the mistakes. Better to find these issues now than to be something like https://en.wikipedia.org/wiki/NOAA-19#Damage_during_manufacture https://en.wikipedia.org/wiki/NOAA-19#Damage_during_manufact...

HarryHirsch 6y ago
Cutting corners here is OK No, it's not. Incidents like this tell a lot about internal culture. Things in highbays aren't supposed to fall over just like that, that sort of accident could have cost a life and certainly has lost the company money and set it back a few months. Safety is very clearly an afterthought at the Starship site. You can't have that in rocketry. They want to go to Mars and can't even get out of the assembly building in one piece! Starship is basic engineering in metal. The thing sat there and just fell over. There wasn't even any kind of load on it. That takes some effort!
- jcims 6y ago
  I tend to agree with you, and that was definitely my first thought. However we don’t know what was going on when this happened. If it was just sitting on the stand and fell over that’s a problem, but they were clearly getting ready to move it and may have been transitioning to a set up where they could lift it. It’s possible (albeit unlikely) that this was a risk that they had calculated and put some countermeasures into place for. They just dropped one of these out of the sky from 8 miles up, but since the risk was managed in a very visible way it generally wasn’t considered a safety issue.
- bogwog 6y ago
  This is a prototype during R&D. It's like when you press F5 and you get a compiler error because you accidentally used `.` when you should've used `->` to access a member function. Does this tell a lot about your abilities as a programmer? Do you hide this incident from your boss so that he doesn't realize you're a fraud? Should you tell your customers about this? It'll be much worse if they hear about it from someone else. The difference between `.` and `->` is basic engineering! You're trying to ship a product, but you can't even get out of the compiler in one piece!
  - HarryHirsch 6y ago
    Some disciplines just don't lend themselves to the "rapid prototyping" mindset. Imagine a process engineer scaling up a chemical reaction in the pilot plant and neglecting the cooling capacity of the reaction vessel. A thermal runaway is encountered, and the vessel contents, 100s of liters of flammable organic solvent are distributed throughout the building. The fire chief will want to have a word with the rapid prototyper, sizing the thermal budget has been practice for 100 years. Software engineers have an odd mindset towards bugs. It's already an outrage that when you activate a new cellphone the first thing the device does is download gigabytes of updates. It's a disgrace.
    - jlmorton 6y ago
      At the same time, out of the rest of the new super-heavy launchers under development, New Glenn, Vulcan, SLS, Ariane 6, and Long March 9, none have flown. Yenisei is expected in 2028. Meanwhile, SpaceX continues welding sections of their new rocket in tents in the Texas brush land, and days ago performed one of the most impressive tests in recent memory.
    - dgrin91 6y ago
      But thats not prototyping... thats testing in production. Two different things. Prototyping would be an engineer in a lab with a scaled down version of whatever process he is working with. He tries 10 options, 9 of which blow up. But since its scaled down the blow up is just a little fizzle, so who cares. The 10th option that worked can be developed more and eventually brought to the full production plant
- dgrin91 6y ago
  > set it back a few months. I'd bet $1000 that by end of next week SN10 will be on the pad, if not a repaired version of SN9. > Safety is very clearly an afterthought at the Starship site. As far as I know no one has died at a SpaceX site. I think clearly we can say that SpaceX is putting sufficient amount of energy into worker safety. I agree with you this issue posed a risk to safety, and I'm sure SpaceX will learn from it and applies the lessons from it in the future. > You can't have that in rocketry. They want to go to Mars and can't even get out of the assembly building in one piece! Literally one day before they just took almost the same rocket up to 12.5k and almost landed it. I'm trying very hard to assume good intentions on your part, but statements like that make it hard. The alternative you seem to be proposing is the Boeing model where everything is extremely slow, methodical and waterfall style. How did that end up? A disaterous demo flight with 2 near misses: https://en.wikipedia.org/wiki/Boeing_Orbital_Flight_Test#Anomalies https://en.wikipedia.org/wiki/Boeing_Orbital_Flight_Test#Ano.... We all agree rocket science is hard. We all agree that going to mars has lots of critical safety requirements. But SpaceX isn't going to mars yet, they are just doing rapid prototyping. The requirements here are WAY less.
  - NortySpock 6y ago
    Unfortunately the number of deaths at SpaceX sites is non-zero. I do not know details of the accident today, but there has been a death in the past. (2014) https://www.spaceflightinsider.com/organizations/space-exploration-technologies/spacex-worker-killed-companys-mcgregor-facility/ https://www.spaceflightinsider.com/organizations/space-explo...
  - baking 6y ago
    The next road closure is December 28. I don't think it set them back too far, but a week was overly optimistic.
- bigbubba 6y ago
  You've been downvoted but clearly heavy objects falling over in workshops where workers might be squished is not meant to be part of their corner-cutting rapid prototyping approach to rocket design. That's an industrial accident, not prototyping. Rockets blowing up on launch pads is fine, that's to be expected. But that's not what this is.
  - HarryHirsch 6y ago
    I just wish some of the downvoters would give their reasons. To me, the Starship project is unusually prone to safety incidents, which tells a lot about organizational culture.
    - bigbubba 6y ago
      This is the first incident I've heard of that seems concerning, but admittedly I've not followed the matter closely. I've heard they had some incidents on the pad, or the one they crashed earlier this week, but neither of those concern me since failure in those scenarios is obviously something they plan for (nobody would be nearby.)
      - HarryHirsch 6y ago
        I found that tank testing incident where a fuel tank exploded during static pressure testing below rated pressure extremely worrying. Pressure vessels are well understood, they aren't supposed to fail like that. Don't they examine their welds at Space X?
        bigbubba 6y ago
        Such a failure seems reasonable to me, such a pressure vessel obviously needs to be as light as possible to maximize payload. But making it lighter will make it weaker. And even today there are limits to computer analysis, hence why they would do pressure tests at all; if the models were foolproof then why would anybody test pressure vessels? Pressure vessels are tested because their danger is well understood. Testing would be a waste of time if outcomes were certain, so failure during tests is not concerning. (This rocket falling over was presumably not part of a test to see whether it will fall over.)
- virgilp 6y ago
  It set it back 11 days apparently... not a few months. They’re moving it to the launch site now: https://twitter.com/BocaChicaGal/status/1341378205962395649?s=09 https://twitter.com/BocaChicaGal/status/1341378205962395649?...
space_ghost 6y ago
"Repairs to the satellite cost $135 million. Lockheed Martin agreed to forfeit all profit from the project to help pay for repair costs" -- yikes! Imagine being the poor sap responsible for removing those bolts. O_O
nerpderp82 6y ago
This is after Lockheed Martin's error lead to the loss of the Mars Climate Orbiter. They shouldn't be allowed near any spacecraft ever again. https://en.wikipedia.org/wiki/Mars_Climate_Orbiter#Cause_of_failure https://en.wikipedia.org/wiki/Mars_Climate_Orbiter#Cause_of_...

HarryHirsch 6y ago

Cutting corners here is OK No, it's not. Incidents like this tell a lot about internal culture. Things in highbays aren't supposed to fall over just like that, that sort of accident could have cost a life and certainly has lost the company money and set it back a few months. Safety is very clearly an afterthought at the Starship site. You can't have that in rocketry. They want to go to Mars and can't even get out of the assembly building in one piece! Starship is basic engineering in metal. The thing sat there and just fell over. There wasn't even any kind of load on it. That takes some effort!

jcims 6y ago
I tend to agree with you, and that was definitely my first thought. However we don’t know what was going on when this happened. If it was just sitting on the stand and fell over that’s a problem, but they were clearly getting ready to move it and may have been transitioning to a set up where they could lift it. It’s possible (albeit unlikely) that this was a risk that they had calculated and put some countermeasures into place for. They just dropped one of these out of the sky from 8 miles up, but since the risk was managed in a very visible way it generally wasn’t considered a safety issue.
bogwog 6y ago
This is a prototype during R&D. It's like when you press F5 and you get a compiler error because you accidentally used `.` when you should've used `->` to access a member function. Does this tell a lot about your abilities as a programmer? Do you hide this incident from your boss so that he doesn't realize you're a fraud? Should you tell your customers about this? It'll be much worse if they hear about it from someone else. The difference between `.` and `->` is basic engineering! You're trying to ship a product, but you can't even get out of the compiler in one piece!
- HarryHirsch 6y ago
  Some disciplines just don't lend themselves to the "rapid prototyping" mindset. Imagine a process engineer scaling up a chemical reaction in the pilot plant and neglecting the cooling capacity of the reaction vessel. A thermal runaway is encountered, and the vessel contents, 100s of liters of flammable organic solvent are distributed throughout the building. The fire chief will want to have a word with the rapid prototyper, sizing the thermal budget has been practice for 100 years. Software engineers have an odd mindset towards bugs. It's already an outrage that when you activate a new cellphone the first thing the device does is download gigabytes of updates. It's a disgrace.
  - jlmorton 6y ago
    At the same time, out of the rest of the new super-heavy launchers under development, New Glenn, Vulcan, SLS, Ariane 6, and Long March 9, none have flown. Yenisei is expected in 2028. Meanwhile, SpaceX continues welding sections of their new rocket in tents in the Texas brush land, and days ago performed one of the most impressive tests in recent memory.
  - dgrin91 6y ago
    But thats not prototyping... thats testing in production. Two different things. Prototyping would be an engineer in a lab with a scaled down version of whatever process he is working with. He tries 10 options, 9 of which blow up. But since its scaled down the blow up is just a little fizzle, so who cares. The 10th option that worked can be developed more and eventually brought to the full production plant
dgrin91 6y ago
> set it back a few months. I'd bet $1000 that by end of next week SN10 will be on the pad, if not a repaired version of SN9. > Safety is very clearly an afterthought at the Starship site. As far as I know no one has died at a SpaceX site. I think clearly we can say that SpaceX is putting sufficient amount of energy into worker safety. I agree with you this issue posed a risk to safety, and I'm sure SpaceX will learn from it and applies the lessons from it in the future. > You can't have that in rocketry. They want to go to Mars and can't even get out of the assembly building in one piece! Literally one day before they just took almost the same rocket up to 12.5k and almost landed it. I'm trying very hard to assume good intentions on your part, but statements like that make it hard. The alternative you seem to be proposing is the Boeing model where everything is extremely slow, methodical and waterfall style. How did that end up? A disaterous demo flight with 2 near misses: https://en.wikipedia.org/wiki/Boeing_Orbital_Flight_Test#Anomalies https://en.wikipedia.org/wiki/Boeing_Orbital_Flight_Test#Ano.... We all agree rocket science is hard. We all agree that going to mars has lots of critical safety requirements. But SpaceX isn't going to mars yet, they are just doing rapid prototyping. The requirements here are WAY less.
- NortySpock 6y ago
  Unfortunately the number of deaths at SpaceX sites is non-zero. I do not know details of the accident today, but there has been a death in the past. (2014) https://www.spaceflightinsider.com/organizations/space-exploration-technologies/spacex-worker-killed-companys-mcgregor-facility/ https://www.spaceflightinsider.com/organizations/space-explo...
- baking 6y ago
  The next road closure is December 28. I don't think it set them back too far, but a week was overly optimistic.
bigbubba 6y ago
You've been downvoted but clearly heavy objects falling over in workshops where workers might be squished is not meant to be part of their corner-cutting rapid prototyping approach to rocket design. That's an industrial accident, not prototyping. Rockets blowing up on launch pads is fine, that's to be expected. But that's not what this is.
- HarryHirsch 6y ago
  I just wish some of the downvoters would give their reasons. To me, the Starship project is unusually prone to safety incidents, which tells a lot about organizational culture.
  - bigbubba 6y ago
    This is the first incident I've heard of that seems concerning, but admittedly I've not followed the matter closely. I've heard they had some incidents on the pad, or the one they crashed earlier this week, but neither of those concern me since failure in those scenarios is obviously something they plan for (nobody would be nearby.)
    - HarryHirsch 6y ago
      I found that tank testing incident where a fuel tank exploded during static pressure testing below rated pressure extremely worrying. Pressure vessels are well understood, they aren't supposed to fail like that. Don't they examine their welds at Space X?
      - bigbubba 6y ago
        Such a failure seems reasonable to me, such a pressure vessel obviously needs to be as light as possible to maximize payload. But making it lighter will make it weaker. And even today there are limits to computer analysis, hence why they would do pressure tests at all; if the models were foolproof then why would anybody test pressure vessels? Pressure vessels are tested because their danger is well understood. Testing would be a waste of time if outcomes were certain, so failure during tests is not concerning. (This rocket falling over was presumably not part of a test to see whether it will fall over.)
virgilp 6y ago
It set it back 11 days apparently... not a few months. They’re moving it to the launch site now: https://twitter.com/BocaChicaGal/status/1341378205962395649?s=09 https://twitter.com/BocaChicaGal/status/1341378205962395649?...

space_ghost 6y ago

"Repairs to the satellite cost $135 million. Lockheed Martin agreed to forfeit all profit from the project to help pay for repair costs" -- yikes! Imagine being the poor sap responsible for removing those bolts. O_O

nerpderp82 6y ago

This is after Lockheed Martin's error lead to the loss of the Mars Climate Orbiter. They shouldn't be allowed near any spacecraft ever again. https://en.wikipedia.org/wiki/Mars_Climate_Orbiter#Cause_of_failure https://en.wikipedia.org/wiki/Mars_Climate_Orbiter#Cause_of_...