Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations waross on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Stiffness of a Concrete Core in a Multi-Storey Building

Status
Not open for further replies.

Tygra_1983

Student
Oct 8, 2021
116
Hi there, everyone,

I have been working on a problem in which I would like to calculate the load shares and deflections of structural elements in a multi-storey building. These elements include moment-resisting rigid frames, diagonally braced frames, concrete cores, and shear walls.

Thus far, I have achieved very good approximations for irregularly shaped buildings that contain just moment-resisting rigid frames. However, now I am trying to do the same but with the addition of concrete cores. I need to know the stiffness of the concrete core so that I can include it in my calculations for stiffness of the entire structure?? This is going to be a long post, so if you can answer my question you can stop here. If you are interested in the procedure read on.

Because my building is asymmetric it will experience a torsion and a translation

Here is the building I am working on:
L-Shaped_Building_Image_plhw6r.png




I start with calculating the shear stiffness of each sub frame that makes up the entire structure in the X and Y directions. Look at the following plan of my building:

Plan_of_L-Shaped_Building_avtafw.png

The X direction is from left to right, the Y direction is from top to bottom, with the wind pressure acting on the structure from right to left.

I hope you can see the frames in the X direction (parallel to the wind load) are made up of 5 frames. Going from bottom to top in the plan we have 3 frames that have 4 columns and 3 beams; then there are 2 more frames that have 2 columns and 1 beam.

To compute the shear stiffness of a frame we can use the formula:

[/indent][pre]Kframe1 = Kb*(Kc/(Kb + Kc))​
where:​
Kc = sum (12*E*Ic)/h^2​
Kb = sum (12*E*Ib)/(h*l)​
E = modulus of elasticity of the steel​
Ic = second moment area of the columns​
Ib = second moment of areas of the beams​
h = storey height​
l = bay width​[pre][/pre]

Thus, for the first frame in the X direction, we would get:​

[pre]Kc = 4*12*E*Ic/h^2​
Kb = 3*12*E*Ib/(h*l)​[/pre]

This procedure is reapeated for all the frames in the Y direction.

So we have the shear stiffness of 5 frames in the X direction and 4 frames in the Y direction.

Lets call these frames:

[pre]K1x ​
K2x​
K3x​
K4x​
K5x​

K1y​
K2y​
K3y​
K4y​[/pre]

Now, a coordinate system must be defined. Taking the bottom left corner as the origin, we find the centroids of all the frames. This is easy as its just the centre of each frame. Hence in the Y direction:

[pre]K1x = 0m​
K2x = 4m​
K3x = 8m​
K4x = 12m​
K5x = 16m​[/pre]

And in the X direction:​

[pre]K1y = 0m​
K2y = 6m​
K3y = 12m​
K4y = 18m​[/pre]

From this I can calculate the centre of stiffness​

[pre]ybar = (K1x*0 + K2x*4 + K3x*8 + K4x*12 + K5x*16)/(K1x + K2x + K3x + K4x)​
xbar = (K1y*0 + K2y*6 + K3y*12 + K4y*18)/(K1y + K2y + K3y + K4y)[/pre]

For my building, these coordinates are:

[pre]ybar = 6m​
xbar = 10.67m​[/pre]

Now the torsional stiffness is required. For this a new cooordinate system must be defined, where the origin is the centre of stiffness. Thus:

[pre]t1x = ybar - 0​
t2x = ybar - 4m​
t3x = 8m - ybar​
t4x = 12m - ybar​
t5x = 16m - ybar​

t1y = xbar - 0​
t2y = xbar - 6​
t3y = 12m - xbar​
t4y = 18m - xbar​[/pre]

The torsional stiffness is obtained by:

[pre]q1x = K1x*t1x^2​
q2x = K2x*t2x^2​
q3x = K3x*t3x^2​
q4x = K4x*t4x^2​
q5x = K5x*t5x^2​

q1y = K1y*t1y^2​
q2y = K2y*t2y^2​
q3y = K3y*t3y^2​
q4y = K4y*t4y^2​

qw = (q1x + q2x + q3x + q4x + q5x + q1t + q2y + q3y + q4y)​
​[/pre]
The total force coming from the wind is:

Fwind = 10 kN/m^2*H*B​

Where:​

[pre]H = height of the building​
B = width of the building face that the wind is acting upon​[/pre]

The moment induced is simply:

[pre]Mwind = Fwind*(8 - ybar)​[/pre]

Now the load share of frame1 can be computed by:

[pre]F1 = Fwind*(K1x/(K1x + K2x + K3x + K4x) - Mwind*(K1x*t1x/qw)​[/pre]

For the other frames use their stiffnesses and plug them into the above equation.

Sorry for this post being so long, but I thought it was a good idea to show you what I am doing. Again, I need to find out how stiff a concrete core is. I have the modulus of elasticity and second moment of area for it, but the equations I have tried like: [pre](8*E*I)/H^4[/pre] are not working!

Many thanks in advance,

Tygra





 
Replies continue below

Recommended for you

I have tried a core made up steel braced frames and I am getting accurate results. So, the method is correct.

Braced_coore_a7fhk0.png


My book by Karoly Zalka gives the stiffness of a braced frame as:

Code:
Kbrace = 2*(((d^3)/(Ad*E*h*3^2) + 3/(Ah*E*h))^-1)

where:

Code:
d = diagonal bracing length
Ad = cross-sectional area of the diagonal bracing
E = modulus of elasticity of the steel
h = storey height
l = bay width
Ah = cross-sectional area of the beams

I multiply it by 2 because in the X and Y directions there are two braced frames.

I am just struggling to find of stiffness of a concrete core. I assume it has EI/H in it somewhere, but using this alone is giving incorrect results.

Once I have the stiffness I can use this formula for the load shares between the different structural elements.

Load_Share_Forumla_dclcbs.jpg
 
You’ll need to make sure you include shear deflection when working out the concrete core stiffness. The force distribution will also depend on if you treat your core as a box section or individual wall panels.
 
I had a thread where I worked through the rigid diaphragm analysis with skewed lateral elements:
Link
 
Hi Celt83,

I am wondering how you calculate the shear stiffness? If we consider one wall of the core is it the shear modulus multiplied by the cross sectional area?
 
Most concrete textbooks should mention the shear deformation for concrete walls, in general it follows Timoshenko Beam Theory.

One thing I missed is if you treat the core as a box section you will also need to include the torsional stiffness, I believe the Concrete Book by S.S. Ray has good coverage on that, it’s an older book so should likely be able to find it online.

Edit: section 2.7 in the Zalka book also addresses the torsional stiffness.
 
Strangely enough I went the library today and looked through the books on concrete and I only found one that had a chapter on shear walls.

From this book I got the formula which I had already tried which was 3*E*I/H^3 + G*A/(1.2*H). Didn't work!

From Zalka he computes the relative stiffness as one divided by the maxiumum deflection for all structural elements. If we consider the formula again:

Load_Share_Forumla_jo6cgj.jpg


I am wondering how would we compute the rotational stiffness in the perpendicular direction to the load? Lets say x is the perpendicular direction to the horizontal load, in direction x the load would be zero, so you wouldn't even have any Kxxj stiffness if you used Zalka's method.

What do you think?
 
Can you clarify what you mean by “didn’t work”?

Is your calculation not matching computer software, example calculation?
 
Yes, its not matching the computer software.

To test the accuracy I compute the maximum deflection under the calculated load shares. When I have the load shares I use the equation delatMax = w*H^2/(2*K).

Where w is the calculated load share, H is the building height, and K is the stiffness of the subframe.

Using the braced core I get the following results in the software:

Load_share_braced_frame_ralnxb.png


And the approximate values as:

Code:
     Deflection
               __________

    yframe1      91.805  
    yframe2      120.42  
    yframe3      149.04  
    yframe4      171.59  
    yframe5      197.63  

>>

As you can see they are quite accurate.

Compare with the concrete core

Load_share_concrete_core_wck4wc.png


And the approximate values using the stiffness of the core as (3*E*I)/H^3 + G*A/(1.2*H) are:

Code:
               Deflection
               __________

    yframe1      9.9562  
    yframe2       36.31  
    yframe3      62.217  
    yframe4      88.125  
    yframe5      105.23

As you can see they are way off.
 
I would recommend going to a simpler model you are getting torsional effects in your current model that are complicating your checks.

Model a structure for one story that consists of just the core walls and a slab spanning between them, this will give you a torsion free system so you can back check just straight translation.
 
Okay, I made a 4 storey symmetric (so no torsion) structure consisting of two shear walls, and using G*A/(1.2*H) I obtained accurate results.

I tried again for our more complex structure with two shear walls instead of the core, but didn't get accurate results.

Perhaps there is an error in my method, but I don't know what. Its strange that it works for a steel braced core, but not when its concrete.

The stiffness equation used for the shear wall assumes a concentrated force at the top of the wall. I am wondering because the force on my structure is central, so half way up the wall, not at the top, perhaps that could be a factor in this problem.

Celt, I can quite easily derive the equation for deflection in flexure using double integration, that for a concentrated force half way along the cantilever. However, I don't have any experience in deriving the shear part. Could you help me with this?
 
I don't believe your stiffness equation, 3*E*I/H^3 + G*A/(1.2*H), is accurate as the stiffness is not the sum of the inverse of the flexure and shear deflections but the inverse of the total deflection.
Capture_gtphjx.jpg


Additionally the shear deformation for a fixed-free beam with a point load on the free end is PH/kAG, where kA is the shear area and k is a factor. For a solid rectangle k is approximately 10(1+v)/12+11v where v is poisson's ratio. For a full box section, while not technically correct, I would approximate kA as the two web areas in the direction of the loading.

The wikipedia page on Timoshenko Beams is pretty good: This book is relatively inexpensive and might be more helpful:
 
Thanks, Celt. I think I will definitely buy the book.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor