Old Thread about P Delta

[OSUCivlEng]

QShake,

In this thread:

http://www.eng-tips.com/viewthread.cfm?qid=368631

I asked a question about P Delta analysis. I never followed up when you said to use a ratio of Icracked = 0.5Iuncracked. How did you determine this? Thanks

 

[IDS]

In my opinion this is something that really needs reconsideration in all reinforced concrete codes of practice.

Codes typically give a single factor for the cracked stiffness, sometimes even higher than 0.5, but when designing for an Ultimate Limit State condition we need to know the maximum value of any action that will increase the stresses in a member.

We should be using the minimum stiffness when calculating p-delta effects, and the cracked stiffness of a column is often very much less than half the uncracked stiffness. My recommendation is to use the fully cracked stiffness for the first case. If the maximum moments from the first run are still in the elastic range (i.e. the steel has not yielded) under ultimate loads then that's OK. If not, the stiffness should be adjusted downwards (if the structure provides an alternative load path), or increase the member stiffness until the steel does not yield under maximum factored loads.

Also check for the maximum stiffness (uncracked EI), and use the worst case.


Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/