Hello,
I have a few questions regarding column slenderness. In my model (v15.0) I am having a few slenderness issues and am trying to resolve them. It appears that the ACI 318 code has changed since 05 in terms of the kl/r limit, which I have found is what RAM enforces, so here are my questions.
1. RAM only flags columns as "failing" in slenderness if the ratio is above 100. Since it appears that this limit is no longer applicable in ACI 318-11, how can I ensure that my column design is per code since this is an old limit?
2. With ACI 318-11, moment magnification is the procedure to account for slenderness effects. Does RAM perform moment magnification? (per 10.10.6 and 10.10.7) If not, would I have to do calculations by hand and/or through another program to check column validity?
Thank you,
Beth
Allen Adams, our chief structural engineer, recently presented a webinar on slenderness requirements for concrete structures. If you are interested in watching a recording of this video (its free), click on the link below and filter "RAM" in the "Select Product" menu. The seminar was titled "Code Requirements for the Analysis of Concrete Structures."
http://pages.info.bentley.com/videos/
The web page covering concrete column slenderness in RAM Concrete will also be useful:
http://communities.bentley.com/products/structural/structural_analysis___design/w/structural_analysis_and_design__wiki/27705.ram-concrete-column-aci-slenderness-design
To answer your questions:
1. Yes, this provision was removed in ACI 318-08. However, it is still enforced when ACI 318-08 or ACI 318-11 is selected as the design code. This is a conservative defect. It is logged as Defect #215695 in our database.
2. ACI 318-08 and later permit the following to evaluate slenderness: moment magnification, elastic second-order, and non-linear second order. These are discussed on the ACI slenderness wiki above. RAM Concrete uses moment magnification for the design of gravity columns and uses an elastic second-order to design frame columns. The sway moment magnification procedure is not implemented in the program.
Kyle,
Yes, RAM Structural System can design sway, frame columns. The forces used to design these columns in RAM Concrete Column are taken from the RAM Frame Analysis (there is an option to use gravity forces from RAM Concrete Analysis). The design follows the elastic, second order analysis (ACI 318-11 10.10.4) and P-Delta should be included in the RAM Frame Analysis. The program does not calculate sway moment magnifiers and design these columns using the moment magnification procedure (ACI 318-11 10.10.5). As noted above, this is a future enhancement that we will consider.
Other notes and limitations are discussed on this web page:
There are second order effects: (1) secondary moment associated with axial load and lateral displacement at the end of a member (P-Delta) and (2) secondary moment associated with axial load and displacement between member ends (P-delta).
RAM Frame accounts for (1) but not (2). ACI 318-11 requires (2) to be included in the design, however, and identifies the moment magnification procedure as one way to account for it. If the P-delta effects are significant, than the column may be under designed in RAM Concrete.
The Defect that is noted in the slenderness tech note has been flagged as a critical known issue. It is something that we are working to correct. The web page below discusses the known issue. We will update the status of the defect on this page as we work on the fix:
http://communities.bentley.com/products/structural/structural_analysis___design/w/structural_analysis_and_design__wiki/31240.aci-318-column-slenderness-second-order-effects-along-member-length
Karl,
Now that the slenderness moment amplifier issue is fixed, I assume the only difference between assigning concrete columns "Lateral-Braced" (non-sway) and "Lateral-Unbraced" (sway) is how RAM handles calculating "k" using nomograph charts.
If I am designing a multi-story, two-way slab structure, it is my understanding that I must override the "k" value because the program is unable to accurately calculate the effective length of columns in the absence of beams. If an effective length override is required, then it makes no difference whether the columns are assigned "braced" or "unbraced." Please confirm.
Thanks!
Yes, RAM Structural System does not account for flexural stiffness of two-way slabs when determining K. If the columns supporting the two-way slab are defined as non-sway, I recommend using K=1. If the columns supporting the two-way are defined as sway, I would determine K accounting for an appropriate slab flexural stiffness and assign that value manually.
There are different slenderness limts for non-sway and sway columns below which slenderness effects can be neglected. These are outlined in ACI 318-14 6.2.5. Those limits can result in different designs for a given K if the column assignment is switched from braced to unbraced.
Your last reply has me questioning this, so I'm hoping you can either confirm and/or clarify. For sway columns in a two-way slabs without beams, my understanding is that k only needs to be manually calculated for sway columns if a first order analysis with moment magnification is used. If an elastic second order analysis was performed (P-Delta on in RAM Frame), then k should still be defined as 1 since ACI 318-14 6.7.1.2 directs to ACI 318-14 6.6.4.5 for moment amplification due to deflection along the member length of non-sway members. ACI 318-14 6.6.4.4.3 then states that k may be taken as 1 for non-sway members. Thank you for your help.
I do not want to speak for Karl / Bentley, but I agree with your assessment that k=1.0 is appropriate for calculating the moment magnification factor associated with Sections 6.7.1.2 and 6.6.4.5. I also agree that the P-Delta analysis performed in RAM Frame satisfies the elastic second-order analysis permitted by Section 6.6.4.6.2. The selected k-value should have no impact on the RAM Frame analysis.
However, an accurate k-value is needed to properly check the slenderness limits of 6.2.5. The problem I see is that RAM does not apply different k-values for the slenderness checks of 6.2.5 and the critical buckling load (Pc) in section 6.6.4.4.2. The user must either apply k=1.0 and risk incorrectly ignoring slenderness effects, or apply k>1.0 and deal with a conservatively large magnification factor which, in some cases, could result in a very conservative column design.
Karl / Bentley - Can you comment on this assessment? Much appreciated!
CEI Engineer is correct. For sway columns supporting a two-way slab with no beams, a K-value greater than 1 should be assigned to check the slenderness limit of ACI 318-14 6.2.5. When accounting for the second order effects between member ends for the same column using the non-way sway moment magnification procedure, K=1 should be used. However, RAM Concrete Column is incorrectly using the larger K instead. This was logged as Defect #1028340, has now be fixed, and will be included in v17.01.00, which will be released within the next few weeks.
Thank you both. I'll keep an eye out for 17.01 to drop and hopefully clear this up. Thanks again.
Karl, I'm running RAM SS V17.01.01. Has this issue been addressed? I was unable to find any documentation by Bentley on this. Thanks!