Space Access Update #145 7/20/15
copyright 2015 by Space Access Society

SpaceX Diagnosis

Technical Background

The SpaceX Falcon 9 launch vehicle's two stages both use kerosene and liquid oxygen (LOX) propellants to feed their engines. Helium gas (He) is used to pressurize each stage's LOX and kerosene tanks to a few tens of psi so the propellants will flow into the rocket engine pumps smoothly. (The lightweight tanks also depend on this modest internal pressure for part of their structural strength.)

This requires a lot of He, as the pressure must be maintained until the propellant tanks are effectively empty. It is common industry design practice to mount high-pressure He bottles inside a rocket stage's LOX tank, because the low LOX temperature makes the compressed He gas significantly denser, allowing more to be stored in a given size He pressure bottle, significantly reducing the (substantial) mass of He bottles needed for a given launch vehicle.

One non-obvious consequence of this is that the main force on the He bottles during launch is upwards, due to the buoyancy of the relatively low-density He bottles in the relatively dense LOX. This buoyant force also increases with increasing G-force (as propellant is depleted while thrust stays constant, late in each stage's burn) since the "weight" of the volume of LOX displaced by each He bottle increases right along with the vehicle acceleration.

The Diagnosis

Elon Musk gave a press teleconference today, going over SpaceX's preliminary diagnosis of the Falcon 9 launch vehicle loss three weeks ago.

The hold-down strut for a high-pressure helium (He) bottle submerged inside the second-stage liquid oxygen (LOX) tank broke at its lower bolt-down point, at about 20% of the stress it should have survived, probably due to an undetected metallurgical flaw.

The helium bottle then broke loose, damaging the helium plumbing and releasing enough helium to over-pressurize the LOX tank, causing the LOX tank to split open, leading to the disintegration of the upper stage and loss of the rocket and the Dragon 1 Commercial Cargo capsule it was carrying to Station.

The hold-down strut was described as steel, about two feet long, with maximum diameter of about an inch, designed to withstand up to about 10,000 pounds-force tension, normally seeing a maximum load of about 3000 to 3500 pounds of tension. It failed near its lower bolt-down point at about 2000 lbs-force tension under about 3.2G acceleration. The initial failure location was identified via triangulation of the arrival-time at nearby strain-gauge sensors of the acoustic impulse from the metal snapping. The total time between that initial impulse and loss of telemetry was 0.893 second.

SpaceX had counted on the fact that these struts were designed to carry three times the maximum expected flight load, plus the (unnamed) outside supplier's guarantee that the steel they were made from meets specification. There is one such strut for every He bottle carried on both first and second stages, hundreds of such struts on each Falcon 9, thousands of such struts have flown previously with no problem. Post-accident load testing of a very large number of the struts from inventory revealed several that didn't meet spec, including one that failed before 20% of design maximum strength. Analysis of the test-failed strut revealed internal metallurgical flaws.

The diagnosis is not yet final - multiple interested parties will need to sign off before it is - and neither is the proposed cure. For now, though, the expected cure is to redesign the struts in question to be even stronger (possibly using a different material) and also to pull-test every strut (plus the other He tank attachment hardware) individually rather than relying on the 3x design factor of safety plus the vendor's material quality certification.

Musk said that he expects the strut redesign and replacement will relatively simple, but that the overall process of reviewing the data to make sure there's nothing else there that needs fixing and going over it all with all the interested parties - the FAA, NASA, USAF, and the various commercial customers - will take some months. The earliest he expects they might fly Falcon 9 again is September, and he emphasized it may take longer than that. (They don't know yet what mission might fly first.)

(Musk confirmed, by the way, that the Dragon 1 cargo capsule came away from the explosion in good enough shape that if they'd been able to deploy the capsule's parachutes, they could have recovered the cargo intact. He said that the Dragon 1 flight software will be modified to allow such deployment for Dragon 1 cargo capsules in future. He also emphasized that the Dragon 2 crew capsule has been designed for intact recovery under such circumstances all along.)

Quick Editorial

We are once again hearing noises about the Commercial Cargo Program needing more NASA "insight and oversight". See the House Space Subcommittee Friday 7/10/15 hearing on the future of Station, Representative Mo Brooks of Huntsville's question (at 55:23, just over a minute) plus Bill Gerstenmaier's mention in his opening statement (at 19:08, just over a minute) that it's easy to overdo "insight and oversight". (We might guess that Gerstenmaier was expecting the question.)

"Insight and oversight" is NASA-ese for overwhelming a given project with NASA supervisors, controlling every aspect in minute detail, and never mind the huge damage this does to cost and schedule plus the diminishing-returns effect (after a point, negative returns) on reliability. It's the Old NASA way of doing business that Commercial Cargo was intended to get away from.

Our read is, Old NASA apparently sees the recent accident as an opportunity to try yet again to assimilate the revolutionary Commercial Cargo (and Crew) programs. We expect we will be seeing more of this renewed "insight and oversight" push in coming months.

We will say once again: NASA insight, of course, within reason. NASA influence as well - as a major customer, naturally there's influence. But detailed NASA oversight of a commercial system with many other customers? Given the horrible 10x costs, endless delays, and multiple failed-programs track record of Old NASA oversight, no way.


Space Access Society's sole purpose is to promote radical reductions in the cost of space transportation. You may redistribute this Update in any medium you choose, as long as you do it unedited in its entirety. You may reproduce selected portions of this Update if you credit this Space Access Update as the source and include a pointer to our website.


Space Access Society

"Reach low orbit and you're halfway to anywhere in the Solar System"

- Robert A. Heinlein