Unconventional Anchor Bolts 1

[DaveVikingPE]

I'm reviewing an anchored retaining wall design with a wale behind the sheet pile. The wale consists of a 3' wide x 18" deep r.c. beam though which tie rods extend back into a continuous deadman anchor. The tie rods are connected to a channel that's embedded in the wale beam. The channel is connected to the sheet pile via "L" shaped anchor bolts, each 30" long and 3/4" in diameter, spaced accordingly between each tie rod (let's say the tie rods are at 20' o.c., their are four anchor bolts between ties rods, then).

Basically, the only connection between the wale and the sheet pile, other than the concrete/steel sheet pile bond (no studs or anything else assist), is the anchor bolts. The designer's calculations do not at all account for tensile stress issues in the bolts themselves (easy issue for me to deal with), but he does appear to address pull-out strength of the anchor bolts or perhaps failure of the concrete in tension. The formula used doesn't look anything like what I'm used to seeing (ACI 318 11.3?) nor does it really make sense. His formula is Tension per bolt in pounds/(embedment length X 3.1). The result is in psi which is compared to another number in psi. I can't really make heads or tails of his analysis.

Any thoughts?

 

[JAE]

Older ACI codes used to predict development length based on an allowable bond stress on the contact area between the rebar and the concrete. Perhaps the engineer is using an old version of this with regards to anchor bolt development. That would explain his use of devlopment length. The 3.1 number is strange....its close to pi and also is close to the circumference of a 1" diameter bolt, not a 3/4" diameter.

The psi allowable that he refers to may be an old allowable bond stress. You might seek out old ACI or AISC manuals.

 

[ishvaaag]

The only thing that comes to my mind is that practical embedment lengths for design are 2.5 or 3 times those of rupture. Then the called "embedment length" should be limit or the rupture, and the check you witness would be one of average required adherence against the limit adherence at service level. Or if you want, it could be a check of the bond.

 

ACI 349 appendix B has a good description on anchor bolt embedments. The 3.1 is a mystery. It usually is a FS of 4 times a strength reduction factor of 0.85 or 3.4. You didn't mention it but I would be concerned about prying on the bolts if the beam settles in the backfill behind the sheet piles.

 

[DaveVikingPE]

Thanks, alboy! I am really concerned about this design since a lot of it doesn't make sense, nonetheless...

I ordered and received ACI 349's AB-81 "Guide to the Design of Anchor Bolts and Other Steel Embedments." I ordered it direct from ACI ($21 for 14 pages, sheesh!) and it looks pretty good. This little guide further reinforces (pun intended) that the 3.1 factor is a PIOOMA factor, though it looks like I need ACI 318-77§12.2 to really get to the bottom of things. Since I have ACI 318-99 at my fingertips...

12.5.2 indicates that the basic development length for a hooked bar is to be multiplied by a factor; the factor or constant carries units of psi. Hmmm...

OK, if the mysterious 3.1 is a constant with units of psi, the result of the T/(3.1 x Ld) will not have units of psi.

The commentary to §12.2(2) mentions pullout failure as controlling failure mode...

eh, you know what? The ACI code has a lot of factors but nothing works out to 3.1 with the right units. The best I come up with is that 3.1 = 2.325 X the diameter of the bolts. The designer hasn't answered my inquiries and the more I look at this the more I'm convinced that something is missing.

Further assistance/comment is grately appreciated, though!

BTW, here's a GREAT web site on anchor bolts:

http://www.decoanchors.com

 

[JAE]

Hey Dave - what is "PIOOMA" ???

 

[DaveVikingPE]

It means...

Pulled
It
Out
Of
My
A...............................

Feel free to fill in the blank as you see fit...

 

[DaveVikingPE]

The plot thickens...

The designer has responded twofold: a) he increased the anchor bolt diameter to 1-1/8", satisfying tension requirements. But...

b) he's checking the bond strength using what must be JAE's &quot;old ACI code.&quot; To wit, Tension/(embedment length X circumference of bolt) =< 4.8 X square root(concrete strength = 4,000 psi in this case)/diameter of bolt. I don't have an older ACI code handy, wait, maybe I do... No, the oldest we have lying around is from 1989 and it doesn't seem to do the trick. I'm on to something, though, as there appear to be formulae looking similar to what's been presented, but I can't resolve dividing (4.8 X fc^1/2) by the diameter of the bolt. Man, I wish he'd included a citation for this equation.

Any assistance is tremendously appreciated.

 

[JAE]

I found an old (1963) ACI Publication (SP-3) which indicates that the allowable bond stress for tension bars is as follows:

Top bars - Plain, max. bond stress = 1.7 * sqrt(f'c) / d
but not greater than 160 psi

Top bars - deformed A305, max. bond stress = 3.4 * sqrt(f'c) / d but not greater than 350 psi (A305 is indicated as being applicable to #3 thru #11 bars)

Other bars - Plain, max. bond stress = 2.4 * sqrt(f'c)/d but not greater than 160 psi

Other bars - deformed A305, max. bond stress = 4.8 * sqrt(f'c)/d but not greater than 500 psi

The last one above appears to be what he used.

 

[DaveVikingPE]

Bravo, JAE!

Well, this 38-year old engineer (born in 1963!) salutes you and your library! I spent not too few hours coming up with variations, some of which got close to 4.8. I tried rationalizing failure cone sizes, looking at nut/bolt geometry - a lot of stuff - and finally lost my mind. Of the good, I discovered a coworker's stash of AISC design guides (LRFD guide to column base plates). Of the bad, I left the office at 8 pm! One thing is for sure: I'll keep my eyes peeled for SP-3. Not only that, but you're tops on my list for the week! Wait, that's two things.

Anyway, again, my sincere thanks. Now, in my journey today I came across a lot of web scuttlebutt from SEAOC regarding the &quot;unrealistic&quot; concept of bond stress - as in &quot;who still uses that&quot;? Well, the designer who used the 1963 ACI equation is still using it (I put him in his late 70's). Personally, I have a lot of respect for engineers who are practicing on through their golden years. However, he loses a little of it for not citing the source of the equation. Especially since it's what he uses to justify his design.

Once again, thanks so much and good night!

 

[JAE]

Dave: Just keep in mind (I do) that someday there will be a whole flock of young engineers coming out of school with loads of &quot;new&quot; concepts and methodolgies and we'll be the &quot;old fogies&quot; sticking to our tried and true methods. We just won't be able to handle their new holographic/dynamic plots or their virtual reality stresses or their George Jetson liquid steel weighing 25 lbs/cy!!

We'll still be using the old archaic PC's.

 

[ronster]

Anchors bolts are typically plain bars not deformed bars and therefor the equation with a factor of 2.4 and a maximum of 160 psi should be used. If you are reviewing a new design, I would recommend to the designer to use a threaded rod with a plate and nut at the end to take the tension neglecting the bond stress of the bar. I doubt the values of the 1963 ACI code are referenced by your building code and would not stand up if a problem did occur and you ended up in court.

 

[DaveVikingPE]

Ronster,

I absolutely agree concerning the old code issue. The design itself, after my &quot;prodding,&quot; calls for custom #9 bars with threaded ends used as anchor bolts. I am of the opinion that the engineer of record (not me, this is a contracted A/E job) is of the mind that he can direct &quot;field changes&quot; on the fly/as necessary/etc. - something that bugs the s--t out of me, but he did pay attention to the issue of acceptability of the bolts in tension I brought up and increased their diameter, so there is hope. The anchor bolts are connected to the steel sheeting by bolts and a nut, though minimal design attention was given to choosing the hardware - something else that bothers me. I would not have designed the waler beam the way this guy did, but who listens to me?