SpaceX launches

A bit of perspective:

Yeah, this really demonstrates the difference between New Shepard and the SpaceX rockets, and shows why it's a tourist toy.

Of course, Blue Origin is also working on New Glenn, which is about the size of SLS.  Currently scheduled to launch next year, but we'll see.

N1 was the Soviet moon rocket, they launched four of them and the longest any of them lasted before catastrophic failure was 3 minutes.

In reply to Keith Tanner :

I've been to 101' in a bucket truck and 140' in a man basket hanging from a crane. Those are more breath taking than you can imaging. Hard pass on 230'. The slightest breeze would be terrifying. 

In reply to codrus (Forum Supporter) :

The N1 was all sorts of bad ideas.

For those who aren't rocket nerds, check the small print under each of those rockets. 
Name of Rocket
years active
payload it can lift to a certain orbit (suborbital, Low Earth Orbit, Geosynchronous Transfer Orbit, Trans Lunar Injection in order of difficulty)
number of successful launches/number of failed launches

And check the (predicted) payload on Starship. 100,000 kg (220,000 lbs) to LEO and a build cost that's orders of magnitude less than the SLS - plus you get to reuse it.

That chart is pretty up to date - it's hard to keep up with running changes in a Starship - but it's missing the Virgin Orbit Launcher One, which has put two payloads into orbit.

It's also missing the Virgin Galactic SpaceShipTwo which is a suborbital carnival ride like New Shepard but may be too "plane like" to get on this list. I suppose if it's on the list, the X-15 and Rutan's SpaceShipOne need to be there as well.

Keith Tanner said:

The N1 was all sorts of bad ideas.

Yeah, I think it's interesting that one of the things that was considered to be a major design flaw (having 30 engines in the first stage with all of the plumbing and control complexity that entails) is being revisited in SuperHeavy/BFR.  Certainly control systems have come a long way in the last 50 years, and I'm sure the quality control in SpaceX's manufacturing and assembly are better than that done in the Soviet Union, but it's still an interesting comparison.

The N1 comparison to SuperHeavy is not really appropriate other than "lots of engines". Those engines couldn't be static fired, they used pyros to open valves so it was a "build it, cross your fingers and light the fuse" sort of situation. I believe that due to the hassle of test firing very few of the individual engines were test fired as well. There were all sorts of vibration and pressure problems with the engine cluster that could only be discovered by trying to launch one - and boy, did they discover failure modes.

It's also worth noting that the Falcon Heavy uses 27 engines split across three cores, and that's actually more potentially problematic than a single core with 29 or 32 because you need to keep the cores attached to each other :) So I don't think the number of engines is a significant risk. The fact that the Raptor is the only full flow combustion engine to be flown is more likely to be causing a headache, but according to Musk the mass production of engines is more difficult than designing them. That mass production allows them to swap out engines if there is a problem in testing though.

I still think the Bellyflop is a bad idea, and will get nixxed in crew vehicles in the future. They might use it for unmanned returns, but not even the bravest astronaut is gonna wanna ride on something that depends on a perfect suicide burn to land without going boom.

There are solid reasons for the bellyflop. And anything with propulsive landing is going to require a burn. The only alternative is to toss a return capsule out the window and bring back the second stage unmanned. 

For crew versions, they may leave more reserve prop in the tank so the second stage can hover longer, and they're already using redundant engines after what they learned in the first attempts. The big difference between Starship and other crew vehicles is going to be the number of flights before they put crew on board. It will be a well-tested platform, unlike the Shuttle or SLS.

Remember that the prototypes were not expected to stick the landing, especially after the first flight revealed some problems that couldn't be fixed before SN15.

In reply to aircooled :

Awesome graphic, thanks.  I had no idea that Starship was so much larger than the Space Shuttle.  What a beast.

pheller said:

I still think the Bellyflop is a bad idea, and will get nixxed in crew vehicles in the future. They might use it for unmanned returns, but not even the bravest astronaut is gonna wanna ride on something that depends on a perfect suicide burn to land without going boom.

You should look into the difficulties in getting drogue chutes and parachutes to deploy safely and consistently from reentry / terminal velocities.  Very much not easy (but lots more data / research on it of course)

What will be interesting is if they add some sort of crew escape system, which for testing could easily be a low powered ejection seat.  For final, higher crew / passenger counts, that would likely not be practical.  Maybe parachutes and a door?  What about Mars parachutes? They would be huge (but could be very lightly built).

I don't believe the bellyflop will be needed for Mars, it's primarily there to control reentry heating. That's less of an issue in the thinner Mars atmosphere.

Also, note that SpaceX found that a bunch of the parachute design data from back in the day was just wrong when developing the Crew Dragon. Parachute development took quite a bit of time there. 

If something goes wrong with the booster, I believe the Starship itself is the escape mechanism the same way that Crew Dragon is the escape mechanism for Falcon 9. Ejection seats wouldn't work after the first stage is done, the craft is moving way too fast and is way too high. But there's more talk about this in the more recent Everyday Astronaut video from what I understand, so I'll have to sit down and watch it tonight. 

I was thinking of the ejection seats more for the crewed testing of the landing of the Starship.  A crew escape module could work, but does add a lot of weight (especially with a larger crew) and I have not seen that in the designs (not that I have seen a lot).  I does seem like there should be (maybe a requirement?) some sort of back up system, but then again, the capsules under the parachutes don't have them.  At some point you just run out of options. The Starship does have multiple engines it could use / switch to, so there is that.

Aerobraking would still be very useful on Mars to slow down from orbital velocities (and maybe trans mars speeds) to save on fuel use, so I would expect them to use it in some form there because if nothing else, why not.

aircooled said:

I was thinking of the ejection seats more for the crewed testing of the landing of the Starship.  A crew escape module could work, but does add a lot of weight (especially with a larger crew) and I have not seen that in the designs (not that I have seen a lot).  I does seem like there should be (maybe a requirement?) some sort of back up system, but then again, the capsules under the parachutes don't have them.  At some point you just run out of options. The Starship does have multiple engines it could use / switch to, so there is that.

Aerobraking would still be very useful on Mars to slow down from orbital velocities (and maybe trans mars speeds) to save on fuel use, so I would expect them to use it in some form there because if nothing else, why not.

I think ejection seats for test flights were of more value when test flights needed to have pilots on board.  Starship belly-flop landings are automated by necessity (pretty sure the timing is too delicate to fly one by hand even if you wanted to), so there's no reason to ever put a person in there until they're completely confident it'll work and then there's no reason to have the ejection seats.

Landing on Mars is HARD.  It's got a lot of gravity for lunar-lander style powered descent, but not enough atmosphere for Earth-style aerobraking and parachutes so you need a complex hybrid of both.

Even if they stick the landing 25, or even 50 times, you want to be the first person riding that horse without some sort of plan B? 
 

Sure. I'll take that over something that has only been tried once or twice like the Starliner or Orion plans - or never, like the Shuttle.

As noted in the most recently released Musk interview, no spacecraft have an abort mechanism for landing. Most (not all, hello Shuttle) have/had one for the first stage. Most do not have one for the second. So the plan is to retire risk with lots of testing and iteration along with redundancy, as with aircraft. Multiple engines. Dozens and dozens of test flights. A willingness to make changes due to the results of unmanned flights. The freedom to be able to have things go wrong is a strength - if you're always flying people (hello Shuttle) you cannot make changes.

I'm only halfway through the second interview but so far it's as good as the first. 

Yeah, I didn't mean to imply that I think BFR is going to have the same sorts of problems as N1, it's just interesting to look at the advancements in the underlying tech since 1970.  SpaceX is trying all kinds of things that nobody else has, or at least that nobody else has ever managed to actually make work.

The nice thing about a capsule like Starliner or Orion (or Crew Dragon) is that the launch abort system is really simple. Explosive bolts to release the capsule, solid or pressure-fed hypergolic rockets to get it away from the rocket stack, and parachutes to land.  The Shuttle had none of that, once you lit the SRBs you had to wait til they burned out and even after that point there was only a very complicated RTLS abort mode where lots of things had to go right for it to work.  Columbia did have ejection seats on the first few flights, but the solid rocket exhaust would probably have melted the parachutes if they'd need to use them.  (Gemini had this problem too)

Starship doesn't have a true launch abort system either, cryogenic liquid engines usually take a bit of "spin up" time so I expect you can't use the Raptors on it like you can the SuperDracos on the Crew Dragon.  It's also not clear if they'd have the performance to do something like a pad abort, the ones with vacuum nozzles are not going to produce much thrust at sea level and a fully-fueled vehicle will be much heavier than at landing.  That said, it doesn't have the inherent safety problems of the SRBs, and being on top of the rocket instead of on the side means that the critical thermal protection system is better protected at launch.  Hopefully this is another area in which SpaceX can do what other people have failed to do!

Of note (not that it really matters): I am pretty sure a launch abort system has never been actually been needed / used.

Even the Russians... well... that they will admit.

I suspect the primary reason they have them is because of all the rockets NASA blew up on the pad in the 60's.  Made getting volunteers a bit easier.

The Russians had a launch abort about three years ago. There was an American astronaut on board so it wasn't much of a secret :) Kind of a big deal, because at the time that was the only access to the ISS. But if you mean a pad abort where the capsule launches without the booster - I don't think so. Certainly not since the early days.

In reply to codrus (Forum Supporter) :

Rvacs will work in atmosphere, just not as efficiently as they will in vacuum. I also wonder what's involved in spinning them up, they're definitely able to light the engines pretty quickly as shown in the flip. Could they put the Starship engines into "idle" on the pad pre-launch? Starship can hover on one engine, a fully loaded hoverslam might be possible with 3 Raptors and 3 Rvacs. But boy, that's going to be some fun to figure out.

SpaceX has moved into a "minimum viable system" mode for now, the goal is to just get into orbit. All the fancy stuff like cargo doors and pad aborts is being back burnered for the time being.

Keith Tanner said:

The Russians had a launch abort about three years ago. There was an American astronaut on board so it wasn't much of a secret :) Kind of a big deal, because at the time that was the only access to the ISS. But if you mean a pad abort where the capsule launches without the booster - I don't think so. Certainly not since the early days.

In reply to codrus (Forum Supporter) :

Rvacs will work in atmosphere, just not as efficiently as they will in vacuum. I also wonder what's involved in spinning them up, they're definitely able to light the engines pretty quickly as shown in the flip. Could they put the Starship engines into "idle" on the pad pre-launch? Starship can hover on one engine, a fully loaded hoverslam might be possible with 3 Raptors and 3 Rvacs. But boy, that's going to be some fun to figure out.

SpaceX has moved into a "minimum viable system" mode for now, the goal is to just get into orbit. All the fancy stuff like cargo doors and pad aborts is being back burnered for the time being.

The launch abort system also saved a Soyuz crew in the 80s when the rocket caught fire on the pad just before launch.

As for the spin-up time, if you watch Falcon 9 launches they sometimes announce "MVAC chill-down" about 30 seconds before staging.  This is where they pump liquid oxygen through the second stage Merlin engine to cool it gradually rather than thermally shocking it by dumping in the full amount of oxidizer all at once.  You could probably do this before launch, although you'd presumably boil off and waste some of the LOX and reduce performance by doing it for longer than necessary.  That said, being a turbopump-driven engine, there's a minimum startup time to get enough initial pressure to spin the pumps to get enough fuel to actually make full thrust.  Raptor is a slightly different design to Merlin, but it's still got turbopumps.  I dunno how long this takes, but it seems unlikely to be in the sub-second range that you really want for an abort system.

The SuperDracos are pressure-fed hypergolic engines, you just open a couple of valves (or use burst discs) and they light with no prep necessary.  This really only works on relatively small and low-performance engines though.  Mercury and Apollo used solid rockets for the abort system which are even simpler.

I'm not sure how much thrust you lose with a vacuum engine operating at sea level -- in Kerbal Space Program it's a lot, but that's a game not real life :).  Googling I'm not finding much about it, except that apparently it leads to flow separation and can cause damage to the engine.

And yeah, it makes a lot of sense to get the core propulsion stuff working without worrying about the fancy bits getting destroyed in the expected test failures.

I'll have to go back to the Starship test flights. I seem to remember seeing the engines doing something to prepare for the flip. I do understand the difference between the hypergolics and the main engines.

IIRC the abort rocket on the Apollo stack was considerably more powerful than the Redstone that Shephard rode into space.

So there you are, watching  a big rocket struggle off the pad, burning tons of fuel per second and you think "there has just Got to be a better way to do this" and you figure an X15 style airplane launch is by far the best way to do this until you do the math. I've been having a similar problem with the Flash Gordon landings. "You mean to tell me it makes more sense to haul a boatload of fuel up, just to burn it so you can land, as opposed to giving your booster stage wings and glide it back?" "Yes, yes it does." 

I . Cannot. Wait. To see Spacex' latest and greatest do its thang....

I hope that somewhere within SpaceX , Elon has a skunkworks team trying to develop an anti-gravity technology - just because.

In reply to Sparkydog :

I've just sorta assumed there has been, for some time now...

Jay_W said:

...I've been having a similar problem with the Flash Gordon landings. "You mean to tell me it makes more sense to haul a boatload of fuel up, just to burn it so you can land, as opposed to giving your booster stage wings and glide it back?" "Yes, yes it does." ...

To clarify, this makes sense because you do haul a boatload of fuel up, because you need it just to get out of the lower atmosphere, resulting in a rocket that most of the weight is in fuel.  That also means that when it's almost empty it's MUCH lighter then when it took off, which means it means a lot less thrust (and fuel) to do a suicide burn landing.  The trick (other then the timing) is the rocket is now way overpowered, so getting it to throttle down is critical which is not super easy in many cases.  Thus having multiple engines is useful, so you can cut some off (and potentially have spares if needed).

I think people assume using wings for launch and landing is a good idea because it's what we're used to seeing every single day. But rocket velocities are in a completely different range, so the air launch isn't much help and those big draggy landing wings are a PITA at all other times.

The Shuttle could have been implemented better, but there were some cross range requirements forced on the designers that forced a suboptimal design. It would be interesting to see how we could meet those requirements today given Starship's 1:1 glide ratio (iirc) that allows for surprising cross range capability.