Відео

The prices of the engines and the solid rockets are essentially fixed, and since the engines are a big part of the price it's very hard to reduce the price. Reuse could help the booster price but a) you still have solid boosters to buy and b) I agree with you that they are drastically underrepresenting the difficulty of the engine pod concept.

@@EagerSpace Even with 100% success with SMART, ULA is still throwing away two very expensive RL10 engines on every flight. The BE-4 engines aren't even that big a percentage of the total cost of flying Vulcan Centaur. At something like $20 million per RL10, even a flight with no boosters would likely be north of $50 million. That's a less capable vehicle than Falcon 9 at probably around three times the cost.

Thanks. I do a wee bit of firmware as a hobby, mostly ESP stuff these days but I've done AVR and even 68HC11 stuff in the past.

I've added that to my list, though it's a very long list these days. I *do* like alternate history topics.

@@EagerSpace Looking forward to it!

I was very disappointed to find that this was not a mashup... ua-cam.com/video/JzS9xzII_Qg/v-deo.html

It's my PowerPoint remote button clicker. It's under my desk and muffled with a blanket and you can still hear it on the mic. Keyboard and mouse are much, much louder. What I need is a new mic but I haven't decided which one I want.

@@EagerSpace Fascinating! What model PowerPoint clicker? And I’d like to suggest that you consider the Lewitt Ray microphone, which automatically adjusts its gain based on a distance sensor. Look at the epos vox review from about a month ago. Link in my reply to this comment. I have a soft spot for UA-cam channels that do great things with PowerPoint. Got a lucky break on an aviation topic a while back and somehow got a load of views with just a PowerPoint presentation. Channels like yours are encouraging me to get back in the saddle and try to launch it for real.

@@EagerSpace fancy mic review ua-cam.com/video/kQ9aR-0YsiQ/v-deo.htmlsi=l4AOD6bsutcgvo5l

Nice. I haven't seen that one.

What numbers are you using? With zero payload, I arrive at Centaur V having around 10.6 km/s of delta v (59.5 ton wet, 5.5 ton dry, 454s), and Falcon having 11.3 km/s of delta v (111 ton wet, 4 ton dry, 348s). With 15 tons of payload, it drops to 4.7 km/s for Centaur V and 6.0 km/s for Falcon.

To what orbit with how much payload? And what numbers are you using for centaur v? If you load a stage with a lot of propellant you can improve your Delta v but end up with a poor thrust to mass, which means you have to stage late. You can

​@@SpaceAdvocateYou're completely right, I messed up calculating exhaust velocity. And this in a vacuum still so the long burn time of centaur V doesn't help out with gravity loses any.

My general opinion about starship performance numbers is that none of them are particularly trustworthy because what we see now is a developmental prototype that is changing along the way and since it's a brand new design, what we see now is not the final version. The thing to note for HLS is that the contractual amount of mass delivered to the lunar surface is much less than spacex has said they can carry, so there's a lot of margin there. Golding's testimony was self serving in my book. He's the reason Artemis looks the way it does and taking attention off of that seems a likely reason for his interest. SpaceX exists largely because his NASA could replace the cargo and crew carrying capacity of shuttle, so we should thank him for that even if his goal was the opposite. The reliability numbers on refueling are the same argument made against large clusters of engines on falcon 9, falcon heavy, and starship, and we can see how those turned out. If you can only get two nines, you aren't doing very well and there's no reason you can't have redundancy for refueling operations. Unlike launch, they are not high energy operations and while you still need to do the development work, I think worrying about refueling is like worrying about docking.

@@EagerSpace Alright, thank you!

Thanks.

Thanks

I have a full video on gravity losses that I probably should have linked to. Gravity losses are roughly proportional to the angle of the velocity vectory; if you are travelling straight up your gravity loss is G while if you are travelling at orbital velocity to the side, the earth is curving away from you at the same rate you are falling towards it, so your gravity losses are zero. To figure this out you need to walk the trajectory path and figure out the gravity loss at each point, then the sum is the full gravity loss. It does depend on the trajectory that is flown and the thrust/weight ratios of the stages. But the Falcon 9 is in orbit in about 490 seconds, and the Atlas V requires about 650 seconds. Seems pretty hard to spend 160 seconds more in a gravity field and have lower gravity losses, especially since you are flying a more lofted trajectory to compensate for the low thrust/weight ratio of the second stage.

@EagerSpace I think you misunderstood something. Gravity losses are NOT proportional to the angle of the velocity vector, and you are still in the gravity field after entering orbit. If you are in a highly elliptical orbit, your velocity vector will be pointing up/out at certain points in the orbit, yet you don't have gravity losses on every single orbit. No, gravity losses are proportion to the angle of the THRUST vector (really the up/outward component of the acceleration from the engine). Early in the flight, when you are trying to get out of the atmosphere, this thrust vector will be pointing almost straight up. But if you look at the graphs, Atlas V actually has a higher thrust to weight ratio at this point in the flight and will therefore have lower gravity losses here. Falcon 9 only has a higher thrust to weight after about 250-300s, when the vehicles will have already pitched over significantly and will be accelerating largely parallel to earth's surface, so gravity losses are strongly suppressed. So, at first glance, it seems like Falcon 9 is likely to have more gravity losses. But that's only if we, I the absence of more data, assume both perform the same pitch-over maneuvering and attitude is the same function of speed for both vehicles. It is possible that the lower TWR of the second stage forces ULA to take a somewhat steeper ascent, which would cause higher gravity losses early on in the flight. But without the attitude data and some actual math, it's just not possible to definitively say which one has higher gravity losses, as there is a lot more to gravity losses than just time to orbit.

@@plainText384 Centaur has to keep the thrust vector pointed at a significant angle towards the ground for quite a while. This is to keep it in space long enough to actually get to orbital speeds. The Falcon upper stage doesn't really have to do this - it reaches orbital speed before reentry is an issue. It's very obvious if you look at the animations from ULA launches.

Gravity losses are based on the velocity vector. Go straight up 10km, aim your thrust fully horizontal, and you will drop our of the sky because gravity will pull you down. Orbit is all about having enough horizontal velocity so that the acceleration of gravity is matched by the curvature of the earth, and you didn't have gravity losses in orbit. At least for circular orbits - elliptical orbits are more complicated.

@@SpaceAdvocate I looked at the "Atlas V GOES-S Mission Profile" animation and there was almost no pitching up shown by the Centaur upper stage.

It's fairly widely believed that the numbers of the launch vehicle performance website are outdated. For Falcon 9 they list the Full Thrust version which was superceded by block 5 quite a few years ago. The Vulcan numbers are lower than the ones in the Vulcan user's guide. Falcon Heavy Expendable on a cheap launch to Mars (C3 = 8) gives a payload of 12,850 kg. SpaceX says 16,800 kg. ULA says Vulcan does less to Mars than the NASA site does.

Part of my complaint isn't that ULA is trying to differentiate their product - that's a natural thing to do - it's that they do it so poorly. There's a good argument to be made about ULA's history of high reliability and being able to do the complex missions that DoD and the Space Force want without doing this weird "high energy" stuff, especially when the rocket you're touting hasn't been certified yet. Better to act like there's a good reason so many USG payloads have been launched on ULA rockets and trade on that.

@@EagerSpace I agree; there is too much _Kool-Aid drinking_ in the rocket business these days…

I used the ones I did because 150 tons to LEO is what ULA chose to use in their charts, so the GTO (not GEO) numbers are in line with that capability. The point was to show that if starship does 150 tons to LEO this assumption that it can't get to GTO is flawed. Capabilities for Starship have squishy numbers because of the nature of the program. By starship, do you mean OFT-3? OFT-4? One of the starship 2 proposals? Where they will be when they first start launching Starlink? Where they will be when they launch commercial payloads? You can make arguments for all of them, so I choose what I think makes the most sense for the context, talk about my caveats, and figure that people understand that Starship is still in development and it's a weird project because it's going to change a lot before it's done.

@@EagerSpace Using the assumptions that ULA did seems justified, but why then are you using numbers for "probably a one way trip" instead of the reusable capacity?

Always questioning only one side. Have you asked TF why he's still claiming Starlink is subsidized even after he was corrected multiple times? Have you asked why he's claiming Soyuz seats are cheaper than Crew Dragon using price tags from decades ago and for tourist flights?

I probably wasn't clear on that. Take the reusable payload to Leo as a baseline. Figure out how much the payload would be reduced if you go to GTO. This is still assuming the heavy reusable starship and fuel for landing. My one way comment is because you may need more fuel to get starship back into a survivable re-entry trajectory.

I don't have any particular insight. My *guess* is that a) pushing Electron to say double their payload doesn't increase their market very much and it adds quite a bit of complexity operationally and b) Neutron is a better place to spend engineering resources because Electron is not the long term plan for the company. Maybe if they had meaningful competition is small launch they'd have a different opinion.

Antares was very clearly a one trick pony, and my guess is it made NASA happy to have a more established company join the program when Kistler dropped out, but one of NASA's goals is to foster new commercial capabilities and the contract to Antares did nothing in that realm. I talk a lot about Falcon 9 development in my recent Falcon Heavy video. They certainly had their eyes on commercial launch because Ariane 5 was so expensive and Proton was so explosive, but they were focused on NASA because it was clear that whoever got the contract would likely be flying cargo for as long as ISS is around. And it gave them a possible gateway to carrying crew, also lucrative and a capability in their long-term plans. But I think that SpaceX is generally pretty good at not getting ahead of themselves.

I don't have any real insight into how the market will shake out with starship flying. SpaceX has flown a lot of transporter missions but their bread and butter is still conventional missions, and I'm not sure that Starship changes that. And yes, there are many scenarios where you might do a partial refuel.

I'm not a big fan of words like "efficient" or "optimized" because they aren't well defined. As for price, I do think there's great evidence that SpaceX bumped up their prices for NSSL 2 considerably because otherwise they would be leaving money on the table, and similarly for NASA. SpaceX is doing extra work for the gateway launch for NASA, but they're getting paid $330 million for a Falcon Heavy launch.

@@EagerSpace I meant "efficient" and "optimized" in terms of bang for buck. We can go by price to customer or by estimated cost to SpaceX which sets price floor in case of competition which is cheaper than ULA or Ariane.

@@zeevtarantov I don't think we have particularly reliable sources on internal costs or profit margins on these launches, and comparing prices is sort of pointless since if they feel like putting in a competitive bid, they both end up at very similar prices (that's just the nature of markets and bidding). Pretty much the only thing we know is that unlike with Delta IV heavy, ULA is able to make bids with Vulcan Centaur that are competitive to (and in some cases even cheaper than) comparable launches on a Falcon variant. Whether that's because SpaceX doesn't want to go lower, SpaceX can't go lower, or because ULA can't or doesn't want to go lower, we don't know. The market isn't that competitive, so if ULA and SpaceX wanted to raise prices and milk profits while their duopoly lasts, there's nothing really forcing them to race to the bottom.

Exactly. With a duopoly, we would expect the two companies to end up roughly at the same prices and that's what we get.

Impulse aerospace is the obvious near-term choice for a lot of payloads.

A Kickstage for Starship (internal on the payload adapter) makes so much sense

The problem is the O. It shows he doesn't really understand what it means.

The height of LEO is "LEO altitude", which is what they chart uses. That's perfectly fine. LEO by itself doesn't mean that, which is pretty clear by all the other uses of LEO in the document.

The "O" means orbit. If "LEO" (just "LEO", not "LEO altitude") also means a location/altitude, then that would apparently mean that suborbital vehicles like New Shepard go to the "location" known as "low earth orbit". Which is just kinda idiotic. Using terminology in this odd, imprecise manner mainly just serves to confuse. And it does so in an area that's already confusing enough to the public: the suborbital vs orbital distinction. And this makes Bruno's double-down tweet-where rather than clarifying that "yeah the graphic title really should have said 'LEO altitude'", he instead insists that it's correct as-is and also that the purpose of all this is to help educate the public about space-frankly quite baffling.

Thanks.

You just can't do propulsive landing with the Vulcan first stage even if you had a cluster of engines. Falcon 9 runs right on the edge of not getting too hot - they melted early grid fins when they were made out of aluminum - and they are still paying somewhere around 25% of their payload penalty. Vulcan is going way way faster, which means a huge amount of fuel to get it to survive reentry. I'd be surprised if the payload penalty was less than 50%.

I don't know if an inflatable heatshield + parachutes is really that much more complicated than a propulsive landing on an ocean platform. Sure, you introduce an additional staging step and inflatable heatshields are relatively new. But if you think about it, the concept isn't that crazy. NASA has been flying tech demonstrations for inflatable heatshields for a while now (LOFTID, etc.), and landing the engines seperately without the disposable tank has already been done with the Space Shuttle. The parachutes and heatshield combo is also a known, uncomplicated method for reentry. It's passively stable, it doesn't need to relight any engines and a splashdown allows for much less precision in the guidance compared to a propulsive recovery. You do need to integrate the engine section with a new rest of the booster every time, but the actual recovery else seems simpler and less complicated to me.

Parachutes are notoriously finicky in capsule applications, with a lot of complexity there. If you look at the amount of parachute testing that Orion, dragon, and starliner have required I do not think the word uncomplicated can be applied. Propulsive landing has turned out to be ridiculously reliable. We have very limited data on inflatable heat shields and we do know that parachutes fail. The big issue is that the reuse requires you to go back to the factory, recertify the whole engine pod, test it, attach it to a new airframe, then ship it back to the launch site. Where with propulsive landing your stage is already in the right location and your recertification work is much less.

The design doesn't exclude using a highly efficient engine like the Centaur (RL10). But it would need more thrust than the RL10 can provide. One option would be the BE-3U. It's similar to the RL10, but it has 6.5 times the thrust. This allows for staging lower. This is the engine used on the New Glenn second stage, and New Glenn also stages fairly low, allowing for booster landings.

@@SpaceAdvocate They likely traded efficiency for thrust, as the specific impulse of the BE-3 engine hasn't been publicly disclosed.

@@xponen It's probably around 440 seconds. Though it's not the only option for a higher thrust high efficiency engine. Another option would be to design a new staged combustion cycle vacuum hydrolox engine. That should enable pushing the specific impulse up towards 460 seconds or more.

@@SpaceAdvocateThe theoretical maximum specific impulse for a hydrolox engine is 460 seconds. To increase thrust or approach this maximum, trade-offs are necessary, such as making the engine heavier or more expensive, like the Space Shuttle's RS-25. Therefore, we should be skeptical of the BE-3U's performance until detailed specifications are published.

@@xponen The absolute theoretical maximum for a hydrolox engine is more like 530 seconds, using only the energy of the propellant. The RL10 is at 465 seconds, but you could increase this by increasing the chamber pressure. The RL10 has a chamber pressure of 44 bar, while if you used a staged combustion cycle engine, 200+ bar is achievable.

ULA isn't doing these papers for us, they are doing them to try to influence decision makers in the space force, DoD, and Congress. Differentiation is a tried and true marketing technique, but it's hard for ULA because the SpaceX holds most of the cards and Vulcan still isn't certified yet. So we end up with stuff like this. My big complaint isn't that they are trying to differentiate, it's that they do it so poorly.

@@EagerSpace But surely decision makers in Space Force, DoD, NRO, Congress, etc. have qualified staffers who can tell them it bullshit and "just look at price per mission" ?

In a speech at the SpaceCom conference in late January of this year Tory Bruno claimed “We run about 34% cheaper on a high-energy mission than the other one, SpaceX, does,”. That's what this is about. ULA has identified a specific category of mission requirements, where they could outcompete SpaceX's Falcon Heavy (probably because SpaceX needs to expend the core stage to complete the mission) and they which think will take up an increasing share of the NSSL market. Yes, this isn't a new capability. Delta IV Heavy was also able to complete such missions, but Vulcan is not just able to do these missions, it's good at them. So it's not just meaningless marketing BS, like "It's Toasted", it's marketing BS based on a real advantage (for a certain subset of customers), like "gluten free" (important if you have gluten intolerance, not so much for other people).

@@plainText384 A fully expended Falcon Heavy likely has a cost in the area of $80 million ($20 million per core, plus $20 million in upper stage, fairing, ground systems, etc). A Vulcan Centaur VC6 likely has a cost in the area of $110 million ($6 million per GEM-63XL, $8 million per BE-4, $20 million per RL10, $18 million in structures, ground systems, etc). The rest is marketing and pricing strategy for maximizing profits. I don't think SpaceX would get out of bed for less than a 50% gross margin. They're just too busy with other stuff.

@@plainText384 But their high energy vs low energy marketing does not advertise they are cheaper it implies that only they are qualified for/capable of those high energy launches, that''s the bs. The only launches that SpaceX can't do (yet) are for payloads that require vertical integration (which might not be that many anyway) but again this is not what they talk about. Also I wouldn't blindly believe Bruno on being cheaper on those high energy missions because it could very well be some accounting "trick" (like when they got paid for development but was separate from launch per se while SpaceX had to tack on extra infrastructure money to a launch task order) or some other not exactly fair/honest comparison

I think this is because SpaceX already managed to demonstrate their approach and their capabilities when they optimized Falcon 9, by incresing its max payload from 9000 to 22800kg (+153%) In order for the Starship to live up to its promises and do 100t reusable, they only need to increase its max payload by 66% and figure out landing on a tower

Whenever I talk about Starship I say that it's hard to do analysis because there are no good numbers out there. All I can do now is project based on the numbers that are out there. Will they reach those numbers? I don't know, but SpaceX is in a better position to know than anybody outside.

@@Neront90 Falcon 9 is a different type of rocket than Starship. Many hyping Starship are still relying on F9 development trajectory.

Eric's main aim here is to critique Tory's specific claims in his "The Secret's of Rocket Design Revealed" article by showing how the claims don't stand up. In fact, it turns out Tory did't make a rather poor case - he made a very poor case. Eric's refutation could stand on its own but the references to Falcon 9 are made for illustration about the paths of rocket design. Starship is a pretty minor point in all of this, and Eric makes it clear those are rough estimates with a foundation on shifting sand.

@@EagerSpace Roger isn't the sharpest or even most honest person there is. When he talks about bias he has a tendency to ignore his regular his own (anything elon=bad) bias to the point where he is more ridiculous than anything.

Love to have comments from the business side.. I agree that organizational inertia plays a big role. And initially with Vulcan one of his big goals was to get out from under having two rockets including the really poor delta IV, and it has/will do that pretty well. I don't think beating SpaceX was ever in the cards, and I don't think LM and Boeing would pony up the money to try to do so. And that's probably the right business decision - it's not clear that they'd ever make the money back. Same problem with reuse. Kuiper is the wild-card here - Vulcan could fly a lot if that actually goes forward but there's been so little happening that it's hard to predict.

All good points. And I think at some point during Vulcan's development, when F9 showed showed how successful it was, ULA decided it didn't have to shift course and try to make a reusable rocket because they didn't have to compete for 1st place, competing for 2nd place would be the most profitable way to go. That's because NSSL and NASA policy mandates two launchers. Another reason for ULA's approach is put best by a quote from Dirty Harry; "A man's got to know his limitations." SpaceX could make F9, and make it work, because they were a lean vertically integrated company with a rocket engine engineering team.

Yes. I think ULA is behaving in a rational way given their current position.

High staging is standard in rocketry because upper-stage engines, which have high impulse, typically have a low thrust-to-weight ratio. This means they need to be staged at a high altitude to have enough time to reach orbital speed.

I know KSP isn't anywhere near close to accurate real life, at least in terms of engineering, but it's funny how I sorta came to that stage low vs stage high conclusion myself when I was making my own reusable rockets. I didn't want to stage my rocket so quickly that the first stage would be coming back in too quickly and burn up or be unable to slow down enough.

It's a challenging question and the only real answer is to do actual trade studies. SpaceX knows the answer to this question but it took a lot of engineering time to get there. I do have a few thoughts... Staging early wins for reuse because a) you spend to get back to a drone ship or RTLS and b) the less velocity & distance you have, the better. This is especially true for RTLS and it's not surprising that starship is moving to have super heavy do less work - the less work it does, the less RTLS costs. The countering force is that a rocket where both stages do 5000 m/s of delta v is easier to build and has a higher payload than one where it's imbalanced - 2000 m/s in the booster and 8000 m/s in the second stage. The low delta v part is easy to build, the high one is really hard to build as it needs to be very light and that's expensive. The limiting case is obviously a SSTO which is a bad idea in my book. It is *very* interesting that SpaceX seems to be moving starship to as close to the an SSTO as practical to get the RTLS costs down as low and to get the booster back as soon as possible. If you want to do reusable, I don't think it's worthwhile to consider any approach where you don't get the first stage back intact, and from what we see, that seems to be the conclusions of everybody except ULA. I don't see a lot of reason to design towards "stage high" and trying to get a ton of delta-v out of the booster as a principle - you are carrying more mass higher and accelerating it more and that's an energy loss. I do want to note that delta-v is one way of looking at things but it ignores gravity losses and they turn out to be quite important. For boosters, thrust is really a dominant factor because you are paying gravity losses during the whole launch.

Yes. Or to look at it another way, you need a big beefy booster to give them enough vertical acceleration so that they don't just fall back to earth.

The most useful comment I've read... Fixed.

My model said the low 20s, I've seen others that say mid 20s. They all depend on mass numbers for starship and those of course aren't trustworthy right now. Shuttle was stuck in LEO because it was (from memory) 27 tons to a very low LEO orbit plus about 136 tons of orbiter + tank, and that gave it a very low payload ratio, maybe 17% Any credible numbers for Starship put it *way* above that. Even if it does 100 tons with a 150 ton stage, that puts it up at 40%. But the GTO numbers are one-way. My guess is that it's going to be a lot more economical to take a kick stage from Impulse and use that on a geo satellite from LEO.

I would expect Starship to survive re-entry from GTO. They want to go interplanetary with it after all. No need to slow down to LEO before re-entry.

​@@debott4538I'm in no position to absolutely agree or disagree, but reentry from GTO will probably be a LOT faster than reentry from LEO (you might be able to estimate reentry speed with vis viva equation? I don't know how much delta v starship would have or where it would make the deceleration burn though (apoapsis?)).

@@Zorba-Ivy If you reenter from geostationary orbit, you'd be going at around 10.3 km/s. If you reenter from the ISS, you'd be going at around 8 km/s. Given that heating increases with the cube of the velocity, reentering from GEO would see over twice the heating.

@@SpaceAdvocate i wasn't aware of reentry heating working like that. I would certainly rather use a kick stage than subject a reusable launcher to that. Or maybe go Kerbal and aerobrake?