huangyhg

slender column
i am using a w27x146 for a tall column. it's web is slender for compression. however, when i go through section e7 of the aisc manual, i find that there is no reduction (qa=1). e7.2 is kind of confusing since they don't tell you what to do when b/t is less than 1.49x(e/f)^0.5. am i missing something here? what do you do when b/t is less than 1.49x(e/f)^0.5?
assuming 50 ksi yield, i get effective width of web of 22.1", (from e7-17). this gives a a effective less than a and thus qa less than one.
my interpretation is that if b/t is less than 1.49*sqrt(e/f) then the element will not buckle locally prior to the column buckling globally and thus qa=1.0. note that if f=fy then the above limit corresponds to the slenderness limit in table b4.1.
i agree with civilperson that the effective web width is 22.1" if you conservatively take f=fy=50ksi. however, for a pin-pinned column over about 15 ft for this q27x146 then effective web width is the true web width and qa=1.
if you assume f=50ksi (as described above) you end up with q<1.0. however, this is incorrect, according to e7-2, f is to be taken as fcr (with q=1.0).
this is how i rationalized it. for a really short column, local buckling of the individual plate sections that make up the column controls. since the column is short, you end up with a high fcr and therefore q<1.0. for a longer column global buckling of the section becomes more critical than local buckling and q=1.0.
unfortunately you have to do the calculation in order to find out what is governing the design of your column.

__________________
借用达朗贝尔的名言：前进吧，你会得到信心!
[url="http://www.dimcax.com"]几何尺寸与公差标准[/url]