It was reviewed a good long time ago (about a year, I think), so it should be out any time. I believe that's probably your best bet if you can wait. Otherwise, asking some specific questions here might help. I really think the idea is very easy. Implementation with current programs might not be, though.
Is reducing the stiffness and adding a lateral (notational) load to the structure the method of performing the second order analysis? or do you reduce stiffness and add a lateral (notational) load and perform a second order analysis?
The second order analysis must be carried out independently.
If using software, you must make sure it captures both P-capital delta and P-lower case delta effects. The commentary in the code gives two example problems that can confirm whether each of these effects is captured by the program.
In any case, you must add intermediate nodes in your members to ensure these effects are captured.
The example in the AISC Design examples seems to imply otherwise.
from page III-60.
"This method requires that both the flexural stiffness and axial stiffness be reduced and that 0.2% notational lateral loads be applied in the analysis. The combination of these two modifications account for the second-order effects and the results for design can be taken directly from the analysis."
That is, you reduce both axial and flexural stiffness, apply notional loads (make sure to apply .003Yi in order to avoid using Tau factors), AND carry out a second order analysis (that captures both p-delta effect).
Your post was fine. I'm just having trouble understanding the logic behind the Direct Analysis Method. If we're still using a P delta (P-little Delta & P-big Delta) analysis, what are the reduced stiffness & lateral notational loads for?
I believe the main reason for this is to account for residual stresses in the cross section. Although the net sum of residual stresses end up canceling out, once the first part of the cross section starts yielding, that steel essentially does not contribute stiffness (E is essentially zero at yield).
I believe this also allows for the use of k=1 in all cases (both sway and non-sway) as opposed to estimating k or using the nomograph. At least that is what I understand from previous postings on the subject.
Are you refering to residual stresses from the manufacturing of the member?
I believe the commentary has some discussion about some factors that contribute to "non-ideal" conditions. These include:
-member out of plumb/manufacturing tolerances
-foundation not completely level
-temperature gradients causing non-uniform deflections
- etc.
Perhaps the reduced stiffness & lateral notational loads are to account for these factors; and the p-delta analysis is still performed to account for the p-delta effects (i.e. member deflection & joint translation), which are in addition to the effects listed above.
The advantage of the Direct Analysis Method is that K is always 1.0, therefore k values don't have to be determined.
