totm nov 2023 SpaceX Discussion Thread

[Racescort666]

Also, I came across this PDF linked on reddit: https://www.nasa.gov/sites/default/files/atoms/files/ccp_nac_july24_2017.pdf It discusses the technical readiness of SpaceX and Boeing for commercial crew operations.

[wumpus]

On 7/26/2017 at 11:46 AM, insert_name said:

High probability of failure on initial f9h launch

https://spaceflightnow.com/2017/07/25/musk-sets-expectations-low-for-maiden-falcon-heavy-launch/

It certainly will have the highest probability of failure for any Spacex launch since the maiden Falcon 9 launch, but I doubt that it is anywhere near as high.  I'm wondering if there are specific things he knows about (pogo, now with three rockets) or simply the inherent issues of anything that complex.  Most of the N1 issues (that blew up the N1, since it never worked we don't know how many more were lurking past these) involved bugs in the control structure that kept shutting motors down (presumably too early).  I'd assume that spacex has a wild advantage in building control logic using modern components that the Soviets never had.

Other rockets simply strap on "moar boosters": the Shuttle and Ariane 5 had many launches and nearly all successful.   Delta Heavy seems to be doing well, with only Long March 5 running 50/50.  I suspect that since each of those rockets has a failure or two, the most obvious time is the first flight.

[sevenperforce]

10 minutes ago, wumpus said:

It certainly will have the highest probability of failure for any Spacex launch since the maiden Falcon 9 launch, but I doubt that it is anywhere near as high.  I'm wondering if there are specific things he knows about (pogo, now with three rockets) or simply the inherent issues of anything that complex.  Most of the N1 issues (that blew up the N1, since it never worked we don't know how many more were lurking past these) involved bugs in the control structure that kept shutting motors down (presumably too early).  I'd assume that spacex has a wild advantage in building control logic using modern components that the Soviets never had.

Other rockets simply strap on "moar boosters": the Shuttle and Ariane 5 had many launches and nearly all successful.   Delta Heavy seems to be doing well, with only Long March 5 running 50/50.  I suspect that since each of those rockets has a failure or two, the most obvious time is the first flight.

Delta Heavy is not a very good comparison, given that we're talking about 3 engines vs 27 engines.

Will there be a 27-engine static fire?

[CatastrophicFailure]

6 minutes ago, sevenperforce said:

Will there be a 27-engine static fire?

Yup.

Musk seemed mighty nervous about it, too, in that release. 

[wumpus]

1 hour ago, CatastrophicFailure said:

Yup.

Musk seemed mighty nervous about it, too, in that release. 

Shouldn't that have been step one of the design process?  Even if you just bolted the three together (regardless of how you held them in place), you would still get a great idea of what you had to do with the 'real design'.

[CatastrophicFailure]

6 minutes ago, wumpus said:

Shouldn't that have been step one of the design process?  Even if you just bolted the three together (regardless of how you held them in place), you would still get a great idea of what you had to do with the 'real design'.

Just guessing here, but I don't think they have the facilities anywhere else for an all-up stacking/fire. 39A will be the first site upgraded to handle FH. 

[sevenperforce]

50 minutes ago, wumpus said:

Shouldn't that have been step one of the design process?  Even if you just bolted the three together (regardless of how you held them in place), you would still get a great idea of what you had to do with the 'real design'.

Not anywhere near my house, you aren't.

[Steel]

1 hour ago, wumpus said:

Shouldn't that have been step one of the design process?  Even if you just bolted the three together (regardless of how you held them in place), you would still get a great idea of what you had to do with the 'real design'.

55 minutes ago, CatastrophicFailure said:

Just guessing here, but I don't think they have the facilities anywhere else for an all-up stacking/fire. 39A will be the first site upgraded to handle FH. 

^My thoughts exactly^

IIRC Musk said they don't even have a way to simulate it, let alone do an actual live firing anywhere but the launchpad.

[1101]

1 hour ago, wumpus said:

Shouldn't that have been step one of the design process?  Even if you just bolted the three together (regardless of how you held them in place), you would still get a great idea of what you had to do with the 'real design'.

Except interesting effects like resonance won't be entirely realistic or relevant unless you do one with the actual vehicle.  For example, STS-1 had some interesting acoustic issues and damage that NASA did not design for or expect, up to the point where one of the pilots said he wouldn't have flown it to space if he'd known what the damage was.

[Kryten]

 Delta IVH had a partial failure on the first launch due to resonance effects with the fuel; the fuel sensors were exposed prematurely, causing early shutdown.

[KSK]

17 hours ago, 1101 said:

Except interesting effects like resonance won't be entirely realistic or relevant unless you do one with the actual vehicle.  For example, STS-1 had some interesting acoustic issues and damage that NASA did not design for or expect, up to the point where one of the pilots said he wouldn't have flown it to space if he'd known what the damage was.

Into the Black by Rowland White is a nicely accessible look at the development and flight of STS-1. Written as a drama (which it was for many reasons) but with a decent amount of technical material too, although I'll admit that I haven't personally checked any of it for accuracy.

If I recall rightly, NASA had test fire data for the STS SRBs and also made estimates of the expected acoustic loads by extrapolating from other solid rocket motors. The real thing turned out to be way worse than they expected, which is sobering given that this was an agency (as was) with fairly recent institutional experience of launching freaking Saturn Vs.

Personal anecdote - I had the great pleasure (and by pleasure I mean 'just managed to refrain from fanboy squeeing') and privilege to meet Chris Hadfield and listen to one of his after-dinner speeches. He had plenty to say about sitting on top of a pair of SRBs.  Paraphrasing slightly but 'if only one of them decided to light off - well you'd probably get an elementary school named after you.' Which was amusing at the time and a lot cooler now I know where he lifted the reference from.  

Edited August 3, 2017 by KSK

[CatastrophicFailure]

1 hour ago, KSK said:

'if only one of them decided to light off - well you'd probably get an elementary school named after you.

Unless you happen to have a crafty, spherical, slightly unstable robot buddy determined to get you into space... 

Chris needs thermal certain failure.

[insert_name]

SES 11 will use a flight-proven booster!

https://spaceflightnow.com/2017/08/04/ses-agrees-to-launch-another-satellite-on-a-previously-flown-falcon-9-booster/

[Silavite]

 

41 minutes ago, WildLynx said:

I wonder about Dragon 2 ... SpaceX don't want landing struts to go through the heat shield ... why heat shield cannot be one big landing strut?

It's how I imagine this:

Heat shield is directly attached to force plate (or structure), Since it not sturdy enough by itself, I think.

Force plate is attached to ship's body by 6 spring-loaded shock absorbers, each of them is locked in UP position by electromechanical lock.

When ship is about to land (into box of soft plastic balls for example), locks are disengaged and force plate and shield are moved by springs into DOWN position.

In case of emergency landing somewhere else shield may be scratched a little, but it would be least concern at zero altitude.

In case of one of the locks failing to disengage ... landing would be a bit harder.

Failure to engage lock ... cannot happen .. it would be noticed in VAB.

Premature disengage ... 2 sets of locks activated by 2 different circuits?

What you're describing sounds very similar to the landing bags used in the Mercury program.

Didn't the Manned Orbiting Laboratory plan to feature a hatch within the heat shield? I wonder how NASA/USAF planned to handle that.

[CatastrophicFailure]

1 hour ago, Silavite said:

Didn't the Manned Orbiting Laboratory plan to feature a hatch within the heat shield? I wonder how NASA/USAF planned to handle that

It welded itself shut during reentry, by design.  Not very reuse freiendly. 

However, the Soviet VA capsule also had a hatch in the heat shield, as was successfully reused more than once (I think). The space shuttle had all sorts of holes doors in the heat shield. 

I got the impression that the problem for SpaceX wasn't so much technical as bureaucratic-- they wouldn't have been able to get the system approved by NASA without extreme effort (cost).

[StupidAndy]

T- what? 8 days? YES!!

[Nibb31]

9 hours ago, WildLynx said:

I wonder about Dragon 2 ... SpaceX don't want landing struts to go through the heat shield ... why heat shield cannot be one big landing strut?

As I said above, I doubt that landing struts were the problem. The problem was paperwork and certification of the whole powered landing concept. The difficulty was getting the LOM/LOC statistics to a figure low enough to be accepted by NASA, which is difficult for an unproven concept.

Edited August 6, 2017 by Nibb31

[magnemoe]

15 minutes ago, Nibb31 said:

As I said above, I doubt that landing struts were the problem. The problem was paperwork and certification of the whole powered landing concept. The difficulty was getting the LOM/LOC statistics to a figure low enough to be accepted by NASA, which is difficult for an unproven concept.

This, you have to satisfy NASA demands, this is following NASA standards and land on water,  you want to make the capsule easy to reuse.

The legs would be pointless in an water landing, add that the holes, legs and shock absorbers would be an leak point making reuse harder.

Having an airbag between shield and capsule would work it would not be easy to reuse and an ground landing might easy damage the shield. protecting the shield is an important function of the legs. 
 

[Nibb31]

You don't understand what I'm saying. It's not about legs or airbags. It's not about designing technological solutions, that's the easy part. It's about probability calculations.

They have a global failure rate for LOC and LOM. The global failure rates are calculated from the failure rates of each system combined with the failure rates for redundancies. The failure rates for each of those systems are calculated from the failure rates of each component, and so on. With that methodology, and no proven track record, it's simply too hard to meet the target requirements based on calculations only.

For example (numbers pulled out of thin air): If the SuperDracos have a 1% failure rate (say they are made of 10 parts that each have a 0,1% chance of failing, this gives each complete engine a 1% failure rate), with 8 engines you actually have a 8% chance of one of them failing during descent. Of course, there is redundancy, but even if they get down to a 2% failure rate of 4 dracos failing, that might still be too high compared if NASA wants 0.1%, which makes the vehicle unable to meet NASA's requirements and unsuitable for crewed landings. Note that none of this is measured by actual testing. It's a purely calculated paperwork exercice. With actual flight testing, you can refine the failure rates for each component.

So in the end, maybe they have determined that even with flight testing, they can't guarantee a lower failure rate for all the components that are required for a safe powered landing (engines, altimeter, radar, sensors, avionics, guidance software, etc...) then there is no point in pursuing development of powered landing for NASA, because they have determined that there is no way they are going to reach NASA's LOC/LOM failure rate requirements.

Edited August 6, 2017 by Nibb31

[Skylon]

Just checking:

LOM is Loss of Mission and LOC is Loss of Crew, right?

Could an empty-of-crew Dragon 2 land from orbit, perhaps from a cargo mission (I think I read this was planned somewhere)

Coudn't they do a drop test where they land with engines? Possibly with some switched off?

[cubinator]

15 hours ago, WildLynx said:

When ship is about to land (into box of soft plastic balls for example),

Never mind everything else, I want to see this!

[magnemoe]

5 hours ago, Nibb31 said:

You don't understand what I'm saying. It's not about legs or airbags. It's not about designing technological solutions, that's the easy part. It's about probability calculations.

They have a global failure rate for LOC and LOM. The global failure rates are calculated from the failure rates of each system combined with the failure rates for redundancies. The failure rates for each of those systems are calculated from the failure rates of each component, and so on. With that methodology, and no proven track record, it's simply too hard to meet the target requirements based on calculations only.

For example (numbers pulled out of thin air): If the SuperDracos have a 1% failure rate (say they are made of 10 parts that each have a 0,1% chance of failing, this gives each complete engine a 1% failure rate), with 8 engines you actually have a 8% chance of one of them failing during descent. Of course, there is redundancy, but even if they get down to a 2% failure rate of 4 dracos failing, that might still be too high compared if NASA wants 0.1%, which makes the vehicle unable to meet NASA's requirements and unsuitable for crewed landings. Note that none of this is measured by actual testing. It's a purely calculated paperwork exercice. With actual flight testing, you can refine the failure rates for each component.

So in the end, maybe they have determined that even with flight testing, they can't guarantee a lower failure rate for all the components that are required for a safe powered landing (engines, altimeter, radar, sensors, avionics, guidance software, etc...) then there is no point in pursuing development of powered landing for NASA, because they have determined that there is no way they are going to reach NASA's LOC/LOM failure rate requirements.

We are simply talking past each other. NASA standards we agree on, you described it far better than I could. 
So they have to land on water, making the landing legs an complication and added weight and was dropped. 
 

5 hours ago, Skylon said:

Just checking:

LOM is Loss of Mission and LOC is Loss of Crew, right?

Could an empty-of-crew Dragon 2 land from orbit, perhaps from a cargo mission (I think I read this was planned somewhere)

Coudn't they do a drop test where they land with engines? Possibly with some switched off?

They could do drop tests from planes and will for the parachutes anyway. This has been done with all capsules. One Apollo boilerplate for drop tests was lost and found by Soviet who returned it. 
Now they could probably do 100 drop tests from planes but it would not simulate reentry heat damage on engines and systems like attitude radar.
 

[_Augustus_]

10 hours ago, Nibb31 said:

As I said above, I doubt that landing struts were the problem. The problem was paperwork and certification of the whole powered landing concept. The difficulty was getting the LOM/LOC statistics to a figure low enough to be accepted by NASA, which is difficult for an unproven concept.

Meanwhile Boeing gets away with their detachable heat shield and airbags, which is also an unproven concept.

Do I smell favoritism?

Edited August 6, 2017 by _Augustus_

[CatastrophicFailure]

2 minutes ago, _Augustus_ said:

Meanwhile Boeing gets away with their detachable heat shield and airbags, which is also an unproven concept.

Do I smell favoritism?

Well, in all fairness, it's not unproven. Dropping the heat shield after reentry has a very long track record with Soyuz, and landing airbags (with detaching heat shield) were used on Mercury.

Propulsively landing a capsule from Earth orbit has never been done, however. 

Like @Nibb31 mentioned, such data, even from unrelated vehicles, probably plays into their probability numbers. 

[Technical Ben]

That and costs. If the paperwork is possible and the costs are possible, but both prohibitively expensive, then they will change for a cheaper/easier option.