SpaceX vs Lockheed/Boeing

[LegendCJS]

If NASA made "the 99% rule", isn't saying they didn't do the realtime calculus themselves a distinction with very little difference?

From my impressions it wasn't so much that the calculation determined that it had a 95% chance of success given the conditions the upper stage found itself in after releasing Dragon. 

It was that NASA and their monte carlo simulation group undertook months of study to determine that if everything was perfect, there was a better than 99% chance that the upper stage could safely relight and put the orbcomm satellite in a safe orbit with respect to the ISS. 

Re-doing the analysis for an upper stage with less fuel than accounted for in the first set of calculations would have taken the same amount of time (order of a month).  So the rough guess is a 95% chance of success, but the real answer was that it would have taken much longer than the upper stage life time on orbit to approve the re-light burn for any situation which fell out of the original envelope of the first analysis.

[docmordrid]

So this all-star group didn't evaluate scenarios that included situations in the F9 playbook (flight plan?)  like engine out?

[LegendCJS]

So this all-star group didn't evaluate scenarios that included situations in the F9 playbook (flight plan?)  like engine out?

My impressions could very well be in error of course.  But no trajectory computed on the fly fully autonomously to account for an unpredictable event can reasonably be considered to be in any "playbook"

[Jim]

So this all-star group didn't evaluate scenarios that included situations in the F9 playbook (flight plan?)  like engine out?

And another slur against NASA.  Face it, Spacex f'ed up and not NASA.

[joek]

Possible game changers:
Ion or Hall upper stages could significantly help SpaceX.

Not upper stages, but the effects of electric propulsion on LV selection are already being seen, as with Asia Broadcast Satellite (ABS) and Satmex; see  this thread.

[Elmar Moelzer]

No, because the obvious implication was that NASA is somehow at fault for what happened to Orbcomm. They didn't calculate anything, they set rules beforehand which SpaceX and/or Orbcomm didn't have to accept. The latter two could have decided to not fly secondaries at all.

Enough of this "NASA didn't allow", "NASA calculated" stuff already. It wasn't NASA's fault it was protecting its expensive orbiting asset.

I never said that it was NASAs fault, nor did I imply that. I simply said that NASAs rules did not allow it. When this decision was made is irrelevant. It is not slander against NASA either. They set the rules and both SpaceX and Orbcomm accepted them. That was my point. Hadnt it been for these rules (that everyone agreed on), the Orbcomm mission would have still been a success, despite the engine failure. It is still SpaceX fault that the engine failed (ultimately), but that was a risk that all parties had agreed on. So I don't think that calling the flight a failure is completely fair.
That was all I was trying to express. People should stop overinterpreting things.

[Hyperion5]

So this all-star group didn't evaluate scenarios that included situations in the F9 playbook (flight plan?)  like engine out?

And another slur against NASA.  Face it, Spacex f'ed up and not NASA.

Orbcomm probably should have waited for the much more powerful v1.1 to send up their experimental satellite.  Given Spacex has said they can lift over 50% more to LEO with the newer version, the engine-out margins for launching both Dragon and the Orbcomm satellite would have been a lot better.  If the same problem had happened on the v1.1 and not the v1.0, I expect Spacex would have had better than a 99% chance of making both orbital insertions.

[Robotbeat]

Time is just as important for Orbcomm. Waiting 6 months or so may not have been a better option for them, considering they did get some testing done.

[LegendCJS]

So this all-star group didn't evaluate scenarios that included situations in the F9 playbook (flight plan?)  like engine out?

And another slur against NASA.  Face it, Spacex f'ed up and not NASA.

Orbcomm probably should have waited for the much more powerful v1.1 to send up their experimental satellite.  Given Spacex has said they can lift over 50% more to LEO with the newer version, the engine-out margins for launching both Dragon and the Orbcomm satellite would have been a lot better.  If the same problem had happened on the v1.1 and not the v1.0, I expect Spacex would have had better than a 99% chance of making both orbital insertions.

False.  If the same thing had happened to v1.1 as v1.0, it still would have been a set of conditions that were outside the envelope of calculation for the simulation team, and the no relight decision would have been taken just the same.

[Robotbeat]

False, you don't know that. If the upper stage had enough propellant and margin for relight, they would've done it on the last flight. But 95% chance of successful burn isn't 99%.

[IRobot]

Time is just as important for Orbcomm. Waiting 6 months or so may not have been a better option for them, considering they did get some testing done.

AFAIK, this was the best thing for Orbcomm, they made most of the testing and insurance payed the launch...

[LegendCJS]

False, you don't know that. If the upper stage had enough propellant and margin for relight, they would've done it on the last flight. But 95% chance of successful burn isn't 99%.

To me it is clearly not a propellant level thing. It is about an envelope of safe probabilities as determined by intensive calculation and simulation:

The mission was not faced with the situation where the probabilities were: 95% chance of getting Orbcomm into the pre-planed orbit and a 5% chance of slamming into ISS because the fuel ran short.  That is ridiculous. 

Now I do not know this for sure, but since conservative (read NASA decision process) engineering is all about defining specifications and meeting them, with clear decisions on go vs no go based on being in spec or not, I conclude the key issue must have simply been about being out side of the envelope of monte-carlo simulated scenarios, and this was the reason that no relight was approved.

I guess our ability to infer what the decision process really was hinges on two things:

1) Was/Is the 95% number explanation just a quick summary way of explaining the fact that the monte carlo probabilistic model simulation data, if not focused on the specific scenario in question, would have had higher uncertainty in the estimated probabilities when it was used to infer this off nominal situation?

2) The definition of "successful burn."  Is success hitting the pre-planed orbit (safely).  Is success hitting any orbit (safely) that might have kept Orbcomm form re-entering as quick as it did  (maybe 6 months would have been useful)?  Or was success with respect to the ISS and its safety alone, and it did not matter where the stage and Orbcomm payload went as long as ISS was safe?  Saying the fuel was too low causing a "95% chance of success" doesn't make a whole lot of sense given some of the possible definitions of success.

[simonbp]

This seems like a rather silly argument, as the vast majority of SpaceX free-flight payloads (dare I say 99% by mass) will not be launched on an ISS-bound F9. Indeed, I doubt if they will sell any more than they already have.

IMHO, ULA/LM/Boeing's problem is not now, but 5-10 years from now. Whether it's SpaceX or Blue Origin or XCOR or the Tooth Fairy, anyone who can get a reusable first stage working will quickly build up a flight history sufficient to have very high reliability. Then the insurance companies will see that ULA's already tiny profit margin disappears. And LM and Boeing's shareholders will drop ULA faster than, well, UTC dropped PWR...

[joek]

AFAIK, this was the best thing for Orbcomm, they made most of the testing and insurance payed the launch...

Insurance hasn't paid for anythig yet.  Orbcomm filed a claim.  How much insurance will pay is TBD.

[joek]

IMHO, ULA/LM/Boeing's problem is not now, but 5-10 years from now. Whether it's SpaceX or Blue Origin or XCOR or the Tooth Fairy, anyone who can get a reusable first stage working will quickly build up a flight history sufficient to have very high reliability. Then the insurance companies will see that ULA's already tiny profit margin disappears. And LM and Boeing's shareholders will drop ULA faster than, well, UTC dropped PWR...

While I agree with your sentiment, I do not believe your arguments apply:
1. ULA has for all intents and purposes ceded the commercial market and is a captive U.S. government supplier.
2. The U.S. government self-insures launches; insurance companies' view of ULA's has little or no bearing on ULA's reliability record--unless and until ULA re-enters the commercial market, which given their cost structure and history, appears unlikely for the foreseeable future.
3. Launch reliability, and thus insurance premiums, are largely independent of whether the payload is government or commercial.  While ULA may have an excellent track record, until and unless they re-enter the commercial market, that record is irrelevant.
4. Launch reliability, and thus insurance premiums, are dependent on track record--obviously the higher SpaceX's tempo--assuming they are successful--the faster the premiums decrease. 

In short, SpaceX's ability to establish a reliable launch record has little to do with reusability and everything to do with launch rate (i.e., demand).  Unless you subscribe to the theory that reusability will create a significant increase in demand and thus in SpaceX's launch rate, leading to a virtuous cycle.  While reusability with lower costs and consequent increase in demand may happen, I think it's a bit early to make such a call.

[Robotbeat]

Reusability Will allow you to fly more test flights for the same cost. It will discourage sharing flights, which will in and of itself increase flight rate. The reduced cost certainly will lead to more people choosing them, so a higher flight rate. That isn't a theory, it is microeconomics 101. That is assuming simply that reusability significantly reduces cost.

And, of course, you have an advantage of getting the hardware back after a flight for examination.

Also, you could fly The reusable first stage a few times jut to wring all the bugs out at low cost.

[simonbp]

In short, SpaceX's ability to establish a reliable launch record has little to do with reusability and everything to do with launch rate (i.e., demand).  Unless you subscribe to the theory that reusability will create a significant increase in demand and thus in SpaceX's launch rate, leading to a virtuous cycle.  While reusability with lower costs and consequent increase in demand may happen, I think it's a bit early to make such a call.

Lowering costs is a secondary factor for reusability. The real reason for it initially is to gain confidence in the vehicle. If a single rocket vehicle has flown ten times without a problem, that's a good sign that it's reliable. And, any weaknesses in the design will accumulate and start to show problems.

It's a bit a paradigm shift from the old ICBM mentality of "mass produce and hope they all work the one time they are fired" to the aircraft approach of "overengineer reliability and get a ton of flight hours before trusting the vehicle". Eventually, I suspect we'll see a spacecraft certification process like the FAA currently does for aircraft, with the prototype earning a type certificate and each space vehicle having to maintain an individual flightworthyness certificate. And without that certificate, you won't get either insurance or government contracts.

So, even if the market never ever grows beyond what it is now, and reusable rockets are just as expensive as expendable, they will still completely dominate the launch market. And unless they start investing their own money in an RLV, LM/Boeing/ULA are going to find themselves completely shut out of the launch market.

[Jim]

So, even if the market never ever grows beyond what it is now, and reusable rockets are just as expensive as expendable, they will still completely dominate the launch market. And unless they start investing their own money in an RLV, LM/Boeing/ULA are going to find themselves completely shut out of the launch market.

Huh?  there is no logic in that

[go4mars]

I think he means 'assuming reliability is proven to be better than with expendable launchers'...

[oldAtlas_Eguy]

For the government customer who has very expensive unique sats (>$1B in costs each) the concern is LV reliability and not cost of the LV. If two LV's have the same reliability then hey will choose the cheaper, maybe. Choice of LV is made at the time the satellite contract is started, 3-8 years for a govenment sat depending on complexity and whether it shares technology with existing other sats such as use of a commercial common sat bus. So the LV must show it is as reliable or better than the other LV choices at that time. So launches in 2015-2020 for ULA is most likely protected from SpaceX total poaching. But as time progresses over that period ULA will loose more and more missions to SpaceX if they show they have just as reliable a LV as ULA.