Scrappy or crappy? Critique by former SpaceX lead engineer

[Robotbeat]

I think he does know what he’s talking about but your point about the churn is correct. If experienced people go start their own businesses, then you’re gonna have to relearn a lot of the same lessons. And you are also correct that the lessons aren’t quite the same as previous falcon launch pads were probably over-built, and this was actually just a one-off launch in this configuration, not planned for many launches.

I think a lot of outside observers are hyper focusing on the dramatic launch pad damage when the main issue with the flight that might hold up future launches is the FTS. Considering the speed of repair, it actually seems increasingly like a justifiable risk to take (unlike the FTS!).

[matthewkantar]

He worked on launch pads, and then moved to landing pads, and now has his own company for both.

And will never be hired by SpaceX to do either. As a competitor bad mouthing the competition, some pushback is expected.

Do you realize he's not badmouthing, but applying lessons learned by SpaceX themselves during his years there? As far as I know the company hasn't pushed back, unless you're a SpaceX representative - are you?

Also, what in the world are you talking about when you state he will "never be hired by SpaceX to do either"? He's already been hired, and he moved on to entrepreneurship on his own - or do you have information that suggests he was fired? EDIT: Actually, if you look in his Linkedin account, there's high praise from then-VP of SpaceX Lee Rosen stating that he was top-of-the-line and decided on his own to leave the company, that he'd "hire him again in a heartbeat" and that he had his full endorsement...

I sometimes wonder whether Elon himself would be subject to this treatment by the unfailingly faithful if he came out to say such things.

Huh? No I am not affiliated with SpaceX.  Are you afflicted with rabies?

His company won’t be hired by SpaceX, SpaceX GSE and what not is done in house.

You read a lot into my post that isn’t there. He is indeed badmouthing SpaceX. Criticizing, second guessing, s@@t talking, Monday morning quarterbacking, whatever you want to call it, his position is that SpaceX is less competent without him.

[john smith 19]

I think he does know what he’s talking about but your point about the churn is correct.

Indeed.

When aerospace companies say "We're a safe pair of hands. We've been doing this since the 60's" my BS detectors pings like a geiger counter in an old Soviet nuclear submarine.

a)The staff from that era are either dead or retired. b)They may not have written down a lot of what they learned.

If so (and for "60's" substitute pretty much any decades since) what you've really got is "Well, we did it back then, so we know it can be done." And that's about all. 

It's called "Knowledge capture" and it's a major issue for all tech companies that don't want to keep reinventing the wheel.

All those old reports aerospace company staff wrote might have told their competitors how they did something, but they also reminded the employer how it was done, in case the author(s) left (or ultimately retired. No one lives forever)

SX is no different from any tech company that's a)Been in business decades b)Has a staff turnover rate above 0 in it's design and implementation departments (the old HP, before it was parasitised by Compaq, was pretty good at this. I doubt it's anything special today    ).

But AFAIK SX doesn't publish much and I've never seen a paper on launch pad construction techniques, but that maybe because  I've been looking in aerospace journals, not civil engineering. Maybe one of the NASA SP8000 series?

Of course if SX require all their key staff to keep a log of their key discoveries (on the corporate server, naturally) then retaining that "corporate memory" just requires a good backup regime and effective search tools. Their current team can pick up exactly where their predecessors left it.

Time will tell how many of those lessons were remembered and how many will have to be learnt. 

[ZachF]

Just going to put this out there. As a fan of SpaceX and spaceflight in general, I find it incredibly annoying how it is becoming increasingly more difficult to ever criticize ANYTHING that SpaceX does, in particular with the Starship program.

It's impossible to do so, because you will just be met with "well that's just how they operate, their mantra is fail fast and learn from it". I appreciate that this is how they operate and it's very refreshing to see how quickly they are able to move and achieve great results, but I just really dislike it when you point out an obvious oversight and people run to the defensive basically say that you aren't allowed to criticise anything they do because they know better, while totally forgetting that Elon Musk himself has admitted to making dumb mistakes at SpaceX.

For example, how in the world did they overlook that Texas regulation about LNG tanks while setting up the fuel farm? Or that they built a 12 metre wide water tank with no reinforcements. Or that they thought somehow that a ~50% thrust static fire was enough to extrapolate that the pad would hold up (once) to a full thrust launch? In the end, none of these things seem to have been showstoppers, but that's not the point nor the implication.

Look, as I said, I think was SpaceX is doing and has accomplished is incredible, and I don't want to take that away from them but anytime you say something with a negative connotation you get called a concern troll.

Perhaps personality has something to do with it. I find overly optimistic people annoying.

It has to do with most criticism of SpaceX being bad faith, some of it incredibly so… but ‘most’ does not equal ‘all’, and some forget that.

But the last point about optimistic people, we honestly need more as modern society seems faaaaaar more heavily biased towards blackpill/doomerist thinking, which is much more obnoxious and leading to civilizational sclerosis in the west.

[ZachF]

Some context; SpaceX vs the rest of the world combined in dV adjusted tonnage to space, 1957 to projected 2023

(Red is RoW, blue SX… I know there’s no labels I stink at mobile sheets  )

SpaceX has 445 adjusted tonnes to orbit so far this year, with the year 37% complete, on track for 1,200t total.

To put in perspective how impressive this number is, the number for Chinese space program during its entire existence is 2,200t, and Europe is 3,200t.

[ZachF]

Maroon: USSR/Russia
Blue: USA ex-SpaceX
Yellow: SpaceX
Green: Europe
Orange: China

[DeimosDream]

When aerospace companies say "We're a safe pair of hands. We've been doing this since the 60's" my BS detectors pings like a geiger counter in an old Soviet nuclear submarine.

a)The staff from that era are either dead or retired. b)They may not have written down a lot of what they learned.

If so (and for "60's" substitute pretty much any decades since) what you've really got is "Well, we did it back then, so we know it can be done." And that's about all. 

It's called "Knowledge capture" and it's a major issue for all tech companies that don't want to keep reinventing the wheel.

All those old reports aerospace company staff wrote might have told their competitors how they did something, but they also reminded the employer how it was done, in case the author(s) left (or ultimately retired. No one lives forever)

SX is no different from any tech company that's a)Been in business decades b)Has a staff turnover rate above 0 in it's design and implementation departments (the old HP, before it was parasitised by Compaq, was pretty good at this. I doubt it's anything special today    ).

But AFAIK SX doesn't publish much and I've never seen a paper on launch pad construction techniques, but that maybe because  I've been looking in aerospace journals, not civil engineering. Maybe one of the NASA SP8000 series?

Of course if SX require all their key staff to keep a log of their key discoveries (on the corporate server, naturally) then retaining that "corporate memory" just requires a good backup regime and effective search tools. Their current team can pick up exactly where their predecessors left it.

Time will tell how many of those lessons were remembered and how many will have to be learnt. 

SpaceX isn't just another tech company. They are an AS9100 compliant tech company whose implementation was found acceptable to be both an NSSL Phase 2 winner and fly crewed missions for NASA. They might have lost documentation on Why, but everything on How Falcon/Dragon need to build, tested, and operated will have been meticulously documented internally.

Of course, Starship != Falcon/Dragon so some wheels will have to be reinvented, some new lessons will have to be learned... and others reinvented/relearned out of an optimistic attempt to build a better mousetrap. Often scrappy vs crappy will come down to risk/reward management. The author calls out the barge wing extensions as crappy, but a low-cost risk-reduction seems pretty scrappy to me. Likewise launching without the pad plate may have led to bad PR optics, but they seem to have judged the damage risk/repairability correctly.

My scrappy/crappy question is why they launched the Starship with the full heatshield on the first flight instead of swapping the order with the next no-heatshield-planned ship.

[woods170]

<snip>
My scrappy/crappy question is why they launched the Starship with the full heatshield on the first flight instead of swapping the order with the next no-heatshield-planned ship.

Because the next no-heatshield ship had stacking clamps that were not compatible with B7. No matter what would have happened to S24, the TPS tiles would still be lost. Had S24 not flown but sent for scrapping instead, then the tiles would have been scrapped too. In reality however the vehicle flew, and the tiles were still lost (regardless of the outcome: had S24 completed the mission then those tiles would have sunk to the bottom of the ocean).
So, keeping THAT in mind meant that there was nothing to gain, nor anything to lose (TPS-wise), from switching S24 for a tile-less replacement.

[volker2020]

For me the big elephant in the room, is the assumption, that Space Flight did work so far and that SpaceX should learn from that.

While that may be true for some technical elements, for me the premise is absolutely wrong. Space Flight does not work, since it is way to expensive. If we ever want to do more in Space, the costs have to sink a lot.
So no, SpaceX don't try to replicate what others tried for decades without real advancement. Continue to blow up things, and maybe at the end we have the means to utilize Space. Otherwise Space will always be a niche market for Communication Satellites and military applications.

[Jim]

But AFAIK SX doesn't publish much and I've never seen a paper on launch pad construction techniques, but that maybe because  I've been looking in aerospace journals, not civil engineering. Maybe one of the NASA SP8000 series?

NASA has nothing to do with SpaceX construction, therefore it won't be any NASA documents on it.

edited.

[Robotbeat]

There’s a bunch of stuff on NTRS.nasa.gov on flame trench/diverter/etc design. Not necessarily *construction techniques* but on the overall design and materials, including one about an angled plate with water cooling orificies.

[john smith 19]

But AFAIK SX doesn't publish much and I've never seen a paper on launch pad construction techniques, but that maybe because  I've been looking in aerospace journals, not civil engineering. Maybe one of the NASA SP8000 series?

NASA has nothing to do with SpaceX construction, therefore it won't be any NASA documents on it.

edited.

No it won't. Directly.

But the SP8000 reports were snapshots of the SoA of various areas of rocketry at the time they were written.

If there's one on launch pads it would have said what was being done in the US (by NASA), what had worked, and what had been proposed but not yet tried.

So while they won't teach you how to design a rocket, they will teach you about what's been shown to work.

But I don't think there ever was one for the launch pad.  Solid rockets, valves, computers (makes quaint reading now of course), navigation sensors of various kinds etc. 

[john smith 19]

While that may be true for some technical elements, for me the premise is absolutely wrong. Space Flight does not work, since it is way to expensive. If we ever want to do more in Space, the costs have to sink a lot.

Correct. The price difference between sending a 20 foot ISO container 200Km holding say 2 tonnes horizontally against the same load vertically IE to LEO, is ridiculous.

So no, SpaceX don't try to replicate what others tried for decades without real advancement. Continue to blow up things, and maybe at the end we have the means to utilize Space. Otherwise Space will always be a niche market for Communication Satellites and military applications.

On that basis the question is actuall wheather or not SX is different enough to make that big a difference to launch prices, not its own internal costs* (which if this works should be substantially lower).


*To an accountant confusing cost with price makes no more sense than someone confusing stress with strain does to an engineer.

[Robotbeat]

They’ve already made a big difference in both. Also, SpaceX is also very vertically integrated, so even if prices don’t change much, a lot of things become possible that wouldn’t otherwise be possible, like bidding about 3000 tonnes equivalent of IMLEO launch capability to NASA for only $3 billion in the form of Artemis III’s lander and its uncrewed demonstrator.

[Jim]

But AFAIK SX doesn't publish much and I've never seen a paper on launch pad construction techniques, but that maybe because  I've been looking in aerospace journals, not civil engineering. Maybe one of the NASA SP8000 series?

NASA has nothing to do with SpaceX construction, therefore it won't be any NASA documents on it.

edited.

No it won't. Directly.

But the SP8000 reports were snapshots of the SoA of various areas of rocketry at the time they were written.

Of NASA funded research

[Jim]

While that may be true for some technical elements, for me the premise is absolutely wrong. Space Flight does not work, since it is way to expensive. If we ever want to do more in Space, the costs have to sink a lot.

Correct. The price difference between sending a 20 foot ISO container 200Km holding say 2 tonnes horizontally against the same load vertically IE to LEO, is ridiculous.
 

Wrong, it isn't.  It is energy expended and not distance.   Fly that 2 tones at 2,400 kph for that 200 km

[envy887]

What i got out of the entire post. The team should have known all the things

That's also what I got out of the article, and it's a fundamentally flawed argument because I see no evidence that SpaceX would have chosen not to launch even if they DID know everything they "should" have known was going to happen with the vehicle and launch mount. There were no safety issues, vehicle anomalies were obviously expected, and fixing the launch mount post-launch won't take any longer than fixing it pre-launch would have.

The only exception is the FTS underperforming, which wasn't confirmed at the time of writing and only got a passing mention as a possible anomaly.

The argument starts with the assumption that the foreseeable results of the flight were an unacceptable result from SpaceX's point of view. That assumption is invalid, so the entire argument falls.

There's a related argument that the foreseeable results SHOULD have been unacceptable to SpaceX, but that argument isn't presented in the article, and is much more difficult to support with facts rather than opinions.

[Vultur]

I agree. At the current rate of repair, the pad damage will be fixed and the steel installed before they are ready to fly again anyway (since they need to wait on FAA approval of the FTS changes for that). So the pad damage won't affect schedule.

Since the water cooled steel protection was already planned, it seems like the only thing the damage actually cost SpaceX was the cost to repair the concrete... which can't be terribly significant on their scale.

[abaddon]

Not sure I agree completely with that; SpaceX will be running static fires that will be enabled by the new pad cooling and will be important risk reduction for flight two.  They don’t need a flight license for those tests, so the pad repairs plus pad qualification plus static test fires might be a longer pole than the flight two license.

[Robotbeat]

Not sure I agree completely with that; SpaceX will be running static fires that will be enabled by the new pad cooling and will be important risk reduction for flight two.  They don’t need a flight license for those tests, so the pad repairs plus pad qualification plus static test fires might be a longer pole than the flight two license.

That’s a good point, but FTS is a much bigger deal.

Additionally, they’re actually on track for achieving the flame diverter on the same timeline as they’d have had without even launching IFT.