rational design of embedments and anchors in concrete 7

[IJR]

What is the rational approach to be used in designing embedded steel components, say for example a bracket attached to a reinforced concrete column, or a non standard anchor bolt made of mild steel rebar, attached to a column subjected to wind uplift.

Thanks in advance

 

[Taro]

For cast-in-place anchors, the state of the art report on concrete anchorage is "Strength Design of Anchorage to Concrete" by Ronald Cook, published by the Portland Cement Association. Other references are the Uniform Building Code section 1923, the PCI Design Handbook, and the AISC LRFD Vol. II Connection Manual.

 

[Ron]

The underlying premise of anchorage design in concrete is tensile/shear-cone failure of the concrete. This is affected by concrete strength, aggregate type/size, depth of embedment, spacing of embedded anchors, mobilization of tension in the anchor material, and edge affects.

 

[IJR]

Thanks Ron, Thanks Taro

Back to Ron, do you have any reference in mind?. Some engineers tend to rely on ACI-based anchorage lengths in designing embedded rods designed for rods, something like 50times the rod diameters. I have also heard of this cone idea but dont know how to implement it.

Thanks once more and again

 

[Ron]

For computing the failure area, you have to consider the method of installation. For epoxy bonded anchors, you have to check two parameters...the first is the bond strength of the epoxy to the concrete. This is done on the drilled diameter of the hole. Compute the circumference and multiply by the depth to get the shear-bond area. Use an allowable bond strength of about 10 percent of the specified minimum bond strength (example: ASTM C881 requires 1500psi min. bond strength).

For wedge anchors and the second check on the epoxy anchor, you compute the tensile-shear failure area based on the area of a truncated cone with height (depth) equal to the fastener length, and having a base diameter of about 50 percent of the fastener length. This is a conservative approach, as the failure cone is usually about the same diameter as the fastener length. I have no basis for this other than it is consistent with the angle used in shear testing for epoxy bond.

Hope this is helpful.

Ron

 

[breaks]

One great reference I know of is "Design of Headed Anchor Bolts" by John G. Shipp and Edward R. Haninger. It can be found in the AISC Engineering Journal, Second Quarter 1983. I know that was awhile ago, but the principles discussed are sound and still pertinent.

It offers great explanations and equations concerning the tensile/shear cone design method.

 

Steel anchor plates attached to a 550 degree steam line insert into large concrete anchor with extensive rebar. Any thoughts on how to prevent expansion of steel from causing longitudinal splits in concrete?

 

[dik]

Would use confinement rebar around the anchorages. In an extreme case, you might want to consider sleeving the anchor with a close fitting steel pipe (also with smaller anchors) in the vicinity of the anchorage plate. This will contain or prevent spalling of the edge of the concrete at the anchor and allow some of the heat in the anchor to dissipate into the concrete mass. I assume that the concrete mass is sufficient to act as a heatsink to dissipate the heat of the anchorages. Also, I assume that the pipe is insulated to some extent. Is it possible to provide insulation at the point of suspension? Teflon or some other high temperature insulator?

 

Thanks for your prompt answer dik
FYI these anchor plates are typically 1"x12"x1" and are pinned into the vault floor steel.
The concrete usually a minimum of 3'x3'x3' and extends down to catch the floor rebar. The anchor itself has horizontal cage type rebar every 4" in vertical height. The anchor plates have a minimum of 4" of concrete to the edge of the anchor, and a minimum of 8" clearance from the exterior of the pipe insulation, and can be up to 20 " from the pipe center line. We still see the splitting occur. Of course as corrosion attacks the first few inches of the embedded plates, the splits more severe.
Was wondering if we should consider a cork layer to prevent the expanding steel from stressing the concrete itself. This would only be used in the first 12" to allow the lower and cooler steel to bind to the concrete. Just thinking out loud along the lines of some type of expansion joint as the answer.

 

[dik]

I don't quite understand the 3'x3'x3' pedestal. It is dowelled to the concrete floor and has anchors embedded in it that the heated pipe connects to; is this correct?

The 1x12x1 anchor? Is this a BAR 1x1x12" long? Are these the anchors that hold the heated pipe? Are they hooked into the pedestal?

I didn't know that corrosion is a part of the problem. Could the cracking be from the expansion by products of corrosion?

Is it possible to have galvanized anchorages? There is no lateral oblation with cork and it would serve as a stress reliever and may force the expansion into the mass of the concrete where it may be resisted; the steel sleeves would serve the same purpose and these could be anchored into the concrete mass with better heat dissipation (The joint between the sleeve and the anchor may allow corrosive material to enter via capillary action). If there is corrosion, then the cork may also serve as a vehicle to cause additional corrosion. It is possible to use a foam sleeve and melt the foam after casting and filling the void with a high temperature epoxy material.

If the thermal expansion stresses are as large as they appear, you might be better served to create a crack path so that it is not 'random' and deal with it. Sort of a thermal expansion joint.

 

Apologies on the anchor plate typo, each anchor consists of a pair of 12"x36"x1" steel plates. They are welded along each side of the pipe at midline on the pipe with a 12" continuous weld. The plates extend down through and are pinned into the two sets of floor rebar (the floor is 12" thick with 2 galvanized rebar layers. The anchor block concrete is formed so the pour extends all the way to the gravel underbed beneath the vault floor.

Corrosion can be an issue, but I think that the majority of the problems revolve around the differential expansion of the steel plates vs. the concrete in the anchor.

Deeply appreciate your input on this subject. Thank you

 

[Ron]

Based on my understanding of your issue, I would suggest a change in anchor geometry. The penetrating flat plates and cast-in-place concrete do not mix well. The edges of the plate create re-entrant corners that promote cracks in the concrete and will allow addition moisture/water intrusion and exacerbate corrosion issues.

Further, the heat transferred by the plate into the concrete creates stress concentrations that are not easily attenuated by the concrete, particularly at those plate corners.

A round penetration cross section would better serve both issues. A "finned" pipe support (using radial fins, not longitudinal ones) would dissipate a lot of heat before allowing it to get to the concrete. If your connection between the pipe and the support were bolted or pinned, even more isolation would be gained. Further, vibrations from the pipe, if there are any, could be attenuated and you would not have such a rigid connection between the concrete and pipe that creates other loading problems.

If you did an infrared thermography scan of both conditions, you would see a significant positive difference in the "non-rigid" approach.

 

[StructuralDzine]

Headed anchors fail in a different way to deformed bar.

The load transfers at the bolt head and this causes cracks to start from the head and propogate up and out in a cone. At some intermediate stage there is likely to be a fully cracked cone around the head with an active cracking zone at the crack tips. When this cone reaches the surface it pulls out, but at no stage can you say there is a failure cone with a uniform stress over its surface.

There is an excellent CEB publication on the subject. I will try to find the title and post it next time I'm on.

I have used reinforcing bar threaded on one end as anchor bolts to overcome these problems. So long as the load is axial I can't see why normal development lengths as calculated for reinforcing bar can't be used.

 

[StructuralDzine]

Try this reference;

CEB Bulletin No. 233
Title:
Design of Fastenings in Concrete - Design Guide - Parts 1 to 3
No.:
233
Year:
1997
Pages:
83
ISBN:
0-7277-2558-0

 

[StructuralDzine]

Fastenings to Concrete and Masonry Structures
State of the Art Report - (printed revised hardbound edition of Bulletins 206 and 207, Telford, London, 1994; ISBN 0-7277-1937-8; 249 pages) Order directly from Thomas Telford Ltd.

http://www.t-telford.co.uk