Rigidity of Steel Column Base Plate

[J1D]

It’s an old topic, I know some threads talking about “hinged” or “fixed” base in analysis. But a specific question occurs to me from time to time: we often use base plates with four anchor bolts outside of wide flange column, and the column base is modeled as a pin. With different details, such as bolt size, bolt spacing, size of column, thickness of base plate, etc, the rotational rigidity of base plates varies a lot. Conservatively designing the anchor bolt, we can still consider a “fixed” base, but this may cause bolt overstress (at least theoretically) by the bending moment induced. On the other hand, once a base plate has some flexibility the bending moment at base is very small in lot of cases. The uncertainty is caused by the difficulty to model the base plate rigidity.

Any suggestion or practical guidance on to what extent/situation we can confidently assume a four-bolt column base is a pin? Thanks.

 

[jike]

The rigidity is dependent upon three things:

Elongation of the anchor bolts
the flexibility of the base plate
Rotation of the footing

I remember using an equation with three terms (matching those items above) that I found in one of the older editions of the PCI handbook. You can determine the approximate rigidity of the column base so that it can be modeled with a rotational spring support, if desired. I wrote an Excel spreadsheet, so that I could see how changing the various components would affect the rigidity.

This gave me more of a feel for the effects of varing the different components. It was more of an educational benefit for me.

In my everyday design practice, I assume either pinned or a fixed base.

 

[kxa]

Jike, When you say that you assume either pinned or fixed, is it based on some criteria or based on the results of your spreadsheet analysis? I come across this situation a lot and normally assume pinned connection.

Can you tell which PCI edition you were referring to? I would be interested in setting up such a spreadsheet. Does your spreadsheet account for soil condition/footing size?

Thanks,

 

[jike]

I normally assume a pin at the base of columns unless I have a rigid frame which may require a fixed base, to control drift.

I must have got the equations from the 1st or 2nd Edition (which I no longer have). I see that it is also in the 4th Edition (which I have at the office)Pages 3-44 to 3-49.

The equations take into account the footing size and modulus of subgrade reaction for the footing rotation.

With a fixed base, I normally would normally want the column to be on a combined footing, grade beam or wall (to minimize footing rotation). If I do not have that situation but still need a fixed base on an isolated footing, I would try to provide a rectangular footing (2:1 ratio) to minimize footing rotation.

 

[J1D]

Good points. I forgot to mention that most of our columns are on piles. And the anchor bolt elongation doesn’t contribute much (no sleeve or sleeve grouted after installation). The bending deformation of the base plate therefore predominates. Some design spec suggests using fixed bases in the analysis model if the foundation is “rigid”. But as I can see, there is a big ambiguity then for the real bending moment before we have better ways other than “hinged” or “fixed”.

Sometimes I use half of the bending moment from the “fixed” case to check the four anchor bolts. I know, with an analysis software, we can actually model a base plate with the size of column and anchor bolt location. But it is a hassle for general structural analysis currently.

 

[Ron]

J1D...you hit the significant point.....no matter how we decide to "model" the condition, the physical aspects will control how the in-place condition will act. Even when you model the base as pinned, there will be some moment induced in the system, particularly with wind loads, and the couple has to be resolved for the anchor bolts.

I prefer to make the base/column interaction rigid, even when the anchor condition might be truly a pinned condition, simply because the flexibility model of a base plate is actually rather complex and this negates the need to consider that.

 

[PowerRanger]

Hi Fellows

I am designing one of this connection now. Just have a question about the capacity of concrete in foundation resisting pull out. Should I only consider the concrete resisting the cone shear failure, or could I consider the tension force transfer to the Reo. bars in pile. (Bars lapped but not weld to bolts). This will create big different of the pile cap.

 

[rholder98]

I've taken to designing for cone failure, but making sure my steel laps also, just to be sure. I have not had a project yet where the added steel makes a significant difference in cost. Be sure and check with Appendix D of ACI 318.