SpaceX Plans To Resume Launches In November (reuters.com)

An anonymous reader quotes a report from Reuters: SpaceX is aiming to resume flights in November following a launch pad fire that destroyed a Falcon 9 rocket and an Israeli communications satellite it was due to lift into orbit, the company's president said on Tuesday. The space services company suspended Falcon 9 flights while it investigates why the rocket burst into flames on Sept 1 as it was being fueled for a routine prelaunch test at Cape Canaveral Air Force Station in Florida. "We're anticipating being down for about three months, getting back to flight in the November timeframe," Gwynne Shotwell, president of Elon Musk's space company, said at a satellite industry conference in Paris. SpaceX previously said a nearly-completed second launch site in Florida, located at NASA's Kennedy Space Center (KSC), would be finished in November. The pad was last used to launch NASA's space shuttles five years ago.

Comments

Re:First they have to find the cause

[Geoffrey.landis]

What is scary is if Musk has already decided they will resume so quickly even if they have not determined the cause.

Everything SpaceX does is always attributed to Musk. The actual article attributes the quote to Gwynne Shotwell, president of SpaceX.

She was speaking at a conference. Somebody asked when they'd be likely to start flying again, and she gave a best guess. This is not a firm commitment to fly whether or not they have found and fixed the problem, it's just a best guess about how long the process will take.

My personal best guess is that a failure review for a non-manned system takes about six months (after their June 2015 failure launches resumed in December, for example) so I think she is a little optimistic, but she probably would prefer to err on the side of optimism.

Re:First they have to find the cause

[Rei]

but now Tesla has had 2 in just over a year,

Huh?

The tactical mistake Musk has made is thinking that people care more about launch cost than about getting their payload successfully to orbit.

In many cases that is correct. AMOS-6 happened not to be one of those, but....

If you have a hyperexpensive multi-hundred million sat as most of them are you will happily pay an extra few tens of millions for ULA-like reliability rather than cheap out and end up on a launcher like Falcon with a 5% chance of desrtoying your payload.

Failure probabilities don't work that way. Every rocket family, and every individual model, tends to get safer as time goes along as problems are remedied and fixed. The cost for innovating (needed to bring costs down) is that you have to start over on that curve. But the more you launch, the more potential problems you fix and the lower the odds of a future failure. There's always a high degree of randomness, of course, but in general you find a problem, you fix the problem, and the rocket is a safer vehicle for it.

Do recall how terrible the Atlas and Delta families used to be in terms of reliability. Things blew up, they learned, and were remedied. Heck, in terms of families, Falcon 9 is almost like a whole family rather than a single rocket thusfar... some of the changes, like switching to densified LOX, are pretty dramatic changes. They're trying to evolve and optimize it very, very quickly. But of course, that faces the learning curve reset problems above.

(Also, on that note, I think it's a bit premature to talk about the spotless record of the Delta-IV heavy, given that it's only ever had 9 launches, vs. Falcon 9's 29 (if you count AMOS-6... which if you're going to count it in the failure category, you should count it toward the total as well).

Some aspects of the Falcon design were designed to speed up the learning curve - and seem to have worked. Namely, the engines seem to have become quite reliable; part of the reason for going with so many engines was not just so that you can keep going after an engine failure, but also so that you're mass producing the engines and going through ten per flight; you're going to retire the risk a lot faster when using something in such large numbers. On the other hand, there's only two stages/pairs of tanks per flight, two COPVs, etc, so the learning curve is going to be - and has been - slower. . Falcon Heavy will help speed it up, of course, since there's four separate cores, all built similarly.

For a totally new (and frequently evolved) branch, Falcon 9's reliability is quite high; there are mature systems in use today with reliability records no better. But everyone wants you to approach 100%. At some point, SpaceX is going to have to stop with working on the "development branch" and offer up a "stable release" - that is, get the same identical cores with a long safe launch record, and stop changing them. And I'm sure they know that. But they seem to have a higher priority that they want to get to first: evolving their rockets to the point where they feel they can change the world. Not just "cheaper than everyone else", but "immensely cheaper than everyone else".

It's a tall order. But I fully sympathize with it.

On the upside from a stability perspective, there's really not much more need for evolution on the F9 production side, now that they're regularly landing cores. Getting multi-mission reliability, however, that's going to be a new challenge.