AISC Direct Analysis Method

[craigmcg]

I'm running STAAD.Pro 2006 (1002) and I am trying to perform a frame analysis based on the AISC D.A.M. I'm concerned that the P-Delta analysis function does not account for the small p-delta effects (member stability) in addition to the big P-Delta. I work with marine structures so we don't typically design a lot with steel, hence I'm a little behind the times in this area.

I guess I just have two questions... Are my concerns warranted? And if so, does anybody have any ideas on how to properly account for these effects with my version of STAAD?

 

[dozer]

Yes, your concerns are warranted and no I don't believe that 2006 does small P-Delta. Supposedly the latest version (2007, build 03 or is it 1003, I've seen it both ways) does. There are two benchmark problems in the AISC 13th ed. manual that you can try. Sorry, I'm not at work so I can't tell you what page they are on.

I don't know of a work around for 2006. If I remember I'll try the benchmarks in 2007 and get back to you. Any reason you don't want to update?

 

[craigmcg]

That's exactly what I found after digging around a few other forums. I ran one of the benchmark problems and confirmed this as well. I guess I have no other choice but to look at upgrading, its not that I don't want to, just that I have to pitch it to management for the funds...

 

[JoshPlumSE]

craigmcg -

I'm not personally a STAAD user, but I know alot about AISC 13th edition and the DA Method. There are some tricks that can relatively easily force a P-Big Delta analysis program to also account for P-Little Delta.

Essentially, all you have to do is take your columns or axially loaded member and split them into smaller pieces. This takes advantage of the fact that a P-Big Delta Analysis is based on NODAL deflections. Therefore, adding in those extra nodes along the height of the column should adequately account for the deflections that occur along the length of the member.... provided that you use enough nodes.

I haven't tested this out in STAAD, but I have tested out the AISC benchmark problems in other programs and it seems to work very well.

I hope that helps!

 

[WillisV]

in addition to JoshPlum's response...it really doesn't take many divisions of the elment either...2 between joints is enough except with very slender elements and even then 3 or 4 divisions will generally be acceptable.