source of load criteria to design compression flange bracing

huangyhg

source of load criteria to design compression flange bracing
for as long as i can remember, the load used to design compression flange bracing of i-beams and/or trusses was 2% of the compression flange load.
i am now being challenged to show where this value comes from.  is it written in code, or is it simply an industry accepted standard.
also, what premise is this 2% criteria based upon.  does it change if the loading condition is due to loads generated while the structure is being crane lifted?
thank you in advance for your response.
from "guide to stability design criteria for metal structures", 3rd. edition, bruce g. johnston, john wiley & sons, 1976, page  148:
"on the basis of a study of a number of hypothetical initial-curvature and load mis-alignment conditions, zuk (reference 1) confirms the customary practice of designing each lateral support for 2% of the total compressive force that exists concurrently in the compression flanges of the laterally braced beam or girder. ......  recent studies by lay (reference 2) have demonstrated the correctness of this rule as extended to plastic design."
reference 1:  zuk, w. "lateral bracing forces on beams and columns," asce journal, engineering mechanics division, vol 82, no em3 (july 1956)
reference 2:  lay, m.g., and galambos, t.v., "bracing requirements for inelastic steel beams," asce journal, structural division, vol 92, no st2 (april 1966) p. 207
jim emanuel, s.e.
what code are you using?  2% has been commonly used in the past but the aisc steel spec now has much more detailed requirements.  these include criteria for both the strength and stiffness of the bracing.  check appendix 6 of the 2005 aisc spec or chapter c of the 1999 spec.
check arema manual for railway engineering, chapter 15, section 1.11.6.  the bracing members used only for reducing the unbraced length of a member need not be designed for more than 2.5% of the force in the   
the 2% rule is not written into any code.  it is an old rule of thumb that worked and gained universal acceptance.  some 30 years ago, researchers began to report that the stiffness of the bracing is more important than the strength, and stiffness criteria were proposed.  from what i re  
for general application, i agree with taro that there are more detailed requirements for both strength and stiffness for bracing (aisc 13th, appendix 6, section 6.3).
for seismic application, 2002 aisc seismic provisions talk about 2% in various sections and 6% at links of eccentrically braced frames.  the new aisc seismic design manual may have updated information but i do not have a copy yet.
jmiec - stiffness usually controls over the strength requirement by a long long shot.  
willisv-
um, did i say (or imply) that it didn't?  or is there something in my post that is incorrect?