RCgothic

Without taking sides in the topic overall, SpaceX just has a better PR department. Their streams are better hosted and in better quality than NASA TV, and they do a lot of their activities in public where we can watch the progress step-by-step in live HD as it happens rather than in still frames weeks after the event after security has cleared the photos for ITAR. Also in the areas of human spaceflight and rocketry (arguably the coolest areas) SpaceX is free to innovate its products to produce things that are modern and exciting, whereas theses days NASA is not. Now that's not an entirely fair comparison, but it is why it's easy to be a fan of SpaceX.

A kilometer-long tether would strain about 3m per strand under load, so that's a pretty effective load-equaliser even without resorting to a tension equaliser, which I have personal experience of designing for an industrial 4-strand lifting application. More strands is just more equaliser complexity, but the equaliser's probably unnecessary. The day/night cycle is an interesting point, but what matters is the rate of change. 7.5m is only a big deal if it takes just seconds to equilibriate, which sounds unlikely.

A gradual start would not produce shocks of the order 1g. Particularly not with a tether which, as you've pointed out, is reasonably elastic. Roughly 3m extension over 1km for a 20x 16mm diameter strand ribbon for 3x Starships as I suggested calculated above. Yes, there may be some small oscillations. Tangential velocity at 1g at 1km is ~70m/s. Assume the spacecraft has a 1 deg deviation from tangent and hits the end of the tether. Radial velocity is ~1.2m/s. Kinetic energy is ~187kJ. Extension is ~ 0.7m. Peak acceleration is ~0.23g. That's a very big oscillation because we're starting from full speed. Very likely oscillations would be damped out by the time it got up to speed. Most would occur as the slack in the tether is taken up at under 1 tenth that velocity. A small damping shock absorber and a brake are sufficient. The motor doesn't need to see any dynamic load at all.

Please explain? The tether can be paid out slowly and then gently braked to a stop. No excessive acceleration required. Start up of the rotation is as gentle as the thrusters that initiate the spin up manoeuvre. This is generally very much less than 1g. A Dragon capsule masses ~15t. A Draco thruster provides 0.04tf thrust. Even as a combination together, this is very much less than 1g. There's no need to do anything quickly. In fact falling objects is a very good reason not to.

There's no reason a gradual and coordinated startup would produce any sort of whip.

No it absolutely shouldn't. Conservation of angular momentum would spin a 2 starship system up to over 60RPM as the tether retracted. 2 Dragons would be much worse due to the closer finishing position - over 10,000RPM. The system needs to be stopped for deployment and retraction, in which case the winch motor only needs to be strong enough to make the tether coil up neatly (which is not zero force - wire rope prefers to be straight). Also I'd also say the tether is not such a significant mass that having a lighter one would be significantly less propellant expended for recovery. But it is relatively trivial to have double or triple redundancy. Quickly get to a situation where failure is not credible. We regularly trust lives to wire rope systems. Elevators, Cranes. It's not a big deal. The only novel factors for space travel are mass budget (which is becoming less important), and vacuum operation, which is just double checking material properties of the tether are appropriate.

A standard steel kilometre long 1x19 wire rope 10mm in diameter would mass half a ton. It could sustain a minimum breaking load of 10t at each end, of which self-weight would contribute 0.125t. With a safety factor of just under 2, there could be a 5t module at each end, a 10.5t spacecraft total. Alternatively a quad-tether of such wire ropes could sustain 2x10t modules with multiple redundancy for 22t. Tethering two 250t loaded Starships together would take a ribbon of 20 16mm 1x19 wire ropes for a safety factor of 2, and would be very tolerant of individual rope failures. They'd weigh 25t (total spacecraft weight 525t). Stronger ropes are available that would reduce the number of strands required at the cost of less redundancy, but flexibility and compactness of the reeling mechanism becomes an issue with larger diameter ropes. Although very large amounts of redundancy are probably unnecessary - multi-strand wire ropes are already self-redundant, and if any failures are detected you'd just reduce the speed of rotation of the combined craft. It's also quite unlikely that a full 1g is required for crew comfort. So tethers are actually quite easy with standard engineering materials if you have the mass budget for the mechanism.

I had a really vivid dream last night (pretty rare for me) that Superheavy launched, cleared the tower, but came down on the beach moments later. Coming on here in a still half-awake state to reassure myself that no, none of that was real, it wasn't even stacked with starship let alone licensed for a launch. (Also I live in the UK, not Texas. There is zero chance I'd be walking around the aftermath). The subconscious can be a strange place sometimes.

So what? Once starship hits subsonic the peak heating is long past. In the subsonic regime it only cares about drag for terminal velocity, not L/D, for glide slope. Also, 3 SL raptors can hold up 6 times Starship's weight. Shuttle's wings peaked at 1.7Gs. Starship definitely has a lot more margin on landing.

Maybe. That's much more important for a vehicle that glides to a landing.

Starship is much less dense than shuttle. It doesn't need as much lift. Nor did shuttle need as much lift as it had. The enormous wings were for cross-range capability, not re-entry performance.

SLS Block 1 payload isn't actually that bad for me. It can send Orion to the moon and in these days of rendezvous that's all it really needs to do (maybe a little more upper stage thrust would be nice). What really gets me is the opportunity cost. SLS/Orion is just so horrendous. Nevermind the cash price which could be better spent on payloads or commercial launchers, I want us to get off world to stay. That requires more than one crewed flight per year and Orion/SLS is completely incapable of meeting the required cadence. It locks NASA into that cul de sac until it's cancelled. They're going to get overtaken by commercial space.

So some guy on Twitter claimed in response to my listing the above that B1B is 50+t to TLI and Orion's mass is 22.8t. Using these figures, 9.4t of propellant gets 24.2t comanifested (47t braked fromTLI). That is a reasonably big payload, and a commercial launcher would have to manage something like 29t after propulsion bus to equal it on equivalent terms. However, the numbers are wildly better than those advertised by NASA, so take with a whole shaker of salt. Cheekily, that's a dry mass of Orion of 13.4t and strictly it only needs 5.2t of propellant to make the round trip, or 18.6t total to TLI. "Falcon Heavy can't launch Orion"

Based on SpaceX's published Mars payload, back calculating TLI throw gives 19.5t. To brake into NRHO takes 2.9t of propellant and let's be generous and say the propulsion bus is 20% of propellant mass - ~0.6t. That gives an Orion Equivalent Payload to NRHO for falcon heavy expendable of 16t, vs Orion 16.9t on B1B. That's not a big percentage increase for Orion (and it's probably neutered by larger margins because of crew on board). How many Falcon Heavies can be launched for the development costs of EUS? I'm betting a lot. Over 6 just for the cost of mobile launcher 2. The ICPS multiple burns is the only argument I've heard that has any validity for the switch, but if the risks are low enough to use that mission mode once they're low enough to use it repeatedly.

Ok, somebody check my numbers. Orion/ESM has a TLI injected mass of 26.5t, of which ~8.6t is service module propellant (I've also seen 9.4t, I'll calculate that too.) The mass with service module propellant expended is 17.9t (or 17.1t). NRHO to/from TLI is ~500m/s and the AJ10 has an ISP of 319s. In order to return from NRHO, Orion needs to reserve 3.1t (3.0t) of propellant for wet 21t, dry 17.9t, DV 500m/s barely. (20.1t/17.1t) That leaves 5.5t (6.4t)propellant to brake into NRHO on the way out. If Dry= Wet - 5.5t (6.4t), then the maximum wet mass that gives a DV over 500m/s is around 37t (43.5t). Subtracting Orion's TLI injected mass gives 10.5t (16.9t) payload. This implies block 2 (>46t to TLI) cannot be fully utilised by Orion, and Block1B (42t to TLI) can't be fully utilised either if the 8.6t service module propellant figure is correct.

How much mass can Orion actually brake into NRHO and return?

You mean the volume?

I still don't see the point in EUS. Block 1 launches Orion and commercial space handle cargo/lander. Blocks 1B and Block 2 launch Orion and not a lander and commercial space still handle cargo/lander. SLS is too expensive and too rare to ever launch probes. EUS is pointless.

For steering at hypersonic speeds grid fins are far more mass efficient, require smaller actuators, less power etc. I believe the proposed New Glenn uses a different "strakes" method of deflecting the airflow.

You mean replacing the grid fins on Superheavy/F9 with KSP style airbrakes?

I thought I'd heard that each heat shield could withstand multiple re-entries, but I can't substantiate where I got that from now ¯\_(ツ)_/¯

Slight reminder that the next SpaceX mission we have a date for is a Falcon Heavy launch - NET Oct 9th. USSF-44, double drone ship landing. That one's crept up! May be a Starlink launch first from Vandenberg and/or The Cape, but not seeing anything definitively scheduled.