Huh? I think that if something doesn’t go right then that is when you need the human to fix the problem.
No, when something doesn’t go right the human is a cloud of red shrapnel.
And the human who will actually fix the problem is sitting safe and sound in an office on the ground, arranging for the replacement flight.
I’ve been worried all along about the prospects of getting the heat shielding to work; as I’ve said more than once, heat shielding makes or breaks the entire Starship concept. Upper stage weight gain is the devil haunting the dreams of aerospace engineers; it torpedoed the X-20 Dyna-Soar all the way back in the 1960s. I started to breath easier once Starship upper stages began surviving to make soft splashdowns, but now I’m not so sure again. Apparently the stages are surviving but too beat up currently to be reused. Even the plans for a stretched Starship Mk. 3 may be less about extra payload and more about making up for anticipated weight gain.
Doesn’t need much for that. A trip to the surface of the moon needs close to a full load of propellant. Getting down to the surface of Mars just needs what’s in the header tanks, which can have extra insulation and won’t share a wall with the outer surface. Though you certainly want to be sure that the small amount you do have is well-preserved.
The worst-case scenario is still much better than the Falcon 9, which is already the most cost-effective rocket available. Even with a fully expended upper stage, it’s naturally cheaper than the F9. The booster stage is everything they hoped it would be. And what they’ve demonstrated so far shows at least partial upper-stage reusability (maybe just the engines). It may take them a few iterations to get rapid upper stage reusability but even if that takes a decade, it will still be by far the most capable rocket in the world in the meantime.
We may certainly hope so. The problem is that Starship has to work all the way if Musk’s plan to colonize Mars is to be attempted; and in turn, the plan to colonize Mars is about the only thing that could create enough boost demand to justify Starship. A semi-reusable Starship might be better and cheaper than anything that has ever been built so far; but if Mars is off the table then we’re left with more booster capacity than the rest of the launch market could support.
It’s to be dearly hoped that an up and running reusable Starship will create secondary and tertiary uses for cheap ultra-heavy boost that no one can fully foresee yet. But space development is still searching for its “killer app”, the path out of the chicken-and-egg dilemma of the last sixty years: that there’s little worth doing in space, so there’s no point to investing more than it’s worth in getting to space, so there’s little worth (at current cost) doing in space; lather, rinse and repeat.
As I mentioned all the way back in post #129, it’s similar to how there was criticism of the 1930s hydroelectric projects, which at the time seemed to some to be little more than a works program boondoggle. It took World War Two to make people glad that the dams had been built. I remember the high hopes of the late 1970s and early 1980s that the Space Shuttle would at least begin to prime the development cycle of cost reduction and further development. The “High Frontier” crowd of Gerard O’Neill and the Stanford group thought that shuttle-derived components could support the beginning of a cislunar infrastructure of colonies and powersats. Needless to say the Shuttle didn’t work. Starship is being developed by the private initiative of a multi-billionaire who has sunk unmeasured billions into the attempt. This really may be the last best chance for it to happen.
I think it’s clear at this point that Starlink will pay for things for the foreseeable future. It’s already profitable on Falcon 9, even without having really expanded yet into every possible niche. They are just getting started with Direct-To-Cell, which will expand its market even further. Starship will enable the true v2 satellites and grow that market even more. It’ll pay for continued Starship development even with no other clients.
And sure, this might “only” provide the forcing function for a $2000/kg → $200/kg reduction and not the second order reduction from $200/kg to $20/kg, but as long as there is a regular client I think they’ll continue to make progress. They haven’t stopped improving Falcon 9, even if the improvements are less dramatic these days (though they’re down to a 2-week turnaround time for their boosters now). As long as they can make improvements, they will.
The risk, I suppose, is that there’s some truly fundamental obstacle to rapid reusability–something that requires a radical redesign. I guess it’s possible but I’m not seeing it. They still have plenty of work to do but there is still plenty of design space to explore even with the current rough architecture.
I guess that’s right. The Martian atmosphere is sufficient for aerobraking to dispose of almost all the re-entry velocily, provided you get the trajectory right. If you don’t, you’re toast, of course, but this is a situation which is fairly well-studied by now, I think?
As for return… the return ship doesn’t have to be the vehicle which lands on Earth.
As long as you don’t mix up metric and imperial units.
I think there’s enough egg on faces to ensure that won’t happen again? 
Even aerobraking into orbit I imagine you’d still want some maneuvering capacity; and probably you’d want to go down to Earth anyway (though maybe as a two-step process).
You’d think so, but the history of “really stupid mistakes done twice” is very long.
Lots of new Flight 7 details. Current date is Jan 10.
A few highlights:
Avionics upgrades include a more powerful flight computer, integrated antennas which combine Starlink, GNSS, and backup RF communication functions into each unit, redesigned inertial navigation and star tracking sensors, integrated smart batteries and power units that distribute data and 2.7MW of power across the ship to 24 high-voltage actuators, and an increase to more than 30 vehicle cameras
While in space, Starship will deploy 10 Starlink simulators, similar in size and weight to next-generation Starlink satellites
On Starship’s upper stage, a significant number of tiles will be removed to stress-test vulnerable areas across the vehicle. Multiple metallic tile options, including one with active cooling, will test alternative materials for protecting Starship during reentry
The Super Heavy booster will utilize flight proven hardware for the first time, reusing a Raptor engine from the booster launched and returned on Starship’s fifth flight test.
2.7 MW is an impressive amount of power! They must have several Tesla-equivalent battery packs installed. The Starlink simulators mean they must have the “Pez dispenser” system operational. And the metallic tile / active cooling test is very interesting. I wonder if the most critical areas will eventually get this system–while most* of the ship is resistant to some tile loss, there are some critical areas that might need something more robust/reusable.
This is in addition to the reshaped flaps that I’ve mentioned earlier. The burn-through we’ve seen on previous flights is basically a result of the previous forward flap design, and the next-gen version should dramatically improve things, particularly at the hinge portion.
I hadn’t realized that that was in doubt. “Pez dispensers” have been used for smaller satellites for decades, and it’s a remarkably simple technology.
It’s a little more complicated here. Unlike cubesat dispensers where a door opens and they just spring out, this one has to slide the satellites laterally from the bottom of a stack. It also dispenses two at a time. So there must be several active actuators, one that lowers the stack, another that pushes them out the door, one to open the door, and so on. Plus it’s large enough that the inherent flex in the system could cause them to bind on the rails and have other issues.
OK, but that’s still not all that difficult. I mean, I probably couldn’t whip up something in my basement that could do it, but it seems like it should be trivial for anyone in the aerospace industry. Or any sort of engineering, really.
Sure, it’s way down on their list of hardest problems to solve. But there’s a first time for everything. And there’s no point to testing this with functional Starlink sats since it’s still on a trajectory that enters the atmosphere. They can make sure the ejection mechanism works on this flight and then actually use it on some future one.
The payload door is probably the hardest part of it. They had trouble with it on an earlier flight. And in general, it’s a weak point in the structure that’s subject to stresses and such that might impact the operation. As far as I know, the last few flights have just had the door welded closed.
Starship Flight 7 scheduled for 2 pm PT / 5 pm ET today. Stream available here:
Somebody dug up an image from the payload bay of (probably) flight 4 during reentry:
So burny! But it survived. Stainless steel for the win.
Crazy that the fwd flaps still managed to work. You gotta think those flap actuators are maybe just a tad above the max operating temperatures.
Sort of a good news/bad news thing. Upper stage reentry is indispensable to the entire Starship concept, and reentry has probably been the single greatest obstacle to making spacecraft reusable. We knew from the Shuttle program that ceramic tiles work after a fashion; but not reliably enough to be acceptably safe for crewed flight, and too finicky to refurbish to support the launch cadence necessary for reusability.
Starship upper stages have been surviving to splashdown, but clearly not in good enough shape to be reused. The revival of active cooling is presented as an innovation but it’s tacit admission of a setback: Starship isn’t going to be viable with the previous tile technology.