Although the conclusions are still preliminary, SpaceX engineers believe they have tracked down the cause of last month’s spectacular failure of a Falcon 9 rocket in the skies over Cape Canaveral. The accident was apparently caused by the failure of a 2 foot long steel strut which was supporting a high pressure helium bottle. The bottle, which is located inside the second stage liquid oxygen tank, would have then accelerated up through the thin walled tank, ripping it apart.
The news came from company founder and CEO Elon Musk, who held a teleconference with reporters to provide an update on the ongoing investigation, and sheds some light on an initial assessment moments after explosion in which he posted via Twitter that the cause of a second stage liquid oxygen tank overpressure was “counter-intuitive.”
In the weeks following the accident, speculation had centered on the high pressure composite overwrapped helium tanks themselves due to the fact that SpaceX has had a number of launch scrubs due to helium leaks, and had recently moved production of the tanks from a vendor in Huntsville, Al, to in-house. The various issues related to the helium pressurization system had previously led Musk to describe it as a “pesky little molecule” and note that the company was looking at alternative solutions for the long run. In this case however, it appears that something far more fundamental led to failure. Designed to withstand forces up 10,000 PSI, the strut in question evidently failed at something closer to 2,000 PSI, causing the second stage to collapse even as the first stage plowed its way into the resulting conflagration. For what it is worth, the struts are not currently made by SpaceX, but provided by an outside vendor.
In the end, both stages exploded, and the now detached Dragon capsule carrying the CRS-7 mission payload descended intact until it impacted the Atlantic Ocean. Future missions will feature an abort software configuration which will deploy parachutes and possibly save the payload in the event of another launch failure. Musk appeared to chide himself for not already incorporating that feature, suggesting that perhaps the company had gotten a bit complacent with Falcon’s outstanding record prior to June 28th.
As far as a return to flight schedule, that is very much in flux, as engineers continue to attempt to validate analysis that came from a triangulation of acoustic signals within the helium tank and the discovery that during the course intensive testing, several similar struts also failed below the required threshold. While Falcon could fly again as early as September, the odds appear to greatly favor a launch after that point.
With the Russian Progress freighter now back in the ISS supply chain after its own failure, and an upcoming launch of Japan’s very capable HTV cargo craft on track, NASA, while concerned, is under no immediate pressure and will likely be more than content to wait for the final results to be announced.
While SpaceX is certainly under pressure from a commercial standpoint to meet a launch manifest which is already heavily loaded, a return to flight sometime this fall would likely not set the company too far back, and as Musk pointed out, there may be some unintended benefits from the accident which cost NASA and the taxpayer over $100 million in lost cargo and experiments. At the time of the last failure, that of Falcon 1 flight 3 in 2008, SpaceX numbered only 500 employees. Today it employs over 4000, meaning that the vast majority have never known the sense of loss which comes with a failed launch. Now however, they do, and that will almost certainly translate into increased vigilance in the future.
As for the press. At the risk of committing the same mistake in the event further analysis reveals a different, or more complex cause for CRS-7 loss, it is worth observing that some in the space press were quick to speculate whether or not factors in the SpaceX culture were responsible for the accident. Perhaps it was due to the insistence on bypassing traditional vendors and bringing production in house. Or, perhaps it was due to long work hours and employee fatigue, or the commitment to iterative improvement which has seen SpaceX continually upgrade its vehicles as competitors maintain a comfortably stable design.
While all those things may turn out to have played a role, at this point is appears that the failure was in fact more closely tied to one of the elements that the NASA and the traditional aerospace establishment love so very much, the widespread distribution of a vendor network which shows as many dots in as many states as possible. In the end, that may be your counter-intuitive data.