Retrofit Holdown - Can't Use Epoxy 1

[eric294yz]

In reference to the closed thread (thread507-420144) I would like to continue the discussion. I too do many retrofit projects that require installing new holdowns on existing concrete foundations. Typical construction here in my area of California is raised floor with 8" wide concrete stemwalls. Using ACI 318 Chapter 17 requirements, you can't get any better than 800# allowable uplift with 2-3/4" edge distance and infinite end distances in an 8" wide stemwall considering overstrength factors since we do not have ductile steel elements with the limited edge/end distances. I used to work for a very highly respected engineering firm that continues to specify epoxy holdowns, turning a blind eye to the breakout requirements of ACI 318. I now work for myself and am not willing to accept the liability of turning a blind eye to this problem and am searching for an alternate solution. It is likely going to get flack from contractors, but so be it as I choose to design my buildings by the letter of the building code and not accept any undue liability.

That being said, here is the problem I'm still trying to overcome. How can I show/calculate the ductility requirements of ACI 318 Section 17.2.3.4.3(c). The solution I'm thinking of using is a steel plate mounted to the inside face of the stem wall with Simpson Titen HD screws into face of wall to develop the needed tension loads in concrete breakout. I have run various scenarios through Simpson Strong-Tie's anchor software and can get allowable loads that handle up to HDU8 holdowns in the face of an 8" stemwall, NOT considering overstrength. Remaining problem is designing the strap from the steel plate to the post in the stud wall to meet 17.2.3.4.3(c). The commentary for this section states, "Similarly, steel design manuals require structural steel connections that are designated nonyielding and part of the seismic load path to have design strengths that exceed a multiple of the nominal strength." So my thought is if I design the strap (considering the expected strength increase versus allowable) to only "be able to transfer" a force that is below the ultimate capacity of the concrete anchors, then I do not need to apply the overstrength (Omega) factor of 2.5 correct? Does anyone know where is the AISC Seismic Design Manual it covers nonyielding connections and this "multiple"? I could not readily locate it.

The ACI Commentary also states, "Option (c) may apply to a variety of special cases, such as the design of sill bolts where the crushing of wood limits the force that can be transferred to the bolt..." This seems counter intuitive to making a building safer, but what if the base of the post the holdown is attached to is notched down to a dimension such that wood fiber crushing would control the design strength of the connection over concrete breakout, therefore eliminating the requirement to apply the overstrength factor (omega) 2.5 to the allowable design strength? Somehow making a connection stronger by making it weaker? Haha, you have to love ACI Chapter 17 (formerly ACI 318 Appendix D)! Thoughts?

 

[radiocontrolhead]

I typically ignore the overstrength factor. I believe the epoxy itself is relatively ductile and has been published in the ICC reports (i will need to look at this again). I've noticed that simpson 3G epoxy provides substantially higher forces than the Set XP and will go with that if my forces are too high.

I typically won't design epoxied hold down for anything greater than a HD4 unless I know i'm working with well known existing concrete conditions. I'll add more wall to bring my HD forces down to those levels otherwise.

I am a fan of the drill through the footing with an underpinned pad for anything greater than HD4 ish forces. I don't care if the contractors don't like it. If I believe there is substantial room and space to get under the footing i'm specifying this detail. I'm not a fan of epoxy for HD devices to begin with so i'll defend this to the end.

Also. I always consider a min of 2 inch edge distance in my epoxy calcs.

 

[sandman21]

This link explains how to provide a ductile solution while anchoring close to an edge.

Simpson Strong-Tie® SET-3G™ Adhesive Offers a Ductile Solution for Post-Installed Anchorage near a Concrete Edge

I typically ignore the overstrength factor. I believe the epoxy itself is relatively ductile and has been published in the ICC reports (i will need to look at this again).

You believing something works is not sufficient for you to ignore how to actually achieve the behavior you desire.

 

[kwinner]

sandman21, I thought this was the answer I have been searching for for all these years, but unfortunately for typical residential stem walls (8" width), this method is not effective.

A couple things that I have been able to get to work is reducing my overturning forces using FTAO or resolving the forces in a stiff beam or rim that can distribute the overturning forces with a larger moment arm.

 

[radiocontrolhead]

How about forcing ductility by reducing the size of the anchor rod. I understand that the deflection calcs of the hold down device includes a component due to elongation of the rod but this can probably be determined and checked separately.

example:

provide a 5/8" diameter threaded rod and epoxy into concrete, install a reducing coupler with a smaller diameter anchorage rod (say 1/2" or 3/8" as req'd). The larger diameter anchor in the concrete allows for higher pullout strength and the smaller diameter of rod strength above the coupler, reduced.

 

[kwinner]

I have considered that option before, but never put it into practice. This might be the best solution though. I don't do as much residential work anymore so I don't battle this everyday like I used to.

 

[eric294yz]

This link explains how to provide a ductile solution while anchoring close to an edge. Simpson Strong-Tie® SET-3G™ Adhesive Offers a Ductile Solution for Post-Installed Anchorage near a Concrete Edge

I have seen the Simpson article regarding the new 3G epoxy. There are several problems with this example: 1) it assumes 3,000 psi concrete, as soon as you go down to 2,500 psi concrete the ductility ratio doesn't work and overstrength factor would apply again. 2) It assumes and 18" deep concrete member with infinite end distances...if you model a typical 12" wide or 18" wide by 18" deep concrete footing, typically seen in residential slab-on-grade construction, concrete breakout strength governs over the steel strength and connection is no longer ductile.

Using the Simpson software, for an 8" wide stemwall with 18" end distances and 2.75" edge distance I get allowable loads of LRFD = 2850 lbs *0.7 = 1,995 lbs (ASD). (See attached) The allowable ASD load, even ignoring overstrength factor, is not enough for a HDU2 holdown (3,045 lbs).

When I run the numbers for a 12" wide x 18" concrete beam with 2.75" edge distance and 18" end distance i get 4,275*0.7=2,990 lbs. Still not enough for full capacity of HDU2 holdown. The Simpson article is very misleading on their solution, considering the calculations they present do not reflect any real life situation in wood framed construction. I just don't see any possible solution with epoxy in tension with minimal edge distances.

 

[kwinner]

Not until we get some research and testing done (which I would think would be valuable for Simpson to pursue)