Design of V shaped RCC columns

[trilinga]

I am invoved in the design of an octogonal shaped RCC building.

Is there a design procedure or a handbook to design V shaped RCC columns (with an included angle of 135 deg.)?

Thanks for the help

Trilinga

 

[ishvaaag]

That sounds something to be treated with component plates or 3D brick elements, far too wide to be understood a line element.

 

[csd72]

Trilinga,

Do you mean V shaped in section or in plan.

Can you post a sketch?

 

[trilinga]

it is an octogonal shaped building. The columns are V shaped in plan.

Trilinga

 

[csd72]

How thick and long are the legs of the V?

 

[ishvaaag]

I took them on elevation. Anyway, a good compatibility of deformations section analyzer for concrete should be able to deal with dimensioning and checking the longitudinal stresses from a strength of materials viewpoint. I actually also remember of the previous concrete code EHE in Spain include a pair of approximations for the buckling issue for general shapes of columns, one more accurate (and usually cheaper) than the less accurate. I would have to look how it stands now. Respect shear quite likely making some approximation thinking along the lines started by csd72 might work depending upon the case.

Apart from FEM, this doesn't end our possibility of action: we can determine also the standing elastic stresses by the usual formulas, then dimension, say, yes, even for shear friction of the worst. Anyway for critical columns like yours likely important you will get better appraisal with the better 3D FEM model that you can afford, since by the approximations and without a reference struts crushing and other things like that that may be critical could go overlooked.

 

[BAretired]

RCC = reinforced cement concrete?

BA

 

[JAE]

I did a building once where the architect wanted 5-sided columns (the building had a pentagon theme).

We simply used PCACOL (the irregular column package that you get with the software).

So I guess I cheated and used a program.

 

[trilinga]

Thanks for the responses from all.

I must apologise for not making my query precise.

The building is a three storeyed structure octagonal in shape.

The columns at the corners of the octagon receive beams from either side of the corner of the octagon.

The columns are 12" wide 24" along the perpendicular direction.

The shape of the column is shown in the enclosed sketch.

I could not locate any design charts or design procedure to design this non-rectangular reinforced concrete section subjected to axial load and biaxial bending.

I need help in getting the design method.

 

[JAE]

I doubt that you will find any design charts for a column shape like that. You could dive in and attempt to do it by hand (recommended) but otherwise take a look at this:

Link to PCA spColumn program

 

[ishvaaag]

You may try to use my Mathcad 2000 worksheet made around 10 years ago that analyzes a section of any number of trapezes in compatibility of deformations for reinforced and prestressed sections; in fact the example is quite close. The case in the worksheet is directly adaptable with the inmediacy concomitant with delving unto a new tool, but it fits your case.

I include pdf and the worksheet itself in the attachment. It may work or not (and then need minor adaptations) in later versions of mathcad.

The problem with this is that you at most can deal with this with longitudinal stresses, but interaction with other torsion and shear forces, buckling itself need be dealt complementarily. So I stand with trying to investigate the member as solid in FEM to ensure you are safely enough free of any limit strength problem.

 

[trilinga]

Thank you guys.

JAE, Prior to starting this thread, I have almost concluded to your suggestion of doing the design by hand from first principles. You have reconfirmed it.

I do not have the PCA software but I can try to get it.

Ishvaag, I will also try to understand the method given in your Mathcad attachment.

Thank you both once again

Trilinga

 

[rapt]

Trilinga,

If you look at a good RC text book (or your engineering course notes) it will give a method of calculating section capacities for concrete sections of any shape fopr any combination of M and N applied.

This is the basis of RC design. You should not have to rely on design charts. I agree they are usefull but you should understand the theory of how they are generated so that you can generate your own for column shapes not covered by charts.

Do it manually, it will be a good learning experience for you.

 

[hokie66]

For a three storey building which wouldn't have really big column loads, I would try designing the corners as two separate rectangular columns and just tie them together. That should be a conservative solution. If the columns were working harder, then I would do further analysis.

 

[rscassar]

Same principles would apply as if you were designing a circular or rectangular column. Only being 12" wide and 24" along the diagonal, I wouldn't expect the results to diverge greatly from that of a rectangular column 12" x 24".

1. Plane sections remain plane after bending (i.e. strain profile is linear).
2. Assume some sort of stress-strain profile for concrete (linear would be easiest).
3. Concrete strain doesn't exceed 0.003 or 0.0035.

I would even be inclined to calculate the pure bending moment (i.e. if designed as a beam), balanced load and squash load and draw a line between the three points and keep within this line.

Because the section is only singly symmetric you should calculate the capacities when the shape is V-up and V-down because the different distributions of reinforcing steel will result in different capacities.

For biaxial bending, I would use code simplifications. I would not even attempt to rotate the cross-section to calculate capacities.

Best of luck.

P.S. FOR PEOPLE THAT HAVE PREFORMED THE CALCULATIONS BY HAND, WHAT STRESS-STRAIN RELATIONSHIP DO YOU USE FOR CONCRETE?

 

[IDS]

P.S. FOR PEOPLE THAT HAVE PREFORMED THE CALCULATIONS BY HAND, WHAT STRESS-STRAIN RELATIONSHIP DO YOU USE FOR CONCRETE?

I can't see any reason not to use a rectangular stress block using the applicable code provisions for the Ultimate Limit State, and linear elastic for serviceability.

Also I agree with checking bending capacity about both axes and using whatever method is in the code for biaxial bending to combine them.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[rscassar]

Thanks Doug, I was playing around with curvilinear stress strain relationships (equations as per the Rangan et al text) but it was just getting too messy. Particularly when trying to analyse a circular cross sections.

 

[ishvaaag]

I forgot to make life easy to say that when using the worksheet all relevant input is in blue, and output in yellow, everything else being ancillary computations.