They aren’t fancy. They were just coined by the French and adopted into English without anglicizing them. They’re pretty much the same sort of word as ‘nosecone’ would be if the French borrowed that one back.
aileron - “fin”
fuselage - “spindle-shaped”
empennage - “tail-plane”
“I have not failed. I’ve just found 10,000 ways that won’t work.” - Thomas Edison
True.
But if you’re going to converge on success anytime soon, you’d better have already avoided fundamental errors and/or have a very open mind about you’re willing to rework to get there. And most of all, you need to have started on a quest that is within, or nearly so, current state of the art.
I have every expectation SpaceX will succeed at making Starship work. If they fail, it’ll be because of something other than engineering. Compared to politics and drugs and insanity, engineering, even of the rocket surgery variety, is trivial. Laborious and intricate and expensive, but trivial. At least in the sense that the problems do always yield to skillfully applied well-informed effort. Most problems over on the human side of the house that’s just not true.
That’s a sort of double-edged sword, though.
If there was ONE systemic problem, it could perhaps be attacked by a significant redesign. However at the moment they seem to be whacking moles, and who knows how many moles are still lurking?
Of course I suppose this could be said of the development stage of any engineering prototype.
Fingers crossed….
I dearly hope so. As I’ve expressed in multiple posts the single most crucial element of getting Starship to work, the one pivotal thing upon which everything else depends, is solving the problem of efficient durable heat shielding.
Shielding against reentry heat has bedeviled aerospace engineering since the late 1950s. The X-20 Dyna-Soar spaceplane, and multiple proposed reusable aerospace planes of the era, foundered when the X-15 rocket plane demonstrated that nickel-steel just couldn’t take the heat. The best technology they could come up with in the 1960s, ablative heat shields, were heavy, expensive and single-use. The Shuttle’s ceramic tiles worked (sometimes barely), but were fragile and finicky enough that they helped turn the Shuttle first into a hanger queen and then into a crew killer.
For Starship to meet its design goal of being as quickly and economically reusable as a FedEx cargo plane, heat shielding has to work; and not just barely either. Which is why the midcourse loss of upper stages has been so frustrating: we need to work on and solve the heat shielding problem, and Starship hasn’t even been getting that far, we’re still farting around with RUDs.
The heat shielding problem is especially irksome because it’s related to another critical problem– reentry stage weight gain. The Dyna-Soar I mentioned was originally planned to be launched on a modified Titan II rocket but soon required a Titan III; that’s how much weight it gained in the design phase. Any increase in the dry mass of an upper stage can end up dropping the payload fraction to zero. Which is why NASA didn’t even try to make the Shuttle a fully reusable two-stage design. If the orbiter proved too much heavier than planned the whole design (both stages) would have to be scrapped and the project started over. The half-assed Shuttle design we got was safely conservative, with an external fuel tank and SRBs that could have been expanded if need be.
Starship will be in trouble if every orbiter barely survives reentry and either won’t be reusable or would require extensive and time-consuming refurbishment. Or if the heat shielding needed ends up weighing twice as much as originally planned.
While I largely agree with your post, what we’re talking about is the difference between a system that costs $10/kg vs. one that costs $100/kg. Both are an immense improvement over F9, which is already the cheapest rocket around at about $1000/kg.
More specifically, I think it’s pretty clear that they’ve solved rapid reusability for the Booster. There’s still some fine-tuning to do and implement all the logistical requirements to launch multiple times a day, but they’ve demonstrated all the major aspects. Already this will put them far ahead of F9. No barge landings, no TEA/TEB, no helium, lands right at the tower, cheaper propellant, etc.
And I think it’s clear that they will achieve some level of reusability for the upper stage. They can get it back to Earth in one piece. And I see no obstacle to them catching it. The engines and most of the structure will be intact.
But of course there’s a difference between reusability and rapid reusability. They need the latter to reach $10/kg. But even if the heat shield requires significant refurbishment, they’ll reach $100/kg. And that’s still much better than the current state of the art. You can build a small lunar colony with that.
The Space Shuttle was expensive not just because it needed refurbishment; if the refurbishment was fast and cheap enough, it would have been less of a problem. SpaceX has already manufactured and destroyed more Starship upper stages than Shuttles that were ever made on a vastly smaller budget and shorter timeline. It’s almost certain that even if they had to totally replace the heat shield, they could do so cheaply. On the order of millions, not tens or hundreds of millions. And in a couple of weeks. Their giant factory could easily sustain refurbishment at 1/day throughput.
So the only thing I think is at risk is their aspirational launch costs. But they’ll very easily improve upon their F9 costs, particularly for Starlink launches for which they’re well-optimized. That will be true even if the heat shield proves troublesome.
I worry that if Musk’s grandiose plans to at least try to colonize Mars never materialize, Starship won’t be able to find enough other uses to support itself. At $10/kg a multibillionaire could afford out of his own pocket to send a thousand flights to Mars as a vanity project, whatever one might think of the wisdom of doing so. At $100/kg he would have to be a trillionaire to do so.
I’m reminded of the dilemma of the Saturn V. Simply by being so huge the Saturn had an economy of scale that made its theoretical cost per kilo to orbit comparatively low; but of course no one but NASA had any use then for putting 120 tons into orbit at a time (circa 1961 the Department of Defense thought they did; by 1967, no). It was cheaper but not cheaper enough, any more than getting a deal on a shipload of bananas is any help to someone who wants to buy one bunch of them. An Atlas/Centaur, a Delta or a Titan III was the payload size anyone was really interested in in the 1970s.
I think it’s clear at this point that Starlink will justify the development. It’s already more than 2/3 of F9 launches. And an improvement to launch costs will make Starlink even more practical over time. It’s already quite profitable for them and that will only improve with launch costs 1/10 the current.
I’m curious what they do to improve the satellite cost. They’ve already made significant improvements in their switch to argon thrusters, low cost solar panels, etc. But they might need to drive the prices down even more. Maybe they can use the Tesla gigacasting tech to manufacture the structures (basically, injection-molded aluminum).
I think there’s also concern about the amount of aluminum going into the atmosphere when the satellites demise. It’s good that they disintegrate but has some effect on the atmosphere. Maybe they need to switch to stainless steel for these, too. Or wood…
Been done (by the Japanese, of course)…
Yes, I’d been thinking about that! It’s cheap and likely to pose no concern to the environment, even with binding agents. Of course, there are some downsides, like poor heat conductivity, but maybe heat pipes or some other tech could solve that problem. And it doesn’t have to be entirely wood, as the example above shows; even replacing 50% of the structural mass would be a win.
The satellites aren’t meant to last more than 5 years, so if the wood degrades faster than aluminum, it’s not necessarily a problem. Just has to last long enough.
I also wonder if it’s possible to reduce the amount of aluminum that completely vaporizes. For instance, suppose you coat small aluminum pellets (~1 mm) with a heat-resistant material like copper or stainless steel. Then those pellets get pressed into the final shape. As the satellite descends, the pellets will become unglued from each other, but individually survive the rest of the trip through the atmosphere (being so small, they have a high ballistic coefficient). Then, a large part of the total aluminum content will just end up landing on the ground (where it’ll degrade quickly) instead of aluminum oxide nanoparticles in the ozone layer.
Turns out it’s a “gigantic jet”, not a red sprite:
The difference is that gigantic jets go all the way from the top of the thunderstorm to the upper atmosphere (100 km). Sprites are disconnected from the storm and there’s no low-altitude portion. But both have a red-looking upper portion due to nitrogen emissions dominating at very low pressures.
Interesting failure mode:
Alpha Flight 6 lifted off and ascended nominally through stage separation. Alpha’s first stage then experienced a rupture milliseconds after stage separation. The pressure wave hit Alpha’s second stage, leading to the loss of the engine’s nozzle extension and substantially reducing stage two thrust. The second stage was able to recover attitude control and continued to ascend to an altitude of 320 km until running out of propellant. The vehicle was three seconds short of achieving orbital velocity and five seconds short of the target payload deployment orbit.
The ground-based video, onboard telemetry, post-flight empirical testing and Computational Fluid Dynamics analysis corroborated excessive heat from Plume Induced Flow Separation as the most probable root cause of the mishap. Alpha Flight 6 flew a higher angle of attack than prior missions. Plume-induced flow separation intensified heat on the leeward side reducing structural margins, causing the booster to rupture from stage separation induced loads.
Fortunately, the corrective actions are straight forward: increase thermal protection system thickness on Stage 1 and reduce angle of attack during key phases of the flight. Corrective actions have already been implemented.
Basically, flow separation caused by blowback caused hotspots. And those hotspots burned through the first stage, which then exploded and damaged the upper stage nozzle. The upper stage actually survived but the efficiency loss meant it couldn’t make orbit.
What’s flow separation? It’s when the rocket exhaust no longer smoothly travels the full length of the nozzle, instead separating at some point further up. And because there’s no longer a sharp edge (the nozzle rim) to attach to, the separation point is unstable and can damage the engine, not to mention causing hotspots.
Here’s an example of flow separation for a different reason–shutting down in an environment where it was already on the edge of stability:
You can see that the flow starts nice and smooth, but then transitions to an oscillating, unstable flow. If something was downstream and designed only to survive the clean flow, it’s likely that the unstable flow would damage it.
I don’t know how real-time or slowed down that was, but I’m impressed the nozzle didn’t RUD, or the fire blow out wo instantly catastrophic consequences.
Said another way, that demonstrates some serious engineering to survive transients, not succumb instantly to them.
Whatever else we might say, SpaceX has friggin’ nailed liquid fueled rocket engines.
It would have been destroyed if it weren’t for the reinforcing rings. They always use those when testing vacuum engines at sea level. Still, yes, impressive that it just takes a bit of reinforcement to survive. And I think it’s mostly to prevent flexural oscillations, where the end deforms into tall-vs-wide ellipses.
The Space Shuttle engines did this at startup, but they were designed to survive it, plus the oscillations weren’t quite so dramatic. Scott Manley has a nice bit on them:
That was great — and not just because of how Manley pronounces “toast.” 
We may get our best view of the interstellar object 3i/atlas…
from Mars!
A nice high-level overview of Rocket Lab’s Neutron rocket:
There’s a good chance that the US has four operational reusable rockets before any other country has even one. Falcon 9, Starship, New Glenn, and Neutron.
Sadly, Rocket Lab is no longer a New Zealand company in any meaningful sense. Founded there, but the majority of their workforce is in the US now and that’s where most further development (and launches) will take place.
What is the X-37b unmanned space plane’s glide ratio and reentry cross range, if that info is publicly available?
Some very cool space exploration pins here:
Apophis will be visible to the naked eye to people in Western Europe and parts of Africa. And there a a cluster of missions being planned to study it.