Can you use anchor reinforcement with Post Installed Anchors? 3

[MJC6125]

I've use the Design of Pedestal Anchorage paper by Windianto in the past (link below for reference). I've only ever used it for cast in headed anchor bolts. I'm wondering if a similar methodology could be used with post installed anchors (specifically threaded rod epoxy anchors with deep embedment)? Is there something different about a properly installed epoxy anchor that would create a different concrete breakout cone or something else that would not satisfy the assumptions of this design methodology?

My specific situation that I'm trying to justify anchor reinforcement for is slightly different than pedestal anchorage, but I'm wondering if I can do something similar with tension breakout out cone anchorage. I have an anchor drilled into the top of an 8" thick concrete wall that has about 5 kips of tension force on it. If I satisfy all of the ACI Chapter 17 anchor design equations except for concrete breakout, can I use the existing vertical reinforcement in the concrete wall to resist the concrete breakout? See image below for reference.

Link to Paper: Link

Image:

 

[Rabbit12]

Following this. I had a rep from Hilti essentially tell me I could but he couldn't provide any sort of research on the subject.

As long as the breakout cone starts at the end of the threaded rod I think it's valid. Would still need to check bond strength of the epoxy. Is the bond strength reported by adhesive manufacturers limitless? I could see that being a sticking point on really deep embedments if so.

 

[Celt83]

have had the same response as Rabbit12 essentially a "sure....but we don't have any testing to backup that statement"

I think for what you are looking it you may find that your Hef ends up exceeding the limits in chapter 17 because it is unlikely that your wall bars are hooked so your Ld,min is going to be in the neighborhood of 20"+ which already exceeds the limits.

My Personal Open Source Structural Applications:

https://github.com/buddyd16/Structural-Engineering

Open Source Structural GitHub Group:

https://github.com/open-struct-engineer

 

[azcats]

Can you assume that the cone starts at the end of the anchor? I'm not sure about that.

Hilti has done research on 'developing' and 'lap splices' using their adhesives. Perhaps some of that can be applied towards a solution here. Even if it's adding supplemental reinforcing.

 

[Rabbit12]

azcats, that's the million dollar question. Does the cone start at the bottom of an adhesive anchor.

I'm not sure either but if I think about the load transfer between the epoxy and concrete it's essentially uniform along the length of the anchor correct? If that's true how could a breakout cone develop let's say halfway up the anchor? In my mind the only way would be for the rod to yield at that point which I see as not feasible.

 

[KootK]

Frankly, I'm not even all that convinced that the breakout cone starts at the bottom for even headed anchors when they are very long unless some kind of special measures are taken to debond some of the embedded length. At some point, I'd think the bond would have to take over as the failure initiating mechanism. But life and anchorage must go on.

Hilti has done research on 'developing' and 'lap splices' using their adhesives. Perhaps some of that can be applied towards a solution here

It's that for me with the post installed anchors. Basically just adapting the Widianto stuff to straight RC concrete principles with the adhesive not doing much other than improving the development length of the anchors. And, given everything else that needs to happen to make a go of the detailing, that isn't really all that much of an improvement over just cast in threaded rod other than the constructability advantages of going post-installed.

 

[AskTooMuch]

Why not? You use that similar logic when you make an existing footer bigger. When you need to develop the new rebar to existing rebar. So why would that not be applicable to anchor bolt.

 

[dauwerda]

There is a decent discussion on this in this thread:

https://www.eng-tips.com/viewthread.cfm?qid=461271

 

[r13]

In the early 90s on one of our dam project, Hilti's sales representative had conducted a non-scientific tension testing on rebars drill and epoxy grouted into unreinforced bulk concrete body. 100% of the failure mode was failure of the rebars through necking, some with observed shallow surface damage of the concrete, but none with deeper cone shaped failure as would be for rebars with bolt/plate at the end.

 

[Klitor]

Was the spacing of legs on his testing apparatus large enough to make cone failure even possible ("unconfined")?
Some old testing info below

http://www.fdot.gov/structures/stru...ior and design of single adhesive anchors.pdf

 

[r13]

The demonstration tests were conducted on single straight bars with embed length as suggested on the Hilti catalogue at that time. The pull force was applied using rebar pretension device.

 

[KootK]

I'm guesting that Kiltor's concern is that the testing apparatus may have pushed back on the concrete substrate within the influence area of the potential concrete cone. Such a setup:

1) Was very common in the past;

2) Still happens a fair bit now and;

3) Effectively tests only development without testing anchorage in a meaningful way that would include concrete breakout.

My wife ran into this on a project only two years ago. They were anchoring tall parapet wind posts to a thin concrete deck and wanted the tension anchorage tested. Anchor guys showed up with the testing apparatus that pushes back inappropriately and said "no problem, your bars will develop fy". As always, it is my life's mission to help the world properly distinguish between anchorage and development. It's a big 'ol world though...

Part of the problem is the anchor guys themselves. 99% of the time, you're dealing with someone who is, effectively, a commissioned sales person. The smart guys that would know better are huddled around a VAX 9000 somewhere in Liechtenstein behind the corporate sales firewall.

 

[sdz]

Yes there is something different about a properly installed epoxy anchor that would create a different concrete breakout cone. For a headed anchor the cone always starts at the anchor head. For an epoxy anchor the cone can start part way along the anchor. Where exactly depends on the bond strength relative to the concrete strength. Generally for interaction between anchors the cone is assumed to start at the embedded end because that's the most conservative.

If you know the bond strength you can calculate the depth of the cone but bear in mind there can be considerable variation so be conservative.

That said you also have the problem of ensuring the reinforcement is anchored on both sides of the failure cone. ACI 318 does allow some modest increase in capacity for edge reinforcement.

 

[Settingsun]

I would work out the depth where the supplementary reinforcement is sufficiently anchored (or sufficient quantity) to take the load and then calculate required embedment beyond that notional depth. Cone failure does not need to be considered as the supp reo handles that. This would cover the (shallow cone)-and-bond combined failure mode which otherwise wouldn't engage the supp reo.

 

[RFreund]

Sorry to bring this back to life...

I'm trying to understand the issues with this idea.

It seems we have the following rejections:

[ol 1]
[li]Celt83 - "I think for what you are looking it you may find that your Hef ends up exceeding the limits in chapter 17 because it is unlikely that your wall bars are hooked so your Ld,min is going to be in the neighborhood of 20"+ which already exceeds the limits."[/li]
[li]A few people - The breakout cone does not start at the bottom of the bar.[/li]
[/ol]

For #2 - Can you just calculate the length that it takes for the epoxy to fully develop the bar? I'm not sure if this would be information provided by the manufacturer or in this thread pbc825 had a fairly reasonable idea on how to do this. In any case, you would take your embedment and subtract out the length needed to develop the bar and start your failure cone at this point. Then you would check your anchor reinforcement above the failure cone.

For #1 - It sounds like you are saying that you are "kicked out" of Appendix D (now chapter 17 anchoring to concrete). Could you then apply non-contact lap splice provisions which to me seem like they would be even less conservative than what is discussed in #2 (meaning yield less embedment).

My trouble with saying that this doesn't work is that you can do this with post installed rebar, but you're going to tell me that I can't do it with threaded rod?

EIT

http://www.HowToEngineer.com

 

[Celt83]

RFreund:
I didn't mean to say it can't be done more that once you get to embedment depth needed to satisfy the development length criteria on each side of the failure cone you are likely beyond the limits of the testing and model basis used for ACI Chapter 17 (Appendix D) and have to work with the adhesive manufacturer which gets to be a bit of a circular loop because they'll say that sounds reasonable and then give you a CYA paragraph about how they don't have test data to back it up.

Snips from 318-14

I believe part of this is why PROFIS Rebar exists now,

https://www.hilti.com/content/hilti...ware/structural-engineering/profis-rebar.html

Could you then apply non-contact lap splice provisions which to me seem like they would be even less conservative than what is discussed in #2 (meaning yield less embedment)

yeah this is how I see a lot of engineers side stepping Chapter 17 (Appendix D) and is a bit of grey area.

My Personal Open Source Structural Applications:

https://github.com/buddyd16/Structural-Engineering

Open Source Structural GitHub Group:

https://github.com/open-struct-engineer

 

[RFreund]

Celt - thanks for the response. A couple questions to make sure I'm clear...
If the anchor reinforcing bars are relatively close to the anchor rods and you don't need the full bar to be developed, so your h.ef is <20da, then you're still in within the limits of Ch17 (appdx D), right? So then you could apply Ch 17 (Section 17.3.2.1) with the Windianto method. The question becomes on where the breakout cone starts.

Regarding Profis - this uses lap splice provisions and not Ch 17 (appdx D), correct? This is similar to the "side-step", right? Except Profis uses rebar instead of a threaded rod?
Sorry, I'm "crutching" a little bit, I suppose I could dig into these questions more.

I guess I'm trying to understand where the breakdown is. It seems like if you satisfy the h.ef constraints then you can use Ch 17 (Section 17.3.2.1) with the Windianto method. If you don't meet the h.ef requirements, then you are looking at non-contact lap splices similar to Profis. I don't know what sort of "modifications" are made in Profis to account for the drilling method, etc. but if you use a threaded rod vs a rebar, are you saying (not you but someone) that this then voids this method?

Thanks again

EIT

http://www.HowToEngineer.com

 

[Celt83]

...and you don't need the full bar to be developed,....

I think like shear friction if you need to bar for anchor reinforcing it has to be fully developed on each side of the failure cone.

On the Profis rebar front yeah I think it's similar to the "side step", I haven't gone to far down that path to give a great answer here.

I guess I'm trying to understand where the breakdown is

From my limited look into the topic it just doesn't seem like there has been enough or adequate testing done to line up with real world construction. I've only every seen papers with test setups in a chunk of plain concrete with the anchor either near or far away from the edge and loaded in tension or toward the edge.

My Personal Open Source Structural Applications:

https://github.com/buddyd16/Structural-Engineering

Open Source Structural GitHub Group:

https://github.com/open-struct-engineer

 

[KootK]

The question becomes on where the breakout cone starts.

1) In my opinion, you always have the option to design the connection via RC concrete principles and, thereby, bypass APP D entirely if you wish. You just gotta follow that logic all the way through which can be a challenge.

2) If you will be effectively lapping the anchor to adjacent rebar, I don't feel that you need to sweat the nature of the breakout cone. I believe that the design methodology for bar development, and therefore lap splicing, assumes and accounts for the development of a shallow breakout cone near the free surface prior to the development of the remainder of the embedded bar. So it's baked into the cake in that sense.

 

[RFreund]

A few further thoughts on this for record keeping sake.

You can either using post installed reinforcement + anchor reinforcement or you can use lap splice / rebar development equations. Using rebar development equations generally leads to fairly deep embedment.


Regarding the breakout cone with epoxy anchors. I take back what I said here:

For #2 - Can you just calculate the length that it takes for the epoxy to fully develop the bar? I'm not sure if this would be information provided by the manufacturer or in this thread pbc825 had a fairly reasonable idea on how to do this. In any case, you would take your embedment and subtract out the length needed to develop the bar and start your failure cone at this point. Then you would check your anchor reinforcement above the failure cone.

When you imagine the breakout cone of an epoxy anchor you wouldn't find a portion of the bottom of the bar that pulled free from the cone, so why wouldn't the cone start at the bottom of the anchor?

Here is the photo for reference:

Regarding developing the bar / lap splice - there is some good discussion here:

https://seblog.strongtie.com/2020/0...-bars-secured-concrete-construction-adhesive/

My hangup has been, why can't you do the same thing with threaded rod. I suppose you could, but another idea would be to use rebar with some sort of threaded coupler at the end.

You just gotta follow that logic all the way through which can be a challenge.

Any notable challenges?
Maybe I just need to work through a problem...


EIT

http://www.HowToEngineer.com