Design of Eccentric Column Connection

[jseng9]

Hi,

I'm an engineer in training and I'm trying to take a first pass at designing a column connection for a 36 inch deep by 18 inch wide by 20 ft tall wood column with a steel plate embedded in the middle of the column, bolted through on each side. The steel plate is welded to a base plate at the base (see attached sketch). The axial load in the column is 15 Kips and due to architectural reasons the plate embedded in the column is offset from the center of the column such that the center of the plate is 8 inches from the center of the column.

Could someone take a look at my attached sketch and tell me if I'm going down the right path? I think the bolts need to be designed to resolve the moment caused by the eccentric load but someone else has told me that the connection can be designed as a simple pin connection. with vertical and horizontal reactions... this doesn't make sense to me.

I think I need to design for shear in the bolts caused by the eccentric load. I would resolve this moment such that the top two bolts and the bottom two bolts provide horizontal reactions (equal and opposite direction) which is essentially a couple that resists the moment. This shear force would be equal to the (axial load * eccentricity)/vertical distance between the top and bottom rows of bolts.

I also think the eccentricity of the support also causes flexure of the column - the column needs to be designed for the moment caused by this eccentricity as well, right? And then isn't there also flexure of the plate as well?

If someone could tell me if I'm going down the right path that would be very helpful!

 

[KootK]

I believe that most engineers would resolve the eccentricity by putting all of the moment into the column and putting none of it into the plate. That, employing the rationale that an indeterminate system will deliver the lion's share of the moment to the component offering the greatest rotational stiffness which, in this case, is the column.

I expect that the connection will need to transfer shear as well, both owing to the eccentric load and, perhaps, direct wind shears etc. That will produce some strong axis flexure in the plate.

 

[KootK]

Yet another argument for no moment into the plate would be that such moment would be transfered into the plate via the bolts forming a perpendicular to grain force couple, similar to what you originally assumed. That load path will naturally have some give, slop, and creep to it which will tend to reduce the moment attracted to that mechanism.

 

[TLHS]

What does the rest of your structure look like?

That eccentricity does need to be resolved somewhere. That somewhere is not /necessarily/ into the connection, but it will depend on the rest of your structure.

From a conceptual standpoint, you are currently thinking about this as a pure axial load in the column that is eccentric to a support that has moment and an axial load. You can flip that and think about a reaction that has a pure axial load and a column that is eccentric to the support and has an axial load and a moment. You can pick your own zero moment point, to some degree.

 

[jseng9]

Thanks for your response, KootK.

I expect that the connection will need to transfer shear as well, both owing to the eccentric load and, perhaps, direct wind shears etc. That will produce some strong axis flexure in the plate.

Yes, there will be a horizontal shear force on the column due to out of plane wind loads. For the shear caused by the eccentric load, do I have the right idea in thinking that the shear in the bolts caused by the eccentric load will be (axial load * eccentricity)/vertical distance between the top and bottom rows of bolts. I know this is likely an approximation but I wanted to see if this would be a good starting point. I think I might need to read up on the design of eccentric connections using something like the uniform force method or instantaneous center of rotation.

 

[jseng9]

What does the rest of your structure look like?

The baseplate bears on a concrete footing and the column spans from the base up to the roof and is braced at mid-height by horizontal members. I would say it is a relatively simple post and beam structure.

 

[jseng9]

So essentially it sounds that the column can resolve the moment and that the plate, being the less stiff element will not likely take the moment induced by the eccentric load.

 

[KootK]

Exactly! And that was pretty much the sketch that I was intending to prepare for you (thanks for sparing me the effort). The only alteration that I'd make would be to delete the M from the right hand sketch entirely. You probably mean for that to represent an internal moment which is valid but it does confuse the sketch a bit in my opinion.

 

[jseng9]

Thanks, KootK! Yes, I meant for the moment to be resolved by the column, I see how it is confusing and will remove the moment from the sketch.

 

[dhengr]

Jseng9:
It looks like it is about 4” from the top of the base pl. to the bot. of the column. I would put an added base pl. right under the wood col. sufficient to take the 15k load in bearing, plus some extra bearing area. I would slot this pl. so it fit and welded around the vert. bolt pl. I would put one or two vert. webs btwn. the two base pls. This takes the vert. col. load and the moment essentially goes away. Now the bolts can be fewer and really only take lateral shear loads, and they are not loaded in several different directions. Otherwise, your analysis of the problem looks about right to me, as a first shot at the problem.

 

[KootK]

X2 for dhengr's detailing recommendation. whenever possible, I like my wood connections to be bearing connections. You'll probably wind up with a big check right down the bolt line anyhow.

 

[jseng9]

Thanks dhengr for the suggestion, I will consider providing more bearing area for the column using the method you suggested. Thanks again!