Elliptical concrete column 1

[Cian_Lynch]

Hi

I'm designing some bridge piers (to Eurocodes) which are to be elliptical in section and am having difficulty finding a reference or design guide as to the best way to assess them. Has anyone undertaken a design of this nature before and would have any suggestions? The piers are subject to bending and axial loading.

(I have already attempted to design them as circular columns inside an elliptical section but they are insufficient to withstand my loads).

Thanks
Cian

 

[Tommy385]

How big is the cross section and how long is the column? Is the bending moment "dominant", probably not...
I would first check the cross section in compression...
can you give us some basic numbers? i should have something in the Office, i will post it, if i will find it

 

[Cian_Lynch]

The section is 1.746m by 0.946m, 5.3m tall. Axial load is 11866kN with bending at the top of the pier of 6265kNm and at the bottom 4377kNm.

Thanks

 

[dik]

Does the column work as a circular column with a diameter equal to the minor diameter of the ellipse? Doe the column work as a rectangular column inscribed within the ellipse... else you have to do some detailed calculations for an interaction diagram... can check max compression value, balanced value and max flexural value to give you a 3 point interaction diagram.

Dik

 

[WARose]

dik has some good advice above. When faced with a irregular column shape in the past (I use to do a lot of architectural work), I'd draw a crude/conservative, lower bound interaction diagram (i.e. consisting of 2 straight lines connecting 3 points (max. P, pure max. bending capacity, and tensile strength based on vertical re-bar capacity)).

 

[dik]

WARose:
See if it works for something 'lesser' for which there may be interaction diagrams... hadn't considered the extra point for pure tension... but, have included it in past. I was just looking into interaction diagrams a couple of days back to unravel a conc wall interaction program I wrote in Delphi about 20 years back.

Dik

 

[JAE]

spColumn from Structurepoint can do column designs with "irregular" cross sections - including ellipticals.

Here's a link to their website: spColumn

I quickly free-handed a rough shape - you could probably do way better than this for sure. It also works in non-English units:

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[Cian_Lynch]

Thanks for the input, yes a simplified interaction diagram is the way I'm leaning. Just a lot of iterations/variables when working with an ellipse. A lesser section within the ellipse is not large enough unfortunately.

Jae - I'm currently waiting to get SP installed on my machine at work so once I do I'll be able to make a comparison hopefully.

 

[Tommy385]

Sorry, unfortunately i haven found it.
There are some good suggestions above.
What kind of system is it, i haven't seen some many bridges with columns fixed on both sides?

 

[ChadV]

You could also go really back-to-basics and do a spreadsheet considering the cross-section as a series of horizontal strips, with a small height to capture strain variation and a width calculated to match your elliptical shape. Rebar would be a set of points. You could iteratively solve for neutral axis depth then use strain compatibility to get the force in each strip and rebar, then integrate to get a moment capacity. If you were including axial load as well, you'd have to consider that in your solution.

Ideally you'll be able to find an appropriate program to do the calculation and/or some literature to back it up. But don't forget there's a first-principles way to handle it too.

EDIT: note also, this is the way programs like spColumn, S-Concrete, etc., are doing the calculation too - some of them will actually output the strip-by-strip calculations.

 

[Cian_Lynch]

No worries Tommy.

It is an integral 2-span concrete bridge. The piers will be in-situ pours fully fixed to the pier table and pile cap. I could model a pin connection at the base but as of yet I don't have SI information. Either way I need to design the pier for some amount of bending. once I get a spreadsheet set up for this it will be easy to go back and re-design if necessary.

Chad that was my first approach, I've not done much integration in a long time and it was quickly expanding to integration involving two variables so I decided to see if there was an alternative, even if it is more conservative.

 

[yakpol]

Concur with the rebar fabricator that it can produce the hoops. More likely you will end up with the corded perimeter hoop and crossties.

 

[dik]

yakpol:

just like JAE's sketch...

Dik

 

[Guest262653]

You could use the Blue5Soft SBeton software for the analysis of concrete sections of any shape:

http://blue5soft.com/en/home/

It's free for three months and then you can pay almost whatever you want.

Disclaimer: I'm not associated in any way with the software provider but I've been using it and I'd like to help the developer.

 

[dik]

Most CAD programs can draw an ellipse...in Bricscad:

Dik

 

[ChadV]

Cian_Lynch: fair enough, particularly considering the relative simplicity of using software that can handle the problem... however I would note, the way I'm describing, you shouldn't need to do any 2D integration - for each horizontal strip you would get a single value of strain, convert to a single value of stress, then multiply by area to get a force... technically it's 1D "integration" when you calculate the axial / moment capacity based on the forces, but really it's just summing forces and moments - again "technically" numerical integration but not calculus-y enough to scare anyone off (I think, anyway!)

Best of luck - I hope you're able to find a solution that isn't too onerous!

 

[Guest262653]

I enclose the results of a test run in Blue5Soft Sbeton using a DXF file created in Bricscad. It uses C30/37 concrete and 30 25mm A500 bars.

 

[Guest262653]

Here's the DXF file.

 

[Cian_Lynch]

ChadV I built a spreadsheet to work out the force and moment equilibrium in the section for a given neutral axis depth. I used this then to generate an interaction diagram, I've attached a couple of screenshots. This in a way, is what you're suggesting? As you rightly point out it's technically integration but in a more literal sense!

Thanks avscorreia, I got SBeton installed and ran a similar assessment, it confirms that the section capacity is sufficient, corroborating my spreadsheet. I've allowed for some bending about the Z axis which has bumped up my steel quantity. On a side note, when you first ran/installed SBeton was it giving you "an unexpected network error" warning? It will not allow me import from dxf or save the file. Apparently I need to dive into the .cfg file and alter the JIT settings. Just wondering if it's a global issue or just something on my machine.

Thanks for all the input again.

 

[Cian_Lynch]

Calcs