While the composite overwrap bears the main circumferential load from internal pressure, the liner is most certainly a structural component. Even though it carries only a fraction of the circumferential load (and typically a larger portion of the spherical tensile load at the domes), it has to take a portion of the load as well as withstand a compressive preload ensuring that the liner remains in intimate contact with the overwrap. If the liner shrinks more than the overwrap upon being loaded with cold gas or liquid then it can develop cracks or be subject to buckling under expansion after it warms up. The liner may also buckle if stresses in the overwrap are not evenly balanced or if the autofrettage process (which creates the preload condition by subjecting the vessel to pressures beyond MEOP for several cycles to yield the metallic liner.
Buckling is definitely not an allowable condition since it creates a condition that will lead to progressive failure with pressurization and thermal cycles. If this is actually a condition that SpaceX has previously observed and dispositioned for flight that that would represent unconscionable negligence. I suspect that the story may be a bit garbled any they may have just found that their margin against buckling was less than predicted by analysis. Although the standard recommends verification by extensive testing including a lead-before-burst (LBB) demonstration, most COPV qualification is done by proof testing a limited number of articles plus analysis and comparison to “similar” designs (although the definition of similarity is often stretched when it using different liner materials, a different resin system, and different overwrap parameters, and the only thing in common is that they’re using the same Toray carbon fiber that everyone uses because it is only domestically-manufactured aerospace grade carbon fiber).
Changing the propellant and ullage gas loading procedures may not be as simple a solution as imagined. SpaceX is dependant upon densified propellants to achieve the necessary performance for the upper payload range of the Falcon 9 v1.1 Upgrade (the articulations they are making on the nomenclature to avoid admitting that the Falcon 9 vehicle has undergone multiple significant redesigns is becoming increasingly clunky) or permitting return of the Stage 1 vehicle, and anything that slows down the propellant loading process will likely compromise densification to some measureable degree. This is why cryogenic propellant densification has been rejected in the past even after studies indicated it could offer marked performance advantages; they safety and reliability concerns with densification were considered too significant. This may end up ultimately contributing to SpaceX abandoning or their reusability concept as costs and limitations of reusability continue to mount.
Thanks for the info. I had been under the impression that the liner is quite thin and meant only to stop helium from passing; clearly, if it is taking a compressive preload, then it is more structural than I had thought.
I don’t get the impression that SpaceX is abandoning their superchilling efforts. Their planned changes to loading ops appear limited to helium.
It does look as though they will take a performance hit through their short-term changes, though I couldn’t say how much. Specifically, they say “this entails changing the COPV configuration to allow warmer temperature helium to be loaded”. Warmer helium means a larger volume, and since the COPVs are immersed in the propellant, that necessarily means less propellant. Maybe the volume is too small to notice.
I dunno what’s up regarding the external nomenclature; it doesn’t seem like the semi-intentional confusion would be targeted at the various certification groups, who aren’t going to spend more than two seconds thinking about the marketing names. Internally they have a “Block N” convention to distinguish the variants. Perhaps it’s to confuse the politicians.
*In the report, the panel said it was concerned that potential risks associated with loading propellants into the rocket with astronauts onboard were not “adequately understood.”
…
The panel said that NASA will also review the September incident, and it urged the agency to focus not only on the cause but also on the “load and go” fueling process.
Despite rounds of testing, issues can often pop up once procedures are actually used in a real-time environment, the report states.
“We strongly encourage NASA top management to scrutinize this issue and ensure that any decision to accept additional risk or novel risk controls with large uncertainties is justified by the value that will be gained.”*
I wish articles would link to primary sources. After a bit of searching, I found the report. A bit more context:
*A number of systems have not yet finalized design or completed testing. Challenges remain in several key systems, such as abort and parachute-related systems, in anchoring the analysis required to certify those systems for human flight. Additionally, there are issues and concerns surrounding the launch systems of both providers, such as the Centaur fault tolerance for Boeing and the “load and go” approach for SpaceX. (“Load and go” refers to a concept of operations in which the flight crew is strapped into the spacecraft prior to final fueling of the launch vehicle.) Both issues represent situations that are ultimately the result of the basic tenet of the CCP that puts the provider in control of the system design. NASA, in the oversight role of certification authority, determines if the hazards have been fully identified and the controls and mitigations implemented, and then decides if the resulting risk is acceptable. In this type of environment, the CCP must work diligently to ensure that acceptable risk is not defined by “the best we can do given the constraints.” The residual risk must be openly acknowledged and elevated to the appropriate level within the Agency for consideration.
One complicating factor for the “load and go” issue is the potential uncertainty in hazard identification and control. Identification of the hazards is dependent on many factors, which include understanding the environment in which the system will operate. In this regard, the Panel is concerned that the dynamic thermal effects on the system associated with loading densified propellants may not be adequately understood, which results in a higher level of uncertainty that must be factored into the risk determination.
On September 1, 2016, during the preparations for a pre-flight static firing of its Falcon 9 launch vehicle, SpaceX experienced an anomaly that led to loss of the vehicle and payload. Although the activity was being conducted in support of a commercial satellite customer, both NASA and the U.S. Air Force (USAF) were invited to participate in the subsequent mishap investigation. The Panel has also been informed that NASA is doing its own independent review. These mishap investigations and determination of causes and contributing factors will not be completed until after this 2016 Annual Report is published. We believe that the focus of the investigation must not be solely to identify and fix the specific cause of this mishap. It must focus also on improving the understanding of how the system functions in the dynamic thermal environment associated with “load and go” so that other previously unidentified hazards can be discovered. This is not a trivial effort. Despite testing at the component and subassembly level, systems often display “emergent” behavior once they are used in the actual operational environment. We are concerned that any determination of risk associated with “load and go” would have significant uncertainty. For these reasons, we strongly encourage NASA top management to scrutinize this issue and ensure that any decision to accept additional risk or novel risk controls with large uncertainties is justified by the value that will be gained. The decision should not be unduly influenced by other secondary factors such as schedule and budget concerns. *
The actual report doesn’t seem quite as damning as the news articles imply. In particular, the bolded line above is completely fair and basically just says that risk needs to be paid for.
Tangentially–the LA Times site required me to turn off my ad blocker to continue. For kicks, I did so. It immediately started loading hundreds of ads and their associated trackers. Ghostery picked up 128 before I manually stopped the page (and those are just the trackers). I can’t believe that people actually browse the web without ad blocking–some sites are virtually unusable.
SpaceX uses superchilled propellant to achieve higher density and thus performance.
For most rockets, the cryogenic components (liquid oxygen, in particular) are at the boiling point of the substance. This is a stable temperature, since any increase just means a small portion boils off. This portion can be topped off easily if the rocket has to sit around for a while (to load cargo or otherwise).
But superchilled propellant doesn’t have this property–it’s getting warmer every moment it sits on the pad. As such, SpaceX has a narrow window (about 30 minutes) between loading and launch. That’s not really enough time to load humans or cargo. So SpaceX has to either load propellant after the cargo, or live with the performance hit of less-dense propellant.
Since the loading process itself is a hazard, NASA clearly doesn’t like the idea of having people on board during this time. SpaceX thinks it’s safe enough, especially given the existence of the escape system. It’s not clear yet which argument will win in the end–SpaceX will have to be very convincing after the latest incident, which arose from exactly the kind of uncharacterized uncertainty that NASA is worried about.
BTW, SpaceX’s return to flight is planned to be tomorrow, from Vandenberg AFB, at 9:54 am PST. It’s a bit of a drive for me (4 hrs) but I think I’ll go for it. The last time I tried was a bust due to the weather.
Yay! Finally got to see a launch–fourth times’s the charm. Last time I went to Vandy, I heard the rocket but could not see it.
The plume is… bright. It’s one of those “way brighter than anything else in the vicinity” things like the sun or an electric arc. Very few things are so bright that they stand out against a bright, clear sky, but this was.
I was just able to see the stage separation and first-stage reorientation before losing it against the background. But that was pretty cool as well, as was the news (only found out much later, since cell service was so crappy there) that the stage had landed successfully.
Nice! I was there too; I didn’t see you. I was very surprised that I could clearly see the separation and maneuvering in my binoculars.
I had a couple radios with me so I was listening to the launch control frequency. I heard them announce “Falcon landed” but they did it with so little fanfare that I was afraid that was a euphemism for “Falcon drilled a hole in the barge.”
There was also a bit of drama up til about fifteen minutes before launch. It seems there was a boat in the restricted zone so someone had to go persuade them to move. I’ll bet that was a fun conversation.
Where were you? I parked, as before, on W Ocean. Parked at the end of the lineup and walked a couple of miles to the base entrance. Lots of people there. Ran into a guy from SpaceX; we talked a bit but didn’t press for anything juicy. He apparently hadn’t seen a launch from Vandy before so I had to tell him where to look :).
I forgot to bring a radio (not to mention looking up the frequencies). Hard to miss once it launched, though!
I’d heard afterward about the ship in the exclusion zone. Wouldn’t be the first time that kind of thing caused a scrub… would have really sucked here since it was an instantaneous launch window.
I was on West Ocean as well, short of 13th, right in front of that NASA building. I didn’t know if it was worth continuing to walk down further. It seemed like most people were just setting up chairs wherever parking dictated they stop. Is there more to see by walking to the gate?
A guy on Reddit who posted clear pictures of the pad gave me this location for next time…
Oh, you must have been close to my car. A black BMW 3-series, in case you were car-spotting :). I was right by the big NASA hangar-looking thing as well.
I don’t think it was a huge advantage to walk closer, but I got there an hour early and figured I’d go for a stroll. Looking at the map, it was only about a mile, and largely parallel to the launch complex, so I wasn’t really closer at all. I guess it was nice to have the palm trees in the shot.
Yep. They call it “one second” although I doubt it’s really as precise as that (given that the vehicle can make small corrections). It really just means that they intend to press the big button at a particular time, and if not everything is ready at that moment, the launch is scrubbed. With longer windows, they potentially have the ability to try again if they think they have some temporary issue.
Polar orbits–such as the one for these Iridium satellites–tend to have an instantaneous window. Equatorial orbits can give themselves a phase shift by perturbing their orbit and waiting for things to align. But for polar orbits this isn’t possible, and adjustments require a plane change maneuver–very expensive in terms of propellant. So the timing is very particular and they only get one shot every ~24 hrs.
Orbital rendezvous–such as cargo shipments to the ISS–also tend to have short launch windows. I believe SpaceX considers their ISS shipments to also have instantaneous windows, though again if you’re willing to wait, there is a fudge factor here.
Geostationary flights can have long windows; over an hour. There is a long coast phase which can be adjusted to compensate for any deviations here. SpaceX is limited by the superchilling here; I’m sure that would end most GTO launch windows before orbital mechanics did.
Yes. As I understand it, that means the orbit they want is overhead right at that moment. If they launch early or late (as in a window), they will have to burn more fuel (which they might now have) to correct back to where they need to be.
Oh, hot damn! I was looking through my pictures again and realized that I actually caught the staging event! It’s a little small and the stages are just a couple of pixels, but you can clearly see the two vehicles separate. Kinda curious why there appear to be multiple points of light in the first two shots; may just be a trick of the light (the exhaust illuminating the surrounding gases, say).
