Discussion Falcon Heavy might match first version of SLS.

[K_Jameson]

Exactly. Falcon 9 reliability is in the same ballpark as most modern launchers so far.

Coff... Atlas V... Coff coff... Ariane 5...

 

[Hlynkacg]

Except that most of the 16 launches had been done by a different configuration and alone the known anomalies during flight, that had to be documented during USAF and NASA launches, suggest a much lower quality than the PR-friendly statistic tells.

Also, remember that the Saturn V also had a much better statistic in its few launches - despite only plain luck preventing it from killing its crew multiple times. Successful missions alone are no indication of quality (The Space Shuttle had 24/24 successes before STS-51L)

Point being that this statement applies equally to most everyone else. ULA is the only competitor that can make a legitimate claim of superiority and their price tag reflects that.

Thus the question stands; Define "reliable".

If "Reliable" is simply "equal to or better than previous/existing manned platforms" I would contend that SpaceX has already met that criteria.

 

[Urwumpe]

If "Reliable" is simply "equal to or better than previous/existing manned platforms" I would contend that SpaceX has already met that criteria.

In the USA. Which is simple.

But what if you want to use reliable in the sense of "Reliable base for a heavy variant, so any schedules or performance estimates have a high probability to be held?" Is the Falcon 9 1.2, that is going to fly, even close to be considered a reliable baseline?

I have doubts there, especially regarding SpaceX tendency to use overly optimistic estimates.

 

[RGClark]

And they still have to develop a reliable Falcon 9.

!

Bob Clark

 

[Urwumpe]

The first launch of the Falcon Heavy was today announced to happen in April or May 2016. Who wants to bet if that is the final delay?

 

[RGClark]

Let me just point out that LEO performance and GTO performance do not scale up linearly, and SpaceX only said that the performance improvement would be 30%.

They've never said WHICH performance was increased by 30%, and it's possible that "performance" is not referring to "mass put in orbit" at all.

The 30% almost certainly is in regards to the payload sent to GTO because it was mentioned in the context of being able to send larger satellites to GTO and still be able to recover the first stage. First stage recovery reduces the payload to GTO and to LEO.

A very rough rule of thumb is payload to GTO is about half to LEO. So we would expect the payload to LEO for the F9 v1.2 very roughly to increase also about 30%.

See for example the list here:

https://en.wikipedia.org/wiki/Comparison_of_orbital_launch_systems

Admittedly though this very rough rule of thumb would be even more uncertain when scaled up to the three core version.



Bob Clark

 

[K_Jameson]

SpaceX is batting 15 for 16 at the moment.

Surely is twice the reliability of a Proton... :shifty:

 

[N_Molson]

Surely is twice the reliability of a Proton...

Proton-M

117 launches
106 successes
8 failures
3 partial failures

Reliability : 90.598 %

Proton-K

311 launches
275 successes
36 failures

Reliability : 88,424 %

(Source Wikipedia)

Falcon 9 (assuming 15s/16l)

Reliability : 93,334 %*

* without the first stage landing attempts failures.

Is the difference between the Proton-M and the Falcon 9 that huge ? No. Plus 16 launches is a bit limited to make stats.

Also, why the "first stage landing attempts" are not considered like partial failures ? This has to work flawlessly in order to make those fantastic reusability saving figures real. Its a part of the design, period.

 

[Hlynkacg]

Also, why the "first stage landing attempts" are not considered like partial failures ? This has to work flawlessly in order to make those fantastic reusability saving figures real. Its a part of the design, period.

For the same reason you don't include wind tunnel and test-stand runs in the successful sorties count.

 

[Urwumpe]

For the same reason you don't include wind tunnel and test-stand runs in the successful sorties count.

Exactly. But still, it lowers maturity of the system.

 

[Andy44]

Yes it's easy to get "mature" faster if you play it safe. The recovery attempts are hard to do, and SpaceX has to love with the loss of face after each failure, but as long as they pose no significant risk to the customer's mission, and as long as they pay for it out of pocket, it's a really good way to gather data on it, even if it ultimately fails.

In any case, the landing attempts have no bearing whatsoever on maximum payload figures, which assume an expendable booster with no landing gear or other equipment.

 

[Urwumpe]

In any case, the landing attempts have no bearing whatsoever on maximum payload figures, which assume an expendable booster with no landing gear or other equipment.

Yes, but even the attempts play into the economics of the company and into the unit costs.

 

[K_Jameson]

Is the difference between the Proton-M and the Falcon 9 that huge ? No. Plus 16 launches is a bit limited to make stats.

Proton isn't a very reliable launch system, never was and probably never will. Soviets had the good sense to never place humans on top of it. The meaning of my statement is that "twice the Proton reliability" is not a great achievement, after all.

Surely, SpaceX numbers are intended to improve. Falcon 9 is not yet a mature launch vehicle. But even when Atlas V or Delta IV were brand new LV, I don't recall so much failures or anomalies in their early missions.
But on the other hand Ariane 5 had a catastrophic failure in the very first launch (although caused by a software error and not a structural failure) and various other problems in the first launches, so I don't see why Falcon 9 can't improve over time as Ariane did. Surely it's not obvious, even for Elon Musk and for now, if I were a customer, I would feel safer to put my payload on Ariane or Atlas, rather than Falcon.

Also, why the "first stage landing attempts" are not considered like partial failures?

They are. But these features are still in development. We don't talk about a mature LV.

 

[Urwumpe]

But on the other hand Ariane 5 had a catastrophic failure in the very first launch (although caused by a software error and not a structural failure) and various other problems in the first launches, so I don't see why Falcon 9 can't improve over time as Ariane did.

Ariane had just two big failures, each with a new configuration of the launcher. Also there had been two partial failures in the interim upper stage. The second flight of Ariane 5 failed because of an excessive roll caused by the inner surface of the nozzle and the weakness of the roll thrusters. And V142 failed because of a propellant line leak in the upper stage, if I remember correctly

The first launch of the ECA variant also had a new core stage engine, Vulcain 2, which failed.

But then, since the 14th flight, there had been no more failures.

 

[francisdrake]

Maybe the FalconHeavie's power can be increased by more energetic boosters.

 

[ISProgram]

Man, wish I could texture like that.

Um, now an interesting question I have to ask is what factor contribute to determining what a "mature" launch vehicle is, and when it is considered "mature".

Mature launch vehicles in the world would include Atlas V, Ariane 5, Soyuz, Delta IV, Delta III, H-IIA/B, PSLV, etc.

I think a reasonable minimum launch count before you start deriving statistics and reliability should be 25-30 launches, most of the LVs I feel are "mature" have at least this much.

 

[Thunder Chicken]

Mature launch vehicles in the world would include Atlas V, Ariane 5, Soyuz, Delta IV, Delta III, H-IIA/B, PSLV, etc.

I think a reasonable minimum launch count before you start deriving statistics and reliability should be 25-30 launches, most of the LVs I feel are "mature" have at least this much.

Does this metric apply to Proton? It has hundreds of launches under its belt in several variants since 1965 - but a less-than-stellar success rate. What is the success "cutoff"? Is Proton "mature"?

 

[Hlynkacg]

As far as I can tell "maturity" is simply a measure of how paid-up the manufacturer is with the media and local powers that be.

Atlas and Ariane both had countless delays and several catastrophic failures that nobody talks about it now. Meanwhile, Proton is a "Mature" system that you'd need a death wish to ride and it's only pure dumb luck that Soyuz hasn't killed anyone lately.

 

[ISProgram]

What is the success "cutoff"? Is Proton "mature"?

Why my previous comment might not seem like I was asking this, that is the fundamental question I'm trying to figure out. It's entirely likely that the "maturity" of a given LV will never be static when you consider it.

On another note, Proton has had over 400 flights, and close to 40 failures. That seems fairly typical, 90% success ratio, IMHO. But when you consider that Proton has variants, it could be said that Proton M is certainly unreliable, and Proton K perhaps marginal. Atlas V and Delta IV don't have this issue, Ariane 5 and Soyuz might. Falcon 9 definitely does though, having gone through 3 reiterations since 2010.

 

[Urwumpe]

Atlas and Ariane both had countless delays and several catastrophic failures that nobody talks about it now.

Citation needed.

Also maturity is determined in simple words like that: Are you doing the same errors twice? In more complex terms, it is a matter of how well you are understanding what you are doing. A mature launcher is not about the flight hardware, it is about your operations from R&D to disposal being managed by you and not by external or random factors.

The two catastrophic failures of the Ariane 5 happened each on the first flight of a new Ariane 5 configuration. It was a flight test, it may fail. Once.

(And one Ariane 4 launch failed catastrophically because of sabotage, since then the security on CSG is really tight around launch)

Lack of failures for a new launcher is not a sign of maturity of the operator. Contrary. It suggests that PR is more important than continued improvement. Russian launches had also been perfect for the first launches of a new configuration, if it was possible to hide the errors. But eventually, all the small failures resulted in visible failures, that needed explanations.