Post-installed or cast-in anchor? 2

[DTS419]

Let's say a 3/4" hole is formed in precast concrete slab so that a 1/2" or 5/8" bolt can later be inserted through the hole to be an anchor. Would this anchor be considered cast-in or post-installed for the purpose of determining factors per ACI 318-19 Ch. 17?

An argument for considering the anchor to be cast is that since the hole is pre-formed, there is no stress induced in the concrete by drilling of the hole, which seems to be a consideration in R17.6.2.6.1.

An argument for considering the anchor to be post-installed is that the anchor does not benefit from cement bond, although for a short anchor with minimal development length, this seems unlikely to be significant.

 

[DTS419]

I will add that I know this is not precisely in the scope of ACI 318 Ch. 17 since 17.1.5 excludes though-bolts, which is what this is. So we need to use some engineering judgment to model some capacity from this situation, and best I can tell this will behave like an embedded headed bolt.

 

[driftLimiter]

I agree but would probably also check bearing and punching shear and compare.

 

[KootK]

Given hole tolerances, I'd think that you'd want a generous plate washer under the bolt head and maybe treat it as a punching shear problem as some engineers do. Less good for shear.

 

[DTS419]

I agree but would probably also check bearing and punching shear and compare.

Given hole tolerances, I'd think that you'd want a generous plate washer under the bolt head and maybe treat it as a punching shear problem as some engineers do. Less good for shear.

So the interesting thing about punching shear, which I know a lot of engineers use for through-bolts, is that is returns a less conservative result. Since 17.1.5 gives us a pass on Ch. 17, I think using punching shear could be justified. My only hang-up with that is that it is counterintuitive that passing a bolt through a hole without any cement bond will give me a higher allowable load than if the bolt was cast in the concrete and evaluated as a headed anchor per Ch. 17.

 

[KootK]

So the interesting thing about punching shear, which I know a lot of engineers use for through-bolts, is that is returns a less conservative result.

Yeah... that.

Conventional punching shear is unconservative for this situation in my opinion. As I understand it, classical punching shear checks benefit significantly from an assumption that is baked into them. Namely, that the punching shear will be accompanied by a significant amount of flexural compression which will enhance the shear capacity across any contemplated shear crack. This makes sense for a typical concrete column to slab connection obviously. Less so for voodoo thru-bolt anchorage.

A work around that I've seen other engineers apply for this is to modify the punching shear check such that it uses the one way shear capacity applied to the punching shear frustum. This will lower the capacity in a way that I feel is fairly rational, since one way shear does not rely on the presence of flexural compression.

This is still not as good as having something backed up by testing but, as you well know, sometimes you just have try to do something reasonable in order to keep things moving forward. In a case like this, I would very much prefer to use a conventional wedge or adhesive anchor. I've no doubt that you have your reasons for what you are attempting however. I've done plenty of weird stuff for plenty of weird reasons.

 

[DTS419]

Yeah... that.

Conventional punching shear is unconservative for this situation in my opinion. As I understand it, classical punching shear checks benefit significantly from an assumption that is baked into them. Namely, that the punching shear will be accompanied by a significant amount of flexural compression which will enhance the shear capacity across any contemplated shear crack. This makes sense for a typical concrete column to slab connection obviously. Less so for voodoo thru-bolt anchorage.

A work around that I've seen other engineers apply for this is to modify the punching shear check such that it uses the one way shear capacity applied to the punching shear frustum. This will lower the capacity in a way that I feel is fairly rational, since one way shear does not rely on the presence of flexural compression.

This is still not as good as having something backed up by testing but, as you well know, sometimes you just have try to do something reasonable in order to keep things moving forward. In a case like this, I would very much prefer to use a conventional wedge or adhesive anchor. I've no doubt that you have your reasons for what you are attempting however. I've done plenty of weird stuff for plenty of weird reasons.

Thanks Koot, you get it! By applying the one-way shear to the punching shear frustum, do you mean using Vc = 2 f'c^(1/2) * bo * d ?

Interestingly, if I look at this as plain concrete (very little reinforcing actually), then I get a result that is very near to what I get by considering it a Cast anchor and using Ch. 17 (Ncb). That plain concrete punching shear strength comes out to 2.66 * f'c^(1/2) so not much more than one-way shear.

 

[KootK]

By applying the one-way shear to the punching shear frustum, do you mean using Vc = 2 f'c^(1/2) * bo * d ?

Yessir. There's even literature precedent for this (sort of). In Bijan Aalami's post-tentioned concrete book, he takes this approach for goofball punching shear checks where the frustums straddle slab soffit steps etc. Bijan seems like a pretty clever cookie. I certainly wouldn't want to compare IQ's. Maybe FTP's....

That plain concrete punching shear strength comes out to 2.66 * f'c^(1/2) so not much more than one-way shear.

Yup, I like that approach too.

 

[BridgeSmith]

I'm not sure how ACI would define it, but it's not an anchor of either type. I would say you'd analyze the tension capacity based on whatever ACI's criteria/method is for punching shear, because that's the failure mode. The head of the bolt is bearing on the concrete; it doesn't matter which side of the slab the bolt is on.

 

[DTS419]

I'm not sure how ACI would define it, but it's not an anchor of either type. I would say you'd analyze the tension capacity based on whatever ACI's criteria/method is for punching shear, because that's the failure mode. The head of the bolt is bearing on the concrete; it doesn't matter which side of the slab the bolt is on.

I don't disagree, but what's the difference between this mechanism, and an embedded headed anchor that is imparting its load on the concrete through the head of the bolt?

 

[BridgeSmith]

I don't disagree, but what's the difference between this mechanism, and an embedded headed anchor that is imparting its load on the concrete through the head of the bolt?

The one difference I see is that in this case, there's no bond between the shaft of the bolt and the concrete. Not sure if that's significant, since the failure perimeter (cone) in the concrete should be similar in either case.

That's assuming the washer under the bolt head is large enough and thick enough to prevent local crushing of the concrete. If that happens, the stress distribution may change in ways that make my brain hurt just thinking about it.

 

[Tomfh]

I don't disagree, but what's the difference between this mechanism, and an embedded headed anchor that is imparting its load on the concrete through the head of the bolt?

No difference. It’s a bolt head pulling out a cone of concrete.

Just make sure you have a decent washer. You don't want the bolt head pulling through the hole.

 

[DTS419]

No difference. It’s a bolt head pulling out a cone of concrete.

Just make sure you have a decent washer. You don't want the bolt head pulling through the hole.

So for the purpose of determining a breakout strength, would you treat the anchor as cast-in or post-installed?