sci.math
[Top] [All Lists]

Re: Has bad mathematics lead to engineering disasters?

Subject: Re: Has bad mathematics lead to engineering disasters?
From: Michael Press
Date: Thu, 07 Dec 2006 01:16:17 GMT
Newsgroups: sci.math
In article <RPGdnUMGmuof_erYRVnyvg@xxxxxx>,
 "Nick" <tulse04-news1@xxxxxxxxxxx> wrote:

> "Michael Press" <jack@xxxxxxxx> wrote in message 
> news:jack-3DAA6E.14374106122006@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
> > In article <UZWdnSp_49ajAevYnZ2dnUVZ8tidnZ2d@xxxxxx>,
> > "Nick" <tulse04-news1@xxxxxxxxxxx> wrote:
> >
> >> "Lighting Rep" <lightingassociates@xxxxxxxxx> wrote in message
> >> news:1165380705.236078.193910@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
> >> >
> >> > Narasimham wrote:
> >> >> Phil Carmody wrote:
> >> >> > "Richard Henry" <pomerado@xxxxxxxxxxx> writes:
> >> >> > > Narasimham wrote:
> >> >> > >
> >> >> > > > Sorry,it is not so as we gather.As Progenoskes also 
> >> >> > > > observed,Tacoma
> >> >> > > > Narrows Bridge was a  new physics/mechanics situation as a 
> >> >> > > > strange
> >> >> > > > unsteady state aerodynamics phenomenon that had not been
> >> >> > > > encountered in
> >> >> > > > engineering situations up till that point of time...naturally 
> >> >> > > > not
> >> >> > > > mathematically modeled to be handed down to practising design
> >> >> > > > engineers.The failure analysis commitee also cleared them of any
> >> >> > > > charges of negligence. Even in late nineties a Boeing aircraft
> >> >> > > > empennage was destroyed at Japan Norita airport due to such
> >> >> > > > causes.However the degree of goofitude in Hyatt Hotel is an 
> >> >> > > > order
> >> >> > > > of
> >> >> > > > magnitude higher.
> >> >> > >
> >> >> > > The Hyatt Hotel walkway was not built as designed, so it is hard 
> >> >> > > to
> >> >> > > fault the designer.
> >> >> >
> >> >> > Not built as *originally* designed. They changed the designs because
> >> >> > they were not practical from the manufacturing standpoint. It was
> >> >> > built in accordance with the new designs. Designers somewhere, and
> >> >> > the people who reviewed and signed off those designs, are to blame.
> >> >> >
> >> >> > And you *can* fault the original designer too for designing 
> >> >> > something
> >> >> > for which the components and construction were practically 
> >> >> > impossible.
> >> >> >
> >> >> > Phil
> >> >> > --
> >> >>
> >> >> If you build 10 stories above the foundation, the foundation has to
> >> >> take the weight of 10 stories in compression.If you hang 10 stories
> >> >> down from the ceiling,the ceiling has to take the weight of 10 stories
> >> >> in tension.What prevented adoption of floorwise tapering I-Beams or
> >> >> welded taper tubes? It is the speed needing to ignore blind
> >> >> spots.Granted things are more clear in hindsight.However insistence on
> >> >> routine elementary but mandatory structural checks by a second party 
> >> >> in
> >> >>
> >> >> large constructions e.g., by finite element analysis would have
> >> >> routinely eliminated such loopholes when elemental modeling would be
> >> >> considered.This is not maths but managerial practice of safety concern
> >> >> in engineering. But"Safety does not sell."
> >> >
> >> >
> >> > Ummm - Remember - Most everything, built, designed, engineered,...
> >> > Is completed by the LOW bidder - You get what you pay for...
> >> >
> >>
> >> My father is a retired structural engineer and he knew, for instance,
> >> someone who was killed in the Melbourne bridge disaster (box girders, I
> >> believe) where the bridge collapsed whilst it was in construction.
> >>
> >> There is the well-known Tay Bridge disaster where the bridge collapsed 
> >> over
> >> the river Tay with a train on it, shortly after it was completed.
> >>
> >> But bridges don't get built by sitting on a sheet of paper or on the
> >> computer.
> >>
> >> There is also the French cathedral (possibly Beauvais) that collapsed in 
> >> the
> >> thirteenth century. The roof was bigger than any previous cathedral and 
> >> if
> >> the proof was in the pudding the pudding collapsed.
> >>
> >> See
> >> http://www.tech.plym.ac.uk/sme/FailureCases/List_Engineering_Successes_Failures.htm
> >
> >
> > Many cathedrals collapsed during the era; mostly from
> > wind load, which is why the ones still standing are all
> > graced with those buttresses, flying and otherwise.
> >
> > The box girder fiasco is an example of
> > over-engineering. By optimizing the strength to weight
> > ratio in each of two dimensions as if the other
> > dimension did not exist, they built in a failure mode
> > that does not exist in one dimension; a failure mode
> > that is complete collapse.
> >
> > From Catastrophe_Theory_and_its_Applications [1]
> > "For an analysis of the stiffened plate, in which
> > equating the loads at which a whole_plate buckling and
> > between-stiffener buckling occur leads to a hyperbolic
> > umbilic catastrophe see [23].
> > (Such plates are a standard component in box-girder
> > bridges whose wreckage was decorating the motorways of
> > Britain and the rivers of Australia a few years ago:
> > but so much else was found wrong by the enquiry that
> > efficient cause, proximate cause, ultimate cause etc.
> > are less than clear.)"
> >
> > [1] Poston and Stewart, 1978. Pitman.
> > [23] Thompson, J.M.T and Hunt, G.W. "A bifurcation
> > theory for the instabilities of optimization and
> > design." In
> > _Mathematical_Methods_in_the_Social_Sciences (D.
> > Berlinski, ed.) Synthese, to appear.
> 
> My father has said that a well-known engineer was prepared to go to the 
> limits of current knowledge.
> 
> In pure maths, noone gets killed...

Search the web for the phrase "logically and morally correct".

-- 
Michael Press

<Prev in Thread] Current Thread [Next in Thread>
Privacy Policy