How would you model this frame? 1

[greznik91]

I have a concrete frame (columns are aprox. 3 m high and 3 m apart).
Columns have circular cross-section with diameter 250 mm.
Beams are 300/400 mm.
Columns are on single foundations.

I know there is no such thing as 100% fixed / pinned in real world, but if you have to model this would you rather choose pinned or fixed support as a colums supports?

Also, at the corner joints (column - beam) would you design this as pinned connection (no moment transfer)?

 

[kellez]

Hi,

I am not that experienced but i am guessing that everyone will suggest to create 2 models

1) model it as fully fixed in order to capture the behaviour of the frame without considering the foundations, even though they are not fully fixed because in reality you will have some sort of soil structure interaction.

2) Since we know that there will be some soil structure interaction you should create another model that includes the foundations in you model and try to model the soil as well. Then you can use both models to dcide on the design of the structure.

 

[ash060]

Typically you would model the column/footing interface as fixed, since a typical concrete column to footing connection with transfer moment.

For the column to beam connection it will be fixed as well. The interior columns would take very little moment since the spans are equal on both sides, but the end columns would get some, usually around a 16th moment.

Your beams could be designed using the coefficients give in the ACI code Chapter 8 (or Chapter 6 in ACI 2014)

 

[rb1957]

not my area, but I'd possibly do both models to bound the problem.

I notice that in both models the columns are fixed to the beams.

another day in paradise, or is paradise one day closer ?

 

[wannabeSE]

With spread footings, pinned bottom and fixed at top. Let the footing rock. If the bases are fixed, the footings need to be very large to prevent them from rocking.

 

[dik]

Is it supporting anything other than self-weight? What sort of snow, ice, or wind? 10" dia columns on 10'high are pretty flexible. 30cm wide beams over 250 columns is a good start...

Dik

 

[EDub24]

Depends on the footings. If they are designed to resist the moment from the column without rocking then you can treat it as fixed. Typically they would have to large enough so that the edges don't lift up to obtain this (or it's on piles). Otherwise I would treat it as pinned.

For the column/beam joints I would treat those as fixed unless you specifically detail it so that there is no moment transfer between the two. This isn't easy to do with concrete. If you do fix it, you can determine the moment transfer based on the deflection of the joint which takes into account the relative rigidities of the columns and beams. The easiest way to do this is to model it using RISA or SAP2000 unless you want to brush up on your structural matrix analysis.

 

[dik]

EDub24... I'd normally treat the columns as fixed at the footings if soil is reasonable... The relative stiffness of a 10" column to almost anything almost guarantees fixity,

Dik

 

[n3jc]

Some of you mentioned 'rocking' of spread footings. Can this be also applied here:

http://www.eng-tips.com/viewthread.cfm?qid=430956

 

[rapt]

For the Beam/Column connection, I hate the term Fixed.

In analysis, Fixed has definite meaning. No Rotation.

In most cases you actually mean 100% moment connection. So the moment in the column is equal to the moment in the beam at edge connections and rotation occurs at the joint. That is NOT Fixed.

For footings, normally you would assume pinned unless you can guarantee that there is no rotation of the footing. Unless you have a rigid pile cap with tension/compression piles or socket into solid rock, there will always be some rotation of the footing, so you will never get full fixity.

 

[mcdermott2]

I would choose pinned for the base. If I had the modulus of subgrade reaction, and was modelling in RISA (or similar program), I would consider calculating soil springs and would use those for my boundary conditions at the footing/soil interface.

 

[Settingsun]

If you're analysing an existing structure, you need to look at the footings to decide whether they are best represented as fixed, pinned or something in between. If you're designing a new structure, you get to choose but then need to design the footings to match the analysis results.

It's a fairly slender structure so the footing/pile needed to fix the bases would probably be out of proportion to the superstructure. This suggests pinned bases, at least for strength design. For serviceability (deflection), you might consider a small fixity. 10~20% of the column stiffness (4EI/L) appears in some design guides for steel structures and these concrete columns are of similar size/stiffness.

Column-beam connection: If your bases are pinned, you need the tops fixed for stability. If the bases are fixed, you have a choice. I'd go fixed column-beam for insitu construction because that's how the real structure will tend to act. Zero moment in the real structure will need significant cracking. For precast, pinned connections might be significantly easier.

Rapt, I think your showing personal preference by not using 'fixed' at the beam-column connection. Fixed in this context refers to the connection between analysis element (member) and the node: there is no relative rotation; the angle between elements (members) is fixed.

 

[rapt]

Steveh49,

It has nothing to do with a preference of mine.

Historically, "fixed" has a meaning for what happens at the end of a member.

If the people who develop analysis software want to use a term for the connection between a beam and a column, it should not be "fixed" as that has a different meaning.

 

[serrojo]

My suggestion to you:
Decide upon consequences.
* Column - Beam: Continuous - - - Means that the column have to take moment (more complex column design/detailing)
* Column - Beam: Pinned - - - Means that the beam is simple supported (more rebar steel to place in the middle span of the beam of height increased)
* Column - Footing: Continuous - - - Means that the Footing have to take moment from the column (the dimensions could be greater if stability issues found)
* Column - Footing: Pinned - - - Means that the Footing transfers mostly axial force (the dimensions could be lesser no stability issues found)

If I had to decide:
* If I had small loads on the beams (say lightweight roof) + Good soil: Column - Beam: Continuous + Column - Footing: Continuous

* If I had great loads on the beams + poor soil: Column - Beam: Pinned + Column - Footing: Pinned

Best Regards,

MSc. Eng. Serguei Joa
Structural Engineer
Bouygues Batiment International, Cuba.

 

[jayrod12]

Serrojo,

How can you have both the column-beam and the column-footing pinned? you need at least one of those connections to resist moment for stability.

There are two options that most of us see, pinned base and moment resisting frame, or moment resisting base and pinned frame connections.

 

[BAretired]

If the frames in the sketch are intended to be moment resisting, the end frames will be critical, not the side frames. If columns are deemed to be pinned at the base, only the four corner columns contribute to lateral stability under wind forces. The central four columns are not engaged.

BA

 

[BAretired]

In response to the question "How would you model this frame?", I would probably seek some shear wall action in the end walls. If that is not available, I would add some diagonal braces. If that is unacceptable, I would consider the columns continuous top and bottom and make the footings large enough to accommodate the resulting moments.


BA

 

[jike]

You could use two conditions to bound the problem and then design for the envelope of forces:
1) Pinned Base
2) Fixed Base

Or you could design the base as a spring to account for the estimated rotation of the footing.

Or you could model the column with a horizontal support at the foundation level and extend the column down a certain distance to a pinned base to provide the same rotational resistance at the foundation level. I believe this method and how to calculate the distance is described in the PCA Handbook (if I remember correctly).

 

[dcarr82775]

Same as everyone else, it could be either pinned or fixed depending on footing size and demand loads. I would lean towards pinned at the base, and then if I needed a little bit of help to make something work better I would estimate a rotational spring stiffness for the base. Fixed at the base is unrealistic in many circumstances for a footing