SpaceX Falcon 9 FT - ORBCOMM-2 - Dec. 21, 2015 (Return To Flight) DISCUSSION

[enzo]

They really need to settle on a vehicle configuration and start cranking out some launches.  The tinkering and refinements are nice and all but they to generate revenue, reduce backlog and show customers they can deliver.

A good point, but it could be wrong. Reusability factors into the business plan. It is more profitable to reuse than to increase production rate. They may have figured that it's better to increase flight rate by enabling reusability of GTO flights via FT, than to do so via increased production. We already know Musk's modus operandi, to take large risks using the full extent of available time and money, in order to accelerate the payoff.

[meekGee]

They really need to settle on a vehicle configuration and start cranking out some launches.  The tinkering and refinements are nice and all but they to generate revenue, reduce backlog and show customers they can deliver.

A good point, but it could be wrong. Reusability factors into the business plan. It is more profitable to reuse than to increase production rate. They may have figured that it's better to increase flight rate by enabling reusability of GTO flights via FT, than to do so via increased production. We already know Musk's modus operandi, to take large risks using the full extent of available time and money, in order to accelerate the payoff.

We had tens of such posts when SpaceX transitioned from 1.0 to 1.1.
The customers will run away, they don't understand about revenue, etc.
Anyone still thinks it was a bad move?  doubt it.

This is just a repeat, and there are less such posts, and they are less abrasive.   Progress.

Wait till the next revision comes around, the one that incorporates changes based on stage recovery and reuse. 

[nadreck]

Indeed meekGee, SpaceX evolving vehicle design extends to their logical conclusion the necessary strategies that lead them to develop of the things they developed internally.

SpaceX manages its development differently than other companies, a lot of the differences come from how vertically integrated their business is, but also the adoption of some of the technologies and practices that they have (gas flow modelling, rapid prototyping, different employee career development paths, as for examples). 

[ChrisWilson68]

Is it reasonable to conclude here on Nov 1, with no confirmation of a successful test at McGregor, and no indication of a range reservation, that Dec 1 is no longer a possible launch date?

No. They have done launches with less than four weeks from stage testing to launch.

Granted, this being the first full thrust stage, with densified propellant, there may be pad-related gremlins to get rid of once the stage gets to Cape, so... who knows.

Before the loss of CRS-7 they were supposedly going to start launching the FT in September.  So they should have been very close to testing a FT stage at the end of June.  Here it is Nov 1 and they haven't.  There is over lapping work with the strut replacement, but the booster is on the test stand so the struts are done and there is not test yet.

They really need to settle on a vehicle configuration and start cranking out some launches.  The tinkering and refinements are nice and all but they to generate revenue, reduce backlog and show customers they can deliver.

Converting to FT is obviously requiring more work than SpaceX has indicated.  Simply because we haven't seen a test of the full FT stage yet.   I think the move to Full Thrust is at least equal if not greater than the RTF effort.

Why do you think work for the new full-thrust configuration and work on investigating the launch failure and fixing it don't affect each other?  It's hard for me to imagine how the same people wouldn't be involved in both efforts.

They got unlucky and had a launch failure right when they were doing a major vehicle transition.  The cause of the failure wasn't immediately obvious, so they needed everyone available working on it for a while.  And then fixing the root cause turned out to require replacing a major structural element.

Given the circumstances, I'd say they've bounced back quite quickly.

As to whether they should be moving to a new full-thrust configuration or not -- it's long term benefit for some short term pain.  I'm glad they're doing it.  It's not clear there would have been any disruption at all from the transition to customers' schedules if not for the launch failure.

[Nomadd]

 The root cause of the failure wasn't a broken bolt. It was SpaceX failing to insure they wouldn't have a broken bolt. The fix is a heck of a lot more than fixing one piece and will apply to the whole rocket and all future rockets, because it's a fix in the way they do things.
 I can't blame them for failing to set an accurate schedule for everything the week after the failure. As always, you plan for the earliest possible rtf because it's a lot easier moving the date out that it is to move it up.

[Semmel]

The root cause of the failure wasn't a broken bolt. It was SpaceX failing to insure they wouldn't have a broken bolt.

Well said. Only SpaceX knows how many other parts of the rocket might fail due to the lack of noticing any errors. This failure is tragic but I am sure SpaceX learned a lt from it.

[shm6666]

Indeed, we all want to see a successful flight. Testing stuff takes time. I myself work in the computer industry. We sometimes have failures but when we do we don´t get spectacular explosions in the server room, for the most times at least. After a failure we test that who failed to get an understanding. It takes time to test.

I think SpaceX is doing the right thing not to rush things, but to take it slow and see to it that they get the testing done. In the end I think they will have a safer and more reliably rocket after this. 

[Proponent]

I thought they don't run engines close to stochiometric - I believe they run fuel-rich to reduce combustion temperature.

Keeping the temperature down may be one reason for a non-stoichiometric mixture ratio, but it's worth bearing in mind that specific impulse does not usually peak at the stoichiometric ratio.  On top of that, maximum specific impulse does not necessarily coincide with maximum delta-V.  Ideally, the mixture ratio would vary during flight.  In princple, the mixture ratio would be set for maximum impulse density (usually oxidizer-rich) at lift-off, changing toward (but not necessarily reaching) maximum specific impulse at later phases of flight.

[abaddon]

The root cause of the failure wasn't a broken bolt. It was SpaceX failing to insure they wouldn't have a broken bolt. The fix is a heck of a lot more than fixing one piece and will apply to the whole rocket and all future rockets, because it's a fix in the way they do things.
 I can't blame them for failing to set an accurate schedule for everything the week after the failure. As always, you plan for the earliest possible rtf because it's a lot easier moving the date out that it is to move it up.

I don't think you meant this, but the implication is that SpaceX doesn't validate anything in their rocket more than they were doing with the struts.

No, the fix won't apply to "the whole rocket", it will apply to any other parts of the rocket that had an insufficient inspection/test/whatever process in place.  Many parts of the rocket presumably already have rigorous process in place, or they wouldn't have made it this far.

How much of an impact this will have is a mystery as we of course have no visibility into any of this.

[kevin-rf]

No, the fix won't apply to "the whole rocket", it will apply to any other parts of the rocket that had an insufficient inspection/test/whatever process in place.  Many parts of the rocket presumably already have rigorous process in place, or they wouldn't have made it this far.

I think better wording would be the fix will apply to any area they found to have insufficient inspection and testing. Other areas might exist that they feel they have enough insight, but may not. Only time will tell.

[baldusi]

Kerosene is a much more molecularly-complex mixture of refined hydrocarbons. Methane is CH4. Methane exhaust will be pale blue/purple flame with very little soot, assuming the engine runs close to stoichiometric combustion. "Smoke" is generally carbon-based particulates made up of partially-combusted carbon and whatever else is in the fuel and/or oxidizer.

I thought they don't run engines close to stochiometric - I believe they run fuel-rich to reduce combustion temperature.

Since usually the oxidizer is denser than the fuel, running fuel-rich usually gives better isp due to the lower molecular weight of the mixture. And yes, running stichiometric would also result in too high temperatures. Heat is good because it can be made into momentum thanks to the nozzle, and usually means higher Pc. In fact, the turbines are mostly a heat engine that converts heat into work.
But both turbines and main combustion chambers are material-limited in working temperature and pressure. So, you have multiple variables and constraints to handle at the same time.

[ChrisWilson68]

The root cause of the failure wasn't a broken bolt. It was SpaceX failing to insure they wouldn't have a broken bolt.

Well said. Only SpaceX knows how many other parts of the rocket might fail due to the lack of noticing any errors. This failure is tragic but I am sure SpaceX learned a lt from it.

Actually, we don't really know to what extend SpaceX did anything wrong.  They've indicated that a part from a supplier (the strut) failed under much smaller loads than it was rated for.  There's only so much a company can do to guard against problems from a supplier.

One could imagine a scenario in which SpaceX had a rigorous qualification testing program for the struts and even a rigorous testing-to-failure of a random sampling of struts on an ongoing basis and just got unlucky because only one bad strut came from the supplier and there was no reasonable way to detect the defect.

It's also possible SpaceX wasn't doing enough checking on the supplier.  We just don't know.

[sdsds]

One could imagine a scenario in which SpaceX [...] just got unlucky because [...] there was no reasonable way to detect the defect.

One can imagine a design that doesn't incorporate parts that can have undetectable defects. But a company doesn't end up with that kind of design as a result of "luck."

[the_other_Doug]

One could imagine a scenario in which SpaceX [...] just got unlucky because [...] there was no reasonable way to detect the defect.

One can imagine a design that doesn't incorporate parts that can have undetectable defects. But a company doesn't end up with that kind of design as a result of "luck."

Since all parts made by human beings and their machines can possibly have undetected defects (unless every single part is tested to destruction, in which case, you still cannot have completely, provably defect-free parts riding upon  your rocket), then I suppose you are advocating that we use parts for rockets made by supernatural beings?  Since, after all, nothing in the natural world can possibly create parts that have zero potential for undetected defects.

Just sayin'...

[sdsds]

There's no need to invoke supernatural beings! Yes, all parts can have defects. Most defects are of kinds that can be detected. In general there is no need to test-to-destruction to eliminate defective parts.

Putting this back into context: what SpaceX is doing now, they could have done before.

[whitelancer64]

One could imagine a scenario in which SpaceX [...] just got unlucky because [...] there was no reasonable way to detect the defect.

One can imagine a design that doesn't incorporate parts that can have undetectable defects. But a company doesn't end up with that kind of design as a result of "luck."

The only way to have caught that particular defect would be by 100% testing of all the struts to full flight loads.

For hundreds of relatively simple structural parts that were over-designed by a factor of 3, that's overkill.

Normal procedure for that sort of part is to test a percentage of random samples from a production batch, usually 10 or 15%, and then certify the entire lot based on that.

Why is it done that way? If a part is going to fail very far below design strength, it usually means there is something very wrong with the production process of that part. If something is very wrong with the production process, then it should have happened to all of the parts that were in that production run. That's the logic behind testing a percentage of a lot of parts.

For a severe outlier failure, there's likely an out-of-process production issue somewhere along the line. That can be VERY hard to detect.

There's no need to invoke supernatural beings! Yes, all parts can have defects. Most defects are of kinds that can be detected. In general there is no need to test-to-destruction to eliminate defective parts.

Putting this back into context: what SpaceX is doing now, they could have done before.

Elon Musk indicated a few months ago that they were thinking of changing the design of the part, either strengthening its design by "an absurd amount", and / or using a different alloy to make it, and / or changing the supplier for the part. From an outside perspective, that largely covers the bases as far as design and materials and production issues, and reduces the chance of an outlier failure, though of course we aren't privy to SpaceX's engineering assessments of that part.

[the_other_Doug]

There's no need to invoke supernatural beings!

Well, um, you did, with the bald statement:

One can imagine a design that doesn't incorporate parts that can have undetectable defects.

And, um, no -- one can't.  Since nature (and humans) cannot, by definition, create parts that cannot have undetectable defects, you left only one other logical possibility.

Glad to see you backed off from that statement.  Always glad to shine a light into small, dark places...

[abaddon]

Putting this back into context: what SpaceX is doing now, they could have done before.

That's almost tautological, and could almost apply to anyone doing anything ever.  Not sure what your point is?

[Kansan52]

IMHO, the point is, if the supplier would have used proven sample testing, the supplier could have stopped this problem. SX trusted their supplier. If SX had not trusted the supplier and done their own sample testing, then they would have info to know there was a problem. Or SX could have someone review the supplier's testing and maybe seen a flaw in the testing before losing a vehicle.

We are guessing. We do not know the testing by the supplier or by SX that went on before the failure.

What has been pointed out in several posts about the site, that there are testing techniques that likely would have caught there was a production problem before it destroyed a vehicle.

[whitelancer64]

IMHO, the point is, if the supplier would have used proven sample testing, the supplier could have stopped this problem. SX trusted their supplier. If SX had not trusted the supplier and done their own sample testing, then they would have info to know there was a problem. Or SX could have someone review the supplier's testing and maybe seen a flaw in the testing before losing a vehicle.

We are guessing. We do not know the testing by the supplier or by SX that went on before the failure.

What has been pointed out in several posts about the site, that there are testing techniques that likely would have caught there was a production problem before it destroyed a vehicle.

Faulty assumption. The supplier most likely DID use standard sample testing procedures. SpaceX may have sample tested the parts when they arrived, too.

But therein lies the rub. ONLY 100% testing, that is, testing of all parts, up to full flight loads could have caught the defect. That's NOT standard testing procedure for that type of part, and it shouldn't have to be, particularly since the strut was over-designed by a factor of 3 to begin with.