View Single Post
Old 03-14-2018, 09:57 PM
Sam Stone Sam Stone is offline
Join Date: Jun 1999
Posts: 27,128
Originally Posted by Stranger On A Train View Post
A few points to address:

The militarization of space may, in fact, be inevitable, and it would behoove the United States to be cognizant of what other players are doing in space and prepared to respond to it, but as previously noted, the Department of Defense already has a significant space component in the United States Space Command (USSPACECOM) and Air Force Space Command (AFSPACECOM) which provides space surveillance, space launch operations in the Evolved Expendible Launch Vehicle (EELV) and Rocket Systems Launch Program (RSLP), operation of the Global Positioning System (GPS), Defense Support Program (DSP), and various other mission support and readiness efforts. Creating an entirely new branch to cover this existing capability is nothing more than a provocational move which signals that the US is not serious about preventing military buildup in space.
You may be correct. I am not an expert on the relative efficiencies of having a separate space command, which is why I said I have no opinion on that particular subject. However, I will note that people whose opinions I respect are against it, so I would defer to them - which includes you as being much more knowledgeable on the subject.

And to be very clear about this, orbital space is a potentially burgeoning area for commercial development as well as providing services such as global satellite Internet access, weather and hazards monitoring, and other services to peoples around the world; military buildup and inevitable conflict in Low Earth Orbit (LEO) threatens that development and could potentially make whole swaths of LEO unusable for an indefinite period fo time.
Are you talking about the Kessler syndrome? Or just overcrowding in general? Musk is currently planning to put 4300 satellites in LEO to provide global high speed internet.

This was kind of my point - with the cost of space access dropping dramatically, the size of the space industry, both civilian and military, can be expected to grow in proportion. Supply/Demand curves, and all that. What was unfeasible or not economical at $10,000/kg launch costs may not be at $500/kg. Or $50/kg.

We have the opportunity now to make agreements with other nations to avoid such a pointless and destructive development that will benefit no one...or we can bluster and arm, denying future generations the birthright of access to orbital space and beyond.
This is where we disagree. Agreements do not stop countries from following their own interests. We have had international agreements before regarding space commercialization and military use of space - and the only reason countries entered into them is because they didn't think they mattered, and were good PR. Kind of like the Kyoto treaty. And like Kyoto, as soon as these opportunities DID arise because of the advance of technology, the agreements were quickly ignored. The same can be said of the nuclear agreements with North Korea and Iran. They abided by them for as long as they couldn't do it anyway. The second they could, the treaties went bye-bye.

If the U.S., China and Russia are launching secret military missions, then as soon as access becomes more affordable more players will join in. That's the way it's going to go. Just as the nuclear agreements North Korea signed became null and void the minute they actually had the capacity to build them. The Soviet Union violated the ABM Treaty almost immediately, and both the Soviets and Americans were in violation of Intermediate-Range missile treaties whenever it suited them.

The Delta IV Heavy does not cost “$600 million per launch”; even for a FAR Part 15 EELV launch which includes full access to all build information for mission assurance and independent verification and validation (IV&V) processes the contract cost comes out to about US$435M.
Numbers vary, depending on the type of mission, etc. The point is that Delta IV and pretty much all other launchers are significantly more expensive than Falcon 9 or Falcon heavy on a per-KG to orbit basis. Even the Chinese have stated that they cannot compete on price with SpaceX.

While this is still far higher than it should be—I worked on a study back in the early 2000s that found that EELV launch costs could be reduced to 60% of the negotiated contract value including a 15% profit margin while still maintaining the same level of mission assurance and IV&V—and I’m morally certain that even with the add-on costs SpaceX can launch a Falcon Heavy for well below that cost (especially since they are on a FAR Part 12 contract with fewer reporting requirements), Boeing and ULA have had a nearly spotless record with the Delta IV with only a moderate anomaly on the inaugural flight of the first heavy configuration. And because of competition with SpaceX, ULA CEO Tory Bruno has been forced to admit that they can do launches as half the cost and has subsequently negotiated down.
One of the reasons I mentioned that space industry execs were watching closely is that I detected a definite shift in tone in the military and space press from other execs after Falcon Heavy. For example, even after Falcon 9 had landed a few boosters, I still saw ESA talking heads saying that they weren't worried because their upcoming Vulcan rocket would be more efficient and cost competitive. After Falcon Heavy, the tone changed more to, "Are we doing enough? How are we going to compete if our rockets aren't fully reusable?" And frankly, the idea of just jettisoning the engines and catching them by helicopter sounds like a half-assed Rube Goldberg scheme after we've seen entire boosters re-land right at the launch site.

As for reliability, consider the difference between a new factory aircraft and one that's been in service for a few hundred hours. The factory aircraft has to be flown by a test-pilot to shake out manufacturing errors. You can't test-fly a rocket other than lighting it on the pad. So instead we need highly expensive extreme inspection regimes and extreme manufacturing controls.

But theoretically, once someone starts flying the same rockets over and over again, they will collect data on common failure points and excessive wear and iterate their designs, and reliability should go up dramatically. When the rockets are thrown away, it's hard to collect data on how they wear. For example, SpaceX found some micro cracks in their engines, and updated the design as a result. Their comment was, 'for all we know, many rockets flying today have the same problem, but there's no way to know if you can't recover the engines.' Reusability should translate into higher safety over time.

Delta IV has had 35 missions, with one partial failure. Falcon 9/H rockets have flown 51 times, with two failures. The latest variant, the 1.1 Full Thrust version, has 31 flights with one pad failure. Which one is safer? Those numbers are way too thin to draw any statistical conclusions from. But at the rate Falcon is flying, if it doesn't have another major incident SpaceX will reasonably be able to say it's the most reliable rocket in the world. Time will tell.

More impressively, other than the center core of Falcon Heavy, which was an extreme performance experimental landing attempt, SpaceX has successfully landed every booster they've attempted to land since Falcon 9 Full Thrust started flying.

As far as the economics of the Falcon Heavy and the Big ‘Falcon’ Rocket, these are far from proven, as is the fiscal viability of reusability.
Well, that's a pretty big statement. SpaceX has internal data on their costs, and they are going full steam ahead on BFR and suspending development of the smaller rockets. They wouldn't be doing that if reusability was a marginal savings, as if it is the company is toast. If BFR flies for even half the cost of say, SLS Block 1, it'll be a disaster for SpaceX. So just from that, I'm going to assume that they're seeing real cost savings. As in, an order of magnitude or more for each reusable section.

As another data point, SpaceX has said that any missions scheduled to fly on a Falcon 9 that require enough Delta V to require expending the rocket will be moved to Falcon heavy in recoverable mode. That tells you that they are saving at least enough to warrant using a much bigger rocket than is technically required, just so they can recover the boosters. They're still throwing away the upper stage regardless, and yet it's still cheaper for them to launch a recoverable Falcon Heavy with three boosters than an expendable Falcon 9 with one. What does that tell you about what they expect to save from reusability?

Finally, with respect to how much effort it takes to re-fly and how well the boosters are surviving, SpaceX as re-lit one booster on the pad just one day after its return, it has re-flown five boosters including the side boosters on Falcon heavy, and all performed nominally. And these were the older boosters without the structural mods to make them more reusable.

SpaceX now considers landing boosters to be a routine part of flight, and no longer considers them experimental. I can't imagine that landing a booster and re-flying it without refurbishment is anywhere NEAR the cost of building a new one. I'd be shocked if it was more than 10-20% of the cost of a new one.

The quoted cost of US$90M for a Falcon Heavy is not the rolled up final flyaway cost; it is the base manifesting cost to by a launch slot on the schedule.
Sure. But that's an apples-to-apples comparison with other launchers. Of course each mission is different and will come with different costs.

Any real payload will require multiple coupled loads analysis (CLA) cycles, some degree of payload processing, likely health and status monitoring while on pad, and various other operating costs on the part of the launch contractor.
Yes. But... If the space industry expands dramatically, we can expect more 'plug and play' satellite designs. Also, the amount of engineering we do on satellites to ensure they function properly is somewhat tied to launch costs - No one wants to blow a $400 million dollar launch because a ten cent part failed. But if launch costs drop dramatically, we should see payload costs drop as well both due to economies of scale and standardization, and because you can afford to engineer to a lower spec if failure costs 1/10 as much as it used to.

The BFR is still unflown at this point and there is no way to evaluate what its final flyaway cost will be, but given that Falcon 9 and Falcon 9H/Heavy costs grew by close to a factor of two, while the advertized savings from the reuse of Stage 1 dropped from an initial 90% to a very modest 10% discount, I don’t think there is good reason to put a lot of faith into the public estimates offered by SpaceX until there is sufficient flight history for the costs to be evidenced.
Falcon costs grew, but so did Falcon capability. The original Falcon 9 was a 318 ton rocket with the capability of putting 3400 kg in GTO. The current Falcon 9 is a 549 ton rocket that can put 8300 kg into GTO in expendable mode. That's more than double the performance of the original rocket. On a per-KG of payload basis, the cost of Falcon 9 is actually lower than the original estimate.

As for the 10% discount... That's not surprising initially. The block 3 and 4 boosters were not really designed for multiple reflights, and were best thought of as test vehicles. SpaceX had no way of knowing if they could be reflown once, twice, ten times, or not at all without heavy refurbishment. Therefore, the discount was small.

Also, previous Falcon 9 landings landed on barges until recently, which adds the cost of the barge recovery, transportation, cleaning salt spray, etc. As a result, the remaining inventory of block 3 and 4 boosters are going to be expended and replaced with Block 5. My understanding is that Block 5 boosters have incorporated the numerous changes required through inspection of previously landed boosters, and SpaceX expects to fly them 10 times without any refurbishment, and 100 times in total before retiring them.

Once that happens, then we'll really be able to see what reusability does to the cost of space flight. But I am very optimistic. Falcon 9/Heavy are not the greatest platform for reusability, as they still lose the second stage and payload fairings, and still have to put the rocket together from many pieces. BFR, on the other hand, will be fully reusable, right down to the payload fairings.

And even a 10% reduction on a rocket that can only reuse half of itself is still pretty impressive. I think Musk has said that the flyaway price difference between a reusable and expendable Falcon Heavy will be about 30-40 million dollars, or not quite half the cost of the entire rocket. Given that they are recovering only 2/3 of it, that must mean the block 5 boosters are expected to fly for maybe 10-20% of the cost when reused.

I remain dubious that we will ever see conventional staged rocket systems that are “almost completely reusable, have fast turnarounds, and be highly reliable like aircraft,” simply by virtue that the loads and environments, hazards, and criticality of space launch operations are inherently greater than those experienced by commercial aircraft.
I think you might be biased by your long experience in the current space industry. There's no reason why rockets have to be that expensive. We can put a cargo container on a train, get it to a port, put it on a ship across the world, put it on another train, and then move the material onto trucks and deliver them for pennies on the pound. I understand that rockets operate in high stress environments, but so what? So do aircraft.

In some ways, the flight envelope of rockets is much more predictable than aircraft, which can fly through unexpected turbulence, make really hard landings, fly in all kinds of weather... If we can build an Airbus A380 and fly it for decades, we have the engineering capability to build a rocket that can fly a dozen or a hundred times. One of the things that made them so hard was the need to shave off every single pound, coupled with the fact that every launch was a one-off using a newly manufactured rocket.

SpaceX has demonstrated the ability to do pinpoint landings. I don't think I've seen one yet that missed the center of the target by more than a few feet. That means they can land them right where the crane is to lift it back to the launch area. Musk says he thinks he can land BFR right back on its pad, but we'll see.

Although responsive and highly reliable space launch capability is achievable, in my opinion, it is going to take some far greater innovations in launch vehicle design and operations, including fundamental advances in propulsion systems and materials.
So what is it in SpaceX's design for BFR that won't work? Because all the components required already exist, or are currently in development. SpaceX has already successfully test its huge carbon fiber tank, and the new Raptor engine has been test-fired repeatedly. They have a heat shield that is non-ablative except in extremely fast entries such as at Mars. So what are they missing?

I've looked at the BFR design in as much detail as is available, and I will say that one of the frustrating things about Musk is that he comes up with good ideas, but the details often escape him. He totally underestimated how hard it is to set up a modern auto assembly plant, his Hyperloop concept is fatally flawed, etc. But when it comes to SpaceX, he's pretty much delivered on everything he promised. Sure, the schedules often slipped, but on the other hand in the end they delivered more rocket than they promised. But I'm fully willing to believe that he is handwaving away some real challenges with BFR, and those challenges will raise the cost and delay the thing.

Even if SpaceX fails (i.e. they build BFR and it explodes. They build another one and it explodes...), someone else will pick up the gauntlet. Jeff Bezos has his own reusable rocket program, and he just announced that he's going to sink his Amazon 'lottery winnings' into it. Depending on how you read that, that could be anywhere from 20 to 50 billion dollars. SpaceX to date has spent about 1.5 billion on R&D. Imagine what Bezos might accomplish with 30 times that money. He'd be limited only by the availability of quality people to work on it. His New Glenn Rocket will be somewhere between Falcon Heavy and BFR in size, and will be mostly reusable. It's supposed to fly within a couple of years.

My take on BFR is that it will probably take years longer than he promises, and will likely cost more than what he says it will cost. His idea of launching for 1-5 million dollars may eventually be achievable in the fairly distant future after development costs are amortized away and assuming the thing can fly 100 times before retiring, but for now it'll be at least an order of magnitude more expensive. But even if he can launch this thing for the cost of a Falcon 9 (say, $60 million), that's under $1000/kg. Some estimates (not SpaceX) are saying more like $7-15 million per launch if fully reusable, which would get the cost of mass to LEO down below $100/kg. That's astounding.

Or look at it this way: It cost $56 billion and thirty six shuttle launches to get the ISS into space. If we had BFR then, the station could have been lifted in two or three flights, and wouldn't have been constrained by the size of the shuttle cargo bay. We could have put that thing in orbit for less than a a hundred million dollars.

So why do you think it might not be feasible with the current tech SpaceX has? What am I missing? Forget Musk's $1 million per launch number. Let's say it's $50 million per launch, which would still revolutionize space access. Why is that not possible, or at least not likely?