[ExI] Musk announces cause of Spacex explosion

BillK pharos at gmail.com
Tue Jul 21 08:33:09 UTC 2015


On 21 July 2015 at 01:51, Stephen Van Sickle  wrote:
> http://www.geekwire.com/2015/spacexs-elon-musk-traces-falcon-9-rocket-failure-to-busted-strut/
>

Failure modes and built-in redundancy are interesting.

In the olden times, if failure of a cable or strut was critical, then
you just built two in. So that if one failed, the other took over the
load and the structure didn't collapse.
e.g. bridges and skyscrapers.

Modern design techniques can research expected loads and make one
strut strong enough so that it is expected to never fail and therefore
the backup is no longer required.
In rockets especially, saving weight is critical.

But this leads to problems of catastrophic failure if an unexpected
load occurs, or as in the Spacex case, some struts have undetected
manufacturing defects.

This may be an unnoticed existential risk. Building more efficient
systems closer to expected design limits could mean that unexpected
circumstances cause a catastrophic failure rather than a soft failure.

The human body has many soft failure modes. Bits can stop working or
degrade performance and the human still keeps going. (Called ageing).
:)
But this is an expensive way to build things.  You don't get soft
failure modes built-in to lowest quote manufacturers.


BillK



More information about the extropy-chat mailing list