How to consider local p-delta in Etabs?? 1

[ayelamayem77]

Dear all

according to CSI recommendations to calculate local p-delta effect
(which is associated with local deformation relative to the chord between member ends)

we have to consider one of these methods:

1-Apply design factors, which ETABS post-processing assumes to be done. These factors are therefore included in design, when applicable.

2-Divide members into segments (at least two per column), then run each load case separately with a different P-Delta load combination for each.

my question is what are these design factors according to ACI-318-08 CODE??
and how to apply them in etabs??

thanks in advance.

 

[johnbridge231]

Under the term "Design factors", I understand the code provisions regarding the checks against buckling and secondly an amplification of end moments that should take place in some cases.
These can be avoided if you perform a second order analysis with divided elements, as you say.
If not, ETABS applies them automatically according to the specified regulation.

Analysis and Design of arbitrary cross sections
Reinforcement design to all major codes
Moment Curvature analysis

http://www.engissol.com/cross-section-analysis-design.html

 

[ayelamayem77]

Dear johnbridge231
thanks for your quick replay, please let me share the following information so you can help me to understand better.

first let me share the following well known information:

The Code recognizes the following three methods to account for slenderness effects:

1. Nonlinear second-order analysis (10.10.3). In this analysis, consideration must be made for material
non linearity, member curvature and lateral drift, duration of loads, shrinkage and creep, and interaction with the
supporting foundation.

2. Elastic second-order analysis (10.10.4). In this analysis, consideration must be made for the influence of axial
loads, the presence of cracked regions along the length of the member, and the effects of load duration.

3. Moment magnification procedure (10.10.5). An approximate analysis of slenderness effects based on a
moment magnifier (see 10.10.6 and 10.10.7) is permitted. In this method, moments computed from first order
analysis are multiplied by a moment magnifier to account for the second-order effects. The moment magnifier
is a function of the factored axial load Pu and the critical buckling load Pc for the column. This method
is discussed in details in sections 10.10.6 and 10.10.7, for non sway and sway columns, respectively.

Previous information from PCA-NOTES ON ACI318-08 Design for Slenderness Effects.

According to ACI318-08 we can use one of the previous three methods for checking and calculating Slenderness.

Back to the manual of etabs ( concrete Frame Design Manual) to understand how the program deals with slenderness

CSI assumes that the design process includes P-DELTA effect, so Etabs uses the second method recognized by ACI318-08 code
(Elastic second order analysis , using modifiers entered by the user in the element sections as per ACI318-08-10.10.4.1 ) so users have to activate p-delta in the program.
but even if we did that , CSI said the following phrase (the effect of P-delta is limited to “individual member stability.” For unbraced components , For the individual member stability effects, the moments are magnified with moment magnification factors, as documented in Chapter 3 of this manual.)
what i understood from many trials on Etabs that for some non sway members where Moment (computed by program is small or zero) the program goes directly to design method number three (moment magnification procedure) to calculate slender based on calculating (delta ns ) even if it is set to 1 by the program.(please correct me if i miss some thing).

now let us go to the definition by CSI on P-DELTA

follow this link:

https://wiki.csiberkeley.com/display/kb/P-Delta+effect

We have two types of P-DELTA:

1-global P-DELTA (that can be done directly by the program by activating P-DELTA based on load combination)

2-Local P-DELTA (That can be done by Applying design factors, in ETABS post-processor)

what i understand that the nonlinear analysis based on p-delta used in etabs cannot capture local p-delta , so user should use factors as CSI described
above.

my specific questions:

1-when users activate p-delta in etabs (assumed by CSI as per their manual to account for slender )they use the second method based on ACI-318-08 definition (Elastic second order analysis) , but to calculate local p-delta they have to use factors in etabs design processor they use
(moment magnification procedure).
so the moment in this case is magnified two times (first based on elastic second order ,second based on the approximate method) how such thing could be happened ????

2-according to ACI318-08-10.10.2.1(Total moment including second-order effects in compression members, restraining beams,or other structural members shall not exceed 1.4 times the moment due to first-order effects.) the previous rule means that the factor (delta ns) in etabs should not be greater than 1.4 for non sway members, and if it is greater a warning message should be appear to increase the column dimension.
why when user enter a value greater than 1.4 in the etabs possessor or when etabs calculate it directly for some individual non sway members as described in the manual and it is found to be greater than 1.4 no warning message appears ??

Please sir help me to understand better and correct me if i miss any thing

thanks in advance.

ayelamayem77.

 

[stressed]

ayelamayem77, ETABS P-Delta analysis will automatically consider both global and local P-delta effects. For local P-delta effects to be fully considered, however, we've learned that you often need to add 1 intermediate joint along the column lengths using draw point if there is no intermediate bracing otherwise adding a joint along the column. The best way to draw the intermediate points is in elevation view with line and/or midpoint snap activated using draw point.. drawing the points one at a time. It can be a bit tedious, but do-able in less than 10 minutes for most buildings. Adding more than 1 point along the length of the columns does not improve accuracy. This method is preferable to dividing the columns into multiple pieces because you get 1 moment diagram for 1 column and the columns are designed using their actual length. The draw point meshes the columns only for internal analysis.

ETABS (to my knowledge) will not automatically perform the 1.4x check, nor will it automatically determine modulus of rupture for cracking. You will have to manually determine that and assign as a property modifier to reduce moments of inertia

 

[ayelamayem77]

Dear stressed,

thanks for you reply, but after adding a point to a frame element using your guidelines
and after the design was done no effect was noticed, but when the element was divided into two parts the design moment was changed based on modifying (delta ns)


check the following link:

http://www.mediafire.com/?vws827w2u3sktma

thanks in advance

 

[stressed]

ayelamayem, please see attached ETABS model (import .e2k file) simple cantilevered flagpoles. There are 3 flagpoles in this model all fixed at the base: 1 divided into 2 sections, another undivided, and another with a joint added as I described in my post above. Note that the flagpole column divided in two matches exactly the results of the column with an intermediate joint added, whereas the undivided column which has no intermediate joint has frame major moments and support reactions significantly different than the other two flagpoles due to P-Delta.

After first running the analysis, you need to go back to the Analyze menu and run the nonlinear static case. Check output results using PDELTANL case step 10 (the final step in the NL analysis).

 

[stressed]

After you import the .e2k file, careful to check if the intermediate joint added to one of the flagpole columns came across. If not, draw one manually on one of the flagpoles which are undivided.

 

[ayelamayem77]

Dear stressed,
thanks very much for your trying to help me , but the type of analysis you did in your file was pushover analysis
i want to use p-delta method as described by csi (which is based on load combination and iterations )
best regards

 

[stressed]

ayelamayem, the nonlinear pushover analysis can be used for P-Delta analysis only which is what the model did. It was a static P-delta analysis, not a pushover analysis as there was no material nonlinear analysis involved, just geometric nonlinearity (P-delta). Please check the settings of that nonlinear case to verify for yourself. Furthermore, it's clear that the flagpole column with the joint added along the length matched results with the flagpole column which was divided in two.

The Analyze menu>Set analysis options is the not the only way to run P-Delta using ETABS. The nonlinear static case method is another way to accomplish this.