Static tension on embedded anchors 1

[grantstructure]

The backspan of a cantilevered roof beam is going to place tension on some anchor rods I'm placing the in top of a concrete column.

Anyone have some tried and true pretensioning methods/notes for anchor rods they'd like to share? I just want enough to compensate for creep.

 

[ishvaaag]

If I understand well the question, I quite likely would apply enough pretension to (say 4 bolts) as to be enough to resist the effects of some tensile (really shearing) force imparted by the beam to the anchor plate atop the column. Then the coefficient of friction multiplied by the axial force actually imparted should be enough to counteract the factored shearing force. I would ensure that the local reinforcement for introoduction of the pretension is enough (as in prestress anchor plates), and that the section of the column can deal with any envisaged solicitation.

A code treatment of the question must be dealt in one "anchors to concrete" viewpoint for which the ACI code has a number of equations, and anchors providers have software for. This would be enough for the introduction of forces, but I would also consider what described above and would ensure that enough column shear reinforcement exceeds the anchors' length since there a complicated stress state occurs, from anchoring the tensioned rods and concurrent directly imparted shear.

If what I describe does not correspond to your problem, please provide a sketch for clarification.

 

[grantstructure]

You've basically got it. It's not shear being applied but pure tension, as the steel beam will sit on top of the column. Unless you're simply referring to the bond stress along the anchor as a shear stress.

My question is really one of what methods are employed to develop pretension in embedded anchor rods. The uplift force is not large (<10k total, serv.) Since in this instance I also am having to detail for future expansion, I'm thinking about placing a lenton coupler on the ends of my main column steel, and then coming back with little projections of weldable rebar which would thread into the lenton splice and be welded off to the beam base plate. That gets me out of worrying about pretension and creep, and gets the force directly into the main column steel.

 

[WARose]

What I’ve done for pre-tensioning anchor bolts is used a product like Rowan’s anchors. They are adjustable and have very large bearing heads at the end that facilitate pre-tensioning.

Here is their web site:

http://www.rlrowan.com/canister-bolt

If the canister doesn’t meet your needs (for whatever reason), you can always use embedded plates. In any scenario (for embedded anchors) you can minimize creep by using a low ratio of the total capacity of the connection (based on the research I’ve seen).

 

[ishvaaag]

OK, since you seem only to anchor some uplift force caused by the cantilever and it seems also be able to readjust any potential variation of the level of the anchor point at the backspan you need a part anchoring some base and then an adjustable device on top of it; this can be combined in a single commercial product as WARose suggests or as per some design that satisfies your need; as you define the case with couplers taking the tension from the baseplate the adjustability it seems would be lost once embedded in the column.

 

[KootK]

Interesting question. I would not have thought to prestress such a connection. What is the nature of the creep that we're concerned with here? How would the prestressing serve to ameliorate it?

 

[dcarr82775]

I assume these are cast-in-place anchors, not post installed. If you really want to pretension the bolts, and not just torque down the nuts, provide a plastic sleeve over the upper portion of the embedded bolt. This will allow a free length so the steel to be tensioned free of the concrete.

I don't see that this will eliminate creep however.

 

[grantstructure]

Kootk, the creep I'd expect is just the concrete "relaxing" under the load, and the only thing I don't want to happen is for a snug-tight nut on the top of the anchor rod to be able to work loose over time. The beam is cantilevered at a roof condition, so uplift on the underside of the overhanging portion of the beam, which is lightly loaded, could reverse the sense of loading at this connection. I'm thinking it could cycle loose, or start to have movement at the connection. I'd like a little more than just spiked threads on this one.

The pretension would just be enough to keep the bolt permanently clamped--that is, that creep over time would not be enough to completely remove the clamping force of the beam connection. Since I have net tension under dead load it's only when the load reverses under wind upift that the bolt could then lose tension. I just want it to stay clamped, even if only slightly.

I'm guessing there's some sort of spring loaded nut I can use to accomplish this as well. Just have to find it.

 

[hokie66]

Like others, I also don't see how shortening of the column would cause the connection to loosen. If the nut is tight, it goes up or down with the top of the column.

 

[grantstructure]

If the column shortens along the length of the rod, but the rod does not shorten, or not as much--it wont' creep, but if it's tensioned will lose some of its elastic extension--, then the nut could lose bearing contact. Am I thinking about that wrong?

Here's my thought process.

1. Don't want the nut to come off, or loose, if connection cycles between applied tension and applied compression. Tension will be normal (gravity load) state.
2. Pretension - how do I do that?
3. If the anchor is in tension, how much do I have to worry about creep? At some level it's negligible. Also at some level of pretension, the creep won't overcome the tensile force in any reasonable time frame.

 

[WARose]

When I have done pretensioned anchors in the past I’ve assumed about 40% loss of pinching force (in the long haul) due to relaxation of the steel in the bolts for my calculations. Most research I’ve seen backs that (although I have seen some that suggests as high as 60% over really long periods).

So given that (if you are worried about the nut loosening) you could set your target pretension force (which I have normally specified as a percentage of the anchor bolt yield stress) to the maximum tension force the anchors will develop in its service life + the % loss I discuss above and I would assume you would be covered.

As an additional safety factor, I have sometimes specified periodic inspection (with critical equipment) of the bolts to be sure it is maintain that pinching force.