Posts:
From:
Registered:

posted 01-15-2011 07:03 PM               

---

It seems to be fairly well published that even a 99.9% success rate in the Apollo hardware would still leave 5,000 pieces to fail (per mission). Has anyone ever tabulated how many things actually did go wrong through the course of the program?

There are many well-known instances, like 13's early center-engine cutoff and later the tank explosion; the broken switch in 11's LM; and 14's loose solder. I also suppose there are many things so minor that few people have taken note of it. It just seems that well over 99.9% was achieved, which is remarkable.

Posts: 1940
From: Fairfax, VA, USA
Registered: Feb 2007

Posts: 2231
From: West Jordan, Utah USA
Registered: Dec 1999

Posts: 1522
From: Greensboro, NC USA
Registered: Feb 2006

Posts: 438
From: NSW, Australia
Registered: May 2005

Posts: 2458
From: U.K.
Registered: Jul 2009

posted 01-16-2011 05:07 AM               

---

I would agree that the Mission Reports provide a good place to look for anomolies. However, it is important to distinguish failures that were more than covered by redundant systems and those that needed a little more thought on how to circumnavigate them.

A good example is the failure of an abort sensor in the sustainer engine of Aurora 7. Two sensors were required to trigger an actual abort. With just one failing the flight continued almost uninterrupted.

I doubt that many minor failures in switches were even recorded such was the built in redundancy, with only what may be loosely called 'major' reaching the flight report and the public domain.

It is worth remembering that the reliability was a major concern. Eventually it was decided that a failure of one every 100 times was unacceptable; a failure of one every 10000 times was too expensive; a failure of one every 1000 times was an acceptable risk. 5000 component failures, on average, represents just a 0.5% failure of the whole vehicle. Quite remarkable for that time.

One other thing to bear in mind. NASA's philosophy was to build components to design specification and assume they were reliable (cheap) not build a component and test it to verify its reliability (expensive).

Posts: 1522
From: Greensboro, NC USA
Registered: Feb 2006

posted 01-16-2011 08:38 AM               

---

One of the more esoteric failures I remember was in the "Range to go/Delta V" display in the Apollo 11 Command Module. One of the electroluminescent segments of the numerical display failed. I think it was in such a way that you couldn't distinguish 9 from 5 on one digit in the display. They had to get the correct reading from context, by watching the numbers change up or down.

If you add the ground support equipment to the list of failures, your catalog of breakdowns would go up exponentially.

Posts: 819
From: Huntington Beach, Calif.
Registered: Aug 2004

Posts: 2458
From: U.K.
Registered: Jul 2009

Posts:
From:
Registered:

posted 01-17-2011 09:18 PM               

---

stsmithva, thanks for the welcome! I'm looking forward to being a part of this great community!

Back to the original question, I guess you have to base the answer off of what NASA considered a failure. If a backup system was forced to be employed, I would assume you would chalk that up as a failure, because something, somewhere didn't work the way it was designed to and expected to.

Posts: 230
From: Maryland, USA
Registered: Oct 2008

posted 01-18-2011 01:16 AM            

---

Little known failure on Apollo 13 - when the crew was setting up their critical burn, they reached for the Command Module's Interval timer to tell them when to stop the burn. It's basically a high-tech kitchen time. Well, when Jack Swigert twisted the knob to set the timer - it fell off!

They managed just fine using their watches, and the original knob ended up in my collections (see it here).