SpaceX attempted to launch their new Falcon 9 Block 5 today, but scrubbed at the last minute due to some ground issue. They’re making another attempt at 16:14 ETD (20:14 UTC) tomorrow.
The Block 5 is the final(ish) iteration of the Falcon 9, with all of their lessons about reusability rolled in. Too many to list here, but in short it will allow up to 10 flights with minimal refurbishment. That is in contrast to the current Block 3/4, which have only been flown at most twice and require substantial refurbishment (though still a lower cost than a new booster).
So all in all it should be pretty nice improvement over the current state, and increase the cost advantage SpaceX has over the competition (which is already substantial) even further.
Block 5 is designed for “24 hour” turnaround, which was not picked because they intend on making that the default case, but because it draws a line in the sand between acceptable and unacceptable levels of inspection and refurbishment. That said, they say they’ll make a 24-hour attempt sometime next year, just to show that it can be done.
Musk has also said that they plan on experimenting with second-stage reusability using ballutes (or some other inflatable) at some point, though I have doubts about the cost effectiveness here. Still, it may provide useful data on the reentry environment. That is likely some ways off–for now, a successful Block 5 flight is the focus.
They’re back to using a fast load for the propellants. Feels a little scary since that probably contributed to the AMOS-6 pad explosion, but they say they’ve made some very serious advancements and upgrades to the COPV (composite overwrapped pressure vessel) tanks which store compressed helium.
Total mission success. Probably could have passed on the Bangladesh propaganda video that they wedged in there, but hey, they’re proud of their satellite and they paid just like anyone else…
Apparently, the first stage has active (water) cooling for some parts of the octoweb. Open-loop, I presume (i.e., just let it turn to steam and dump it overboard). Sounds like they had some hotspots that would have reduced the structural lifetime if left unchecked. Would be interesting if they could reduce the reentry burn (allowing the cooling system to compensate for higher temperatures) and thus the payload impact of reuse.
Flawless indeed. And they managed to land the booster even after a Geostationary satellite launch. I guess that’s due to the increased power of the engines on block 5?
They’ve landed several other GTO boosters. Just last October they landed both the boosters for KoreaSat 5A and SES-11 (which was itself being re-used from an earlier mission). I’m sure the extra margin from the new engines helps, but it wasn’t strictly necessary before.
They’ve landed boosters for GTO missions before–and heavier ones at that (the record being the 5,300 kg SES-10 bird). Bangabandhu-1 was only 3,500 kg. Still, Block 5 probably allowed a more gentle landing–likely using a single-engine landing burn instead of a triple-engine one (triple-engine is more propellant efficient, but riskier and harder on the components).
It’s possible that some of the efficiency improvements on Block 5 (such as the uprated engines) were eaten by reusability components, like the larger titanium grid fins, cooling system, additional thermal protection, etc. Hopefully we’ll find out more as time goes on.
I was sorta hoping we’d see RTLS (return to landing site) missions more often, but SpaceX has said they prefer drone ship landings since they’re easier on the vehicle. I wonder if that’ll be the new standard, and if so, if we’ll see more barges for Falcon Heavy missions. Then again, the boosters on Heavy missions have an easier time (earlier BECO) than on F9 missions.
A Shortfall of Gravitas, of course. We wouldn’t want to present the illusion that there is a significant quantity of gravitas at play here. There is in fact very little gravitas present at all, and any claims that said gravitas was handed out or otherwise transferred to the ships in question are quite false.
There’s a falcon-9 launch planned in approximately 40 minutes. This one is interesting in that they’re going to try launching satellites into two different orbits. This means they get to a certain altitude and park it for a while (at a billion miles an hour or so) and then relight the second stage to get to GTO.
Haven’t really updated the thread with every last launch, but the one tonight (just happened) was a milestone–first reuse of a Block 5 booster. They’re still at month+ turnaround levels, but that’s expected to change.
One thing they confirmed during the broadcast: they plan on using the same booster for a third time later this year. SpaceX, as of yet, has only reused their boosters once so far. Block 5 was designed for 10 reuses without significant refurbishment, and 100 uses with more involved reconditioning, so the booster should continue to have a long life ahead of it.
The Falcon Heavy is the most Kerbal of SpaceX’s rockets. The BFR has the engine count but otherwise falls short of the KSP aesthetic. Maybe one day we will get a BFR superheavy with 7 cores, 217 engines, and lots of struts.
That new picture of BFR worries me. That’s a radical design change from the last one. I would not expect much other than very small incremental changes in a rocket that’s supposed to be already under construction. This looks like a completely different BFR than the last one, which was already a complete design overhaul from ITS.
I am worried that the BFR concept may not have even had a proper feasibility analysis. There shouldn’t be such large design changes at this point.
A “proper feasibility analysis” can’t have been done yet–the BFS design is driven by the requirements of landing on Mars, and that’s just not something that’s well understood yet. The requirements are also driven by the design itself, and that’s something that only iteration can solve.
For example, one change appears to be the placement of the landing struts–they’re more widely spaced than on previous iterations. That suggests that they wanted to improve landing stability. But the required landing stability is driven by all sorts of factors, such as the terrain of the landing spot and the overall precision. They may have realized that they can’t guarantee they won’t land on a large rock, for instance. The landing precision is driven by the degree of control authority, and that’s not something you would know until you’ve made quite a lot of progress in the design.
The Raptor engines themselves don’t seem to have changed (just the engine bells), and the outside diameter looks to be the same, suggesting that the composite propellant tanks are more or less the same as they were. I don’t see any changes that would invalidate the work they’ve done so far.
The BFS is going to be progressively built anyhow. The first version will just do Earth-bound hop tests and be missing a ton of stuff (heat shields, most aerodynamic surfaces, etc.). They’ll move onto Earth orbital tests and only after a while have something that can actually get to the Martian surface.
They haven’t shown an updated version of the BFB (booster) yet. I’d expect that it’s a bit closer to last year’s version, since SpaceX knows a lot more about landing booster stages on Earth than they do about landing orbital craft on Mars.
This is a mostly nonsensical statement. A feasibility study is a top level analysis that indicates whether top level design requirements for a desired capability can be met by existing or proposed technology and within some acceptable budget. Requirements are not “driven by the design itself”; by definition, the design is a result of meeting requirements, which is true whether you do a top down or bottom up requirements scheme. You don’t build something and then make up a requirement to match it, because then you end up with a jumble of mismatched requirements.
Both the necessary capability, and the loads and environments experienced in the entry, descent, and landing (EDL) to Mars for a large payload are actually pretty well understood since we have extensive atmospheric data from the Mars Climate Orbiter. The problem is that there is a large gap of capability between the capability demonstrated to land payloads the size of the Mars Science Laboratory (~1 ton) and the larger payloads required for even a minimal human mission (>30 tons), which is not a surprise because among engineers who have worked on Mars landing capabilities it is considered the most difficult solid body in the solar system to controllably land upon. It isn’t clear that SpaceX has ever done detailed feasibility studies of Mars EDL capability; back when they were proposing the Red Dragon as a lander, it was abundantly clear from just the size of the propulsion module and the amount of propellant it could carry that supersonic retropropulsion wouldn’t be feasible, and it isn’t clear exactly what work SpaceX has done to validate any requirements or “key design parameters” they have for the BFS.
Anyone developing a vehicle for a crewed landing on Mars would certainly approach it in an incremental fashion, but the ambition of SpaceX is to land an uncrewed ‘cargo’ mission in 2022 and human mission at the next opportunity in 2025, which doesn’t really give any time for incremental development, and no detail has yet been provided about how SpaceX will deal with the multitude of physiological, sociological, and logistical problems beyond the challenges of EDL.
That sounds like old-school, waterfall-style development. The kind that SpaceX tries to avoid.
It’s very important to give yourself flexibility in trading off requirements. At a very high level, for example, one can trade landing precision for robustness. A very robust lander can get away with low precision; a high precision lander can be relatively fragile. It is not necessarily obvious in advance what the true costs are of each one. It would be silly to set a hard requirement for one if it is possible to cover a shortfall with the other.
Dan Rasky of NASA has done some interviews where he talks about smaller-scale tradeoffs, and gives SpaceX a lot of credit for having this kind of institutional flexibility. He gave an example where (IIRC) one component went over its volume budget. A nearby module had spare volume, though, and so there was no need for an expensive redesign–just a conversation with the other group. Holding the overbudget module to a hard requirement would have been ridiculous.
To be clear, I was speaking solely of the style of EDL that SpaceX is using: lifting-body reentry combined with supersonic retropropulsion on a heavy, medium-ballistic-coefficient vehicle. This is not a well-understood regime on Mars. AFAIK, SpaceX is the only organization with a operational supersonic retropropulsion system even on Earth, so they are at the head of the pack, but there is still a large gap between that and landing things on Mars.
They have done all kinds of simulations, obviously–we’ve seen those in years past. But I suspect this is an area where there is still much to learn. For example, we know that the scale height of the Martian atmosphere has significant seasonal variations. This clearly has a huge effect on reentry–not just whether it is possible or not for some vehicle, but whether you can land in a particular spot with some precision and some probability of success. Increase the modeling fidelity and you may learn that you need more control authority than you hoped to get away with originally.
One other thing: it’s likely that SpaceX has several possible designs going at a time. What we see at any given moment is just their currently most plausible design. SpaceX may have bee thinking about wingtip landing legs and canards back in 2016 or whenever, but didn’t show it because they didn’t think they would need them back then.
