unbraced length

huangyhg

unbraced length
i have always used inflection points for determining unbraced lengths at negative moment locations on a beam.  i was just informed that this is not okay.  can anyone either lead me to a code or standard that discusses this?
thanks in advance,
drew
"steel structures - design and behavior", second edition, by salmon and johnson, pages 481 - 482, gives the best discussion of this issue i have ever seen.  they point out that using a point of inflection as a braced point is okay if cb is not used in calculating fb (i design using asd).  however, the correct way to design is to not assume the point of inflection is braced, and use the cb factor.
daveatkins
a point of inflection by itself does not provide for a bracing point agianst ltb (lateral torsional buckling).  a good discussion on this issue can be found in modern steel construction january and november 1997 issues.
i think that amokhta is technically correct, but we have done, essentially, what daveatkins states above - assume an unbraced length from the column to the inflection point, add about 20% to that length for extra safety, and use cb=1.  if you do this, you will find that your designs are very similar to that using the full span as your unbraced length along with calculating an exact cb.
see "bracing for stability" by dr. joseph yura and dr. todd helwig.  this is notes from a seminar that they present through aisc.  dr. yura has a good example of why it can be unconservative to use inflection points. also, you can search this site for more threads on the subject.
my vote? don't consider the inflection point to be braced and if you want to be more precise compute cb.