Continuing discussion of SpaceX launches [edited title]

Miscellaneous and Personal Stuff I Must Share
621

Currently, most satellites that you would use for internet are in geostationary orbit. This is useful, because they stay in the same place and you can point your dish at them, and they are far enough away that one can really cover an entire continent. This has a downside in that the are orbiting at 24k miles. This means that you need a stronger transmitter receiver on both satellite and the ground, and also introduces noticeable lag from light speed limitations.

A large constellation of satellites in lower orbit will solve those problems, unfortunately, they will be moving, so you can’t just point an antenna dish at them, and they will be closer, so they won’t have as much coverage, and you will need more of them. Keep in mind, they will be spending about 3/4’s their time over ocean, not all that much population (though not zero) there.

Only to feed people in orbit. The cost of growing in orbit would be far higher than anything that can be done on the ground, so growing it to send to earth would be financially ludicrous. OTOH, getting stuff from earth to orbit is also expensive, so it may be that you can find a way to grow food in orbit cheaper than you can launch it, but that only benefits those who live in space, currently, a very small demographic.

It’s not all that resource rich. Plenty of oxygen if you are willing to split it from the regolith, but nothing too special outside of that.

622

That’s not really true. For instance, the moon has billions of gallons of water, sitting in a gravity well 1/6 as deep as Earth’s. If we move into space in a big way, lunar water could be a very lucrative thing to mine.

The moon is also covered in meteor inpact sites. billions of years worth. On the Earth, meteorites usually burn up in the atmosphere, or if they do impact, natural geologic processes erase them and bury the actual meteorites. We may find that meteorite prospecting on the Moon is quite lucrative.

Then there are the Lava Tubes. Not only are these potential real estate sites for millions of people (the Moon’s crust is 12% void spaces), but we may find sealed lava tubes filled with all sorts of valuable volatile chemicals, or even tubes filled with liquid water or ice. We really aren’t going to know until we go look.

The lunar reglith is also full of the same kinds of material the Earth’s crust has. Aluminum, Iron, Magnesium, etc. The difference is that A) again, these materials being there means we don’t have to transport a lot of mass from Earth, and B) the location may make it easier to extract the materials. You could make a smelter on the moon with just s big solar reflector focused on a target. regolith goes in, gets smelted. The volatiles bubble out and sublimate away in the vacuum, and a simple centrifuge could separate out the solids by density. All the stuff we have to do on Earth to prevent pollution would be unnecessary, and the final products would already be on the moon for use in constucting habitats, building spaceships, whatever.

The moon likely has trace amounts of all the valuable elements we have on Earth, but the low gravity and vacuum environment might make them significantly easier to mine than on Earth. A big meteor hitting the moon would vaporize and distribute its heavy metals and rare earth elements across a large area. Repeat for billions of years, and there might be all kinds of surprises waiting for us when we start to seriously explore the moon.

Again, you have to completely change your thinking in an era where we can put a 10 ton lander on the moon for half a million bucks. Things that were not cost-effective in the era of $10,000/kg payload prices take on a whole new value when the cost drops to $100/kg. Or $10/kg.

623

My esteemed doctor: you were right and I was wrong.

I talked to a former SpaceX employee, who I would say was in a high enough position to know about these things, and I got set straight. The F1 cost about $90 million to develop, the F9 cost about $300 million, and Dragon was somewhere in the $460 million range (though we didn’t talk about this as much). So I was definitely wrong on that.

We also talked about NASA’s contracts, and he said that at one point NASA was seriously considering terminating SpaceX for default. NASA believed SpaceX did great on meeting milestones for things like raising private investment, but was doing terribly on meeting any technical milestones. There was a mad scramble inside the company to turn things around, as a termination of the contract was viewed as being the end of the company. I don’t recall from the conversation when this happened, but I’m thinking around 2008? And it was also perfectly clear that other NASA technical assistance was extremely valuable.

So from this formerly inside source, I’m pretty confident that I was entirely wrong on SpaceX under reporting it’s development costs, and totally correct on how much SpaceX needed NASA’s help - and money - in the early years.

624

It was December 2008 when SpaceX won a huge NASA contract that basically saved the company. source

ETA: my impression is that Elon Musk is extremely grateful to NASA for this support that arrived at such a critical time for him and his company

625

Thanks for digging into this! My sources have been just NASA documents and SpaceX public statements, so it’s nice to see some insider confirmation.

I don’t think I’ve ever denied that SpaceX benefited immensely from NASA support. And while we’ll never know if a terminated contract would have truly been the end of SpaceX, I think it’s clear that they would not have survived in their current form. The NASA contracts continue to be of immense (mutual) benefit.

And yes, the technical assistance was also valuable. As I mentioned earlier, the SpaceX PICA-X heat shielding material was derived from NASA’s PICA material, but optimized for cost and performance. I think this is one of the areas where NASA’s influence shines and it hearkens back to the NACA days.

If only NASA could stop getting dicked around by the changing administrations, and focus on what they do best–exploration and research. Not driving trucks to space.

626

He is. One example:
http://shitelonsays.com/transcript/elon-musk-youre-safe-if-my-spacecraft-crashes-2014-06-11

Or:
http://shitelonsays.com/transcript/making-humans-a-multiplanetary-species-mexico-city-2016-09-27

627

Elon Tweetstorm today.

On the naming of a new East-coast droneship (yet another Culture reference):

On why the center core landing failed (I assume the “pretty obvious” solution is “add more igniter fluid”):

On partially reused flights:

Throwing shade at ULA’s plans for Vulcan + Centaur:

628

Nice interview with Gwynne Shotwell:
2017 Satellite Executive of the Year: Gwynne Shotwell, President and COO, SpaceX

Musk gets credit for driving the technical and long-term vision of the company. But Shotwell deserves all the credit for actually running the company. They would not have recovered from the AMOS-6 failure, or launched 18 missions in 2017 (5 of them on reused boosters), or racked up a backlog of $12B in orders without her.

629

nm

630

As I think I pointed out earlier, and as I think most of us here know, three-engine retroburn for landing was a novelty. My suspicion is that someone missed a red-lined process doc or checklist and therefore loaded the center core with only enough TEA-TEB for a single-engine return, as would have been customary before SES-16’s (excessively) successful return.

This was almost certainly a Mission Assurance failure, to use NASA parlance.

631

That was my original thought as well–and it’s still a likely scenario, I think. However, at one point (maybe during the post-launch conference) Musk mentioned that the center core went through “several” three-engine burns. Normally a core just does a boostback, a reentry burn, and a landing burn. Perhaps he misspoke, or isn’t using the word “several” the way I would, but two burns prior to landing doesn’t sound like “several” to me.

Perhaps there’s another explanation: since the center core was on a trajectory they haven’t attempted before (higher and hotter than normal), there may have been more burns than normal. Possibly even a variable number based on how the control systems are behaving.

SpaceX talks about their success rate based on “sigmas”, which is just another way of saying that there is some nominal behavior, and then some probability of being a certain distance away from that. 2-sigma represents a 5% chance, so (for example) they may say that based on the variability in weather, or instrument error, or otherwise a 20% increase in propellant use is a 2-sigma event. If they give the booster a 30% propellant margin then we would know (again, just an example) that we have a greater than 2-sigma, or 95%, success rate.

Well, it may be that igniter fluid follows the same pattern. Perhaps they give enough for a nominal four re-ignitions and that usually they only need enough for three. But then there’s some probability of some ignitions using more than their share, so that there’s only really enough for three. And then some probability of needing an extra correction burn, which eats another. And if you get unlucky and these things happen together (whether a 2-sigma, 3-sigma, or greater event), then you run out of fluid and crash.

So it could well be that it just got unlucky in some fashion. That said, since the fluid doesn’t take much volume and they aren’t really constrained on first-stage mass, it’s trivial to increase their odds by adding a bit more fluid, which would push them out a couple more sigmas worth.

632

Would a booster do more than one burn launching the payload, and do several on re-entry?

633

That’s a good point. They have, given full consideration of what they knew in advance of an edge-of-the-envelope situation, under-estimated consumables before.

634

Nice cite. I’d forgotten about that one. Yeah, I’m thinking it may be the same kind of deal–they put what they think is enough in, plus some margin, but either they got unlucky or their modeling is off and it wasn’t enough.

Since they didn’t really care much about recovering this booster, maybe they just didn’t bother modeling it that accurately. It’s not even representative of the final config, not being a Block 5.

There’s just one for the launch–that’s easy to see from the feed. And all previous landings have had exactly three additional burns: boostback, reentry, and landing. However, it’s possible they would need more under some circumstances. For a particularly hot reentry, they may have to perform two. Or it may be more efficient in some cases to split a single burn into multiple (generally it’s best to do just one, but heating or aerodynamic forces may dictate otherwise).

That said, intuitively I’d expect their control system to work with the three different burns and just vary the parameters on those, but we really just don’t know.

635

A nice Falcon Heavy “trailer” that happens to include some previously unseen footage of the center core crashing near the barge (1:10):

Ouch! But their strategy of aiming the stage off-barge before the burn seems to have paid off; they lost the stage but not the barge.

636

Emphasized text added for context, since I’m bump-quoting a post from January.

It’s official: Northrup Grumman’s modified payload adaptor is the reason the classified Zuma spacecraft failed to separate from the Falcon 9 upper stage, leading to the loss of the spacecraft. And NG’s original snark at the time of the failure blaming SpaceX is looking shit-headed and petty now.

637

I would say that boot-licking industry toadie Loren Thompson should be ashamed of his original reporting of Zuma, casting the usual FUD at SpaceX, but boot-licking toadies don’t have shame. I love how he allows that it might have been Northrop Grumman’s payload adapter that failed, but then immediately goes on to say that it’s really SpaceX’s fault anyway.

638

What snark? I recall that NG didn’t comment at all, and SpaceX ran out to declare their innocence.

ETA: For example: "“This is a classified mission. We cannot comment on classified missions,” Lon Rains, communications director for Northrop, said in a statement to The Verge. "

639

I may have conflated Northrop-Grumman proper with wholly-owned-subsidiary-of-Northrop-Grumman commentator Loren Thompson, as the good doctor mentions.

640

In (what I can only imagine is) an extra bit of salt in NG’s wound, SpaceX is planning to reuse the first stage from the ZUMA mission to launch the next few Iridium satellites next month:

ETA: kind of like an extra “we told you our rocket was fine”