What's absurd is the notion that a bunch of bulk hydrocarbons and liquid oxygen would cost anything close to the high-performance aerospace hardware containing them. The propellants are mostly LOX by mass, and that's literally cheaper than dirt. The remainder is RP-1, a kerosene blend similar in cost to jet fuel, or liquid methane, easily separated from natural gas. There's no shortage of references out there showing the Falcon 9 propellant costs are about $200k-400k per launch. For Starship, it'll be less than $1M...possibly much less depending on the bulk pricing they can get, methane being cheaper than specialty kerosene blends.

If you won't take Elon's word for it, here's a graphic from ULA laying out the typical costs: https://twitter.com/torybruno/...

Never mind that it was clear long ago that F9/Heavy would be successful and your "proven" SLS still hasn't launched, the original proposals for distributed launch architectures used EELVs: Atlas V and Delta IV. Reuse and orbital propellant transfer are not necessary, orbital assembly is proven, and the old "multiple launches" excuse, really? SLS payload to LEO is a fiction. SLS payload to NRHO, co-manifested with the Orion/ESM that are required to actually deliver the payloads there, is...guess what...the same 10 t that Falcon 9 v1.0 could manage to LEO.

SLS requires multiple launches and orbital assembly, it just does them far more slowly, at vastly greater cost, and in a more risky location.

This is a contract *modification* adding engines to an existing contract. R&D and production line startup were in the original contract along with 6 engines. The modification adds 18 more engines at ~$100M each, or $400M for a full SLS core's set of engines.

https://arxiv.org/pdf/2108.002... page 9.

Looks like each edge is two AD5220 digipots, two comparators, and an XOR gate and a flip-flop to increment or decrement the digitpot settings at each training step based on the applied and training signals. No manual intervention or computer control, it's a hardware implementation of a simple training rule.

> that's 50% more than Falcon Heavy in *expendable* mode

At many times the cost of two Falcon Heavy launches. Even in 2011 they had to pretend doing multiple launches was unacceptably risky to justify SLS over Falcon 9 or even the EELV launchers.

And they of course then designed an architecture composed of 10 t components assembled over multiple launches, because that's what can be co-manifested on a launch to NRHO along with an Orion. The cost of doing things this way has resulted in a single Falcon Heavy now being contracted do the work of two SLS launches by launching the Gateway PPE and HALO elements together.

...no. Even for the Falcon 9, propellant costs are about half a percent of the total. Reusing the first stage just once saves enough money to pay propellant costs for about 100 Falcon 9 launches. Even Starship's propellant costs will be less than 10% of the total. Propellant is cheap, aerospace hardware is not. Neither are gigantic hypersonic maglev launch tracks.

SpaceX put about 335 t of payload into orbit last year, more than anyone else by a wide margin: next up was China at 180 t. They did that with 31 launches, only two of which used new boosters. You can't pretend they're some kind of scam or don't know what they're doing any more.

No, SpaceX's plan is to mine water ice and get the hydrogen from that. Hydrogen is bulky and difficult to store for long periods, they'd be sacrificing hundreds of tons of useful equipment and supplies to carry it instead.

You're producing rocket propellant from its combustion products. Of course its going to be energy intensive, that's what the multi-megawatt solar farm is for.

"And LEO is closer to the Moon than to the Earth, when you look at energy/fuel required to get there...."

Only if you ignore the energy/propellant required to stop there. It takes about as much to brake into low lunar orbit as it does to go to Mars, where you can use the atmosphere for braking and only need enough propellant for a short landing burn.

And of course, lunar propellant requires a massive lunar propellant mining/processing facility, as opposed to just sending up a tanker from Earth. Lunar propellant is mainly going to be useful for lunar operations, it's of little use for Mars.

Er...NGSS doesn't have any heavy lift experience. The closest they get is that Northop Grumman bought the company (a few mergers removed) that made the Shuttle boosters. They're making the SLS boosters as well, but the core that does most of the work is Boeing's. For a while they worked on a Shuttle-derived medium-heavy vehicle, but gave up on it when the USAF decided not to use it.

ULA and SpaceX are the only US companies with currently operational HLVs (SpaceX having the only operational superheavy launch vehicle), and ULA's Delta IV has only a handful of launches left which are all spoken for (plus they didn't actually develop it).

And there's nickel-iron meteorites constantly raining down on Earth. A 2.9 t battery pack's a drop in the bucket.

He pulled political strings to get himself on the Shuttle and was nicknamed "Ballast" by the crew. He also helped make sure Constellation got resurrected as SLS and was fully funded while Commercial Crew struggled along with half its requested funding, in fact arguing for devoting all of the Commercial Crew funding to SLS.

What do you expect an investigation to show?

"These placeholder legs aren't very good."? You need an investigation for that?

"It landed a bit hard." And one of the main reasons for doing the test flight was to gather data to refine the landing parameters. They got that data.

Again, this is a test article. Much of the stuff on there is the minimum sufficient to support the tests, and will be replaced with something actually intended for regular operation in some future prototype. They only care that it does its job well enough for a few tests. They obviously don't *want* it to explode afterward, but it's not a major setback if it does. A lot of these systems are not redundant, are not robust, are not reliable, because designing them to be these things will take time, and there's no reason to hold up testing that can yield data that informs their design.

It's not a "priority 1 discrepancy", there's no disruption to business or operations. It's a test outcome that they already anticipated and prepared for, as evidenced by the prototypes already queued up for further testing.

This is not an operational vehicle, it is a test article. It's meaningless to talk about "discrepancies", most of the systems required for flight exist only in very early and basic forms, and many required for actual operation don't exist at all yet. There's a near-identical one already on its way to the pad because they expect to lose them in testing...rocket flight involves narrow margins and a great many things that can result in loss of a vehicle. They actually planned for several more, they scrapped SN12, SN13, and SN14 early in construction because after SN8, it was clear they wouldn't need so many.

Yes, I'm sure. Your "only 3260 m/s" is almost half the total delta-v of a fully fueled Starship, and almost 50 times the propulsive delta-v needed for the landing.

Multiple passes through the upper atmosphere take time, aerobraking slowly enough to avoid the need for heat shielding could take literally months of slowly dropping apogee down through the radiation belts. The same problem applies to low-thrust electric propulsion systems. If you're carrying people, you want to directly reenter and land.

As for nuclear-thermal, it doesn't allow dramatically faster trips (orbital refueling will allow Starship to make the trip to Mars in around 90-120 days), and in fact much of the performance advantage is eaten up by the need to propulsively brake at the destination. It very much does not allow ignoring the launch windows.