Author Topic: Starship!  (Read 57005 times)

Offline smartcooky

  • Uranus
  • ****
  • Posts: 1959
Re: Starship!
« Reply #105 on: December 25, 2020, 02:57:18 PM »
I just skimmed the video of Starship assembly, so everyone please forgive me if I'm misrepresenting it.  But it looked like some parts of assembly and stacking were being done using general-purpose construction equipment.  Naturally almost none of that would be acceptable in the development of any other crewed space vehicle, if done under the auspices of a NASA contract to do so.

Genuine question... why?

If a $5 million "general purpose" crane meets and exceeds all the specifications required to do the job, why is it necessary to design, build or buy a $50 million "special purpose" crane to do the exact same job? 
« Last Edit: December 25, 2020, 02:59:51 PM by smartcooky »
If you're not a scientist but you think you've destroyed the foundation of a vast scientific edifice with 10 minutes of Googling, you might want to consider the possibility that you're wrong.

Offline Zakalwe

  • Uranus
  • ****
  • Posts: 1588
Re: Starship!
« Reply #106 on: December 25, 2020, 03:10:17 PM »
Because pork barrel politics.
Shelby has mouths to feed dontcha know...
"The strain of anti-intellectualism has been a constant thread winding its way through our political and cultural life, nurtured by the false notion that democracy means that 'my ignorance is just as good as your knowledge.' " - Isaac Asimov

Offline jfb

  • Mars
  • ***
  • Posts: 396
Re: Starship!
« Reply #107 on: December 26, 2020, 11:08:32 AM »
I just skimmed the video of Starship assembly, so everyone please forgive me if I'm misrepresenting it.  But it looked like some parts of assembly and stacking were being done using general-purpose construction equipment.  Naturally almost none of that would be acceptable in the development of any other crewed space vehicle, if done under the auspices of a NASA contract to do so.

Genuine question... why?

If a $5 million "general purpose" crane meets and exceeds all the specifications required to do the job, why is it necessary to design, build or buy a $50 million "special purpose" crane to do the exact same job?

Special-purpose equipment may be designed to make the process safer or more efficient, or to bake in a known level of reliability.  A general-purpose mobile crane can handle a lot of different tasks, but may not have a level of precision that a special-purpose crane can provide.

But...

SpaceX’s answer appears to be to tailor their design and process to accommodate commercial equipment rather than vice versa. 

Offline raven

  • Uranus
  • ****
  • Posts: 1637
Re: Starship!
« Reply #108 on: December 26, 2020, 11:14:37 AM »
There's a quite good Kerbal Space Program fanfic that explains the distinctly bodged appearance of the parts being due to the early Kerbal space program being operated out of a junk yard. I can't help but be reminded of that.

Offline smartcooky

  • Uranus
  • ****
  • Posts: 1959
Re: Starship!
« Reply #109 on: December 26, 2020, 07:09:08 PM »
I just skimmed the video of Starship assembly, so everyone please forgive me if I'm misrepresenting it.  But it looked like some parts of assembly and stacking were being done using general-purpose construction equipment.  Naturally almost none of that would be acceptable in the development of any other crewed space vehicle, if done under the auspices of a NASA contract to do so.

Genuine question... why?

If a $5 million "general purpose" crane meets and exceeds all the specifications required to do the job, why is it necessary to design, build or buy a $50 million "special purpose" crane to do the exact same job?

Special-purpose equipment may be designed to make the process safer or more efficient, or to bake in a known level of reliability.  A general-purpose mobile crane can handle a lot of different tasks, but may not have a level of precision that a special-purpose crane can provide.

But...

SpaceX’s answer appears to be to tailor their design and process to accommodate commercial equipment rather than vice versa. 

Thanks for this answer

I'm not necessarily buying the first part. I said if the commercially available GP equipment meets ALL of the required specifications (and that would necessarily include being able to do the job just as safely as the SP equipment would) then why the need to spend the extra to buy the SP equipment?

I am well versed in using specialist test equipment from my avionics days so I am certainly aware that having something built to do a specific task can make the job easier. For example, one of the pieces of equipment I use to work on was the APN59B weather radar out of a C-130. It had a test rig which plugged into an internal test port on the RT289 that allowed all the internal voltages and circuit board test points to be monitored with the turn of a rotary switch on the test panel. That job could just as safely be done by using a meter or oscilloscope and directly probing those points. In fact we had to do that sometimes because we only had one test rig, but would often have two RT units in the workshop at the same time.

As to the second part, yes, that seems to be the intent here. I am looking forward to seeing how they are going to go about stacking Starship on top of Super Heavy - "Bluezilla" they used on SN8 doesn't look as if it will lift high enough. Maybe they are going to build a Super-High Bay?
If you're not a scientist but you think you've destroyed the foundation of a vast scientific edifice with 10 minutes of Googling, you might want to consider the possibility that you're wrong.

Offline JayUtah

  • Neptune
  • ****
  • Posts: 3787
    • Clavius
Re: Starship!
« Reply #110 on: December 27, 2020, 02:16:48 PM »
So many great, thoughtful contributions here.  Again, this is hard for me to debate because a lot of it is my industry and I have to represent it fairly in public.  It's hard to distill insight away from its mixture with opinion, but I assume a fair amount of my opinion is part of why you guys would want me to answer.

But mostly, you folks have touched on items that would result in another wall of text were I to address all of them carefully.  I spent a lot of time yesterday baking, and thinking about how to respond in a way that's honest, respectful of others' opinions, informative, and shorter than book length.  I literally taught a class on this (tooling for advanced manufacturing) in college, so there is a plethora of thoughts to collect.  There is also a lot of historical background that we can't just ignore.  And some business stuff that I frankly let other people be experts in whereas I am not.
"Facts are stubborn things." --John Adams

Offline molesworth

  • Mars
  • ***
  • Posts: 349
  • the curse of st custards
Re: Starship!
« Reply #111 on: December 28, 2020, 04:27:42 PM »
So many great, thoughtful contributions here.  Again, this is hard for me to debate because a lot of it is my industry and I have to represent it fairly in public.  It's hard to distill insight away from its mixture with opinion, but I assume a fair amount of my opinion is part of why you guys would want me to answer.

I'm sure a wee disclaimer along the lines of "not speaking in my professional capacity.." would suffice  ;)

I think all of us in the space industry have the same problem, and I certainly wouldn't want any thoughts I might have on the efficiency or fiscal wisdom of any agency, company or project to be seen as representing the views of my company...  :D
Days spent at sea are not deducted from one's allotted span - Phoenician proverb

Offline jfb

  • Mars
  • ***
  • Posts: 396
Re: Starship!
« Reply #112 on: December 29, 2020, 12:26:00 AM »
I’d like to take back my earlier statement somewhat, because the more I think on it the less happy I am with it.

SpaceX are building their rockets and their rocket manufacturing facility concurrently.  They were stacking rockets before they had buildings in which to stack them.  They are using general-purpose construction equipment to build and stack starships because it was the quickest path to get hardware on the pad.

That is changing - they’re in the process of installing a permanent crane in the high bay.  They’ve bodged together a crawler-transporter out of ganged SPMTs, but at some point they may roll out a bespoke CT.  There are a number of jigs and customized welders in the tents for building the barrels, domes, and nosecones.  But even so, plenty of components are built off-site (I believe in Hawthorne) - the flaps, header tanks, downcomers, and of course the engines. 

And these are all still prototypes, not production vehicles.  They’re test beds for validating design and procedures.  Things are still in flux.

I can imagine as the manufacturing side gets tuned, we’ll see more specialized equipment and fewer United Rentals decals. 

Offline JayUtah

  • Neptune
  • ****
  • Posts: 3787
    • Clavius
Re: Starship!
« Reply #113 on: December 29, 2020, 12:27:08 PM »
Indeed, just last night I saw a short video about the Starship launch pad. Along the way someone quoted someone at SpaceX saying they didn't know how to build Starship yet.  So yes, that casts the whole process of constructing flight-test articles in more favorable light.  You can dig up photos of early assemblies from the lunar module and space shuttle showing them being assembled with hand tools on plywood fixtures.  It's a matter of understanding where we are in the process.

Without delving into a lengthier response...

Generally you have to develop the thing you're building and the way you're going to build it together.  Fabrication and assembly methods place constraints on the design, and it's costly to ignore them until later.  So I'm just not sure where the Starship is in its development process.  If the contention is that one company outpaces all others, then you need to take careful stock of what they might be skipping and why.  My impression is that Elon Musk prefers a Silicon Valley style engineering development process, which is hardly surprising.  You work very quickly to build prototypes and proofs of concept, then you go back later and fill in the gaps, redesigning as necessary to accommodate things you learned during your initial work.  This seems to work well for software and electronics, although those of you who specialize in those fields can correct me.

But the manufacturability issues don't resolve themselves easily that way when the product is a costly physical engineering article.  The Falcon 9 acceptance was delayed because NASA inspectors kept finding nonconforrmant manufacturing and assembly processes.  The Tesla Model 3 was delayed for the same manufacturability reasons.  I drove a demonstration model and was impressed enough with it to plunk down the deposit for one.  A year went by, and then another year.  The price kept going up and the feature list kept getting shorter.  This was Tesla trying to figure out how to build it after it had already been "designed."  And lest this turn into a one-sided criticism, it's also the reason the Dreamliner was late.  Boeing tried to cut the same corners and paid the price in a costly redesign effort and delayed delivery.

That said, if you can afford to learn from failure then it's worth it to explore other development processes and see if you can find a better way.  So just because you aren't doing the same thing everyone else does doesn't mean you're doing it wrong.  It just means it falls somewhere else on the cost-benefit spectrum.  Now in the case of Falcon 9 you can clearly say that the manufacturing issues are resolved, because the delivery tempo is enviable.  But the question is how big a check Elon Musk had to write to fix the manufacturing variances?  And what if it had happened in a company where that sort of correction wasn't an option?

More importantly, manufacturability is not strictly an economic argument.  If you're going to do work for a national space agency, acceptance testing based on a manufactured product is the requirement.  And it's the same for some commercial customers (or more appropriately, their insurers).  A one-off, hand-built article isn't eligible for acceptance testing because a lot of the other requirements of a development and manufacturing process (which I'll get into in a separate post) have to be baked into the process.  The tests aren't just for the article, but for the way the article is being designed and built.  You have to demonstrate that you can make an article that passes various tests using a process that also satisfies certain criteria that pertain to the process.  You have to prove you can do it that way every time, not just for the article submitted for test.

So the answer to why Starship prototypes are currently being assembled with general purpose construction equipment is simply that it's not as far along in the development process as I had previously thought, and SpaceX have elected to accelerate to flight testing.  And from what I understand the objectives of these test flights to be, I have far less a problem with the construction methods now that the evidence seems to show not only that they're building up the actual production line, but that they're doing it fairly early in the process.  This is good, because it means they seem to be addressing a problem that has plagued Musk's companies before.
"Facts are stubborn things." --John Adams

Offline smartcooky

  • Uranus
  • ****
  • Posts: 1959
Re: Starship!
« Reply #114 on: December 29, 2020, 04:02:18 PM »
Very good reply Jay, thank you

It seems obvious to me that SpaceX are using a "conceptual engineering" philosophy in the designing of Starship... they are trying things that are leading edge, never done before concepts, such as a 100% reusable rocket, a closed-cycle methalox engine and the flip maneuver to land upright, so this really is innovation on a grand scale. The idea seems to be to prove the concepts work before you spend billions on refined engineering, only to find that the concept is flawed and never going to work or is too difficult to reliably and safely engineer.

Was the video you watched the one by Primal Space? If so, you might remember they talked about how one of the static test fires kicked up material that damaged stuff inside the engine bay. This could be fixed by having water deluge and/or exhaust channels to prevent the shock-wave from doing this damage. However, the point needs to be made that the primary purpose of Starship is to fly humans to Mars, land them on the surface and then return them to earth. There will be no specially built launch pads, exhaust channels or water deluges on Mars. They are going to have to learn to handle this debris issue, so why not start now.
« Last Edit: December 29, 2020, 04:17:43 PM by smartcooky »
If you're not a scientist but you think you've destroyed the foundation of a vast scientific edifice with 10 minutes of Googling, you might want to consider the possibility that you're wrong.

Offline Zakalwe

  • Uranus
  • ****
  • Posts: 1588
Re: Starship!
« Reply #115 on: December 29, 2020, 05:02:14 PM »
Its also worth considering that Falcon 9 was primarily designed to win contracts for SpaceX. Starship is being primarily designed to get to Mars. It's secondary purpose will be a heavy lift vehicle that may be used to lift Starlink satellites in order to build the network and generate the revenue that SpaceX needs to get to Mars. In other words, Starship's primary customer will be SpaceX.

Being their own customer means that they do not have to jump go through the myriad of hoops that, say, a contractor to a government agency will insist on.
"The strain of anti-intellectualism has been a constant thread winding its way through our political and cultural life, nurtured by the false notion that democracy means that 'my ignorance is just as good as your knowledge.' " - Isaac Asimov

Offline smartcooky

  • Uranus
  • ****
  • Posts: 1959
Re: Starship!
« Reply #116 on: December 29, 2020, 05:36:19 PM »
Its also worth considering that Falcon 9 was primarily designed to win contracts for SpaceX. Starship is being primarily designed to get to Mars. It's secondary purpose will be a heavy lift vehicle that may be used to lift Starlink satellites in order to build the network and generate the revenue that SpaceX needs to get to Mars. In other words, Starship's primary customer will be SpaceX.

Being their own customer means that they do not have to jump go through the myriad of hoops that, say, a contractor to a government agency will insist on.

True, but to be fair, NASA  have awarded them a $135 million contract to advance the design of the Starship for potential use as a crewed lunar lander in the Artemis program.
If you're not a scientist but you think you've destroyed the foundation of a vast scientific edifice with 10 minutes of Googling, you might want to consider the possibility that you're wrong.

Offline JayUtah

  • Neptune
  • ****
  • Posts: 3787
    • Clavius
Re: Starship!
« Reply #117 on: December 29, 2020, 08:11:50 PM »
The idea seems to be to prove the concepts work before you spend billions on refined engineering, only to find that the concept is flawed and never going to work or is too difficult to reliably and safely engineer.

Yes, Silicon Valley has been quite successful at large by following this model.  And you might ask why Boeing, Lockheed, Northrop Grumman, and others don't follow suit.  And the answer is long, but boils down largely to them being established, publicly-traded companies.  Private companies have a different risk landscape, and this factors into deciding who will build what.

Quote
Was the video you watched the one by Primal Space? If so, you might remember they talked about how one of the static test fires kicked up material that damaged stuff inside the engine bay.

I shuffled back through my history and can confirm it was theirs:  "SpaceX's Launch Pad Problem."  I think it might be the first video from them I've seen.  First, the idea of a coated concrete pad is a good advancement.  But without the coating, the heat spalling was pretty spectacular.  And I get the wisdom of using that behavior to figure out what a field landing and takeoff on Mars will need in terms of engineering.  They have money, an already-wrecked launch pad, and privately-funded prototypes to hammer to pieces if needed, and aren't really accountable to anyone for how that investigation goes.  This would have to be planned for much differently if NASA had wanted an established company to investigate it.  "Move fast and (literally) break things," has considerable value, but it's not always an option for complicated reasons, and that raises thoughtful questions.

"Facts are stubborn things." --John Adams

Offline molesworth

  • Mars
  • ***
  • Posts: 349
  • the curse of st custards
Re: Starship!
« Reply #118 on: December 30, 2020, 04:59:28 AM »
My impression is that Elon Musk prefers a Silicon Valley style engineering development process, which is hardly surprising.  You work very quickly to build prototypes and proofs of concept, then you go back later and fill in the gaps, redesigning as necessary to accommodate things you learned during your initial work.  This seems to work well for software and electronics, although those of you who specialize in those fields can correct me.

I'll volunteer to be the one to correct you on this  ;D  As a software engineer of <ahem!> years, and working for a company that does everything from chip design up, I'd say it's widely accepted that careful design and testing results in much better quality for the end results.  There's an old saying that the earlier you find a bug, the cheaper and easier it is to fix.  (A couple of exceptions would be experimental prototypes where you want to quickly try ideas, and in my previous life as a game developer where getting the "gameplay" right was often more experimental - although the underlying "engine" tech had to be solidly designed.)

However, to my non-physical-engineering eyes, I don't think SpaceX is using a "throw it at the wall and see what sticks" approach at all.  To have gotten this far with what appears to be fairly mature spacecraft and engine designs leads me to believe they've spent a lot of time working on them.  Using off-the-shelf ground equipment like cranes and forklifts is perhaps more down to wanting to get things moving quickly rather than spending a lot of time building the infrastructure first.  I'll bet that there was quite a bit of similarly creative equipment use in the first years of most space programmes, including early and pre-NASA.
Days spent at sea are not deducted from one's allotted span - Phoenician proverb

Offline molesworth

  • Mars
  • ***
  • Posts: 349
  • the curse of st custards
Re: Starship!
« Reply #119 on: December 30, 2020, 06:44:12 PM »
Oh, and if there was any doubt that Elon is quite, quite mad...

https://twitter.com/elonmusk/status/1344327757916868608
Quote
We’re going to try to catch the Super Heavy Booster with the launch tower arm, using the grid fins to take the load
:o
Days spent at sea are not deducted from one's allotted span - Phoenician proverb