anchor rod in tension embedded in concrete

[bjcure]

For a threaded anchor rod in tension, embedded in concrete, ACI 318 Appendix D gives methods to check the steel anchor failure as well as various concrete failures (breakout, etc.).

But is it necessary to check the washer bearing on concrete? Are standard washers sized so that they always work for bearing (if so, for what concrete strength)?

In this case, I am wondering about a 1 1/4" diameter F1554 Gr.55 anchor with double nuts, and a standard washer conforming to ASTM F436 (washer diameter 2 3/4")

 

[ToadJones]

I do not think you need to consider the concrete bearing stress on the area of the washer. If you did, I think you'd find very high stresses in that region.
I believe this is accounted for in the break-out cone.

 

[ToadJones]

I take that back...I'm looking through a similar calc now and I see where I have checked bearing area's as given in AISC Design Guide One (boooo).
A 1.25" rod with nut only has a bearing area of 2.24 sq. in. and a concrete pull out strength of 50.2 kips for 4 ksi conc.; from AISC DG1 table 3.2

 

[Ron]

You have over 4-1/2 square inches of contact area for the washer. If you have 3000 psi concrete, that will take over 14,000 lbs of contact load. If the compression in your bolts is greater than that, you might need to consider more bolts or a larger washer. Usually not an issue.

 

[Lion06]

If you're talking about a washer embedded in the concrete, you definitely have to check it. This is the pull-out failure per App. D. That being said, a standard washer is probably worthless in this case, the bearing stresses will be so high that it will probably fail in bending. If you need more contact area then provided by the nut itself, put a plate washer on it and size the thickness for bending. I've rarely encountered a pull-out controlled anchor.

 

[WillisV]

Pullout tends to be the controlling case for very deeply embedded high strength (GR105) anchors in which you are trying to get the failure to occur in the steel for ductility concerns. You will find you need plate washers to force the failure into the steel anchor once the embedment is deep enough to prevent breakout from controlling.

 

[ToadJones]

I should make my earlier poor posts a little more clear.
Section D5.3 (I believe) has an equation for pullout. Part of that equation is the bearing area of the anchor.

I thought you were going to check the bearing ONLY based on the area of the washer, which would cause the stress to be very high most likely.

 

[bjcure]

Thanks for the input everyone.

In response to Ron's comment, I'm checking someone's existing design and the load is well above 14 kips (about 26k). The only way I know to check bearing is to use the formula in ACI 318 10.14.1 and I'm coming up with about 15k ultimate capacity.

I'm not talking about pullout, I'm talking about localized crushing of the concrete at the plate washer. Using the 10.14.1 formula, it doesn't work, but is this the appropriate way to do it?

 

[ToadJones]

That was what I thought you were doing and I do not think that is the correct approach, but I could be wrong.

Personally, I have had the same question & I have always checked it as straight bearing pressure the way you are saying and used plate washers to accommodate the bearing stress for anchors with large tensile forces.

 

[Lion06]

"pull out" is a localized bearing check of the concrete. Unless I'm missing something this is pretty straightforward. You get a much higher
bearing capacity because it is local and extremely confined. Look in App. D. It seems pretty clear cut.

 

[JLNJ]

I'm trying to picture how concrete completely confined would crush. You'd need to make the washer thicker than the nut to get the concrete to crush before the washer gave out. It may happen, but I have a hard time picturing it in my mind.

 

[Lion06]

The thickness needs to be designed for bending......... As noted above.

 

[ToadJones]

I can never see the need to use a standard washer as they aren't much bigger than the nut.
A 1.25" nut has a width-across-flats of 2" I think. The width at the points is about 2.3". So, your bending check would be on a 3/8" cantilever at the most.
I also don't see the need for double embedded nuts. I'm pretty sure one nut will develop the strength of the rod.

 

[ToadJones]

I'll take a guess here (no books around) ....
I think the pull out check is:

Np = (0.7)x(8fc')x(Area of Bearing)x 1.4

The 1.4 factor only applies if the concrete is found to be uncracked by yada yada....I never used the 1.4 factor unless I had to. I'd just check first with out the factor.

SEIT- The bearing stress is assumed to act over an inclined plane in this case, no?

Hopefully someone can confirm what I have said here.

 

[Lion06]

Toad-

You have it right. That value (while I agree it gives a very high bearing stress) is the stress at which the concrete will begin to crush under the bearing area of the nut (or plate). I'm not sure what you mean by an inclined plane.

From what I can gather, the reason the bearing stress is so high is because of confinement and the failure mechanism is actually crushing of the concrete. I rarely see cylinder breaks where the concrete crushes. If there were a confining force to the cylinders (in cylinder tests) such that the concrete was forced to crush instead failing in whatever manner it does, I guess the f'c values would be through the roof.

 

[ToadJones]

Good point with the cylinder breaks...we break all of our own and I know I have never seen one crush.
As far as the inclined plane, I must be mixing things up here (easy to do with App. D, no?). I think I was falsely assuming that the bearing stress was spread out over the area of a "pull out cone".

 

[PMR06]

I have always wondered what would happen if you poured a concrete cylinder inside say an XXS steel pipe, then loaded it into the press and tried to crush it.

ToadJones, you have access to such equipment? Maybe you could run a little experiment and post some photos?

 

[Lion06]

I did something similar with my son for science last year. We put loose sand inside a toilet paper tube and loaded it with everything we could find. We got it up to over 700 pounds without failing it. We had to stop because the stack of stuff we had on top was getting too high and too dangerous. That was loose sand!

 

[PMR06]

SEIT, that's awesome. Assuming 1.5" diameter, you got almost 400 psi out of loose sand and cardboard. With concrete and steel pipe, I imagine one would reach the limit of the test press.

Now extrapolate that to the anchor bolt bearing. Assuming there were enough confinement, I imagine the concrete bearing would never physically govern. My guess would be that one of the other failure modes would be reached first.

Anyone know of tests set up to force failure in concrete bearing?

 

[bjcure]

What everyone is saying makes sense. I don't see how the concrete could "crush" when it is so confined. That's probably why Appendix D doesn't mention "bearing", but instead gives you all the breakout and blowout failure modes.

The equation in 10.14.1 is interesting. I can imagine how this equation applies to a beam sitting on a concrete column, for example. The failure mode would be some kind of "blowout" where a large chunk of the top of the column spalls off on an inclined plane. If the area of the support is greater than the area of bearing, it lets you multiply the capacity by as much as a factor of 2. I wonder why they limited it to a factor of 2.

I also just noticed that the commentary says "this section deals with bearing strength of concrete supports". Since it is for supports, it makes sense that this equation would not apply for something like a steel column baseplate on a concrete footing. The concrete failure mode in this scenario would always be one or two-way (punching) shear, or flexure, not "crushing". I don't even think crushing is possible, if the concrete is confined. So the bearing equation doesn't apply to a plate washer embedded in a footing, either.

This is one of those situations where I realize maybe I've been overconservative for the last 10 or so years. It makes me wonder, why use plate washers at all? Like Toad said, the standard washer isn't much bigger than the nut. Why not use the nut (and standard washer, but the nut primarily) to "grab" the concrete when the anchor is in tension and start the failure cone that way? The pressure on the nut would be very high, but nuts are thick, and either way, plate washer or not, the nut has to transfer the force through the threads onto the anchor.