31 May 02 17:33
Noticed that you're posting some questions about concrete columns and how to practically and economically design them for various Pu and Mu values related to Pdelta and alternating live loads.
Prior to the advent of 3D analysis, engineers would typically model a 2D frame cut through the building and alternate live loads in patterns to get single-bending values (rarely have I ever done a biaxial design as I'm not convinced it ever governs).
With 3D you now have a lot more "knowledge" of your structure and the broad CYA language that ACI uses can be taken to an extreme.
I would suggest one way to handle the data coming out of your frame analysis would be with a spreadsheet. With RISA (I think that's what you said you were using) you can copy and paste the Member Forces right off the screen and drop it into a spreadsheet that can be set up to do the following:
1. For each column, under each separate load combination, calculate the value sqrt((Pu)^2 + (Mu)^2). This is a square root of the sum of squares and gives you a fictitious radial measure of the intensity of the P/M combination.
2. With all of the SRSS values (these will all be positive values), divide them into two parts, the upper, maximum half and the lower, minimum half.
3. The upper half will be the only ones to ever govern a column design. Think of your scatter plot of Pu/Mu values on your column interaction diagram. The upper half values will always be closer to the limit state line than the other half.
This doesn't reduce the analysis time but does minimize the number of Pu/Mu combos that you have to work with. You could also, perhaps, take just the upper 25% and still be OK.
As far alternating live loads, I'd still work with as many different span-load combinations that would drive up the design. The code (ACI 318 and others) are always written as a minimum - and as an engineer you should always, obviously, be on the "right" side of that minimum.