Options for dealing with pryout failure (CSA A23.3) 1

[EngDM]

Not sure if this is how it works for others, but where I practice pre-engineered building suppliers provide us with reactions and the EOR designs the foundation for these loads. I have several frame lines with insane kickout forces (700kN) and the pre-eng supplier has provided (10)-1 1/4"Ø anchor bolts. The problem is, these bolts do not work in pryout even with taking 3 out of 4 edge distances as code max 1.5hef and playing with the hef. One side has to remain a set edge distance to align with building edge (edge of grade beam).

Question is, what kind of detail can I provide that effectively eliminates pryout? thread507-455245 seems to give differing opinions of if you can reinforce for pryout, but I'm wondering if I could possible add weldable rebar or some other type of restraint to relieve the shear. So far, any anchor designer software I try to run in does not consider pryout as a tension reinforecable failure which leads me to believe that there is no real way to "reinforce" the concrete for pryout by adding ties or vertical rebar.

I'm solely solving pryout at the moment. We are considering tie-rods from frame to frame or dywidag anchors to deal with the shear loads.

Edit: I checked it with infinite edge distance to trigger 1.5hef as my area on all sides and it still fails.

 

[jayrod12]

I'll respond just to bump this back to the front page. I don't have much in terms of true code guidance. I feel if you had both stirrups, and ties, those would eliminate the chance of pryout, thoughts? Reinforcing crossing the shear plane is normally the fix no?

 

[EngDM]

I feel if you had both stirrups, and ties, those would eliminate the chance of pryout, thoughts?

I agree that stirrups and ties could likely resolve a pryout failure, however in the thread I linked, KootK's post showed a failure mode where the anchor bears against concrete and almost looks more similar to a tension failure, in that the breakout cone comes up to concrete surface. I'm not sure how I'd determine what tensile force would need to be resisted when just a shear load is applied, and the failure looks close to the surface - would I even have a stirrup yet if my top cover is 2"? I'm not sure if the failure seen in figure c) and d) is only seen when anchors are shallow since the crack appears to propagate from the embedded stud head.

I'm worried about a high stress concentration on the concrete near the top of the anchor just under the baseplate due to such a high shear.

 

[sticksandtriangles]

The go to paper on anchor reinforcing in concrete says that you can ignore concrete pryout failures for embedments deeper than 12*Ø. Any chance you can get longer embedments from your PEMB to meet this guidance? 12*1 1/4 = 15, not too bad.

https://www.sefindia.org/forum/file...nforcement_a_petrolium_project_papers_199.pdf

I typically just go deeper and ignore pryout.



S&T

 

[EngDM]

The go to paper on anchor reinforcing in concrete says that you can ignore concrete pryout failures for embedments deeper than 12*Ø. Any chance you can get longer embedments from your PEMB to meet this guidance? 12*1 1/4 = 15, not too bad.

Thank you for the link. I'm just not sure when it is acceptable to ignore certain types of failure when it is not directly specified by the code that you may do so.

I can easily go 15" for depth as I specify the depth. Depends on if I need to rely on the verticals to develop my anchors or if they work in plain concrete or not, may need to cheat the anchors up a bit to ensure that the lap length below is sufficient.

 

[icebloom]

I've encountered this same problem when designing large anchorages and have had to dig through the codes and literature before. Once you eliminate tension cone failure and edge breakout either via large edge distances or reinforcing, pryout becomes the governing failure mode (especially when you also consider the combined shear and tension in concrete interaction equations).

Codified guidance world wide on anchoring to concrete is based on the CCD method which has mostly been developed in Europe. Pryout failure is considered as a failure mode in all these codes (I'm not too familiar with CSA however based on your comments sounds like it also considers it). I'd be careful with ignoring pryout as per the Widianto paper - some studies can be found where longer headed shear studs have been found not to exhibit pryout failure past certain h/d ratios, however I note that these studies have been around for almost as long as the CCD method yet omission of pryout has not been adopted in any code. The experiments for these are also typically small groups of studs, and may not reflect a global pryout failure mechanism for larger group anchorages, which is what the standards check. For deeply embedded individual anchors things point to pryout not being a concern, however I've never found any such guidance for anchor groups. You will also not find the 12d rule in any code either as far as I know. The Widianto paper does have some useful and practical examples for detailing and helping see a strut-tie load path.

Some other threads on here have postulated that since pryout capacity is based on tensile cone capacity, reinforcing against cone failure will also reinforce against pryout failure. Intuitively and also if you construct a strut and tie load path, reinforcement of the kind you'd detail to resist tension cone failure would surely do something for pryout - unfortunately as to what that something is, there seems to be no guidance available.

Another fact supporting this is that pryout failure is generated by catenary tension in the anchors due to deformation and rotation of a baseplate under shear force - so the forces causing pryout are a sort of indirect tension. The experimental tests and non linear FEA that the code equations are based on found that tensile force in anchors prior to pryout failure is about 40% of the applied shear load, and the fracture surface is about 60-70% that of the tension cone of an anchor under concentric tension (see book Anchorage in Concrete Construction by Silva et al - great resource). Intuitively based on this, what seems a logical conclusion would be that if you design for this tension force then you can preclude pryout. In the past I've used hooked rebars welded to the back of cast in plates which then lap to other reinforcement (eg closed ties) in the concrete and done a conservative strut and tie design, based on the premise that if you design the rebars to have enough strength to take this tension the pryout will not occur. Note you need reinforcement in the concrete to then develop a feasible strut and tie load path to spread this tension out to your supports (analogy is hanger reinforcement when you are anchoring to the tension side of a beam - the hanger reinforcement takes the load back up to the compressive side). Without this extra reinforcement you've just got a bar in tension which will break out of the concrete. fib bulletin 58 (can be easily found online with a google) is another authoritative source including commentary where you can find more info on pryout.

There is no codified method on checking the struts for these kinds of anchor reinforcement. Recently investigations have begun and can be found in the literature, yet none have made their way to codes yet (and unfortunately as much of the research comes out of Stuttgart, some useful paper are in German). If you try using hooked bars, you can find info about the stresses inside the hook from various resouorces. Theres also a paper "Curved Bar Nodes" by Klein which is freely available from the ACI on this if you really want to get into it.

An easier way to avoid pryout is to simply increase the size and embedment of your anchorage to generate a larger tension cone till you can get it to work based on pryout alone, using say headed cast in anchors. Have you got a sketch of the kind of anchorage you are considering?

More recently there was a paper by Eligehausen et al studying supplementary reinforcement design for anchorages. In the experiments, to preclude pryout failure they had an assembly which prevented rotation of the baseplate and used extra fixings to resolve the tensile force. I'll dig it out tomorrow at work, might be useful for you. Another guide on detailing would be ACI section R25.4.4.2 - its for headed reinforcing bars anchoring into walls or columns however it shows a good strut and tie analogy and how the local phenomena is reinforced against to then resolve via strut and tie into global bending/shear of the concrete element.

 

[EngDM]

Have you got a sketch of the kind of anchorage you are considering?

Was planning to use 1 1/4"Ø (diameter is governed by PEMB supplier) F1554 Grade 105 anchors with plate washers held in place by threaded nuts at the bottom.

The way CSA calculates pryout has a factor of ANC/ANCo where ANC is the tensile breakout area based on your given edge distances and ANCo is the tensile breakout area assuming all edge distances are not governing, and instead is goverened by 1.5hef of the anchor. Because one of my edge distances cannot get bigger due to building size restrictions the ANCo term far exceeds the ANC value, which tanks the capacity of my anchors. If I go deeper with the anchors, it is making the equation worse.

There are PEMB's out there I'm sure that have had similar loads but did not really worry about pryout as a failure mode, without having insanely sized pileasters.

Essentially, the deeper I make my anchor the larger I need my edge distance to be, and having a 6'x6' pileaster and still fail the pryout check is just insane; the contractor is going to lose his mind seeing that in comparison to other builds.

 

[TLHS]

I don't have it in front of me, but one of the recent ACI 318s (maybe 2019?) made changes to pryout. I think over a certain length the strength increases. I would look at that. I have generally used ACI as a secondary reference for the CSA Annex D stuff because it's a situation where they're effectively the same formulation except for some of the seismic thought-process and ACI had a more complete treatment and implemented things like clear direction on how to actually do reinforcement when CSA didn't.

While I think it's reasonable to supplement with ACI for this, you are technically out on your own with it if there's ever a problem.

That being said, I don't personally see how you can have a catastrophic pryout failure on a long anchor that's also been reinforced against tension failures with some reasonable thought behind the sizing of it. When that block separates you're, at worst, going to have a shear friction style situation with bar crossing the plane. If you have that, plus some stirrups/ties holding the section together then I can't see where the failure mode is other than potentially a local failure around the tops of the bolts.

 

[TLHS]

I think I might have imagined that ACI revision. I'm not seeing it. So ignore that part.

 

[icebloom]

The paper I was referring to is "A new model for concrete edge failure of multiple row anchorages with supplementary reinforcement—Reinforcement failure" by Sharma et al (it is a fib Technical Paper,

https://doi.org/10.1002/suco.201700002).

They use an uplift restraint to preclude failure due to rotation of the baseplate inducing additional tension.

If my understanding of your connection is correct you are trying to use a a group of 10 anchors to take 700kN of shear? Not sure if these are individual anchors or connected via a common endplate - if common endplate then you'd use the group Anc/Anco ratio which should always be > 1 (i.e. Anc > Anco). As you mentioned, to the code the only way to increase the pryout is by increasing the area of your tension cone either via embedment or playing around with the anchor spacing if it is a group. Unfortunately there is nothing in the codes or literature explicitly detailing how you can reinforce against pryout - I think that reinforcing in a manner similar to reinforcing against tension breakout would be the way to go, combined with deep embedment. This is based on the literature indicating that pryout failure resembles tension cone breakout, and also other studies that indicate pryout becomes less of an issue at deeper embedment (although as I noted before this is not codified).

You could also look at studies on pryout by the PCI in Link.

Your best bet might be to have hooked rebar welded to the back of your plate and then reinforce the hell out of the concrete.

 

[EngDM]

My baseplate is 12" x 27.5" sitting on a pileaster with an ideal max embedment of 22". From my understanding on the figures you put above and prior responses, if I took 40% of the anchor shear and applied it as a tensile force it would be roughly equivalent to the pryout force. I'd then have to follow the breakout cone up on a 35° angle and develop the bar above and below this intersection with the reinforcing I provide.

Your best bet might be to have hooked rebar welded to the back of your plate and then reinforce the hell out of the concrete.

I wish I could do this, but the baseplate comes pre-welded to the large pre-engineered frame and they just stand it up.

 

[icebloom]

I've attached the derivation of the pryout capacity equation where it talks about how the typical pryout capacity is obtained including the 40% tension and 60-70% cone size etc, which is what my answer above was based on. Based on a semi tension cone failure surface (note this might also indicate that reo in a certain zone may be less effective) being what forms in pryout failure, I'd say that reinforcement similar to supplementary reinforcement that you'd detail for tension cone failure would be the way to go, although as mentioned theres no code or guide anywhere that I know of that tells you how to do this. Probably worth having reinforcement in all directions in your concrete to help with confinement.