An always cool Falcon 9 launch (and other Space News)

[Uttamattamakin]

1 hour ago, Taf the Ghost said:

As Starship gets closer to a functional platform, I'm not surprised we're going to see more and more that simply don't understand SpaceX's testing approach for this project.   

I understand plenty I also understand that history doesn't repeat but often rhymes. 

 

A rocket like starship was tried before... it had all these problems and ultimately was a failure due to certain flaws in this type of design. 

 

The Soviet Union tried this, with the same kind of situation. 
 

There is a certain instability inherent in all rockets.  All rockets of any design.  Rocket engineers need to minimize that.  The fewer the moving parts the better.  The fewer the engines the better.  This is complex enough without that. 

4 hours ago, HenrySalayne said:

Not to mention an actual working flight termination system is a HUGE step in the right direction.

Demonstrating that is probably a big relieve to anyone.

Furthermore, Scott Manly points out that Starship was well within it's flight envelope and so this would NOT have been an automated flight termination system caused explosion. 

(I know this was posted above but this is time stamped to where I am pointing out).

So no this may not have been a successful test of the flight termination system I was a step in the right direction but not HUGE.  At least it's to where it blows up over the gulf of Mexico now. 

[Kisai]

7 hours ago, Uttamattamakin said:

Simple rule don't fly on a spacecraft that a billionaire wouldn't. 

Maybe don't get into vehicles, even if the CEO would, if the CEO complains about safety regulations.

 

I would rather "build it right the first time", the method that engineers used before computers were a commodity, than "keep hitting that compile button, eventually, one of these builds will work" strategy that Google uses. In the chase for AI, we're really getting as far from "build it right the first time" and instead "build as many broken things as possible and simulate it until it self-figures out how to do do it in one try", that can be very expensive, very quickly.

 

 

[HenrySalayne]

13 minutes ago, Uttamattamakin said:

I understand plenty I also understand that history doesn't repeat but often rhymes. 

 

A rocket like starship was tried before... it had all these problems and ultimately was a failure due to certain flaws in this type of design. 

 

The Soviet Union tried this, with the same kind of situation. 
 

There is a certain instability inherent in all rockets.  All rockets of any design.  Rocket engineers need to minimize that.  The fewer the moving parts the better.  The fewer the engines the better.  This is complex enough without that. 

The video is extremely poor. "Largest explosion ever" - absolutely not.

And no, using a bunch of smaller engines is not at all worse than a few larger engines; nor are these failures inherent to the design.

[manikyath]

9 minutes ago, Uttamattamakin said:

 

The Soviet Union tried this, with the same kind of situation. 

they tried this in an era where computers essentially didnt exist. spaceX is working with a computer system that can auto abort 0.1 second before launch if one pump didnt quite spool up as fast as it should have. (that actually happened on a starship test flight, which they later decided to run anyways, and the underperforming engine ended up running engine-rich on landing. which, yes, is another meme term.)

 

you're still stuck on the much many engines being a problem.. while none of what we saw today was a "much many engines" problem.. in fact, spaceX has demonstrated with the starship testing that they can use the "much many engines" as a form of reliability. the ship's landing burn only needs 2 engines to light, but they spool all 3 and then rapidly select the best performing 2 out of 3. these are things that are calculated MUCH faster than any computer system the soviets could have dreamt of.

you're essentially saying that because <that french airplane that is debated to be the first real airplane instead of the wright brothers> wasnt really all that controllable, the douglas DC-9 could never fly.

 

and again.. the N1 problems are not even vaguely related to starship. the N1 was deeply flawed, and because "soviet technology cannot fail" they just elected to not test things, so problems only ever showed up when it fell off the launch pad. essentially.. you're using N1's lack of testing and design iteration as proof that starship's potentially excessive testing will never work.

[Donut417]

18 minutes ago, Kisai said:

I would rather "build it right the first time", the method that engineers used before computers were a commodity, than "keep hitting that compile button, eventually, one of these builds will work"

The funny thing is, we have a design that works. The Saturn V, it made it to space, the moon and back with only a few issues, that being what happened to Apollo 1, which caused a slight redesign and Apollo 13, which I'm assuming we would have better quality control now days.

 

Now Im not saying we build a 1960s spacecraft. BUT we can borrow from its design and make a 21st century version.

[Mark Kaine]

1 hour ago, manikyath said:

they tried this in an era where computers essentially didnt exist. spaceX is working with a computer system that can auto abort 0.1 second before launch if one pump didnt quite spool up as fast as it should have. (that actually happened on a starship test flight, which they later decided to run anyways, and the underperforming engine ended up running engine-rich on landing. which, yes, is another meme term.)

 

you're still stuck on the much many engines being a problem.. while none of what we saw today was a "much many engines" problem.. in fact, spaceX has demonstrated with the starship testing that they can use the "much many engines" as a form of reliability. the ship's landing burn only needs 2 engines to light, but they spool all 3 and then rapidly select the best performing 2 out of 3. these are things that are calculated MUCH faster than any computer system the soviets could have dreamt of.

you're essentially saying that because <that french airplane that is debated to be the first real airplane instead of the wright brothers> wasnt really all that controllable, the douglas DC-9 could never fly.

 

and again.. the N1 problems are not even vaguely related to starship. the N1 was deeply flawed, and because "soviet technology cannot fail" they just elected to not test things, so problems only ever showed up when it fell off the launch pad. essentially.. you're using N1's lack of testing and design iteration as proof that starship's potentially excessive testing will never work.

actually no lol... the more engines you have the more chances of failure... this just sounds like a typical musk thing, computers everywhere and needlessly complicated (and probably inefficient af just like his cars) 

 

 

[divito]

Given what SpaceX is attempting compared to anyone else before them, this is a solid step, even if the media probably doesn't understand it or thinks less of it. They've been pretty open that they're mostly trying to learn and their R&D should yield efficiencies and savings that NASA probably never attempted, given the bureaucracy. 

[Taf the Ghost]

6 hours ago, Donut417 said:

The funny thing is, we have a design that works. The Saturn V, it made it to space, the moon and back with only a few issues, that being what happened to Apollo 1, which caused a slight redesign and Apollo 13, which I'm assuming we would have better quality control now days.

 

Now Im not saying we build a 1960s spacecraft. BUT we can borrow from its design and make a 21st century version.

Very technically, the Saturn V is "lost tech". It is not possible to remake it. You could build an approximation of it, but the introduction of any new systems would require fully re-engineering huge chunks.  Even the F5 engines couldn't be remade in the same methodologies, though they've thankfully pulled a museum one apart and deeply 3D scanned everything. With new 3D metal printing and other advanced manufacturing techniques, a similar model could be produced with a lot less parts needed.

 

The other detail about the Saturn V is that it's likely the most complicated machine ever built by humans, and will likely remain that way for a rather long time. It was the actual pinnacle of technology of its era. Which is also why it was so incredibly expensive.

 

 

6 hours ago, Kisai said:

Maybe don't get into vehicles, even if the CEO would, if the CEO complains about safety regulations.

 

I would rather "build it right the first time", the method that engineers used before computers were a commodity, than "keep hitting that compile button, eventually, one of these builds will work" strategy that Google uses. In the chase for AI, we're really getting as far from "build it right the first time" and instead "build as many broken things as possible and simulate it until it self-figures out how to do do it in one try", that can be very expensive, very quickly.

 

 

 

Cost. The "design it, build it, test it until it explodes" approach is going to save the entire program for Starship such a massive amount of money & time it isn't funny. SpaceX will have a flight ready vehicle and a fully scaled manufacturing system ready likely a decade before they would in the classic spaceflight approach, while at the same time having vast quantities of real world performance data.  This approach isn't new or even all that technically interesting.  It's a tried & true method for rapid development, and actually used to be a staple of missile testing.

 

The difference is that when the decision makers are politicians, they're far more concerned about the optics than the outcomes.  This is actually how the Space Shuttle program ended up costing more than if they'd just kept improving the Saturn V and made various mission modules for it.

[Kisai]

6 hours ago, Donut417 said:

The funny thing is, we have a design that works. The Saturn V, it made it to space, the moon and back with only a few issues, that being what happened to Apollo 1, which caused a slight redesign and Apollo 13, which I'm assuming we would have better quality control now days.

 

Now Im not saying we build a 1960s spacecraft. BUT we can borrow from its design and make a 21st century version.

Indeed.

 

It would be different if we just needed to get payloads into space and assemble something bigger there that doesn't need to land again (eg the ISS) , but we're largely in a position where we already know what works, and not much can be improved on that without inventing something like a mass driver or space elevator to get things into orbit without sitting on a stick of dynamite.

 

It's like how we've had gasoline engines for nearly a century before any kind of fuel efficiency was even thought of. That's why every car built since 1980 started getting streamlined, so they went from boxy 80's cars to "bubble" 90 cars to somewhere where they are right now where the compromise between aerodynamic drag on the car, and the electronicly controlled fuel injection systems made things much cleaner and more fuel efficient. That basically meant that the "muscle car" was no longer economically viable.

 

At any rate, back-of-the-napkin "would this work?" to me would suggest let's just figure out how to build spacecraft "in space" first before trying to get them to Mars. The fact the ISS wasn't really built to last 100 years should have been clue enough that we need to figure out how to do this first.

 

Trying to launch things from the ground that ultimately can't be very big, and have high rates of metal fatigue (this is why airplanes have to be retired after 135,000 hours, or about 30 years.) Even if the Space Shuttle was still operating, it's very likely it would have had to be retired as well with the space station. 

 

As a side, I also don't think the Soviet Union would have built additional Buran orbiters and their version of a larger space station. A lot of what we've seen from communist countries is about faking things to the West. Not everything, but I doubt the space program would have continued much longer. The space program for the US and for the USSR was just cover for military uses, just like every country says they are doing "Peaceful" , yes... but they wouldn't be building rockets themselves if they didn't need the technology to build ICBM's "if the need arises"

 

It's kinda funny overall what happened with the ISS, we ended up relying on the remnants of the soviet union's space program to service it to it's end of life because the shuttle was older than the ISS.

 

Anyway. We've always had options on the table, just not ones that made economic sense if there was not a military purpose.

 

[Donut417]

10 minutes ago, Kisai said:

t would be different if we just needed to get payloads into space and assemble something bigger there that doesn't need to land again

We probably could do this today. If all you needed was a ship to carry a crew to orbit, you could easily build something that could be reusable like the shuttle. Then use heavy rockets to get the sections of the craft in to orbit. Possibly use the ISS as a staging area to help build the thing. The biggest issue is fuel. I know that I have seen info on electric power engines or other methods where you can use a nuclear reactor. 

[wanderingfool2]

16 hours ago, Uttamattamakin said:

Congratulations to Space X for doing a little better.  Bear in mind, for comparison.  No Saturn V ever blew up during testing.   No Space shuttle did during testing, sadly two did fail while manned.   In modern times SLS has not blown up once.   Falcon Heavy has not blown up once.  The numbers are small but for this number of launches this number of failures is not good.  

That is the wrong mindset and ignorant statement to the development approach of SpaceX.  Your comparison is also comparing apples to oranges, but even worse you are taking datapoints at different times in development.

Lets start with the flawed "SLS has not blown up" argument.  The SLS borrows on the design and experience from the previous boosters along with using the same type of fuel as well, which most certainly did blow up during development.

The SLS itself, has been in development for 11 years and has cost $27 billion...for that time and price they better make the first attempt successful...and even then it failed it's wet dress rehearsal which lead to months of delay.  (And had the same problem arose during flight it would have been a RUD as well). [For reference starhopper the first "test" article from SpaceX using the new type of LOX and Methane engine occurred in 2019]

The SLS was a rocket that was using a known rocket fuel, and old design.  They had to abort the first flight test so many times due to issues that arose (One of the last ones being the stuck valve).  That is also why it's going to cost like $4 billion per launch of the SLS (vs at the moment $100 mill per Starship)

 

The difference is we are talking about the SLS which is only producing 8.8 mill lbs of thrust...compared to the ~17 mill lbs of thrust of Starship; and a ship that was designed to eventually be recoverable.

 

There is also the different design approaches, where you test to fail while also collecting as much data as you can (simulations of things can only take you so far).  NASA has the approach of testing each individual component and essentially designing at a turtle's pace but having essentially more assuredness to it.

 

The SLS itself had it's own setbacks, where the engine shutoff during the green run at the 22 second mark due to a major failure within the engine.  It took nearly 10 months to identify and rebuild to the point of the actual SLS launch  (btw the SLS decimated the launch tower...which itself cost like a billion).  If SpaceX had gone the NASA way, they could have clamped down the Starship...but at 17m lbs of thrust it would have been difficult and required building out the entire flametrench and even then it wouldn't be able to simulate the boost back burn or the hot staging.  The boost back I suspect being the cause of the failed engine ignitions on the booster, probably cracking the pipes and springing a leak...or getting an airpocket in the fuel line.

 

So yea, you are incorrectly associating a different design philosophy as some means of comparison of quality while it doesn't mean a thing...as they effectively are trading equipment for saved time.  It's the same type of philosophy which has lead to the VERY successful Falcon 9

 

As for the Falcon Heavy comparison, a Falcon Heavy uses the same types of engines as the Falcon 9; and the boosters are very very similar...and Falcon 9 had it's fair share of blow ups in it's beginning stages.

 

The thing to remember, while it was talked about since ~2011; the first real designs for it never became a thing until Starhopper in 2019.  (And include an additional year of waiting for the FAA to do their review).

 

The SLS has it's fair share of problems, yes it hasn't blown up...but that again is because they have been doing the old school method of testings (which SLS had it's own fair share of testing failures).  The stuck valve which caused premature shutdown of the engines (which had it been a test flight it would have lead to the RUD as well).  The effectively tested each component separately and then assembled and tested as a whole (and that itself took like the decade to complete).

 

Essentially the SLS approach traded time and money for the "right on the first go"...which effectively means a lot of testing and reworking until things are "perfect"...and even then it can fail (SLS green run failure).

 

 

16 hours ago, Uttamattamakin said:

Seriously though its kinda sad.  I'm sure someone can make the multi-multi engine heavy lift work.  The USSR tried and failed with the N1 even after multiple launches.   Many smaller engines might just not be as good as having 3-5 biguns. 

And that thinking is why the SLS is costing like 40x more to build than each Starship (and also is limited to only a flight a year).

 

Each engine on the SLS costs apparently about $40 million to build.  This most recent flight of Starship was estimated at $100 - $150 million.  To create that amount of power with engines you would need about 6 SLS.  That doesn't even factor in that it takes months to make each of those engines.  SpaceX's approach of mass producing one (so if one fault was made it can easily be replaced) reduces the cost of the engines on Starship down to like $100k per engine (albeit needing 33 of them).

 

It also lets then iterate on the design more, Raptor V1 vs Raptor V2 vs Raptor V3...they look quite different (with each time they are simplifying down the engine to have less parts and being more streamlined).  They gain that benefit by effectively doing destructive testings, and yes by doing test flights like these to see how it reacts under true loads.

 

 

8 hours ago, Kisai said:

I would rather "build it right the first time", the method that engineers used before computers were a commodity, than "keep hitting that compile button, eventually, one of these builds will work" strategy that Google uses. In the chase for AI, we're really getting as far from "build it right the first time" and instead "build as many broken things as possible and simulate it until it self-figures out how to do do it in one try", that can be very expensive, very quickly.

Except it always has been a "test to see if it works".

 

Back in the old days, they would test an engine and see it failing in a certain way and they would go in and make tweaks to it (adjusting it in their notes).  It's actually why it's so hard to rebuild the old engines and designs...because the people who made them actually ended up making lots of tweaks to the designs and such so that in the end if you build it strictly on the "build it right the first time" design specs it would fail.

[HenrySalayne]

6 hours ago, Mark Kaine said:

actually no lol... the more engines you have the more chances of failure...

On the very surface, by multiplying a constant failure rate, this would be true. In reality this is a completely pointless statement.

It rapidly falls apart because engine A ans engine B have different failure rates. Not to mention all other parameters.

[HenrySalayne]

1 hour ago, wanderingfool2 said:

If SpaceX had gone the NASA way, they could have clamped down the Starship...but at 17m lbs of thrust it would have been difficult and required building out the entire flametrench and even then it wouldn't be able to simulate the boost back burn or the hot staging.  The boost back I suspect being the cause of the failed engine ignitions on the booster, probably cracking the pipes and springing a leak...or getting an airpocket in the fuel line.

They are clamping down spaceship. They not only did a static fire test, they need to clamp down spaceship until engine startup is complete and all engines are running at full power or the rocket might tip over or slip off the pad.

 

[manikyath]

9 hours ago, Mark Kaine said:

actually no lol... the more engines you have the more chances of failure... this just sounds like a typical musk thing, computers everywhere and needlessly complicated (and probably inefficient af just like his cars) 

 

 

if we presume a small engine and a big engine to have the same failure rate - yes. but the idea behind many small engines is that they can be "mass produced" (hundreds, as opposed to douzens), and then in theory QC can weed out the underperforming ones on the test stand. there's also the engineering challenge of actually making and transporting, let's say 3 big engines as opposed to numerous smaller ones.

 

multiple engines also gives an amount of redundance, which the rocket is actually made to account for. on the landing phase of starship it will light all engines, and then select the best performing ones to use for the landing itself.

 

or in other words.. this is the way we've been keeping aircraft in the air for decades. all key systems on an aircraft are 'doubled': two (or sometimes even four) hydrolic systems, 4 smaller engines as opposed to 2 bigger ones, 3 gyro's, etc.. there's one bigger aircraft that even has two separate rudders. making more smaller parts does in theory create more points of failure, but often the benefit of smaller parts (and thus, easier shipping) and more redundancy outweighs this theoretical increase in points of failure.

 

as for computers everywhere.. do you think saturn V is not full of computers then? 

 

also - what happened to free market competition? space is very commercial, and space is the reason we have worldwide communication and our phones can give us directions to our destination. any business striving to make that process cheaper is only a benefit to our technological advancement.

 

and PS; if NASA's rockets are so great.. why are they launching classified satelites on 3rd party launch providers? and i'm not talking just SpaceX here.. there's rocketlab, there's astra who have an entire launch system that can be shipped around the world in regular shipping containers, etc. competition is good, even at this level.

 

11 hours ago, Donut417 said:

The funny thing is, we have a design that works. The Saturn V

the trabant is a working car, no need to engineer anything that could be better. we're done. fire all engineers.

 

if everyone thought the way you're thingking here, we would actually still be walking everywhere on dirt paths. why invent cobbled streets when dirt paths work just fine?

[Mark Kaine]

1 hour ago, manikyath said:

if we presume a small engine and a big engine to have the same failure rate - yes. but the idea behind many small engines is that they can be "mass produced" (hundreds, as opposed to douzens), and then in theory QC can weed out the underperforming ones on the test stand. there's also the engineering challenge of actually making and transporting, let's say 3 big engines as opposed to numerous smaller ones.

 

multiple engines also gives an amount of redundance, which the rocket is actually made to account for. on the landing phase of starship it will light all engines, and then select the best performing ones to use for the landing itself.

 

or in other words.. this is the way we've been keeping aircraft in the air for decades. all key systems on an aircraft are 'doubled': two (or sometimes even four) hydrolic systems, 4 smaller engines as opposed to 2 bigger ones, 3 gyro's, etc.. there's one bigger aircraft that even has two separate rudders. making more smaller parts does in theory create more points of failure, but often the benefit of smaller parts (and thus, easier shipping) and more redundancy outweighs this theoretical increase in points of failure.

 

as for computers everywhere.. do you think saturn V is not full of computers then? 

 

also - what happened to free market competition? space is very commercial, and space is the reason we have worldwide communication and our phones can give us directions to our destination. any business striving to make that process cheaper is only a benefit to our technological advancement.

 

and PS; if NASA's rockets are so great.. why are they launching classified satelites on 3rd party launch providers? and i'm not talking just SpaceX here.. there's rocketlab, there's astra who have an entire launch system that can be shipped around the world in regular shipping containers, etc. competition is good, even at this level.

 

the trabant is a working car, no need to engineer anything that could be better. we're done. fire all engineers.

 

if everyone thought the way you're thingking here, we would actually still be walking everywhere on dirt paths. why invent cobbled streets when dirt paths work just fine?

oh i get the reason they use spaceX, the idea is to have cheaper rockets - that of course makes sense, and other than the cars (environmental cost is gigantic and not really justifiable) they seem to have success with that, im not denying that... as for multiple small engines, the advantages are clear, its just nobody ever got it to work, and with musks apparent "throw stuff at the wall and see what sticks tactics" i remain highly skeptical. 

[manikyath]

2 minutes ago, Mark Kaine said:

oh i get the reason they use spaceX, the idea is to have cheaper rockets - that of course makes sense, and other than the cars (environmental cost is gigantic and not really justifiable) they seem to have success with that, im not denying that... as for multiple small engines, the advantages are clear, its just nobody ever got it to work, and with musks apparent "throw stuff at the wall and see what sticks tactics" i remain highly skeptical. 

before falcons, no one had first stages land for re-use. being skeptical is good, but you need to recognize that SpaceX has proven they can do things that seem extremely unlikely.

 

also - i prefer to call these achievements SpaceX achievements, because at most elon is just "one of the many engineers" that make up the teams at SpaceX. falcon reusability is an achievement of a large team of engineers that work at a company owned by musk. attirbuting this to the man himself is like saying linus did a great job on the screwdriver. he was involved, but it's the merch team that actually made it happen.

 

as for "throw stuff at the wall and see what sticks" at SpaceX... their general methodology is to engineer things to the point where more can be learned from a physical prototype being smashed into a wall than from more simulations.. then take the input of that prototype and add the real world results to the plans on paper, rinse and repeat.

 

none of the stuff made at boca chica are "finished rockets" - they're prototypes engineered to the point where the paperwork checks off all the requirements for the next stage of testing. if that test goes well - good. if that test doesnt go well, it's at the very least an example of what issues to account for. in a sense of irony, if the first test flight didnt drill a hole in the launch pad, we likely wouldnt have known the problems with the FTS, and they wouldnt have been resolved until much later in the process, and potentially in a far more risky scenario.

 

and clearly.. multiple engines work, the booster made it trough the entire launch trajectory, it didnt land.. but neither did any of the existing "single engine big rockets", because none of them were made to land. so.. SpaceX has got multiple engines to work to the same extent big engine rockets work.

[Kisai]

5 hours ago, wanderingfool2 said:

 

Except it always has been a "test to see if it works".

 

Back in the old days, they would test an engine and see it failing in a certain way and they would go in and make tweaks to it (adjusting it in their notes).  It's actually why it's so hard to rebuild the old engines and designs...because the people who made them actually ended up making lots of tweaks to the designs and such so that in the end if you build it strictly on the "build it right the first time" design specs it would fail.

No, you're not quite getting the picture.

 

NASA, and various other space agencies always tested things in labs, not building the the entire thing and watching it explode, destroying all traces of data. That was part of the problem before computers were mainstream, all they could do was film things.

 

With the kinds of things we have today, we shouldn't need to be "putting stuff on the launch pad" that might explode before it even gets off the pad. It makes far more sense to only put stuff on the pad after simulations, and prototypes have already proven to "Work as designed". A lot of pre-computer stuff was modeled by making miniature 1/10th scale versions.

 

Anyway, I'm not suggesting SpaceX is wrong for what they are doing, just that they are making an extreme mess and it looks horribly embarrassing seeing things explode. That shatters confidence in the people who want to put payloads, or even people on them. "Yeah, you could go up on a 50 year old Soviet rocket design, or you could go on one of Musk's every-second-one-explodes models", I think you'd have all the space tourists prefer that soviet rocket.

 

And I'm not saying the Soviet rocket was good, just it's been refined enough that people are willing to risk their life on it.

 

Reading the wikipedia entry on the Soyuz MS, apparently one of the largest improvements over time was shrinking the computer, which used to be 70kg down to 8.4kg. That's like an entire person-worth of weight reduction.  So it's not like the Russians are using  the vintage 1970's designs still.

 

Anyway what I see happening probably in the next decade is "simulation smashing" by using AI models of the actual designs, and simulating the intended parts at the material level to see how long before it fails before actually trying to build it. That would save a lot more expensive launch failures.

 

[Mark Kaine]

11 minutes ago, manikyath said:

before falcons, no one had first stages land for re-use. being skeptical is good, but you need to recognize that SpaceX has proven they can do things that seem extremely unlikely.

yes, i did... they actually succeeded to make rockest more economical, especially due to reusability.

 

what i don't get is why do they keep testing... i thought its a done deal lol... when did they decide to use multiple small engines? 

 

14 minutes ago, manikyath said:

none of the stuff made at boca chica are "finished rockets" - they're prototypes

ah, ok, so what is the endgoal here?  

 

I thought i know most about this stuff, but apparently Scott Manley failed me... i knew they would need to do more testing i just thought the design was finalized.

 

i mean if they can get it to work (eventually) , fine, i guess. 

[manikyath]

14 minutes ago, Kisai said:

 

With the kinds of things we have today, we shouldn't need to be "putting stuff on the launch pad" that might explode before it even gets off the pad. It makes far more sense to only put stuff on the pad after simulations, and prototypes have already proven to "Work as designed". A lot of pre-computer stuff was modeled by making miniature 1/10th scale versions.

we do though, because there's always things that dont scale as expected, things that werent accounted for in the computer simulations, etc. the reason they put stuff on a pad to blow it up, is because that ends up being cheaper than a simulation at that point in the development cycle.

 

12 minutes ago, Mark Kaine said:

 

what i don't get is why do they keep testing... i thought its a done deal lol... when did they decide to use multiple small engines? 

because they deemed that for starship's booster, making that amount of power out of a small amount of engines makes those engines very expensive, very difficult to manufacture, and very difficult to transport.

 

it's kinda difficult to think of because the scale of starship creates weird contexts.. but these "small" engines are still very much "enormous" engines.

for context.. falcon9 is a 3.7m diameter rocket, starship is a 9m diameter rocket.

falcon9 has done 7 tonnes to orbit, starship's plans aim for at least 100 tonnes to orbit.

 

which - by the way, falcon 9 already uses 9 merlin engines, and has been doing so with shocking reliability.

 

in theory they could follow the SLS route and have solid fuel boosters carry a large portion of the uphill route, but i feel like SpaceX doesnt want to deal with the engineering witchcraft that is throttle control on SRB's (the power output of an SRB is dependant on how much surface area the fuel has, which changes as it burns up, so they're cast with intricate patterns to create a thrust profile to match a specific flight plan.)

 

the booster has 33 "sea level optimized" raptor engines, the upper stage / starship has 3 vaccuum optimized and 3 sea level optimized raptor engines.

 

for scale: sea level raptor on the left, vaccuum raptor on the right, human in the middle:

 

as compared to the merlin engines on falcon 9:

[HenrySalayne]

1 hour ago, Kisai said:

Anyway, I'm not suggesting SpaceX is wrong for what they are doing, just that they are making an extreme mess and it looks horribly embarrassing seeing things explode. That shatters confidence in the people who want to put payloads, or even people on them. "Yeah, you could go up on a 50 year old Soviet rocket design, or you could go on one of Musk's every-second-one-explodes models", I think you'd have all the space tourists prefer that soviet rocket.

I honestly doubt that. SpaceX with the Falcon 9 has proven to be one of if not the most reliable launch provider out there. If you are spending more than 50 million USD for a lauch (Starship - despite the claims of rapid and full reusability will probably be more in the 200 million USD region), you are hopefully sensible enough to tell test flights and operational flights apart.

 

The Ariane 5 was a great success despite its maiden flight ending with flight termination. The Shuttle program suffered two complete losses and was still the workhorse for all sorts of launches, while it was known that many design flaws could not be fixed. Starship might not live up to the Falcon 9's reliability, but since there is nothing with the same launch capability around, it has to do.

[wanderingfool2]

7 hours ago, HenrySalayne said:

They are clamping down spaceship. They not only did a static fire test, they need to clamp down spaceship until engine startup is complete and all engines are running at full power or the rocket might tip over or slip off the pad.

Not entirely true, for that static fire they only are at half power.

 

On the ignition and liftoff, they actually light the engines unclamp then throttle up the engines.  It's actually why there was a difference in the time it took to get off the pad this time.  They learned information from the first full power-up that they were able to push the sequence of powering up quicker, which is why it only stayed like 2 seconds on the pad this time.

 

The clamps "clamping down startship" also weren't designed for testing a full stage burn either (but the bigger argument with that amount of thrust it means they would have to build a full on flame trench and other custom equipment as I'm not sure there is a flametrench in existence that could even support it currently).  Also keep in mind that the booster has ~3600 tons of fuel on board with an additional weight of ~1200 tons ontop.  Which means you actually would have to design the holddown and the booster points where it holds down to hold an additional ~3600 tons when the booster becomes empty and spend money on a ~1200 ton mountable weight to be put ontop.  You then also need a mechanism test rotational force on the thing.

 

3 hours ago, manikyath said:

also - i prefer to call these achievements SpaceX achievements, because at most elon is just "one of the many engineers" that make up the teams at SpaceX. falcon reusability is an achievement of a large team of engineers that work at a company owned by musk. attirbuting this to the man himself is like saying linus did a great job on the screwdriver. he was involved, but it's the merch team that actually made it happen.

Not really an engineer.  He knows enough to be understand talent and finds a way to attract talent.  He then gives them his idea, and he doesn't act like seniority matters.  Seriously at lets say Tesla, if you are an engineer and you think you have a brilliant idea and your boss shoots you down; you can appeal it directly to him (but you better be right).  It gives the engineers a certain set of freedom, where you don't necessarily have to do things the old way just because it's the old way.

 

That is why I think things like Tesla and SpaceX essentially became successful; because he at least pinpointed the major things of importants (Tesla it was manufacturing), SpaceX it was reusability to minimize the cost.

 

3 hours ago, Kisai said:

No, you're not quite getting the picture.

 

NASA, and various other space agencies always tested things in labs, not building the the entire thing and watching it explode, destroying all traces of data. That was part of the problem before computers were mainstream, all they could do was film things.

I never said they didn't test things in the labs YOU were the one who said "build it right the first time"; which if you didn't catch on is why I said they do test to see if it works.  I never said that they test it as a whole.  I did clip a longer paragraph explaining that they did test things in labs before going to the full scale...but I think the general sentiment that I was saying also existed in my post.

 

Why do you think the designs that were used couldn't be reused today...it's because those small tweaks were caught in testing and written in notes/shorthand very few engineers could read.  The biggest change between the two, SpaceX and NASA, is that SpaceX does full on destruction testing which tests the system as a whole and that their system is doing a whole lot more than anything NASA could have hoped for.

 

It's key to remember that SpaceX is hitting uncharted territories in terms of some of applied physics and stuff going on; they can't simulate things perfectly and some of the things you can't test before hand without making expensive test chambers or burning through time and resources to check if you were right.  That's a major difference between the old school approach and SpaceX's approach.  SpaceX is about trying an idea to see if it works quickly, rather than spend years trying to analyze and test if it will work.  The very principles is why Falcon 9 is so successful now.

 

3 hours ago, Kisai said:

That shatters confidence in the people who want to put payloads, or even people on them. "Yeah, you could go up on a 50 year old Soviet rocket design, or you could go on one of Musk's every-second-one-explodes models"

No, it is just something that the media likes portraying as a failure.  Anyone who is sending or designing payloads will know that the final rocket that launches the payloads isn't represented by the failures of it's predecessors (unless they actually set it up as a "final" build and it explodes).

 

What would rattle would be if lets say a Falcon 9 explodes, as it's a rocket that is human certified.  No one in the space industry actually expected the thing not to explode....there was even the recognition that there was a good chance of the hot staging to destroy the booster or still a good chance the thing blows up before leaving the launch area.

 

3 hours ago, Kisai said:

simulating the intended parts at the material level to see how long before it fails before actually trying to build it. That would save a lot more expensive launch failures.

They already are simulating things, you just can't get the simulations all correct; which is why they do realworld testing as they are pushing what is currently known (and ultimately improving future simulations).

 

2 hours ago, HenrySalayne said:

I honestly doubt that. SpaceX with the Falcon 9 has proven to be one of if not the most reliable launch provider out there. If you are spending more than 50 million USD for a lauch (Starship - despite the claims of rapid and full reusability will probably be more in the 200 million USD region), you are hopefully sensible enough to tell test flights and operational flights apart.

Honestly, the people who will use Starship I think will be ones that require the massive fairing size...and at that point quite frankly there isn't really many other options.  (Unless the group literally has billions to spend on a single launch of an SLS)

[Uttamattamakin]

3 hours ago, manikyath said:

which - by the way, falcon 9 already uses 9 merlin engines, and has been doing so with shocking reliability. ....

 

This and 

2 hours ago, HenrySalayne said:

I honestly doubt that. SpaceX with the Falcon 9 has proven to be one of if not the most reliable launch provider out there. If you are spending more than 50 million USD for a lauch (Starship - despite the claims of rapid and full reusability will probably be more in the 200 million USD region), you are hopefully sensible enough to tell test flights and operational flights apart.

This is the reason I think Space X and the Artemis program and the Dear Moon flight should be done using a Falcon Heavy.  Falcon Heavy is built on top of the well tested tried and true Falcon 9.  The Falcon Heavy configuration of these has proven that it can launch heavy payloads to space without blowing them up. 

Starship gives a large fairing and high theoretical payload but, as I have said, a single rocket having that many engines has been tried.    It has inherent issues.   A Starship redesigned to have the same engine config as the Saturn V did would work much better.    Even better. 

Just take the falcon 9 and to an extent scale it, and it's engines up.  Just make the thing that works bigger. (Yes yes I know it is not that simple). 

 

https://timesofindia.indiatimes.com/world/us/ussfs-secret-spaceplane-set-for-historic-launch-aboard-spacexs-falcon-heavy/articleshow/105264995.cms?from=mdr

 

 

This has 27 small engines in total ... but it bypasses the many small engines problem by not plumbing them into one system that has to all work in harmony.  They could use this. 

 

SLS is great but it is 1 Billlion dollars a shot or there about.

 

[manikyath]

1 hour ago, Uttamattamakin said:

should be done using a Falcon Heavy

falcon heavy comes up way short on payload to orbit, and cant provide the same fairing size.

1 hour ago, Uttamattamakin said:

Starship gives a large fairing and high theoretical payload but, as I have said, a single rocket having that many engines has been tried.    It has inherent issues.   A Starship redesigned to have the same engine config as the Saturn V did would work much better.    Even better. 

the engines are not a problem. you just watched a 33 engine rocket get to space.

you're assuming that a 5 engine layout (saturn V is 5 engines in an X as far as i can find?) would perform better.. but even if it did, have you considered the cost, the way they have to be transported, the throttle flexibility required for landing the booster for recovery, etc? starship is *NOT* a 1:1 replacement for the saturn V, it has different design goals, which result in a different rocket.

1 hour ago, Uttamattamakin said:

Just take the falcon 9 and to an extent scale it, and it's engines up.  Just make the thing that works bigger. (Yes yes I know it is not that simple). 

if this works, they'd have done it. and again.. falcon 9 is a 9-engine design, falcon heavy by extension is a 27 engine design. The number of engines is not a problem.

 

1 hour ago, Uttamattamakin said:

This has 27 small engines in total ... but it bypasses the many small engines problem by not plumbing them into one system that has to all work in harmony.  They could use this. 

they have 3 systems that have to work in harmony. this is not a simpler design, not by a long shot. reusability on falcon heavy is also more complicated because they have 3 separate stages to land, so if they want to recover all of it, they *have* to RTLS.

 

i'm not a rocket engineer, neither are you.. let's just assume that the people who are a rocket engineer have determined that feeding 33 engines from one tank is not a problem have done the engineering required to make this decision. and those engineers are not the nutcase billionaire you clearly have severe problems with.

[HenrySalayne]

2 hours ago, Uttamattamakin said:

This is the reason I think Space X and the Artemis program and the Dear Moon flight should be done using a Falcon Heavy.  Falcon Heavy is built on top of the well tested tried and true Falcon 9.  The Falcon Heavy configuration of these has proven that it can launch heavy payloads to space without blowing them up. 

The Falcon Heavy might have an impressive (theoretical) payload to LEO, but it drops off rapidly. The Falcon Heavy has not enough performance / payload to send people and equipment to land on the moon. That's why we need SLS and Spaceship.

[Uttamattamakin]

10 minutes ago, HenrySalayne said:

The Falcon Heavy might have an impressive (theoretical) payload to LEO, but it drops off rapidly. The Falcon Heavy has not enough performance / payload to send people and equipment to land on the moon. That's why we need SLS and Spaceship.

Even in fully expendable mode?  

Well maybe it points a way towards a better design for a star ship.  Rather than plumbing together 33 engines in one big system... break it into several clusters of plumbing systems.  Separate tanks and internals....   There is a reason the one big rocket with a lot of small engines approach hasn't worked for anyone yet.  

I hope I am wrong because I want to see us go back to the moon and the landing system is supposed to be based on star ship.  NASA may have to find a way to do it with just an SLS and forget about starship.