Return to index: [Subject] [Thread] [Date] [Author]

Re: rho

[Subject Prev][Subject Next][Thread Prev][Thread Next]

Sorry Dennis, but I didn't say "statistical studies" I said "probabilistic
design method", something which is quite different.  Fundamentally there are
several approaches to design.  There is "rule based design" which is what fire
sprinkler designers use.  There is "deterministic design" which is what you and
I know and love.  And there is "probalistic design" which is, to my mind, a
somewhat wierd way to approach design that assumes that anything could
potentially fail under any load, but what is the risk?  While you and I
calculate a load capacity of a beam and perceive that in theory one ounce
greater load could induce failure, probalistic designers perceive that if you
tested a hundred or a thousand beams you could develop a curve which would
define the risk of failure for that particular situation.  It is used in the
nuclear industry to quantify risk and represents a fundamental basis of the LRFD
design approach, which is something else which we sometimes love to hate.

By the way, when I said "bashing" I was being somewhat tongue-in-cheek.  I know,
I know, I should type a silly happy face when I do it, but I refuse to type the
darned things.  If I could communicate more clearly I would not need such
things, it seems to me that typing them is like giving up on language.

Martin