Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations pierreick on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Concrete edge breakout for multiple anchors in a single row (drag load) 2

StrEng007

Structural
Aug 22, 2014
537
I don’t know the best way to go about introducing this topic, so I’ll give a specific scenario before I try to explain the applicable part of ACI Chapter 17.

Let’s say you have a load that is being dragged along a concrete beam and you’re providing post installed anchors to resolve that loading. In this scenario the beam is an 8”x24” concrete beam, and the anchors are provided along the centerline of the 8” dimension (top of beam).

The governing load failure for a single anchor in shear (parallel to length of beam) ends up being concrete edge breakout. This can achieve a strength design reaction of 2,000 lb.

Per Hilti Profis analysis, when providing groups of anchors loaded in the same scenario, the load capacity is also 2,000 lb regardless of the number of anchors.

Single anchor fails from concrete edge breakout at 2,000 lb.
Screenshot 2025-03-10 120427.png
3-anchor group fails from concrete edge breakout at 2,000 lb.
Screenshot 2025-03-10 120449.png
5-anchor group fails from concrete edge breakout at 2,000 lb.
Screenshot 2025-03-10 120514.png

I’m not sure of the right way to interpret this information. For instance, if you had a single angle and single plate using this attachment, you’d develop the 2,000 lb reaction. If 2 ft away, you had another anchor on a separate plate, I believe you’d achieve another 2,000 lb reaction. However, when you put these anchors on the same plate, you get limited to a single breakout capacity based on a single anchor (regardless of the number of anchors). This is due to the limiting geometry perpendicular to the edge, all described through Avc and Avco per ACI 17.7.2.

Screenshot 2025-03-10 120833.png

So, what is the right answer? Do I base my numbers of the single anchor, and determine the required field spacing? Limiting multiples groups of anchors to a single anchor's concrete edge breakout seems too conservative.

The reason why I even attempted to put these anchors on the same plate was to make sure their proximity to each other didn’t have some sort of influence.
 
Replies continue below

Recommended for you

It's basing the group capacity on the rear anchor only, once the shear failure plane is developed there the rest of your anchors will just go along for the ride. I've only ever had situations like this is new construction and have always use shear reinforcement.
 
You beam is narrow, you will not be able to increase your breakout capacity with additional anchors.


If you have top rebar available, you can justify this with anchor reinforcing:

Refer to ACI chapter 17.
For examples, refer to "Companion to ACI 318-14" and "PIP STE05121".

1741626500676.png
 
It's hard to tell from the graphics, but note that my shear is parallel to the length of the concrete, not perpendicular. I've run into this condition with an existing beam, so it's difficult to apply shear reinforcement after the fact. I don't know how to justify using edge reinforcement when my loading is NOT perpendicular to the free edge.

Let's say I have a drag load of LRFD 1,000 lb/ft and a single anchor can achieve 2,000 lb. Is it incorrect to say if my anchors can be spaced as such:

Capacity/Demand = 2,000 lb / (1,000 lb/ft) =2FT O.C.

My loading is in lb/ft. This is why I was saying the group interaction might be too conservative? Am I missing something?
 
Probably yes, but no way to determine what they are without scanning them.

How does the tie work to relieve the thrust from the struts? In an 8" wide beam, the tie would not developed the required hooked length for something like this.
 
My original question still stands though about approaching this for a linear reactions. Let's pretend for a minute this is unreinforced concrete.

Why can I not base my anchor spacing off off the the capacity of one anchor then spacing accordingly?
 
StrEng007, that seems logical to me. Per your profis output, ca1 is 16". I don't see why you couldn't space the anchors at 16" o.c.. The true edge distance is a much higher number.
How else would you anchor drag strut elements to walls? They need to be designed as a linear load.

If you wanted to do 24" o.c., I would probably do a single anchor in profis, with an edge distance somewhere between 12" and 24", with the load being 2' times the linear load.
 
I’m inclined to agree with @CDLD. Back in 2014, there were cases to determine how the shear got split up between the anchors and whether one anchor (front or back of the line) was limiting the capacity. I haven’t kept up with how this has changed in 2019, but I don’t think it’s reasonable to simply assume that no single anchor in the row will see more load than what will produce a breakout failure.
 

Attachments

  • image.jpg
    image.jpg
    4.9 MB · Views: 11
ANE91:
but I don’t think it’s reasonable to simply assume that no single anchor in the row will see more load than what will produce a breakout failure.
OK, I think I can see some truth in that. This is like saying, you can have an entire baseplate of 20 anchors in a line, but if the location of your load is concentrated at a specific point along that plate, a single anchor or a couple anchors are going to feel a majority of the load.

Now would you say the same thing for something like a column bolted to a wall. Let's say you have gravity load at the top of a column, but that column is bolted to the wall at intervals. The initial axial load is in the column... does the first anchor closest to the top of the column feel all the load?

At a certain point, it seems crazy to think that you could have 100 anchors supporting that column, but the entire system fails due to a local affect. What's the middle ground?
 
I think one key assumption here is that these are not oversized holes since this is a post installed anchor. If you have a 30' long drag strut element, with anchors at 2' o.c., it seems unreasonable to assume that 1 anchor out of 15 will take all of the load, or only 3 anchors out of 15 since Profis says up to first 3 rows for the second option in shear load distribution per ACI 318.

1741636269312.png
 
What's the middle ground?

It depends on the situation. Engineering judgement comes in.

Nothing to intercept the final failure cone? Go with the conservative answer the software is giving you.

Sufficient longitudinal reo to restrain the anchor? Then share the load
 
I noticed that your Phi.parallel,V is 1.0. I think it should be 2.0. Try deleting the frontmost anchor / increasing anchor spacing.
Now would you say the same thing for something like a column bolted to a wall. Let's say you have gravity load at the top of a column, but that column is bolted to the wall at intervals. The initial axial load is in the column... does the first anchor closest to the top of the column feel all the load?
Shear is directed away from the edge, so 17.7.1 or 17.7.3 governs. I take your point, though. It’s a complicated stiffness problem. For perpendicular cases, you can weld the baseplate to the anchors to mobilize them “equally.” These provisions boil down to fracture mechanics and concrete capacity design. This method is overly conservative to your case. Remember that the underlying research is relatively new. Try running a hand calc per 17.7.2.1.2.

Generally, if you’d like to justify any shear distribution that isn’t Case 1, 2, or 3, then you’d need to sharpen the pencil.
 
Last edited:
One thing to be mindful of is for lets say the base of a shear wall you are really designing for a shear per unit length where in Hilti's tool you are telling it a single concentrated load to then be distributed to the anchors using the plate. If this was truly a concentrated load case then I think you would be limited to the single anchor capacity as Hilti suggests but in the distributed load case the load path to each anchor is different and rather than try to trick Profis to check this I would work out the design of the elements in the load path to the anchors independently and then check a single anchor in Profis.
 
So, what is the right answer?

For me, this goes onto the heap of things classified as "Yeah, problems. But since there are no solutions and life must go on, don't sweat it".

- The holes will be bigger than the bolts so there's really no way to guarantee which bolt sees load first.

- There really is not ductility in the system such that one might lean on redistribution. Particularly with respect to single anchor pryout which often governs.

So what do you do? Embed a bunch of single anchor plates and weld a plate on top of them? That's got its own problems:

- Wet setting quality control in general.

- Heat of welding probably elongates the plate and tears everything up.

The reason why I even attempted to put these anchors on the same plate was to make sure their proximity to each other didn’t have some sort of influence.

Depending on your situation, I actually would prefer to not have all the anchors on the same plate. at 24" oc, I wouldn't think that you would have issues of overlapping influence. Maybe you do it in groups of two anchors per plate just so you have rotational stability.
 
There really is not ductility in the system such that one might lean on redistribution.

If we're concerned about brittle edge failure, I agree. But if we have reinforcement, then additional anchors can be mobilised.
 
If we're concerned about brittle edge failure, I agree. But if we have reinforcement, then additional anchors can be mobilized.

I was referring to pryout of each anchor individually. Only the stirrups would help with that. And the kind and spacing of stirrups that I'd expect in an 8" wide beam wouldn't normally be of much help with that.

But yes, longitudinal top bars can be used to restrain the shear breakout of the front anchors if the detailing checks out.
 
I was referring to pryout of each anchor individually.

I was considering the edge breakout of the front anchor, as that appears to be the governing factor in this case, which we seem to agree on, regarding the longitudinal bars catching it.

However, regarding pryout, does the same principle of no load sharing apply? The PROFIS software seems to distribute the load across multiple anchors when calculating the pryout load per anchor, as opposed to edge breakout where it spreads the load across the outer anchors only. I take your point that in theory pryout is brittle too, like edge breakout is.
 
Last edited:

Part and Inventory Search

Sponsor