Column Base Plate - Fixed or Pinned 3

[PEFLWI]

I am involved in a dispute with a contractor about a base plate connection to a concrete pier. The base plate is anchored to a concrete pier similar to the attached detail taken from Design Guide 1. The contractor's engineer is adamant that this can be a pinned connection. He is suggesting that the anchor bolts will stretch to relieve the moments. What do you think?
Can the moment be relieved by stretching the anchor bolts? If so, how much stretch of the anchor bolt would be required?
The anchor bolts only extend 5 inches from the top of the pier to the top of the base plate. I don’t think I can count on stretching the anchor rod inside the pier.
A complicating factor is that the concrete piers are already poured. Adding anchor rods would be expensive.

 

[GeorgeTheCivilEngineer]

In a braced frame the braced locations will be much stiffer (order(s) of magnitude) so the pinned/fixed distinction won't matter.

For a moment frame, most nominally pinned baseplates will act as fixed for most 'real' loads - that is, under SLS loadcases. But if these nominally pinned connections did see ULS loads then they would not sustain fixed behaviour for very long and would become like a pin. If the structure relied on nominal pinned baseplates being fixed for stability at ULS then the structure may be liable to fail.

If the concern is a about parasitic fixed behaviour then I think it would need consideration of all the stability systems and what the load path is.

.

 

[human909]

Most of the time things work because stiffness is a relative thing. If you design assuming a pinned connection then your frame generally is stiffer so the rigidity that does exist in the column base doesn't really come into play. Likewise if you have a relatively flexible frame and rigid base connections. The base connection will attract plenty of load precisely because there it is the stiffest path.

I've seen failure of an end bay of a portal frame where the roof buckled and thus the stiffness of the frame against lateral movement dropped dramatically. The pinned baseplate was the next stiffest path and attracted the load. Unfortunately the earth around the bored pier wasn't strong enough and the pier underwent rotation. The drama halted when the large roller door flew off its tracks and reduced the load on the frame.

 

[PEFLWI]

Thanks everyone for your help.

 

[CrabbyT]

I'm not sure what this column and base plate are supporting, but it looks pretty fixed to me.

I'm not sure I buy the bolt stretch argument, but it's definitely possible for this configuration to be pinned if the column were sufficiently flimsy.

 

[CANPRO]

it's definitely possible for this configuration to be pinned if the column were sufficiently flimsy

Are you sure about this? I think this is backwards. The more "flimsy" the column (lower bending stiffness) the stiffer the base plate is in comparison, and which would lead the column to "feel" the fixity of the baseplate more than it would for a stiffer column. The stiffer the column is compared to the baseplate, the less impact the base plate fixity will have on the column behavior. In the extreme example, if the column is infinitely stiff and the base plate is as proposed in this thread, the column is infinitely more stiff than the base plate and the base plate will essentially be pinned as far as the column behavior is concerned. On the other side of the extreme, if the column is essentially a cooked noodle, it doesn't take much baseplate stiffness for the end of the column to behave in a fixed manner.

 

[bones206]

I think one rational way to classify the connection is to calculate the secant stiffness Ks of the connection at service loads, then compare to the AISC classification scheme for fully restrained (FR), partially restrained (PR) or simple connections. Basically, if KsL/EI ≤ 2, it's classified as simple. If KsL/EI ≥ 20 it's FR. In between those limits it's PR. L and EI are the properties of the column. This classification scheme is laid out in Chapter B commentary in the ISC specification.

There are various analytical ways of determining Ks, which is equal to Ms/Θs, but I would probably use Hilti Profis since it would be quick and easy and accurate. Model the base plate and anchors then run it in Profis with Ms applied, then back calculate Θs using the elongation of the anchors (strain in the anchors x the stretch length). The stretch length in this case would be the thickness of the grout + base plate thickness + plate washer thickness.

The above only consider anchor stretch, but you could also consider base plate deformation (which Profis outputs) and foundation tilt. Not sure it's worth going through all that just to settle a dispute in a project that's already under construction, but it's an interesting from a theory standpoint.

 

[CrabbyT]

Are you sure about this? I think this is backwards.

I'm pretty confident about this. If you form a plastic hinge at the base of the column (near the point where it attaches to the base plate), the joint can act like a hinge.

Imagine (4) columns in a 20' x 20' bay. Imagine they are connected by W21x55 beams. Imagine the column base plates are 2' x 2' x 6" thick and connected to the slab with 16 anchors.

Now imagine the columns are made out of 1/2" round stock. It doesn't matter how fixed the plate is to the concrete or how fixed the column is to the plate. If the column is very small compared to the beams/loads that it needs to support, the column itself can act like a hinge by bending at some distance about the base plate.

 

[CANPRO]

CrabbyT, I see what you're saying and I agree. The point I was trying to make is better explained in the follow up post by bones206. In terms of classifying the base plate as either pinned or rigid, a stiffer column requires a stiffer connection to be considered fixed. Using your example above, if your column is 1/2" round bar it doesn't take much of a base plate to be considered fixed - in which case you can form that plastic hinge above the base plate before the base connection rotates excessively. If you have an incredibly stiff column, your base plate may exhibit pin-like rotations long before the column develops any significant bending moments at the base (compared to the bending capacity of the column). What is considered fixed for the 1/2" round bar column might be more appropriately classified as pinned for a W14x120.

As your column stiffness (EI/L) increases, you need a stiffer base connection for the column to exhibit fixed-base behavior. The stiffness of the base connection being determined by the plate flexibility, bolt stretch, foundation movement, etc...