Shear wall forces ! 2

[Murali27]

Hi all

In one of the project, 15-storeys building consists of two elevator core walls running from foundation to roof and 4 planar walls from foundation to 3rd level. The core wall forces are increasing from top to 4th level and thenceforth it is reducing to a minimum value.
ETABS is used for this project.

The 4 planar walls are at the periphearls of the building. Large forces are induced in these walls compared to core walls at lower level.

Can i proceed with this resultd for wall design / foundation design ?

Your input will also be appreciated.

 

[billyrayjimbob]

I am assuming that the planar walls are much longer than core walls and therefore much stiffer. In this case then they would theoretically attract the bulk of horizontal load at 3rd floor and below.
However this means that the horizontal forces must be transferred via diaphragm action in the slab at level 3. These forces must be considered.
Depending on the actual stiffness of the slab (and openings etc) it may not behave as a perfect diaphragm. Therefore it may be prudent to allow for a larger percentage of horizontal loads to remain in core walls all the way to foundations. Then design the planar walls for the maximum forces. Conservative but safe.

 

[andiron]

murali:

look closely at the diaphragm action if the load distribution is
adequate..

how are modelling the diaphragm in etabs though..i have 7.0 version and
i know etabs floor meshing has bugs..are u meshing to get the floor
slab? let me know please

gurlate

 

[Murali27]

Andiron:

You are right, i am meshing the floor slab.. Easy for applying the gravity load too...

I have now 8.2.6 version.

I tried the other option of not defining the diaphragm to the floor slab... In this case also i got more or less the same results.

Let me know your opinion on the above
Thank you

Murali

 

[andiron]

murali:

just got the etabs 8.2.6 version but it largely is similar to 7.0 with a few changes..since i had big probs with meshing the diaphragm before, i was using etabs floor only as rigid diaphragm only, obviously, for lateral (no gravity) analysis. the reason is, meshing was almost nonexistent unless i am missing something...besides, only times meshing worked for me was from the template with regular layout (useless for comlicated geometry anyway)...even here, i tested the meshing by making big holes in the flat plate and the results looked suspect..these features are unchanged, as i see it, in etabs 8.2.6 ..so my question to you how are u meshing the floor geometry..(in staad, which is far superior in this area, one can literally mesh element by element connecting column/wall nodes as the case may be...in etabs i tried using draw area/slab1..but the modelling doesn't work /results are suspect...(confirm that by displaying shell stresses M11/M22 etc and the contours have no continuity as you would expect)..so how do you model the slab in etabs for complex geometry?? i know now etabs has ability to import dxf file also..letme know anyhow..appreciate it...


thx

gurlate

 

[Murali27]

Andiron:

You are right. Only for lateral analysis, we can assign rigid diaphragm.

But we used to get the foundation forces from the same etabs model only. So we apply gravity loads also as correctly as possible.(90% accurate).

Floor can be literally meshed element by element connecting column/wall nodes as the case may be...

You may need to have SAFE to print the slab forces....etc

In complex geometry we import the *.dxf file (slab extent, columns, walls).... Meshing element by element...(No Auto mesh)

My question to you

What parameters (like sway/Non sway, Lateral drifts)need to be studied in the lateral analysis.(20 storey bldg)

 

[amadeus]

Area meshing is far superior in ETABS V8 that in was in V7. Select the desired element(s) and then do Assign -> Shell/Area -> Area object mesh options and on the dialog box that appears select "Auto mesh object into Structural elements" and "Further subdivide Auto Mesh with maximum size of: " and I normally indicate 1 mt. Note that for walls it is more convenient to manually mesh all walls into 1 by 1 elements using Edit ->Mesh Areas -> Mesh quad /triangles into 1 by 1 areas. To maintain continuity between meshes in walls and slabs, select one area object (any one, it doesn't matter) and then do Assign -> Shell/Area -> Auto line constraint and on the dialog box, select "Create ..." and check "Apply to full structure". That should solve all of your meshing problems.
One more thing. DO NOT apply rigid diaphragms to the structure if there is considreable load transfer through one or more diaphragms. If you do that, the axial forces of the element lying on the diaphragms and the in plane deformations of the diaphragm itself will be ignored.


Hope this was useful.

Amadeus

 

[andiron]

murali/amadeus :

on drift etc:

L/400 seems like an acceptable drift in wind analysis..earthquake i think explicitly stated in the code by drift ratio..in any case the cladding should be properly checked for the drift..

code also has a "Q" factor to suggest if the gravity columns fall under braced condition or not...some code also stipulate wall stiffnesses to be some multiple (6) of column in order it to be considered braced..

on etabs:

meshing:

i was comparing etabs 7 with etabs 8 and i still think some bugs were addressed smartly but some still remain..habibulla in his article does say categorically that you don't need to manually mesh the slab and nor do we have to have matching wall meshing (for displacement compatibilty)..in his paper the deformed shape is indeed compatible (which was not the case on version 7)..but check the shell stress (display/shell stresses : if the fem analysis is correct then slab M11 etc contours should NOT jump suddenly from element to element, but this is exactly what it does..say sudden jump from green in one element to red in the next (belying the continuom nature of the slab)..statics are still ok though...so habibulla is NOT correct in what he claims that etabs could do automatically...

2. rigid diaphragm and wind loading:

i think rigid diaphragm should be explicitly stated esp if so (concrete/steel deck)..i did the analysis without and some stresses do NOT make sense...axial deformation are NOT significant for floors anyway...

more importantly..how do specify the wind load now..automatic wind load that is..if the diaphragm is not meshed and used as a rigiddiaphragm only(no gravity analysis here) then automatic wind load generator recognizes the extent of the diaphram (one single huge nonstructural element) and puts the auto generated windload in the centroid of the slab at each floor (purists may not like it as this is very general, but i guess it is quick..how do you people specify wind load (not node-to-node /floor-to-floor manually , i hope)
Anyhow, the problem is: when the floor slab is meshed, the program, i believe, confuses the extents of what is now an assemblage of diaphragms (meshed shell element w/ rigid diaphragm constraint) and auto geberation of wind load is NOT correct...please share if you have some other view in this case...

Anyhow, the main point was do not beleive FEM analysis is correct until take a look at the shell stresses/contours..if continuity of colors is not observed across the elements then the FEM analysis is suspect...

 

[Maraveas]

ok all this, but is anyone there who is checking
his software results by hand calculations?
I think this is the best way to sleep without
problems the night.