When a rocket launch fails, typically launches for that rocket type are halted while an investigation takes place to identify and correct the root cause of the failure. However, it is possible that the cause of the failure couldn't be conclusively identified, or that it was misidentified through misleading data.

This has happened before in aviation, most notably the De Havilland Comet which was brought back into service following an initial investigation in early 1954, only for a subsequent crash to finally ground the fleet until the true root cause of fatigue cracking was definitively identified.

Has something similar ever happened in the field of rocketry, where an incomplete or inaccurate investigation has led to further flights and failures of a launch vehicle?

  • 5
    $\begingroup$ The Soviet N1 moon rocket with its four consecutive launch failures is a likely candidate. $\endgroup$
    – dan04
    Commented Apr 28, 2023 at 16:31
  • 7
    $\begingroup$ @dan04, each of the four failures of the N1 was from a different cause (although there's a fair chance that the cause of failure #2 was mis-identified). $\endgroup$
    – Mark
    Commented Apr 28, 2023 at 18:45
  • 2
    $\begingroup$ @uhoh: The cause was not mis-identified. In fact, on the first airplane that crashed, the same thing had happened the day before, and was correctly identified and averted by the flight crew. The second crash happened before the investigation of the first was finished, so the identification of the cause cannot possibly have influenced it. In both cases, the cause was correctly identified as a malfunctioning MCAS system due to faulty sensor data (due to a lack of redundancy) which was not correctly understood by the flight crew due to insufficient training guidance by Boeing. $\endgroup$ Commented Apr 29, 2023 at 17:27
  • 2
    $\begingroup$ The ultimate root cause was that Boeing had underestimated the effect a malfunctioning MCAS system could have and thus designed it without sensor redundancy, thinking that it was merely a "convenience" system but not safety-critical. The FAA, in turn, did not spot this due to the fact that owing to a long and – until then very successful – history of cooperation, they essentially allowed Boeing to perform their own safety certification testing, merely reviewing Boeing's procudures but not the actual results. $\endgroup$ Commented Apr 29, 2023 at 17:29
  • 6
    $\begingroup$ @JörgWMittag That's a very Boeing friendly version of the story. Afaik, management had downplayed the importance of MCAS to avoid full recertification of the aircraft, and consequently requiring pilots to obtain special license for being allowed to fly it. The entire story bears all the signs of systematic management failure due to political goals getting more important than safety considerations. Quite comparable to the cause for the Challenger disaster and the VW "defeat device" invention, imho, (though the VW case was "just" pollution, not safety, it was the most plain criminal). $\endgroup$ Commented Apr 29, 2023 at 23:01

2 Answers 2


The Taurus XL launches of the Orbiting Carbon Observatory and the Glory satellites in 2009 and 2011 both failed due to fairing separation problems. Orbital spent two years after the OCO launch working on what they believed to be a fix for the issue, only to have the problem recur on the very next flight.

I don't think it's been publicly disclosed what the exact fixes they attempted were, but the actual problem turned out to be defective extruded aluminum components from Sapa Profiles, Inc that had had their test results falsified.

  • 3
    $\begingroup$ Man, the US industrial-military complex sure is a sweet gig. SPI's falsifications went on for nearly 2 decades, caused a loss to the government of over $700 million on the aforementioned satellites alone (ignoring all of the other contracts that SPI won during that period), yet all that the company had to do to "make it right" was pay a measly $46 million fine. $\endgroup$
    – Ian Kemp
    Commented Apr 30, 2023 at 22:44
  • $\begingroup$ @IanKemp not to mention all the engineering resources spent on trying to figure out why the fairing separation failed. $\endgroup$ Commented May 2, 2023 at 3:06

The root cause of External Tank foam shedding, which initiated the chain of events leading to the destruction of the Columbia orbiter on re-entry during the STS-107 mission, was initially mis-identified as "defects in the application of the foam insulation".

We informed the foam technicians at our plant in Michoud Louisiana that they were the cause of the loss of Columbia and then worked them overtime in training with new and exhaustive techniques on how to apply foam with no defects.

(Wayne Hale, How We Nearly Lost Discovery)

The next launch, a massive chunk of foam was shed, fortunately not destroying the vehicle and killing the crew this time.

Another long investigation and stand-down followed. Which resulted in:

It turns out that the thermal cycles associated with filling the tank could crack the foam, especially in areas where there were two or more layers of foam....Finally, this explained the Columbia foam loss. And the Discovery foam loss. And it had nothing to do with the improper installation of the foam.

  • 1
    $\begingroup$ "We informed the foam technicians at our plant in Michoud Louisiana that they were the cause of the loss of Columbia and then worked them overtime" - this is downright dystopian when taken in context with the rest of the answer! Imagine your boss comes to you and says "You personally murdered 7 people, so we're going to need you to work 80 hours a week from now on." and then it turns out it wasn't actually your fault. $\endgroup$ Commented May 3, 2023 at 3:01
  • $\begingroup$ @user253751 Yeah, I agree. Mr. Hale says in the article that he personally went and apologized to them, but still. $\endgroup$ Commented May 3, 2023 at 3:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.