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

Re: HSS vs. TS Properties For Design (WAS: Re: Copyright Infringement)

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

"David I. Ruby, P.E." wrote:

> Davis, LRFD is only the ASD with mystic factors. What benefit to the
> industry? the profession? The load is increased by 1.4 or more, the results
> are then factored down, Actually the  original LRFD factors were developed
> to insure the same results as ASD; in order that the method would be
> accepted more readily.

LRFD apportions the factor of safety better. A higher factor of safety is
accorded the live loads which to an extent are "a best guess". If I remember
correctly, the load factors were established by an ANSI committee based on Dr.
Galambos' papers.

> The method has been actively advanced by AISC since 1986, continuously
> taught in seminars and endorsed by all university Civil Engineering
> programs; Why has it not been accepted????? Can it be that it does not suit
> the needs of the design profession nor the steel industry.

When ACI adopted "Ultimate Strength Design" (LRFD for concrete) in the 1971
code, the engineering community went through the same turmoil about change. For
several years afterward, there would arguments between junior engineers and
senior engineers checking their work about whether or not the reinforcing steel
provided was adequate. I know since I was one of those junior engineers.

As an older engineer said to me years ago: "The hardest thing to do is to change
the mind of a set-in-their-ways engineer".

If using load factors is a problem for structural engineers, then are we going
to ignore the loss of the 1/3 stress increase for wind? Based on the way some
engineers run the numbers, not being able to use the 1/3 stress increase in
combined loadings will become a problem fo them. I haven't heard any complaints
in this change of design methodology.

I like LRFD since it takes into account the full strength of the cross-section
using 0.9 yield instead of designing members for .6 yielding on the extreme
fiber of the member.

> There is value in knowing the failure mechanism of your structure, but it is
> hard to imagine that multiplying loads by factors then applying phi factors
> to predict the ultimate load capacity gives anyone a better feel with a
> material that is routinely furnished to yields 10% higher that advertised. I
> understand that this is of concern in high seismic areas and the procedure
> maybe of great value their; but try to apply the method to the design of
> crane runways, industrial buildings and stacks where serviceability is the
> prime design criteria.

We need to introduce a new paradigm - Design members for serviceability and
connectability, then check for strength. With 50 yield steel, I am finding that
serviceabilty requirements control the member sizes.  This is a radical
departure from the classical design methods tught in school. I think you will
find that you don't have to do so many calculations.