sevenperforce

They probably are / already have. Strong suspicion that FAA simply asked SX what happened (beyond what everyone could see) and SX gave them open access. No need to shut SX down for months for some hyped up Congressional Inquiry like dog-and-pony show. Smart folks from the FAA look at what happened and what SX plans to do about it, write a report... done. Yes, they are absolutely already investigating. The suggestion by Bob seems to be that because there was something he thinks was a failure, that means we can infer a failure in the FAA's oversight process, and that means we can infer a fundamental flaw in the FAA oversight process, and that means we can infer that FAA is incapable of rectifying the fundamental flaws in its oversight process, and that means we can infer an independent commission is necessary to fix the inferred problems in the FAA.

So you’re saying the FAA should investigate itself? I'm saying the FAA should investigate the incident. Because that is its job. There is no indication or suggestion (other than wild unsourced and unevidenced speculation) that that FAA ought to be investigated as an institution. Failure investigations and process lapse evaluations happen all the time across industry and regulators.

Putting the capsule at the bottom is very Kerbal, very New Space, and very clever. I'm still unsure where the engines are. Can this be used to physically drop cargo pods, or would the vehicle need to land and offload cargo robotically? Fingers crossed. Re-entry from orbit is just an inherently hard problem. I'm not sure how they do it. The four options seem to be (1) the DC-X/FHUS concept, (2) the Starship body flap approach, (3) the Shuttle/X-37/Dream Chaser approach, and (4) the Stoke/Chrysler SERV approach. Number 4 clearly won't work for a New Glenn upper stage (unless they do a radical redesign and replace the BE-3Us with BE-7s), and you'd need separate landing engines for Numbers 1 and 2. So a winged, wheeled vehicle seems like the only possibility.

Nuclear pulse propulsion works in space because the pusher-plate works as a shield. In atmosphere, the heat an radiation can bounce off of air in addition to conduction and other methods of heat propagation that are not an issue in space. While a ground-launch Orion is more realistic than a star-trek shuttle, you would need a fairing that probably weighs more than the rest of the ship put together to have anything survive to orbit, and even that would be highly questionable. In-atmo-launched nuclear pulse propulsion would actually be slightly MORE efficient because the air provides an additional reaction mass. However, getting off the ground initially is...challenging. Most designs would use solid boosters or similar tech to get moving before the pulses start dropping. The pusher-plate would still act as a reasonably good shield in-atmo although the standoff distance might need to increase. Alternatively, it might be better to decrease the standoff distance and use a smaller nuke in-atmo, thus trapping more of the radiation behind the shield and relying on the increased impulse from aerodynamic remass to make up for the smaller nuke size. Average acceleration isn't significantly different than an ordinary vehicle so the fairing is not a big deal.

I certainly hope that's not true, because it would be pretty careless of them if so. The Shuttle and STS booster segments are steel, so using a similar FTS charge would make sense. It wouldn't necessarily be careless if they ran the numbers and they looked appropriate. Shocking how much my prior experience comes into play here, LOL. Crack propagation in steel is a well-studied phenomenon in the field of transportation pipelines, although it primarily focuses on stress corrosion cracking (SCC) and other fatigue-based crack propagation. That's what you end up with in thick steels at ordinarily temperatures which are subjected to large internal pressure gradients. Rapid crack propagation in rupture scenarios can follow existing crack structures, but it is much more likely to burst at one point and then have a directional propagation through the steel at the local speed of sound. I remember one instance (I won't name the provider) where fatigue-based cracking weakened a particular bend in a pipeline, and then one day they had to abruptly shut down flow in an offshoot line several miles downstream. The spike in oil pressure from that shutdown propagated backward along the pipeline through the flow and intersected the existing standing-wave pressure gradient at that joint, causing a constructively-additive pressure excursion at the precise point of the SCC fatigue. The resulting rupture released all that energy in a beautiful sinusoidal wave for about 8' along the length of the pipeline. You could actually see where the crack propagation dropped below the speed of sound and stopped. Of course we only figured all of this out AFTER an entire LAKE of crude oil got dumped into someone's farmland, but I digress. All that to say: the relationship between rapid crack propagation, ductility, and tensile strength is a complex one, made all the more challenging by temperature. They may have been expecting crack propagation that happened in ground tests but didn't happen under cryo conditions, or they may have simply been relying on the rupture to cause tank implosion. SpaceX uses the same FTS explosive initiator, detonator, and linear shaped charges used by NASA on the Shuttle and STS main tank and boosters. Although the shaped charges are linear, they do not run the full length of the tanks.

Unfortunately it won't. You need a spark plug and a plasma source like Fogbank, and neither of those are going to be approximated by a giant-reflective-sphere arrangement. I don't particularly feel like running the numbers atm but I am guessing that you would need at least multi-megaton-class nukes before the accessible energy from thermonuclear fusion pulses would exceed the output of simply shoving the fission materials into a modern reactor.

Good point. Are there any kind of information whether it was some problem transmitting the FTS signal (maybe not redudant via starlink ?), processing the signal, igniting the charge or time / pressure for enough structural damage to rip it apart ? Would we have seen the charge going off on the video ? I mean on the test tank it looked pretty obvious. But if it wasn't going off, was dynamic pressure or integrity really an issue ? Or maybe antenna / processing ? Starship+Superheavy, like Falcon 9, uses an AFTS. So there is no signal at all. Rather, the ship itself knows where it is and detonates itself if it flies outside of an acceptable corridor. What happened with the Starship test launch was twofold. First, the range of acceptable launch profiles was quite high (since they had contingencies for thrust shortfalls) and so the acceptable corridor was extremely broad. Second, when it finally did fall outside of its acceptable corridor, it was at 38 km and tumbling slowly with relatively low aerodynamic stress. The FTS charges popped holes in the tanks, but there was enough residual head pressure in the tanks that they maintained structural integrity for almost a minute before finally imploding and then ripping apart. Scott Manley has a video where he highlights the moment that the FTS charges go off and you can clearly see propellant spewing out while the tanks remain intact. Had the AFTS fired lower in the atmosphere or while under significant acceleration (either due to gravity or thrust), the FTS charges would have caused immediate tank collapse and breakup. But being so high in the atmosphere while traveling relatively slowly was a different matter.

To be clear, I was saying that the FAA's lapse was over-delegation. The more serious lapses were within Boeing for sure.

Eh, there's no resistance to oversight to be found here. I've worked in the federal government -- specifically in the field of accident investigation, root cause evaluation, and process lapse mitigation. Specifically with methane, liquid hydrocarbons, and cryos, in fact. The feds on the ground performing certifications and evaluating process trees and double-checking industry work aren't political appointees. These offices are staffed with absolute nerd lifers whose entire personalities revolve around personal pet peeves that just happen to align with public safety. Bureaucracy moves slowly, but that can be a good thing in the world of safety, because political pressure doesn't often transfer downstream very well. Where you see problems are political and administrative overrides, like Challenger. All of those people on the ground doing good work are useless when the higher-ups simply overrule their objections. If that happens, it's definitely a reason for Congressional or other oversight of the regulatory agency, but fortunately that sort of failure is quite open and notorious. Calling for an "independent review" of something which still has not been shown to even constitute a failure doesn't seem like oversight; that seems like political posturing in its own right. I'm reminded of that scene from The Phantom Menace: Chancellor Valorum: "The point is conceded. Will you defer to allow a commission to explore the validity of your accusations?" Queen Amidala: "I will not defer. I've come before you to resolve this attack on our sovereignty now. I was not elected to watch my people suffer and die while you discuss this invasion in a committee!" If there's an actual problem, then we should identify the problem, characterize the problem, and determine whether it was a lapse in oversight or an unknown unknown...not create an independent commission to tie things up endlessly with no actual goal or objective. I haven't looked at the environmental review closely, but given that a pad RUD was one of the possible failures, I can't imagine what actually happened being WORSE than what the environmental review was prepared for. If the whole stack had blown up on the pad it would have been a lot more destructive than what actually happened, and with a much bigger radius, too. The 2020 Beirut explosion had a yield of approximately 1.1 kilotons TNT equivalent and did this: A Starship+Superheavy pad RUD would have a yield of at least three Beirut explosions and possibly up to nine Beirut explosions. Again, I cannot imagine that the pyroclastic pad launch (which blasted chunks of concrete as far as a kilometer but only sent fine particulate matter and dust significantly beyond that) was as bad as an RUD would have been. Dust and particulate matter certainly went beyond the expected debris field, which is cause for investigation, but that's already being investigated by the FAA. You and me both. In hindsight it was more of a gamble than they thought, although I think it was more of a program/schedule gamble than a public safety gamble. Thick air, yes, but how thick? 38 km is certainly much lower in the atmosphere than the altitude where separation was intended to take place, but it is still MUCH higher than ordinary aerodynamic altitudes. At a speed of 500 meters per second and an air density of 0.41 g/m3, the dynamic pressure is 1.95% of the dynamic pressure on the SR-71 and a scanty 0.24% of the dynamic pressure on Falcon 9 at MaxQ. It's the equivalent of a semi-truck doing slow donuts while sliding across an icy parking lot at an average speed of 20 mph -- definitely out of the ordinary, but nothing you'd expect to crumple steel (at least not from aerodynamic forces). Has there ever been an FTS activation at that altitude? Challenger blew up at around 15 km. Activation of FTS at 30+ km altitude might simply be something we've never experienced before.

IIRC, the FAA's lapse over the 737 MAX came from delegating too many of the certifications to Boeing and just taking their word for it. If the delayed effect of the Starship AFTS is considered a failure (and again, that's something I'm assuming, not something that's necessarily evident), then there's value in the FAA looking at whether a similar lapse was involved. Did the FAA simply trust SpaceX's assurances that the AFTS would produce an immediate RUD without verifying? Or was the delayed effect an unintended consequence of enhanced structural integrity that no one at SpaceX or the FAA could have anticipated? If the former, then it would be worthwhile for the FAA to look at its review process to see if it is unjustifiably delegating certifications in other areas. If the latter, then it falls into the "unknown unknown" category and increased oversight isn't going to make a lick of difference. And ^^THAT^^ is how you properly handle root cause evaluation and mitigation of process lapse during a failure investigation. Not by demanding an undefined, ill-conceived "independent review" with no meaningful goals or realizable objectives.

No experience with bones outside the one inside me and in food. But wooden shafts has lots of problems. I don't understand all of this insistence on using these new-fangled wooden tools. That's a fast way to get splinters, if you ask me. Fists and dirt clods were good enough for my grandpappy and they should be good enough for me.

Indeed. Made even more advantageous if the bell alone gimbals rather than the entire thrust structure.

The LM didn't gimbal. The descent propulsion system engine had 6° of gimbal in pitch and yaw, with roll handled by the ascent propulsion system's roll control thrusters. Only the ascent propulsion system had no gimbal. There's some suggestion that the BE-7 has a bell gimbal with a fixed thrust chamber.

I haven't seen it depicted with any decent scale other than in some of the Blue Moon mockups which were inflatable and thus likely not to be trusted. I believe it's an dual expander bleed cycle, so I'm expecting it to have slightly better thrust/area and thrust/weight characteristics than an RL10, but slightly worse specific impulse. It pushes 44.5 kN. Pixel counting from this image suggests an expansion ratio of around 88:1 which makes sense and is comparable to the RL10-A-4 line's expansion ratio of 84:1. The RL10-A-4 has a thrust/area ratio of 53.9 kPa and a thrust/weight of 60.3, so plugging in these numbers for the BE-7's lower thrust would suggest a nozzle diameter of around 1.03 m, an overall length of 1.25 m, and a weight of 274 kg. Given the expected improvements in thrust from the dual expander cycle, we can expect these numbers to be overestimates (although the length will likely remain about the same, especially given that the BE-7 uses an aggressively cylindrical thrust chamber). If it's not a bleed cycle, the numbers may only be slightly improved over the RL10. One thing that's been discussed with the BE-7 is that the O/F ratio is a carefully guarded secret. It's possible, then, that they are going with some unique architecture that deviates heavily from typical O/F, like a partial bleed cycle. For example, they could have a closed full-flow hydrogen expander cycle on the hydrogen side, but an open split expander cycle on the LOX turbopump in order to squeeze out higher chamber pressures at the expense of an O/F ratio that is closer to stoichiometric. That would help with reducing the bulk and mass of the hydrogen tanks. On the other hand, they could be doing a closed split expander on the LOX side but an open split expander on the hydrogen side, dumping the hydrogen turbine exhaust downstream. This would both help with film cooling (note that Tom Mueller specifically thought the BE-7 was running with too much film cooling) and increase the efficiency a little over a normal expander bleed without sacrificing improved chamber pressures; it would require a much lower O/F ratio. Maybe it's got a very high O/F ratio and they're only expecting to take on LOX from SpaceX and will launch extra LH2 with New Glenn. I would be very surprised if there was no gimbal at all.

How was launching without a flame diverter "unsafe"? The FAA safety review included as one possible outcome a complete pad RUD and had assured safety measures in place for that eventuality. What part of launching without a flame diverter made the test launch more unsafe or could have led to more catastrophic results than a complete pad RUD? Let's assume, without concluding, that the AFTS delay (caused by greater-than-anticipated structural margins) represents a failure by the FAA safety review. What sort of review failure is this, and what sort of independent analysis would be required in order to mitigate the second-order impact of such a failure? If you're proposing some sort of solution, then it's not enough to say "X failed, therefore Y." You need to properly characterize the nature of the review process failure and explain what sort of corrections to the review process are necessary to prevent not only this specific failure but any other known or unknown unknowns.

Arguably the bulk of the tonnage of a modern oceanic cruise ship is, well, ship, and not passenger quarters or other spaces. Simply having a steady platform on the high seas takes a lot of structural mass. A ship with artificial gravity will be heavier than something like the ISS, but still vastly lighter than an ocean liner.

I'm really liking this design. Keeping that lightweight hydrogen up top will (among other things) help keep the whole thing much more bottom-heavy, which is an advantage over prior designs and over Starship.

I think we’ve had similar discussions…in fact I think I made such a thread once

The actual amount of water being used may not be THAT extreme, all things considered. It will be a lot, of course, but possibly not as much as in some other designs. They are going to use blow-down from a pressurized gas reservoir to force the water through rapidly during ignition and liftoff.

That is where you are getting confused. The use of a compass and straight edge has nothing whatsoever to do with forcing the use of whole numbers.

@RCgothic, are we thinking that the pipes running lengthwise through the assembly will have holes in the side to allow the entire assembly to fill with pressurized water, which then would exit through the holes on top? EDIT: Wait, no, now I see the pipe holes coming up underneath. Is it a flow-through design? I'm having trouble visualizing the water path.

Oooh, I get it now. Yeah I guess I was wrongly assuming that the plates were solid steel.

Some interesting developments: With this closer view, I can see that it's actually ONE steel plate with channels cut into it, not two plates with an internally-pressurized space. That should help significantly with the structural integrity of the plate and prevent ballooning at the center.

I strongly suspect they are less efficient. Very strongly suspect that. Propeller efficiency drops with blade count, and in many ways these double the blade count. The reason that propeller efficiency drops with blade count is due in part to things like turbulence and blade interference, which in turn is a factor of blade tip vortices, and so if the toroidal propellers more evenly distribute blade tip vortices to reduce noise, they might also reduce turbulence and interference, increasing efficiency in certain RPM ranges. Or at least canceling out losses. But yes, I would suspect that they are most likely less efficient across aviation RPM ranges.

@tater already made many good points, but interjecting to note that this is the fatal error in your reasoning. SpaceX isn't claiming that the per-kilogram cost of launching on Starship makes them the cheapest. SpaceX is claiming that the per-launch cost of launching Starship makes them the cheapest. If I am out in the city on foot and need a ride home, and I pull up Lyft, and I see a Prius for $50 and a Cadillac Escalade for $22, and both of them have the same pickup and dropoff times, then I will take the Escalade, thank you very much. Even though I might not NEED that much room, I will choose the bigger vehicle if the bigger vehicle is cheaper. Even if I'm all by myself and don't have anybody riding with me. I will take the cheapest option. If it is cheaper to launch a single Cubesat on Starship than it is to get a dedicated launch with Electron or a rideshare launch with New Glenn, and Starship is available, then the Cubesat owner will launch it on Starship. Every time. This is a lot of words and not enough numbers. Most of the New Space companies are not focusing on full reuse, and the ones that are don't yet have anything remotely resembling full-stack hardware. If Stoke Space comes along and is able to launch smallsats cheaper (on a per-launch basis) than Starship through full reuse, then Stoke Space will capture the smallsat market, right up to the limit of its payload capacity, and Starship will have the rest of the market.