AISC Second Order Effects

[DaveAtkins]

I confess I have been slow to transition from the AISC 9th Edition to the more recent editions (but I am working on it!). I have a copy of the 14th Edition, and am learning how to account for second order effects. It appears there are four options available:
* Direct Analysis Method
* Effective Length Method
* First-Order Analysis Method
* Approximate Second-Order Analysis Method

Question 1: If you set up your computer model (I use RISA-3D) to include second order effects, you can use the Direct Analysis Method--is this correct?

Question 2: The method which is most similar to the method in the 9th Edition would be the Approximate Second-Order Analysis Method--is this correct? It uses amplification factors (actually it also appears to be similar to the method used in ACI 318 for amplifying column moments).

DaveAtkins

 

[PicoStruc]

GENERAL COMMENT ABOUT SECOND ORDER EFFECT AND FEM SOFTWARE :

I am outside US, I won't comment about AISC..

Be caution with "Second order effects or P-Delta checkbox" feature in FEM software, it's generally means 'Consider Geometric Stiffness'

If it is the case with your FEM software :

1) The internal forces from analysis generally account only for Big P-Delta (drift effect).
2) Small P-Delta (local moment amplification due to axial load) will be considered only if the axial member is subdivided into smaller finite elements.

Don't know about RISA3D.

 

[JAE]

Dave:

Question 1: If you set up your computer model (I use RISA-3D) to include second order effects, you can use the Direct Analysis Method--is this correct?
[red] Yes - Direct Analysis method is the best way to go with software such as RISA.[/red]

Question 2: The method which is most similar to the method in the 9th Edition would be the Approximate Second-Order Analysis Method--is this correct? It uses amplification factors (actually it also appears to be similar to the method used in ACI 318 for amplifying column moments)
[red] The Direct method does not artificially amplify moments using B1 and B2 factors from the 9th Edition (which is similar to ACI's Chapter 10 moment magnification method using [δ] magnifiers). The Direct method essentially uses small notional lateral loads and reduced section properties along with a good second order program (like RISA) to magnify the lateral sways and bending in axial loaded members to produce second order effects.

Read the commentary on AISC's section on the Direct Analysis Method - it is very comprehensive and helpful.[/red]

[blue]PicoStruc - for RISA, it does satisfy the code in including P[Δ] and P[δ] effects.[/blue]

 

[DaveAtkins]

A follow up question--RISA accounts for adjustments to stiffness (if you program it to do so), but not initial imperfections. You must input notional loads to account for this. Is that correct?

DaveAtkins

 

[JAE]

Yes - the Direct method's basis is tagged to AISC's maximum tolerances for out-of-plumbness, sweep in sections, etc.

 

[WillisV]

Question 1: NO. Setting up your computer program "to include second order affects" does not constitute a direct analysis procedure, though it is part of it.

You must:
1. Account for large (sway) p-delta effects either using computer methods (this generally corresponds to the p-delta on or off option) or B2 modifiers.
2. Account for small (member)p-delta effects either using computer methods (splitting axially loaded members into 3 or more pieces in the model) or B1 modifiers.
3. Reduce stiffnesses with Tau factors.
4. Introduce notional loads (or model the building with an initial imperfection - generally easier to automate the notional load process).

 

[structSU10]

You can have access, for free, to a webinar on stability from AISC through its website here:

http://www.aisc.org/content.aspx?id=31628

It discusses all of the possible methods that can be used with the current code.

 

[RFreund]

I think this was covered pretty good in the above posts however I will add the following -
You can use the approx 2nd order method with the direct analysis method (DAM). Basically I would say that you need to follow a procedure to verify stability which in AISC there are the three you listed above:
* Direct Analysis Method
* Effective Length Method
* First-Order Analysis Method

However a second order analysis is only part of these procedures. For DAM you must considered a reduced stiffness, imperfections and run a second order analysis. For the other 2 procedures you run a second order analysis but you must find your K factor. You may run a rigorous 2nd order analysis (RISA does not technically do this however they have an excellent white paper explaining why it is not typically necessary) or an approximate 2nd order analysis (usually acceptable). In all 3 cases you perform the second order analysis however with the effective length method and first-order method you must find an approximate K factor (alignment charts) but in DAM K=1.

EIT

http://www.HowToEngineer.com

 

[JAE]

WillisV - I agree with your statements but Question 1 above was not whether a software like RISA with second order effects DOES the DAM but rather they asked whether using RISA with second order effects can be used WITH the DAM...at least that's how I read it.

 

[WillisV]

JAE - no worries, we read differently