Falcon Heavy

Leonard David has a story on its prospects for initial success, with quotes from Yours Truly.

To expand on the point, while he didn’t include it, I told Leonard in email:

The other issue is not launch reliability, but schedule reliability. SpaceX has aborted launches of the Falcon 9 when one or more of the engines was indicating performance issues on ignition. Three times as many engines means a lot higher probability of having an issue with one of them. For example, while I don’t know what the ignition reliability of a Merlin D is, suppose it’s 99% (that is, there’s a one in a hundred chance of failure to perform up to spec on ignition). For nine engines, that means the probability of a Falcon 9 aborting on the pad due to an engine issue would be one minus 0.99 to the ninth power, or about 8% per flight, or about once every dozen flights (which doesn’t seem that far off from their record). I don’t know what their flight rules will be for the Heavy, but if they have the same rule that they can’t take off with an underperforming engine, the reliability for twenty-seven engines will be one minus 0.99 to the 27th power, or 24%. That is, if they’re only 0.99 reliable per engine, and require all engines operating properly to take off, they have about a one in four chance of aborting a Falcon Heavy every single flight. That says to me that either they think Merlin reliability is greater than that (which it could well be) or that they’ll relax the rules to allow engine out from liftoff, or perhaps both.

Regardless, here’s hoping for a successful flight in less than three months.

15 comments on this post.
  1. Godzilla:

    I give it a 35% chance of success. They will do a hold-down test right? That one should be interesting to watch.

    I don’t doubt they could get this to work. But I think it was a really lame idea not to have facilities for full testing of Falcon 9 Heavy at Texas.

  2. Godzilla:

    I say this because I assume the reliability of the boosters can be increased if they have pre-flown stages which are known to work. Also I’m ignoring cancelled flights which don’t result in a loss of vehicle. And yes, it’s a guesstimate.

  3. Godzilla:

    Also what I suspect will doom it will be booster-booster interactions like vibrations or whatever.

  4. DarkstarSF:

    Wow. I never considered this until now. Thanks!

    Unless an FH was fully laden, or a Dragon, they should be able to go with fewer than 27 engines.

  5. DarkstarSF:

    After much additional thought, I’d suggest that a failure in the central core cannot be tolerated, but one in a strap-on can if it is balanced by an engine being shut down in the opposite core. This would depend on the remaining engines being able to support the weight of the payload.

  6. ken anthony:

    A failure in an F9 engine still allows a mission to succeed, so why would the same failure in a FH central core not?

  7. DarkstarSF:

    These actions would have to be gamed out in advance. “If engine #3 in core 2 fails, shut down engine #9 in core 3 and extend strap-on burns by 16.3 seconds.

  8. DarkstarSF:

    Just getting into orbit should work, but maybe not something time-critical like reaching a particular target like the ISS.

  9. DarkstarSF:

    Ya know, it might be prudent to just plan on launching with a pair of engines in the strap-ons turned off, as spares. This will solve potential orbital-mechanics timing issues withou scrubbing the launch.

  10. DarkstarSF:

    I don’t think this logic can apply to the central core, but I would love to be proved wrong!

  11. Paul D.:

    One possibility the F9/FH have is saving a launch by using some of the recovery propellant and expending some cores instead.

  12. ken anthony:

    The value of ignorance: He never thought it would take 15 years to get here. I wonder what the next decades will bring? Something out of left field probably.

  13. wodun:

    Rand Simberg is a consultant in space business and technology, and a self-described “recovering aerospace engineer” based in Jackson Hole, Wyoming.

    I thought you were based in California. As a libertarian, you might get a bit of a pass from the native people of Wyoming.

  14. David Spain:

    I also suspect SpaceX believes the reliability of the Merlin 1D is far better than 99%. But achieving multiple launches of F9H’s will be a sure way to find out. Another key factor will be reliance on reused boosters. At the beginning, if the customer will allow it, I’d try to make as many F9H launches using reused boosters as possible. In order to figure out the aging curve on boosters. I can get there more quickly if I have successions of F9H launches with booster recovery.

    Also I wonder if eventually F9H launches will feature routine throttle back of core booster to conserve fuel for orbital insertion and recovery of booster further down range? Would throttle down also extend the life of a core booster? Interesting speculation that “core boosters” would be ear-marked separately for this specific purpose due to “enhanced reliability” due to a less stressful operating regimen. On the one booster where reliability really counts.

  15. peterh:

    Whatever the current engine reliability is, I expect it to improve as the engineers catalog failure modes and devise ways to prevent them.

Leave a comment