A LInkedIn Discusion on the Correct Stiffness Modifiers for Wall Pier or Spandrel Per ETABS

For the WALLS modelled as a shell element, the stiffness modification factor used is 0.7 ( according to ACI 10.10.4.1). IT SHOULD BE APPLIED TO ALL THE MODIFICATION FACTORS OR ONLY M11,M22 & M12?     

Mirza Asif Ali BaigStructural Engineer at Hashem Contracting and Trading Co. Ltd. Like (9) Comment (44) Follow Reply Privately November 12, 2013 Comments farideh ibrahim, Mirza Asif Ali Baig and 7 others like this 44 comments Jump to most recent comment

 ibrahem ibrahem nasreldin Structural Engineer at A.S.Company

o o o o

According to CSI , at default local axis settings you should only modify F22 for wall piers and only F11 for wall spandrels Like (3) Reply privately Flag as inappropriate November 12, 2013 mehdi mazloum, farideh ibrahim and 1 other like this

 Mohamad R Mohamad R Fatemi Structural Engineer at Shiraz Municipality

o o o o

The shell element have in plane or membrane and out of plane or bending stiffness. therefore according to ACI you should modify F22 , F11 for in plane and M22 , M11 for out of plane bending stiffness. Like (3) Reply privately Flag as inappropriate November 13, 2013 mehdi mazloum, Mirza Asif Ali Baig and 1 other like this

 Mirza Mirza Asif Ali Baig Structural Engineer at Hashem Contracting and Trading Co. Ltd. so f11,f22 m22 m11 as per ACI 318-11 section 10.4.4.1 should be 0.35 for cracked wall

1   

o o o o

Like Reply privately Flag as inappropriate November 13, 2013

 Mirza Mirza Asif Ali Baig Structural Engineer at Hashem Contracting and Trading Co. Ltd. o o o o

sorry the ref should be 10.10.4.1 Like Reply privately Flag as inappropriate November 13, 2013

 ibrahem ibrahem nasreldin Structural Engineer at A.S.Company

o o o o

Shear walls resist only in plane moments refer to this link https://wiki.csiamerica.com/display/etabs/Modeling+cracked+shear-wall+behavior Like (2) Reply privately Flag as inappropriate November 13, 2013 Mohamed AbuShady, Avinash Agrawal like this

 Mohamad R Mohamad R Fatemi Structural Engineer at Shiraz Municipality

o o o o

Out plane behavior of shear wall same to concrete column under bending moment. for example out plane bending stiffness of shear wall with 5*0.35 meter dimension equal with bending stiffness of ten column with 0.5*0.3 dimension (around weak axis) and obviously cracked under bending moment. Csi is a company that produced the software for structural engineers and not the reference for design of structure. Like Reply privately Flag as inappropriate November 13, 2013

 Peter

2   

Peter Placzek Design Director at Meinhardt (VIC) Top Contributor I use M11,M22=0.3 to 0.6 (depends on thickness and building height) and M12=0.1 as default. To spandrels I apply externaly F12=0.3 - this is based on matching behaviour of finely meshed spandrel compared to coarse meshing as well allowance for shear cracking. (Note F12 =1 in unmeshed spandrels gives excessive spandrel stiffness) I than factor F12 externally in zones of walls to release any excessive local stresses as necessary. I believe that since the weakness of the walls is shear, this is what should be factored not F11 and F22 since you canot factor for tension without factoring compression as well.

o o o o

The F12 reduction automatically reduces F11 and F22 in the most efficent way. Like (4) Reply privately Flag as inappropriate November 14, 2013 Ravi Arjun, Ghazala Farooq and 2 others like this

 Peter Peter Placzek Design Director at Meinhardt (VIC) Top Contributor I use M11,M22=0.3 to 0.6 (depends on thickness and building height) and M12=0.1 as default. To spandrels I apply externaly F12=0.3 - this is based on matching behaviour of finely meshed spandrel compared to coarse meshing as well allowance for shear cracking. (Note F12 =1 in unmeshed spandrels gives excessive spandrel stiffness) I than factor F12 externally in zones of walls to release any excessive local stresses as necessary. I believe that since the weakness of the walls is shear, this is what should be factored not F11 and F22 since you canot factor for tension without factoring compression as well.

o o o o

The F12 reduction automatically reduces F11 and F22 in the most efficent way. Like (1) Reply privately Flag as inappropriate November 14, 2013 farideh ibrahim likes this

 Ravi Ravi Arjun Senior Structural Engineer at Al Shandagha Architects & Engineering Consultants

3   

o o o o

I use f11,f22,f12,m11,m22 & m12 =0.70 in my shear wall design. Like Reply privately Flag as inappropriate November 14, 2013

 Faisal Faisal Habib Software Support Manager at Computers and Structures Inc

o o o o

In ETABS, shell or area element has two types of stiffnesses i.e. inplane stiffness refers as f11, f22 and f12 and out-of-plane stiffness refers as m11, m22 and m12. Refer to the Figure 40 of CSi Analysis Reference Manual which shows the direction of local axes and their corresponding stiffnesses: For shear wall (both piers and spandrels), the flexural and axial behavior is modified by either f11 or f22 depending on the orientation of the local axis and the shear behavior is controlled by f12. In column and code terms f11 or f22 would correspond to modifications of EI or EA and f12 would correspond to modifications to GAshear. The code recommendations in Section 10.10 of ACI 318 code are related to slenderness effects where flexural deformations govern so they have recommended modifying EI (corresponding to f11 or f22 for shear walls). There is no recommendation about reducing the GAshear. You should, however, note that some of our users use modifiers for f12 also, where they expect deterioration of shear stiffness and want to be realistic in their modeling.The above discussion applies assuming the local axes 1 and 2 of the shear wall area object are either vertical or horizontal. This is under user control. When drawing in ETABS the default is to have the 1 axis horizontal and the 2 axis vertical. This means that the flexural modifier for EI should be applied to f22 for wall piers and to f11 for spandrels. If you apply the modifier to both f11 and f22 it hardly affects the results.For slabs where bending is always in the out-of-plane direction, modifiers m11, m22 and m22 are required to model cracking behavior.Summary:Assuming beams and columns are modeled as frame then the stiffness modifier table is as follows: ACI ETABS Beams........................................0.35*Ig I22 = I33 = 0.35Columns....................................0.70*Ig I22 = I33 = 0.70Walls-Uncracked.................0.70*Ig modeled as membrane – f11, f22 = 0.70 modeled as shell – f11, f22 = 0.70Walls-Cracked......................0.35*Ig similar to Walls-Uncracked (use modifiers of 0.35)NOTE:Walls are generally not designed for out-of-plane bending to avoid excessive longitudinal reinforcement. In this case, use a small modifier say 0.1 for m11, m22 and m12 so numerical instabilities could be avoided. However, use m11, mm2, mm12 = 0.70 (or 0.35) when considering the out-of-plane bending in wall.Flat Plates Like (10) Reply privately Flag as inappropriate November 15, 2013 Muhammad Kamran Bashir, Suhrid Sharma and 8 others like this

 Faisal Faisal Habib Software Support Manager at Computers and Structures Inc Please use the following link which provides better formatting: Thanks

o o

https://wiki.csiamerica.com/x/AoBF Like (5) Reply privately

4   

o o

Flag as inappropriate November 15, 2013 ibrahem nasreldin, farideh ibrahim and 3 others like this

 ibrahem ibrahem nasreldin Structural Engineer at A.S.Company o o o o

thanks Mr.Habib Like Reply privately Flag as inappropriate November 16, 2013

 Suhrid Suhrid Sharma Associate at Meinhardt (Malaysia) Sdn Bhd o o o o

Thanks Faisal for this info. Was looking for this for a long time. Like (1) Reply privately Flag as inappropriate November 17, 2013 Mirza Asif Ali Baig likes this

 Peter Peter Placzek Design Director at Meinhardt (VIC) Top Contributor Faisal, good to see you get involved. I see no issue with F11 factor on spandrels, but I reiterate that unless the spandrels are meshed very finely F12 factor is required to get realistic stiffness.

o o o o

I see problem with F22 factor on piers. The F22 reduction softens the walls in compression - the dominat load and potentially changes load distribution in tall buildings. I think you should factor the column axial stifness to match. Like (4) Reply privately Flag as inappropriate November 17, 2013 Muhammad Kamran Bashir, Suhrid Sharma and 2 others like this

5   

 Suhrid Suhrid Sharma Associate at Meinhardt (Malaysia) Sdn Bhd

o o o o

Fully agree with Peter on wall stiffness modifications. But as long as walls are drawn conventionally, Local 2 is the vertical axis and thus f22 does not require any qualification as we do not want to change the axial stiffness. But situation can be different if walls or columns are subject to pure tension during later loading. Also there is no recommendation to modify the shear stiffness, I am not sure should we modify f12 for spandrels. Maybe Faisal is better person to clarify. Like Reply privately Flag as inappropriate November 17, 2013

 Muhammad Kamran Muhammad Kamran Bashir Section Manager (Civil/Structural) at DAR ENGINEERING

o o o o

further to Suhrid sharma's question, in case a pier at a paticular storey is Over stressed in shear, is it a correct practice to reduce the shear modifier considering that this shear will be re distributed to the other shear walls at the same storey level? Like (2) Reply privately Flag as inappropriate November 18, 2013 Mohamed AbuShady, ibrahem nasreldin like this

 Peter Peter Placzek Design Director at Meinhardt (VIC) Top Contributor

o o o o

Muhamad, I do it all the time. For example in a case where you have door offset from centre in the wall leaving narrow pier on one side. If the pier is very narrow (~1 m) I factor it now automatically even before any results (F12=0.3). Like (1) Reply privately Flag as inappropriate November 18, 2013 ibrahem nasreldin likes this

6   

 Muhammad Kamran Muhammad Kamran Bashir Section Manager (Civil/Structural) at DAR ENGINEERING

o o o o

Thanks peter. but can F12 be as low as 0.1, if the pier is displayed as O/S against F12=0.3? i have manually checked some of the piers, displayed as O/S in ETABS, and found them to be within the allowable 10x phi x sqrt fc' x 0.8xLxd. why is it? am i missing out something? Like (1) Reply privately Flag as inappropriate November 19, 2013 Mohamed AbuShady likes this

 Hamid Hamid Radmard Rahmani Founder of Behsazan Co.

o o o o

I want to say that I am agree with Peter about considering F22=1 for concrete walls and I want to add some detail to the discussion. According to ACI, the stress in the wall should be analysis and check with crack criteria. then you should estimate the factor you should use to decrease the stiffness of the wall. Also in CALTRANS code and AASHTO there are very useful graphs that defines the reduction factor to the reinforcement of the columns and the P and M. ( they are about Columns but useful for walls) Like Reply privately Flag as inappropriate November 19, 2013

 Jonathan Jonathan Wirthlin, SE LEED AP Engineer, Structural at Coffman Engineers Muhammed, I generally agree with Peter, but i would be careful before doing this. It used to be common practice, but post 2006 IBC, i think this became against the code for new buildings. Directly, the code expects you to distribute forces based on rigidity under the same E (not assuming some of the piers will crack up and have some ratio lower of E). You can call a pier a secondary structural element and not count on it at all, but then it needs to meet deflection compatibility requirements and maintain gravity capacity.

o

Whatever you decide to do I suggest the following...tie the pier closely for ductility, and ensure without a doubt gravity load carrying capacity will not be lost. Assuming the pier cracks up 70 or 99% is a lot of assumed damage and the pier will be useless for gravity. Maybe embed steel posts in the pier so the gravity capacity is there no matter what. Like (1)

7   

o o o

Reply privately Flag as inappropriate November 19, 2013 Suhrid Sharma likes this

 Peter Peter Placzek Design Director at Meinhardt (VIC) Top Contributor

o o o o

Actually reading the clause in ACI it clearly says Area modification factor =1.0 Unfortunately for piers F22 influnces both Ig which requires factor 0.7 and Area which requires factor 1.0. You can't do one without the other. I do see merit in factoring both columns and piers axial stiffness by 0.7 for consistency. After all, when you consider creep, the long term shortening of walls and columns will actually be closer to 0.7A stiffness than 1.0A stiffness (0.5 to 0.7A). Like (2) Reply privately Flag as inappropriate November 19, 2013 Mohamed AbuShady, Suhrid Sharma like this

 Faisal Faisal Habib Software Support Manager at Computers and Structures Inc Hi Peter & Suhrid, Unlike frame element where axial and flexural stiffnesses are uncoupled, wall axial and inplane bending stiffness depends on f22. Some users add columns at wall ends which are pin-ended at both ends to compensate the change in axial stiffness. However, recent papers suggest reducing the axial stiffness as well as bending stiffness for seismic analysis of walls. You can refer to the following paper for further details on this item: Effective flexural stiffness for linear seismic analysis of concrete walls by Ahmed M.M. Ibrahim and Perry Adebar. This paper can be purchased from the following location: https://article.pubs.nrccnrc.gc.ca/RPAS/RPViewDoc?_handler_=HandleInitialGet&journal=cjce&volume=31&calyLang=eng&article File=l04-014.pdf&secure=true

o o o

Spandrels where flexural bending governs i.e. span/depth ratio > 2, modifying f12 may not provide significant difference. However, spandrels with span/depth ≤ 2 have shear deformation and using a shear modifier may yield significant difference. In both cases, spandrel meshing is recommended to capture the double curvature bending in spandrels. Like (4) Reply privately Flag as inappropriate

8   

o

November 19, 2013 Kirankumar K.L, Suhrid Sharma and 2 others like this

 Jonathan Jonathan Wirthlin, SE LEED AP Engineer, Structural at Coffman Engineers

o o o o

Good point Faisal.I suppose i could sum up my previous comment by saying that I don't believe a typical static or rspectra new building design should try to guess and assume inelastic behavior of shear walls to design reinforcement. You should take an assumed cracked section modifier globally and design some piers as secondary members if you'd like.If you are doing an in-depth model and truly designing considering inelastic behavior of all elements that is different and ok. Of course at this point you should actually calculate what stiffness modifiers to use. See Faisal's paper. Like (2) Reply privately Flag as inappropriate November 19, 2013 Mohamed AbuShady, Seifeldin Marzouk like this

 Peter Peter Placzek Design Director at Meinhardt (VIC) Top Contributor

o o o o

Faisal, for practicing engineer it is important to have as simple model as possible. Adding columns to offset pier area stifness reduction is just cazy. Spandrel meshing also complicates things - extends run times and causes instablities, makes changes more difficult, makes spandrel labels illegible........ As you say for deep spandrels the K12 makes no differance, threfore it does no harm either. Deep spandrels can also handle the loads. For the shalow spandrels however it achieves the right result (at least one of the right results) without the onerous meshing. The exact K12 varies from 1 to 0.3 depending on the span/depth ratio - 0.3 is the best conservative shot. But of course line object is always better option for shallow spandrels. Like (2) Reply privately Flag as inappropriate November 20, 2013 Seifeldin Marzouk, Kirankumar K.L like this

 Jonathan Jonathan Wirthlin, SE LEED AP

9   

Engineer, Structural at Coffman Engineers

o o o o

Peter, i would suggest spandrel meshing at a few conditions in the building to check results and behavior of the other modeling. I try to pick at least 3 locations to finely mesh with each model. (that way run times stay reasonable). Like Reply privately Flag as inappropriate November 20, 2013

 Jonathan Jonathan Wirthlin, SE LEED AP Engineer, Structural at Coffman Engineers

o o o o

Peter, i would suggest spandrel meshing at a few conditions in the building to check results and behavior of the other modeling. I try to pick at least 3 locations to finely mesh with each model. (that way run times stay reasonable). Like Reply privately Flag as inappropriate November 20, 2013

 Faisal Faisal Habib Software Support Manager at Computers and Structures Inc Peter, I totally agree with you that adding additional column for compensating the axial stiffness is just a crazy idea (it was just an observation instead of recommendation). I do not think that meshing spandrel should cause instabilities. The shear stiffness modifier (f12 or K12) ranges differently depending on shear deformation vs. axial deformation. FEMA 356 provides some guidelines for using bending and shear modifiers.

o o o o

Jonathan, changing stiffness would cause larger secondary moment due to P-delta analysis since displacement increase due to reduce stiffness. Like (2) Reply privately Flag as inappropriate November 20, 2013 Mohamed AbuShady, Suhrid Sharma like this

 Peter Peter Placzek Design Director at Meinhardt (VIC) Top Contributor

10   

Jonathan, Faisal, I have done extensive testing (as you Jonathan suggested) with meshing for generic lobby and single door headers with different span to depth ratios. I just dug my work out - I have not looked at for some time. Details: The test models are one spandrel between two equal walls - it is a H shape structure. The walls are hinged at base and load is applied at top. The changes to sway at top of the walls is the measure of the spandrel stiffness. Based on the test I arrived at following conclusions: For unmeshed spandrel use F12 as follows 1