Eccentricity Question

[moyseh]

Hi,

I'm looking at designing columns including the eccentricity acting on them by their connections. This is simple enough to understand for a simple pinned connection. However, for a moment connection, do you still need to allow for eccentricity since the connection isn't fixed to the shear centre?

Cheers

 

[dik]

I don't understand the question. Can you post a sketch? Columns and cantilevers are two things I don't 'skimp' on...

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[HTURKAK]

May be not so simple ? The eccentricity of a bolted beam to column connection with end plate, web cleats and so on) to web of an open column may be neglected but considering the eccentricity of a connection to a column flange will be better approach ..

[quote..... However, for a moment connection, do you still need to allow for eccentricity since the connection isn't fixed to the shear centre?][/quote]

If the conn. is rigid , you will simply model the conn. as rigid and perform the analysis . If the conn. is semirigid , the modelling is another story.. I hope you are not planning to consider moment of moment ?

I will suggest you to look to the following doc. for modelling issues ..










I cannot give you the formula for success, but I can give you the formula for failure..It is: Try to please everybody.

 

[GeorgeTheCivilEngineer]

Am I misreading the diagram or are the (b) and (c) labels the wrong way around?

 

[HTURKAK]

Yesterday when i posted , i saw the mistake.. You are right.. the (b) and (c) labels the wrong way ..Shall be ( b)beam connected flange and c) beam connected web)..











I cannot give you the formula for success, but I can give you the formula for failure..It is: Try to please everybody.

 

[JoshPlumSE]

I've seen many people do this. I have some thoughts (based on computer software experience):

a) My tendency is to believe that few people did this prior to the advent of RAMSteel which popularized the calculation.
b) Though, you can certainly argue that for gravity only corner columns, this can actually make an important change to the design. This is because the eccentricity is not opposed by a similar load + eccentricity on the other side.
c) Where it starts to gets complicated is when you start considering multi-story columns with beams framing in from both sides at multiple levels. When I did this (for the development of RISAFloor), we used a moment distribution type of scheme to distribute the loads along the height of the column. It ended up being kind of tricky.

Again, to me this is very much worth doing for corner columns where it will have more of an effect on the final design. Maybe for columns where the beam on one side is much more heavily loaded than the other. But, for columns where the beams on either side are pretty much the same span.... I just don't see it making much of a difference.

 

[human909]

I've seen many people do this. I have some thoughts (based on computer software experience)

Do what exactly? I'm curious. Do you mean allow for eccentricity on a moment connection or not allow for it?

To my mind it makes no sense to allow for eccentricity on a rigid connection as this would likely involve double counting of static forces.

Having never used RAMSteel I'm not sure what calculation you mean by popularising it. Could you elaborate?

Thanks in advance.

 

[structSU10]

Josh is just talking about the simple connections. I would have to check what it does for moment connections - it may take the moment due to the shear force eccentricity into the column and all the beam moment stays in the beam, basically double counting, but I am not sure. when there is continuity over the column I think it is fair to say the shear is delivered through the centroid of the column and any bending in it is just a direct correlation to the relative stiffness between the beams and columns - the effect of 'eccentricity' is accounted for directly.