Use 1 Bent Instead of Whole Building For Stability Check per AISC DAM?

[PengStruct]

When I run my model for a 4 story building with 5 moment frames the Pdelta and DAM takes an hour for the initial run.

Does this sound like a normal run time? Is it acceptable to analyze 1 bent only for the stability check? All frames are the same.

 

[JoshPlumSE]

For a 4 story building? An hour seems like too long of a solution time. When investigating this sort of thing I usually do the following:
1) Try to determine which load combinations are taking so long to solve. If your analysis program has a solution status window that tells you what it's doing then this should be easy.

2) Try running the model without P-Delta to see how long it solves.

3) Use the alternate approach with DAM where you don't iterate the flexural stiffness adjustments. See if this makes a major difference. Or, just turn off DA method temporarily (while using P-Delta) and see how long it takes it to solve.

4) Look at other potential issues for iteration or long solution times. Tension only or compression only members. Link elements and such.

What I'm trying to do with the above is determine the source of the long solution times. So, that I can take a closer look at the cause and determine if it is legit or if there is a modeling problem.

 

[PengStruct]

Thanks JoshPlum. I did each of the things you suggested. Running it without P Delta did help but was not the cause of the long run time. I found some overlapping beams and diaphragm connection issues that fixed it. Now it will run in only a few minutes and I was able to compare second and first order drifts to determine that only gravity combos required notional loads and then found that even with notional loads my design steel sections did not need to change from what I needed before checking with DAM. With a tall first story the drift requirements drove the steel design and the resulting column sizes are already big enough to handle the extra loads from the stability check.

I very much appreciate your input as I've read many of your posts and you seem to have a good grasp of the topic.