Cast-in vs Post-Fixed Anchors - Equitable or not 2

[Greenalleycat]

Howdy all

A trend here in NZ is for almost all fixings (at least in residential) to go to epoxy post-fixed anchors rather than cast-in.
As engineers we much prefer to spec cast-in bolts for capacity, but contractors push for post-fixed as they give significant tolerance
This is particularly useful for portal baseplates (or for epoxying in wall starters because they inevitably did not put enough care into getting them into the right place to start with).

However, a topic of hot discussion in our office is whether post-fixed anchors really are an equitable substitute at all and whether we, as an industry, should be backing off our use of these in primary structural elements.
Our concerns are that the actual load path of an epoxy anchor is substantially different to a cast-in fixing
Cast-in fixings generally rely on a washer etc at the bottom as the origin of the cone, meaning that the failure cone of the anchor will spread out and engage bars around it
The reliability of the load path can be checked by assuming a failure cone angle and checking that steel is developed within this cone

Epoxy anchors, by contrast.... we calculate say a shear and tension force on the portal baseplates, apply these to our Hilti Profis model (or similar) and get a tick saying that our proposed 2/4 anchor arrangement is satisfactory - job done, design out the door, black box software happy

However, epoxy anchors generally fail by pulling out a relatively shallow cone that starts at an unknown point above the base of the fixing
This means that there is a reduced ability to develop reinforcement bars within the depth of the cone
The strength of the anchor therefore ultimately relies on concrete in tension - something you aren't allowed to do in our codes

This problem becomes even worse when considering the tendency for contractors to stuff up their starter bars and try to epoxy in new starters to substitute
The original hooked starter has well-understood performance in terms of strength of connection, formation of concrete strut on the inside of the hook etc
The epoxy bar...does anyone know how these fare when you have a mixture of epoxied and cast-in starters?
What is the strut and tie model for an epoxy bar? When the capacity of the epoxy is exceeded, what happens (presumably the whole thing fails very brittly)?

There are also practical issues such as the fact that the QA on these is generally non-existent - you're reliant on the contractor following the spec, not anchoring into cover concrete, etc

I am curious to know what other engineers' views are on this topio as we are starting to view it as one of those unspoken issues being pushed by the manufacturers and contractors and not resisted by engineers

Cheers!

 

[phamENG]

If I need to develop anchor reinforcement, I do not allow post installed anchors. I check for the anchor when I check for rebar. If it isn't there, in place, and tied securely you don't get your sign off on the pre-pour inspection. If it doesn't need to develop the reinforcement and isn't especially close to reinforcement (read: they won't cut bars when the drill for the anchor), then I'm willing to entertain the idea. If it's a high load condition and I know the cast in is really work hard...I'll tell them to provide a sealed alternate design and I'll consider it...but don't pour the concrete until I've approved it writing. That's usually enough to get them to take the 10 minutes to cut out a plywood template.

On most residential work, I let the post installed stuff go in 98% of the cases. If it's standard code compliance, then I'll go for it all the time. If I've engineered a specific solution to a unique problem...I fall back to my statements above. It's usually pretty rare that I have to deal with loads that meet those requirements in a house, though.

 

[dik]

Use them all the time... I think that cast in are better with a caveat that if they are being 'sticked'... bond could be compromised. I also require that rebar be located prior to drilling... have confidence in both methods...

I just had a recent project for curtainwall support on an L4x3x3/8 with 2-1/2" from the top of the curb to the anchor and 1-1/2" from the bottom. Load was vertical and horizontal in the worst combination... with all sorts of prying action.

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik

 

[phamENG]

dik - good point on the 'sticking' or 'wet setting'. That's a definite no-no. Goes for anchors and rebar.

 

[dik]

I've seen it done too often... generally the merit for using pre-installed.

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik

 

[skeletron]

For my projects (residential, upgrades to existing, limited scope repairs) it's almost 100% epoxy anchors. The tolerance issue trickles down to speed in the field which is a huge cost benefit. This is even considering that the industry standard epoxy is (to the best of my knowledge) a one-shot tube so the contractor has to reload a new mix every anchor (or something similar). Crews are either moving too fast or independently of the framers, making cast-in anchors tough to line up if you are working with skeleton crews, unfortunately.

As for the QA/QC on the epoxy installation, unfortunately I leave that in the contractor's ball of responsibility via general notes. They need to get trained personnel, they need to be responsible to follow the manuf. instructions, they need to do it right. In very basic terms: I'm looking for goop coming out of the holes and straight, tight bolts installed. If they want to pay me for handholding, checking their certs, watching their first few installs, I'm happy to do it for a fee that is otherwise non-existent at my project level.

Cast-in is way better but you need to know who is doing the work and what they are capable of. There's a product out there called BoltRail that has been said to speed crews up. If I had the chance to use it, I would, but I rarely dip into that scope of project right now.

One flaw of epoxy bolts that I've seen a couple colleagues spec blindly is the depth required...it's really easy to forget that you need that length of drill bit to get in there and sometimes THAT is the cost driver.

 

[BridgeSmith]

We've learned from messy experience to not rely on contractors to proper install epoxy anchors, especially if they're subjected to tension. They work fine if the epoxy is properly and thoroughly mixed and if the holes get cleaned really well. More often than not, at least one of those things doesn't happen.

On one job, we had girders attached down at the abutment caps, since we estimated about 5 kips of uplift at the ends of the girders once the center girder piece was in place. We specified cast-in anchor bolts, but the contractor conveniently 'forgot' to cast them in. They epoxied them in, but didn't clean the holes well, and they popped up when the center girder piece was released from the crane. I bet they had fun trying to balance 3 tons on the ends of each of the girders. After that, we detail cast-in u-bolts, so they have to do it right.

Short version - epoxy anchors are fine for shear, but the QC/QA better be darn good if you're going to rely on them in tension.

Rod Smith, P.E., The artist formerly known as HotRod10

 

[dik]

with shades of the 'Big Dig' failure... not just the contractor in that case.

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik

 

[phamENG]

I had a project that had a walkway along a masonry kneewall that had 16"x16" pilasters in it. Shortly after construction started, the architect decided to extend those pilasters up a couple feet above the wall and put cantilevered shade canopies on top (we're talking a 10' cantilever, in a coastal area with decent winds). It was a nightmare. Had to switch from masonry to concrete for the pilasters, and specified a rather crowded (but feasible) rebar/anchor arrangement at the top to set the steel canopy column on. Well...the contractor failed to cast them in, thinking he could just epoxy them later. These things had about 14" of embedment called out! We had a fun little meeting where I proved to him that no, they were not "over engineered" and, as a matter of fact, if you try to put a core through there large enough to meet the epoxy specs it will cut a bunch of rebar.

We went back and forth for a while - I wanted them to chip out the concrete and repour the upper few feet. They fought it long enough to set the deck and pour the concrete slab next to it so ripping out the column wouldn't be possible without ripping all of that out, too. So I designed this fantastic 'clamp' complete with pretensioning procedures, field welded side plates, and other fun things to engage the top few feet of the column externally. We've had some pretty good storms since then and they're still up there...

So...yeah...sometimes contractors will put you in a pretty tough place and, whether it's intentional or not, make you look like the bad guy.

 

[Teguci]

Probably do you good to do an epoxy anchor hand calc just to gain confidence. True the anchors require a development length to engage with the concrete but so does rebar. As for quality assurance, especially for critical structures, have them proof tested. Usually the anchor guys will have a testing crew that can give you the warm and fuzzy for an installation.

Developing the tension from an anchor into, say, a drilled pier foundation cage where the pier is 3' dia. and the anchor is in the center area should be given careful thought. A strut and tie model would be appropriate where the tension is transferred with compression struts and confined with hoops, or sim.

 

[Greenalleycat]

Cheers all, interesting to get everyone's views on them.

@Teguci - I have done my share of hand calcs so I am familiar with the processes - NZ Codes have their own method for cast-in fixings but just reference ETAG guidance for post-installed.
My/our concern is that, particularly when using software, the tendency as the designer is simply to have capacity > demand and call it a day.
As you say, a proper strut and tie model (or at least a half-arsed one showing load spread engaging stirrups/main steel) is never really done for post-installed fixings

 

[STrctPono]

I run into this issue all the time. We use CIP anchors where appropriate and allow post-installed adhesive anchors where appropriate. I actually took the time to get ACI post-installed adhesive anchor inspector certified since we had one particular job that had over 10,000 adhesive anchors.

In my experience, most Contractors are just really really bad at properly installing adhesive anchors and it has given me a bad taste in my mouth for blanketly allowing them. They don't read the adhesive manufacturer's printed installation instructions, they don't burn the initially mixed adhesive, and can't seem to clean a hole to save their life. I've seen Contractors trying to clean holes by putting their mouth on the end of a 1/4" tube and blowing air through it! Obviously, the bigger the job, the better the Contractor, and the (hopefully) better QC on their installations.

Either way, I have a very similar approach to allowing adhesive anchors as phamENG does.

 

[Enable]

I live in the world of restoration and if we couldn't rely on epoxy then we'd be totally screwed. We drill and epoxy all manner of things, including: primary reinforcing steel, shear dowels, and threaded anchors.

It works. Though as mentioned above, contractors may or may not have a tendency to short-change the epoxy (just because it looks oozed out on the outside doesn't mean they put it all the way into the hole) and are annoyingly bad at cleaning the hole. So, sure, QA can be an issue but if you have doubts about the contractor's ability to follow the literature, mandate proof-loading via an independent testing agency.

But don't just say no to epoxy. You be crazy!

 

[Greenalleycat]

@Enable, believe me we use our share of epoxy!

I think most of the comments in this thread have skirted around the root of the question though, which is that the load path of an epoxy anchor is not equitable to the load path of a cast-in anchor (or a bearing plate system such as phamEng retrofitted).
Numerically we can make epoxy anchors work, but NZ codes at least have underlying assumptions, intentions, and requirements for beyond ULS loading - failure hierarchies, not brittle failure, no concrete in tension, valid load path, etc

The only load path for the majority of epoxy fixings is concrete in tension. When this is exceeded, you have nothing.

 

[Enable]

EDIT - I'm an idiot

 

[Tomfh]

Cast in anchors (and reinforcement) rely on concrete in tension too.

 

[Greenalleycat]

Yea we have provisions for non-contact laps. In that case though you are typically lapping within a beam that is either stirruped or has fibre reinforcement in it, so there is still a transverse and longitudinal tension path. With a post-fixed anchor the embedments are generally 120mm to 200mm so, by the time you take out cover to reo (40-50mm) and the fact that the epoxy failure cone starts 20mm or so up from the base of the bolt.. you might only get 50mm to 150mm lap with your reo, which doesn't allow for much development. Deeper drilling is an option, but as already said in this thread, consideration has to be given to the practicalities of actually doing that (and the cost of the drill bits!)

 

[Greenalleycat]

@Tomfh, yes I agree that both rely on concrete in tension to some degree. Being accurate, a concern that I have with cast-in vs post-fixed really is the ease of providing and verifying the load path. With cast-in anchors you can generally get them deeper, with more reliable QA, and a more readily verifiable load path to confirm that you have developed steel within the pullout cone.

 

[BridgeSmith]

with shades of the 'Big Dig' failure... not just the contractor in that case.

Yeah, kinda similar, but that one was an issue with the poor long term creep performance of the epoxy material. The material was not suitable for sustained tension of the magnitude exerted by the panels suspended overhead.

Rod Smith, P.E., The artist formerly known as HotRod10

 

[Tomfh]

The same mixed-failure cone occurs in a piece of embedded straight rebar.

Epoxy has its shortcomings, but you can’t apply a black mark against epoxy for general behaviour of embedded rebar.