Anchor Bolts in Shear

[CN-EIT]

Context: Baseplate at a braced frame transferring shear to anchor rods; that is, no frictional shear transfer. Anchor rod layout is square consisting of 3/4" diameter anchor rods in 1 5/16" holes w/ 2" plate washer (not welded to baseplate)

Discussion:
1) AISC DG-1 recommends that only two anchors be used to resist the shear; subsequently, concrete failure mechanisms are calculated assuming only 2 anchors are utilized to resist shear.
2) DG-1 goes on to say that if 4 anchors are to be utilized, it is recommended that a plate washer with internal diameter 1/16" greater than the anchor bolt diameter be used and welded to the baseplate.
2a) If one were to weld a washer to the baseplate, my presumption is that it limits your slip before engaging four anchors to 1/16" of an inch. It does not expound upon sizing the weld; my presumption is the weld would need to be designed such that the weld strength is sufficient to carry the full shear from baseplate to anchor rod.
3) To my knowledge, there is no code provision saying that unless you weld a plate washer to the baseplate, you can only use two anchor bolts in resisting shear.

Example for Reference: 10"x10"xthickness baseplate with 4 3/4" diameter anchor rods in 1 5/16" holes. Assuming no welded washers, if all anchor bolts were perfectly placed, the slip before engaging all four anchors would be 5/16". In the worst case scenario, the slip would have to be 9/16" but the baseplate where it bears against the anchor rods would need to yield or tear.

Question 1: Can a valid argument be made to distribute shear to 4 anchor rods without welding plate washers?
Question 1a: Say the baseplate is dropped below finished floor elevation and subsequently encased in concrete. The concrete is not monolithic to the pier or the adjacent slab on grade (column isolation joint), however, can an argument be made that this provides a degree of confinement sufficient to distribute the shear force to all anchors?

 

[canwesteng]

Do you think the bars will yield 1/4" in shear? And there are code provisions in ACI 314 relating to load distribution on anchors for shear breakout, you could review those.

 

[CN-EIT]

That is a good point about bar bending. I suppose it is quite possible that the anchor will yield prior to tearing or yielding of the baseplate...sounds like a pretty good limit state to me! And forgive my lack of due diligence, it seems like ACI 318-14 is pretty thorough with the treatment of the distribution of forces to anchor rods in shear. They do a pretty good job of dictating the three cases of load distribution as it seems the distribution is a function of anchor spacing vs edge distance of nearest edge anchor (17.5.2.1). I may need to find the studies for this since I'm not sure that solves the problem of how all four anchors are resisting at once if there is play in the holes involved. Must be from testing I guess

 

[dhengr]

CN-EIT:
The 3/4” A.B. in a 1&5/16” hole is to allow for some A.B. placement tolerance vs. the holes in the base pl. which should be to a tighter spacing tolerance. The question is, can you tolerate 1/4” to 9/16” movement in the connection before the A. Bs. all come into play. And, what does the lateral load do to only one or two of the A.Bs. since that’s the way it might be reacted before all of the A.Bs. are brought into play. With the worst case movement, the first loaded A.B. is probably destroyed before the others come into play. That’s a pretty poor way to transfer a significant shear force or lateral load. The washer pls don’t do much other than cover the bigger holes, unless they are welded. But, then the lateral loading is applied to the A.B. with an inch or so more canti. length (the base pl. thk.) which makes the A.B. more difficult to make work.

 

[JoshPlumSE]

dhengr (IMO) has it basically right.

The ACI provisions start with single rod capacity values and expand this into group values and such. Good procedure, but maybe a bit complex compared to how we used to do it.

That AISC design guide is specific to a particular type of construction (steel base plates). They are merely acknowledging an inherent issue with the oversizing of holes in a base plate.... that this can lead to uneven distribution of shear load. The welding of washer plates seems to be a reasonable solution to this. Not sure that it's fully demonstrated in testing. But, it is a reasonable way to approach this from a design perspective.

Does welding the washer plate to the base plate necessitate checking other failure modes (like bending of anchor rods? Not something that particular concerns me personally. But, that's not unreasonable to check... However, I would suggest that this is probably okay. Deflection / bending of the anchor rods will allow better re-distribution amongst the anchors.... which is what we want.

 

[r6155]

I suggest not to weld on the anchor system (bolt, nut and washer). Welders in the field are often unqualified and tack welds are of poor quality. Some bolts, nuts and washers are made of special materials. Galvanized parts are welded without precautions.

Regards

 

[phamENG]

r6155 - in these applications, the washer in question is made from A36 plate steel, 1/4" is usually the minimum thickness. These aren't tack welds and there's no welding of the anchor itself. Just the plate washer to the base plate with a fillet weld. As for qualifications, most projects I've been involved with that have a requirement like this are big enough to have a robust submittal review phase during which the erector is required to submit the welding certs of everyone who will be on site. Granted, they can slip somebody in if they wanted to, but most of the erectors I've worked with would rather not have that mark on their record. Good point on the galvanizing.

I agree with dhengr and Josh. While there's no code provision that dictates you have to do it that way, the way it works in the real world dictates you do. And in terms of 'codes' and liability, if something went wrong and it came down to this, you'd probably fall outside of the standard of care and be responsible for the failure. If you don't want to use a shear key, embed the baseplate in the concrete and grout it in, or use friction (I'm guessing there's an associated uplift?), then shear through anchors is all you have. Shear through the anchors means assume only two or weld on washers.

 

[RWW0002]

I know the OP specifically mentioned the use of the super-oversize holes from T 14-2, but I wonder how the use of 1 1/16"Ø holes ("old" anchor rod hole sizes) with USS Standard washers changes the discussion. See Footnotes in T 14-2. Previous versions of DG1 (prior to the super-oversize hole recommendations) had no mention of the only use 2 anchor requirement that I am aware of.

In my market, I find detailers still like to use the 5/16" oversize hole sizes for anchors smaller than 1"Ø. I'm sure the concrete guys don't like it much, but it seems to work out ok.

 

[dik]

Unless heavy loaded in shear (not defined) I usually design 2 bolts (out of 4) for shear and check combined stresses. I use welded plate washers because anchor rods are not stressed as an A325 and holes are oversized. If HDG touch up the connection with zinc rich primer. I also only use weldable anchor rods, except for special occasions where I have to use very high strength anchor rods.

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

-Dik

 

[CN-EIT]

JoshPlumSE:
I agree that the ACI provisions are provide some rational basis for distributing anchor forces. However, that distribution being contingent on the equality of anchor spacing and edge distance does not seem to me to address the inherent problem of distribution as addressed in DG 1. DG 1 does a pretty good job of addressing the bending moment in the anchor rods, and they do comment that the ACI reduction factor when using a grout pad (which is pretty typical around here unless its a PEMB) of 0.8 and so I am thinking the ACI reduction probably is sufficient to account for the bending. Even if they provide zero justification for it.

phamENG:
The standard of care is something I was hoping would be brought up; mostly by me reading what other engineers are doing. There seems to be some argumentative cover provided in the ACI code provisions if something did fail. But if the general consensus among practicing engineers is that, barring the use of weld washers, shear force should be distributed to half the anchors, and there is a structural failure, you're correct in that the design fails in terms of standard practice. And in such a case as this, where a general code provision does not jive with recommendations tailored to the connection, it is my opinion that the standard of care carries more weight.

Dik:
I may be a bit confused, are you saying that if its low shear that you design for shear in two rods and if high shear (relatively speaking) you are welding washers to baseplates? Or for high shear are you welding anchor bolts? If you are specifying weldable anchor bolts in all cases, I would infer the latter. Also, what is the benefit in specifying weldable anchor rods in all cases? I believe the weldable anchors are F1554-55 which costs more than the F1554-36.

 

[CN-EIT]

As a follow up question, what are people using to design anchorage? It seems that among the various programs out there (RAM, Tekla, Simpson AD, Idea Statica, etc), there is serious variability in the distribution of shear forces. I myself have seen this when designing baseplates in tekla (distribution to two anchors only) vs RAM connections (limit states checked with distribution of shear following case 1 and case 2 of ACI) vs. idea statica (distribution to four anchors only based on stiffness)

https://www.structuremag.org/?p=438

 

[RWW0002]

I think you have the weldability mixed up. F1554-55 designates Grade 55, which may not be weldable but has a higher yield strength (55 ksi). If you want Grade 55 and weldable it needs to meet supplementary requirement S1 for carbon content.

F1554-36 on the other hand is generally considered weldable but has a lower yield strength (36 ksi).

Specifying weldable anchor rods can sometimes help if an anchor is broken or has projection issues.

 

[dik]

Sorry, I wasn't clear... if low shear, I use anchor rods, else use a lug welded to the U/S of base plate.

If I use anchor rods for shear and there are 4 of them (common), I only use 2 for shear resistance. I weld the plate washers to the base plate, not to the anchor rods. Unless it's a repair, I don't weld to anchor rods unless I need to. Because of past issues, I always like to have anchor rods weldable... too many misplaced ones. With the 1554-55 make sure you spec S1 else it may not be weldable... the cost difference between 36 and 55-S1 is small (in this neighbourhood), and the strength gain is significant (almost for free).

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

-Dik

 

[dik]

In the last 50 years, the number of misplaced/bent anchor rods I've encountered can be counted on the fingers of two hands ... using binary, of course.

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

-Dik