Though there are still PR benefits to being able to say “The delays weren’t our fault; they were totally because of the FAA, and if we’d had our way we totally would have been ready”. Even, or especially, if they weren’t actually ready and would have delayed on their own if it weren’t for the FAA.
So far I haven’t seen anything quite like that, probably because SpaceX doesn’t really “owe” anything to the public in terms of timelines. And their development so far has been largely self-funded.
It’s possible they’ll use the delays as some kind of leverage with NASA, in case of missing deadlines for the HLS (Human Landing System, which is the component of Artemis for landing people on the moon). But I’m not sure it’ll really account for much in the long run. They’d be launching out of Florida anyway (hence the second tower), so really the only question is if the delays here impact the program as a whole. And really, they aren’t going to be the long pole here.
Musk has said that the FAA isn’t really suited for regulating the current space industry. They’re institutionally set up to regulate the airline industry for maximum safety. That’s not what we need right now for rockets, where the vehicles themselves are undergoing radical design changes. Trying to overemphasize safety right now would be a mistake because it tends to involve locking in the current design. We need the opposite; to explore a huge variety of designs to see which work best.
One could argue that they’re doing a bad job with aircraft, too. Obviously we want commercial flight to be very safe, but it’s probably almost impossible to build a plane that isn’t the standard tube-with-wings. It’s not a bad design, but it would be nice to have more variety. There’s no reason to think the current designs are optimal.
There is right now a monster ferment between FAA and EASA (the EU equivalent) about how to regulate the design and operation of what are generally called E-VTOL or AAM vehicles. Think mostly quadcopter hobby drones scaled up to carry 4 people, but there are a lot of unconventional semi-winged designs as well. Most are mostly electric powered.
Are they helicopters? Airplanes? Or something totally new? Even in terms of just number-of-9s reliability standards, should we apply big jet standards, bizjet standards, helo standards, or <9 passenger prop plane standards? How do we regulate electric propulsion systems design & redundancy? Hell, how do we even define what “redundancy” means for these things and the software that drives them? Then there’s the can of worms about defining and regulating levels of autopilot that start with simple stability augmentation up to approaching autonomy, much less achieve actual full autonomy. The economic case for these things requires either no pilot or a pilot whose controls look like a map app: click the destination or press the “emergency land” button and that’s it.
Right now the two big regulators are going in 17 disparate directions while the nascent industry is screaming for a single harmonized (which is not “identical” but is also guaranteed “not mutually exclusive”) standard and one that leaves lots of room for experimentation. IOW: specify the result = desired level of safety, not the process = how to build e.g. the windings on the motors.
It’s an unholy mess.
Oh, they’re not crude enough to state it like I did. And you and I aren’t the primary target of SpaceX’s PR, either: That would be directed mostly at the big investors. But I’m sure that they’re privately relieved that it was the FAA causing the delays, not themselves.
Good points. Even just trying to shoehorn a novel design into one of the broad categories of aircraft may be impossible. I agree about the “specify the result” approach, but how does one do that when failure rates are so low that you can’t measure them directly? You can start looking at a component level, but then you need some model for combining the component-level rates together, and those models might not be the same across different aircraft types. You mentioned earlier about placing outboard engines farther out to reduce the chance of shrapnel damaging a nearby engine, and you could come up with some failure rate based on that happening as a function of spacing… but that might not be true of an electric system, for instance. Any failure model has to take all these details into account.
Booster 7 static fire today:
There seems to be a lot of social media hangwringing about whether it was expected, but Musk tweeted that it was, and I don’t see any problems. There was a bit of a methane fireball, but nothing remotely enough to cause serious damage and probably semi-expected. Booster and tower appear to be fine. Only casualty is a flimsy plastic ventilation hose.
It’s hard to see, but there appear to be bright exhaust plumes coming from the booster, so I’d guess that the engines fired as expected. No green, either (which would indicate the combustion chambers vaporizing).
ETA: @Dr.Strangelove two posts above.
Precisely. It’s a legitimately difficult task engineering-wise. My new wife is a retired aerospace reliability engineer. What little she can share (mostly worked DoD projects at Yoyodyne) is both eye-glazingly complicated and utterly fascinating at the same time.
Then you get the bureaucratic inertia, and the difference between the culture of FAA & EU, and the fact that after Boeing’s famous problems the FAA is under the Congressional gun to be ultra-stringent and ultra-conservative and ultra-cautious …
And of course we could use the principle of least change and label them as e.g. helicopters, then tweak in some exceptions. But that’s building in some very square corners for the future as these things mutate and mature. Said another way, we can go for early stage regulatory stability and late-stage techno-stasis, or we can go for lots of uncertainty and delay now in exchange for a greater degree of future-proofing.
All that sounds utterly familiar to me doing significant scale IT work after a corporate merger: quick and dirty while building in a crisis that’ll ripen in a few more years. Or do it “right” but end up irrelevant as events pass you by.
Well … the stew stinks, is insanely thick, and will take a long time to be ready to eat.
Do you have any sense of there being jurisdictional power grab issues? I.e., if you’re the head of the FAA department that does helicopters, you stand to gain a lot of power and influence if you can get the new batch of personal drones classified as helicopters. That might make you more willing to widen your classifications, but that has to fight against general inertia and such.
I have no personal FAA connections to ask directly. The trade press is handling this particular aspect with kid gloves, which is itself telling.
With those factual limitations my IMHO answer is I have no doubt there’s a decent dollop of exactly that. There’s going to be a lot of budget for whichever department gets to scale out to manage these new vehicles. Lots of hiring and growth of middle management ranks, etc.
OTOH, there’s also going to be a lot of controversy and stubbed toes & mishaps and Congressional interest. Some bureaucrats relish the upside, others fear the downside. I don’t know enough about the personalities in charge of the respective branches to have an opinion on who leans which way.
Looks like it wasn’t quite nominal:
Still, it seems like it was just a little more energetic than expected (especially since there was a shockwave). Tower, stand, and booster are all largely intact, though.
Yeah, I watched the test, and it didn’t look planned to me. But it also doesn’t look like there was much damage.
E-VTOL
Here are the issues I see:
-
No positive safety. Unlike a helicopter, these things can’t auto-rotate. They will spend most of their time down low where ballistic parachutes don’t help, and air taxis will spend most of their time over built-up areas. They are going to probably be certified through redundancy, which means at least six or eight motors, redundant control computers, etc.
-
Range/endurance. Have any of them demonstrated an endurance of over an hour? How are they supposed to meet VFR or IFR reserve requirements? Or are we just going to waive that?
-
Certifying motors/batteries. What does maintenance look like? How do we establish a TBO? How do we certify a control system for an unstable airplane that will crash if the computer fails?
-
Landing. Where can you land these things? FOD will be a big issue, as will noise and risk to people on the ground. Are we going to need special e-pads for these, or will they be forced to use helipads?
-
Pilot certification. What does it mean to be a ‘pilot’? Do you need one? if so, what are they supposed to do, who trains them, and what does certification look like?
So far, these things look like they aren’t even close to being ready for prime-time. Certainly far away from commercial use flying passengers. Unless there’s a breakthrough in batteries. I think the range is still too short to use these safely for anything other than recreational flying over small distances.
There is a huge gulf between scaling up a toy quadcopter to carry peoole, and a certified vehicle that can fly commercially. If it’s going to work at all I suspect we’ll need a hybrid that transitions to traditional winged flight after takeoff. Joby’s electric Vtol is like that, and can fly for just over an hour and go about 130 miles. But that doesn’t include any reserve.
Yeah, shockwave is a bit of a giveaway. It didn’t look centred on the stand to me. So one wonders about a leak and a near stoichiometric mix of gas floating about at just the wrong moment.
Not nearly as exciting as Musk is promising.
Yeah, I think a hot test of some kind was planned, but a methane leak (or just bad luck with the mixture) made the fireball more energetic than expected. The test happened at 4:20, plus there was Musk’s initial tweet (which he deleted, but indicated that he wasn’t really surprised). There’s no way this was just out of the blue.
At any rate, that’s what you want a failed test to look like. No serious damage, just replace some equipment and try again. I’m sure they’ll have a couple more of these before fully loading the tank, let alone going for an orbital attempt.
Well, it looked to me like at least one engine was firing. Very strong downward plume of fire that lasted a couole of seconds. Maybe they planned a test fire, but a leak caused an oxygen-rich environment around the engines causing a small explosion when the engine(s) fired up and immediately shutdown?
You’re kinda just reinforcing LSLGuy’s point. You cited a bunch of things that were developed over time for conventional craft. There’s no inherent reason why they should apply to a new type of craft. Requiring an endurance of an hour is crazy; there’s zero reason for that if you can ensure that a safe landing is always only a few minutes away. Pilot certifications are crazy for totally autonomous vehicles. Etc.
We need to come up with equivalents to these, but setting things in stone too early will force them down a suboptimal path. We need to explore the design space before forcing any requirements. Allow manufacturers to demonstrate that they aren’t endangering the public through whatever means works best.
As to the booster 7 test I don’t see much reusable hardware remaining on that test stack. Yes, it’s not a pile of twisted wreckage.
But everything in that stack was exposed to loads undesigned-for. Quicker and cheaper to throw it away and build new than to figure out how to recertify that nothing was overstressed by a chaotic explosion you can’t characterize in any engineering detail.
Even if none of that was intended as flight hardware, all as ground test hardware, there’s no reason to assemble a hodgepodge of fresh test hardware and possibly-damaged-somehow test hardware for a future ground test. If something gives way next time, you can’t ascribe the cause in a trustworthy way.
Maybe. But they have a significant goal of making the entire stack robust. In the interview I mentioned earlier, Musk mentioned that they were reusing hardware from engines that underwent some “engine rich combustion” (largely due to cooling failures). They just swapped in a new combustion chamber and tried again. And even for this particular booster, it previously had some kind of anomaly that collapsed the fuel downcomer pipe. They ripped out the old pipe and put in a new one with more reinforcement.
Maybe there’s enough hidden damage that they’ll scrap it, but I don’t think they’ll let the mere possibility of unknown damage will prevent them from using it in further testing. Just keep patching it up until they have a problem they can’t easily fix.
It helps that it’s all stainless. No weird failure modes like internal delamination or weakness against heat soak. If they have a suspicious area, they can cut out a sample to “biopsy” and weld in a patch if it looks ok.
Sure, but that’s my point: All of that has to be developed and proven out. With new aircraft engine types, for example, the TBO is often set quite low until enough real-world data comes in to allow the TBO envelope to be expanded. For example, the Rotax 912 was originally given 600 hour TBO. It took four years of data before the FAA increased the TBO to 1200 hours, then 1500. Today, it has a 2000-hour TBO like the O-200. Certifying electric motors, speed controllers, control systems, batteries… That’s gonna take a while.
And how much reserve do you need in one of these? With a total endurance of an hour or so, clearly traditional reserves can’t work. Helicopters, which have the same ability to make emergency landings on short notice, are required to have VFR reserves of 20 minutes. If we apply that to our Quadcopter, that leaves a real range of maybe 40 minutes, assuming there is a charger at each end of the route. If you are landing somewhere and then flying back as a taxi would, and you have a total endurance of an hour, you couldn’t fly more than 20 minutes to a destination and still return home with VFR reserves.
None of these are technical showstoppers, but they will be quite limiting and will take years to prove out and formalize. Csrtifying even a conventional aircraft can bankrupt a company- Lancair almost went bankrupt trying to certify the fairly traditiinal Columbia 300 which used an off-the shelf certified engine. Cessna bought them out and finished certification and produced it as the TTx. Cessna had to put the composite airframe through 171,000 cycles before the FAA was satisfied.
Now imagine certifying a vehicle that uses all new concepts, new engines, batteries, and control systems, and which has no ability to land safely in the case of a system failure. No one is going to be selling certified versions of these any time soon. Experimental versions maybe.
On all this, IANA expert. I just read the trade press and have for decades. These things are just now breaking out of the Popular Science gee-whiz phase in the minds of the public, but around the industry these are decade-old projects backed at least partially by some of the biggest names in the traditional aviation business. And in IT. It’s not just sleepless grad students tinkering in dorm rooms.
As @Dr.Strangelove said above, you don’t.
The goal isn’t to be able to autorotate: 737s can’t do that and they got certificated. Tiltrotors really can’t either and they’re being certificated as civilian vehicles in addition to the military models. The goal is to be confident you’ll kill less than X people per million flight hours. Autorotation (as an example) is a means, not an end.
Yes, redundancy in rotors, control systems, power wiring, etc. And enough excess available power that a good fraction of the motors can quit and a) there’s still enough total thrust/lift to maintain at least a controlled survivable descent, and b) the control system can maintain stability with whatever mixture of live and dead it’s faced with. Deciding whether it has to be able to fly on any 2 rotors, or 4, or any 5 as long as they’re not all on the same side, etc., are the kinds of tradeoffs we’re talking about.
A 737 can’t fly with 2 engines out either. At least not for long and for a great portion of many / most flights, it’s highly unlikely that a no-engine situation will end happily. So we design to make those situations rare enough and call that good enough.
Range / endurance. The design assumption today is the things are operating over urban / suburban areas exclusively, at least in the erly generations. The average hop is 20-30 minutes, then a quickcharge to top up while boarding / deboarding then do it again. So a machine with a 1-hour endurance has 50-66% more than the strict mission requirement. We routinely fly big jets over the ocean now with reserves equal to as little as 5% of our planned flying time. Why does a planned 20 minute flight need another 200% of reserves just because an hour made sense in 1940?
Battery certification and maintenance: we already have some experience certificating and maintaining small Li-ion batteries on big jets. These will be larger, and worked harder. Yes there are unknowns. But it’s not a great gray mass of unknown unknowns. Yes, mistakes will be made and some specs will prove to be too loose and others too tight. It’s an engineering problem. They’re working it. They being both the manufacturers and the regulators. The automotive and railway industries are gaining a lot of hours of in service experience on what works and what fails in practice. All that can be, and is being, leveraged to inform this work.
An unstable airplane that will crash if the computer fails. You mean like an Airbus A320 / 330 / 340 / 350 / 380? Or an F-16, F-22, or F-35? Or their foreign equivalents? All are ballistic darts when the last control computer gives up the ghost or runs out of electricity. Or hydraulic pressure. That was a conceptually solved problem in the 1970s. The rest is just implementation-specific details. Not simple by any stetch, but eminently doable.
Landing: the expectation is to build dedicated landing pads. That will cost serious investor coin. But they’re about the size of the playing field of a tennis court, and the machines are noiseless enough the pads can be put in suburban parking lots at big box stores, etc. And downtown on buildings or atop parking structures. At high density destinations, the archetypal example being airline hub airports, you’d build a dedicated parking structure-like building with a ticket counter, lounge, security if needed, etc inside and a bunch of E-VTOL parking spots on the roof . A key assumption is that much like a taxi stand, any given vehicle won’t be tied up loading & unloading for long.
Pilots: The whole thing is about limited skill pilots. You don’t fly it conventionally. You tell it the lat/long and away it goes. if things start going badly it ought to land itself on any flat surface it knows about. A 2-lane road is plenty wide enough for most of these things. And they have zero landing roll. I suppose the largest role of the “pilot” is really backup “navigator”. If there’s an unexpected obstacle, be that a crane or another air vehicle, the operator will take control of heading and altitude much as one flies a quadcopter drone now until past the threat. With the goal that in a few more years the expensive human employee occupying an otherwise saleable seat can be dispensed with. These folks’ knowledge of conventional aviation may be about like that of powered parachute pilots. IOW, not much.
I’m not saying this last stuff is all magic cavorting happy unicorns. There’s going to be a shitload of blown deadlines getting this to work. But what it won’t have to do in maintain any compatibility with the legacy ATC system, VHF radios, approaches at airports, etc. It’s all greenfield meant to operate around the fringes of the existing aviation system not within the existing ATC system.
Precisely because this new system is coming of age in the netowrked distributed AI world of the 2020s, not the Morse code on HF radio world of 1940 that we can design something from the git go that lacks the limitations of backwards compatibility to Charles Lindbergh.
Or at least that’s the hope. Which one hopes is not wholly replaced by that other “h” word: hype.
I can safely predict fortunes will be made and lost on this industry. I just can’t predict whose fortune.
Your whole post just above mine wasn’t there when I started, so my earlier reply was aimed at your first one.
The snip of yours I quote here is dead right on. The whole industry has realized we can’t keep spending 171,000 cycles to prove something just because it wasn’t built from WWII aluminum.
One of the largest points about the whole exercise whether to regulate them as airplanes or helos is to regulate them, and the process of designing them, and of building them, and of operating them according to a totally new paradigm much more based on fresh thinking than on tried and true (read as “WWII era”) ways. Fresh need not be sloppy. And cannot be sloppy. But the old ways are strangling innovation for both traditional airplanes and for new ideas. To no benefit that we can’t, at least with fairly high confidence, obtain with less BS.
Again, or so’s the hope. People who do the every day and have for a decade have answered your type of straightforward objections based on traditional processes a thousand times. If the Feds reject out-of-regulatory-box thinking, some other country will build these things and we’ll be on the outside looking in. I for one hope they don’t.