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

Re: Steel Yield Strength, A36 vs A 572

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

On June 3, Keith Fix  wrote:
Doesn't this create a dangererous situation wherein "yield" occurs much later
than anticipated?  And are you telling me that my double angle brace might
experience a rapid failure in an earthquake, as opposed to the elongation I
would prefer?

The answer to your question is, basically, yes.  SEAOC first became aware of
this back in 1992, when several of the new smaller producers of shape started to
advertise dual certified steel for wide flange material, in which you could
order either A36 or A572/50 material, for the same price - what a deal.  This
also meant if you ordered A36 you could get A572.  SEAOC Seismology became very
concerned about this becuase it meant, eactly as you surmise, that structures
would have much greater overstrength than anticipated.  Overstrength is
generally a good thing, as long as you are aware of it, and act accordingly.
SEAOC's initial concer was for moment frames and EBFs. We were worried that:
 1- in moment frames unanticiapted overstrengh in beams
    would lead to unintentional weak column frames
 2- in EBFs, unanticipated overstrength would defeat the
    effectiveness of the link as a fuse.

With the assistance of Hank Martin of AISI, Nabih Youssef and I made a
presentation to the Structural Shape Producer's Council in 1992, to tell them
that stronger was not always better.  They had no clue.  At our urging they did
a statistical study of variation in yield strength for the 1993 production year,
and it was clear that yield strengths were much higher than assumed by most
designers.  STill nothing happened to help the designer.  It was only after
several years of SEAOC threating to introduce code provisions that would put an
upper limit on yield strength for material, that AISC worded with SSPC to solve
this problem for wide flange shape.  IN 1995, AISC came out with a Supplemental
Requirement to limit yield strength in A572 Gr 50 material and there was  a
general recommendation to use A572 Gr. 50 rather than A36 for beams in lateral
systems.  Finally, two years ago, the industry moved a new specification A992
through ASTM.  This is bascially the A572 Gr. 50 material with the supplemental
yield str
ength requirements and some additional control on trace elements, also a good
thing that can improve weldability.

All engineers should now specifiy ASTM A992, not A572 Gr 50 or A36 for beams and
column in the lateral force resisting system.  However, be prepared to accept
A572 Gr 50 as a substitute.  Eventually, A992 will be generally available and we
won't have to do this.  Also, eventually, A36 may go out of production for wide
flange shape.

Unfortunately, tubes, angles and other misc. shape are not avaialble in A992 and
may not be that available in A572.  Also, there is no statistical data, like
that performed for wide flange shape in 1993, to tell us what the probable
variation in strength is.

It is unlikely that braces in a concentric braced frame will "break", however,
there connections might, because the brace is inadvertently stronger than the
connection.  Also, it is possible that columns could be overloaded, because the
braces deliver to much overturning load to them.

Until this sorts itself out, wise practice would be to specify grade 36 material
for angles and assume you have A572 material for purposes of determining
connection and column demands.  SEAOC may need to consider increasing the Omega
"overstrengh" term for braced frame systems, to account for this overstrengh.
Frankly, we hadn't worried about braced frames before.  But the problem is