How to model A C-shaped shear wall in 3D frame 1

[IJR]

I was just curious, how any of my friends can model a chanel shaped shear wall by frame elements in a 3D finite element model. I am familiar with simple shearwall modelling using rigid links and column equivalents but have never tried replacing a large channel shearwall with a simpler model. Imagine a shear wall around elevator shafts

Thanks in advance.

 

[Riz]

IJR,
U have not mentioned about sw to be used. I have modelled a 15 storey braced structure, lift shaft together with shear walls on Staad. The modelling is elemental for the walls and load placement do not differ from your stick models.
Riz

 

[IJR]

Thanks Riz for this "Gonzalez" fast response. I usually model walls in SAP2000 using shell elements. But this makes me loose feel for wall forces (Moment, Shear and Axial) or makes the model too complex to check out. so I prefer equivalent frame modelling.

For a simple shear wall I would use a column of equal stiffness and attach this column to beams using rigid links.

I would like to do this with other shear walls, eg. 4 walls connected to form a rectangular section in plan or 3 walls forming a channel section in plan.

Thus what I am looking for is not really software dependent rather an easy equivalent which will retain behavior.

Thanks once more Riz and keep up lending us a hand.

 

[Riz]

IJR,
U r treading on difficult ground I suppose.
Shear walls are basically there to counteract lateral forces. The stiffness of each wall is different in each direction. A manually carried out analysis can suffice in 2D. Results are appreciative.
With a 3D model and if I have understood u correctly, the combined liftwalls for example can be projected as a boxed section and forces calculated. Thereafter, u need to create a spreadsheet that has moment of inertia of each unit. The inertia calcs will lead u to stiffness calcs of each wall that would counteract the forces.
Hope this helps
Riz

 

[IJR]

Thanks Riz

 

[pankaj]

IJR,

There is a simple way in stick modelling. Lste consider a l-shaped shear wall (you can take this analogy & extend to any shape).
Lets consider a L-shape with points a,b & c, with b being the intersection of the walls. so you assign 3 columns, 1 each at a,b & c. calculate the properties of both the walls a-b & b-c. apply half of the properties of the wall a-b to a & half of b-c to c. Then add half of a-b to half of b-c to get the properties of column b (all the properties like A, IX, IZ etc). This way you can have the shear wall of any shape, broken up in columns with equivalent properties.
I hope this helps

 

[Riz]

Sorry Pankaj, I tend to differ for the case IJR is proposing to simulate.
The stick model u have created combines a multi-lateral stiffness scenario. This will not help in calculating the extreme fibre stresses if there is tension build up at ends.
What IJR has to carry out is to allow about 25% lateral force to be taken by frame and 75% by shear walls. Shear walls than can be manually analysed as a 3D structure.
Riz

 

[ishvaaag]

I don't know if all have been talking of the same. One can model a C lift core even with lips with line column "frame" elements. Autocad AME will give you position of cog in the shape and Ixx and Iyy, and J and shear areas you can derive from Pilkey if such refinement you want. Then you put your core in 3D where its cog, and rigid links connect it to anything there meets.

Then conventional sectional analysis gives you the stresses; and other things can be checked if necessary precisely as per clause stated (think of one tip of shearwall required reinforcement).

Of course, some things are lost in this model, but in any model are lost, and normally I wouldn't proceed nonchalantly if the core would be critical... where I have worked with them cores were not, jsut contributors to stiffness. I practice in Spain and the percent stiffness allocation is for now of no concern, what is not to say ther are no clauses pertaining to aseismic design.

On the other hand, I also model walls and slabs with plates, and it is not difficult to come to some reasonable reinforcement scheme...what does not mean it would be nice to have a very good automatical way of such thing doing. So programs do...but do what you would think is correct? (assume similar and good knowledge in both the programmer and the user). In Spain one program reinforces slabs, true, but the output is as if made by one mad draftsman.