#5 vertical rebar dowels post-installed with HIT HY 200 to develop horizontal shear

[Ingenuity]

I am reviewing a sketch with a call-out of #5 grade 60 dowels @ 6" c/c, with 5-5/8" embedment with HILTI HIT HY 200 resin.
4000 psi concrete, assumed UNcracked, non-seismic, no edge distance effects.

I wish to get a quick confirmation of ultimate capacity using the HILTI tables (withOUT direct calc of Appendix D of ACI-318).

So Table 19, #5 rebar, 4000 psi uncracked concrete, ΦN_n = 11,925 lb.f:

Spacing factor of 0.68 for 6" spacing of dowels:

So effective ultimate capacity is 0.68*11,925= 8,109 lb.f

Assuming this is correct, this detail is a long way short of developing F_y of the #5 (18.6 kips), as required for shear-friction. Agree?

Intuitively, 6" embedment for a #5 is what I may use for a non-structural curb, or a dwarf wall, not something 'structural'.

What am I missing here before I piss of the engineer on the other end?

 

[JAE]

I would agree with you. You cannot reduce the development if you are required to achieve fy (which shear friction requires).



Check out Eng-Tips Forum's Policies here:
faq731-376

 

[Ingenuity]

JAE: Thanks for the 'sanity' check/agreement.

I had proposed #4 rebar with twice the embedment depth!

 

[MacGruber22]

Of course it is a long way from using it for shear friction. You can't reasonably develop a #5 bar in existing concrete without it being cast or diamond cored and grouted. Do they provide loads or do they just have a note that says 'shear friction' bars or similar, and you (rightful) became suspicious? They may be misspeaking (or don't understand). There is no way they tried to calculate actual shear friction and then landed at that solution.

Hopefully, the "real" shear force can be transferred by providing shear dowel action (right-hand side of Hilti table) rather than by trying to justify the resistance on shear friction. Of course dowel action is much more sensitive to edge distance and spacing than shear friction. Though, I am not certain that it is always reasonable to substitute dowel action for shear friction bars in every type of concrete cold joint.

Why would you piss them off? Just ask for clarification or what the real loads are. If they get pissy from that, it sounds like more of a "them" problem rather than a "you" problem.

One in the hand is worth two in the bush.

 

[jayrod12]

Hilt has a Web program for rebar called profisrebar I believe it will design shear dowel embed mentioned depth for you. The few times I've used it the embed mentioned is far in excess of 6"

 

[KootK]

What am I missing here before I piss of the engineer on the other end

Something rather important although I doubt that it will change the outcome. In your classic shear friction scenario, the shear friction mechanism doesn't produce a net tension in the concrete on either side of the joint. As such, it's purely a bond/development problem and not an anchorage problem. There are no concrete breakout cones etc to be concerned about. And that means that the Hilti table which deal with those failure modes don't apply.

As far as the need to develop for Fy in shear friction applications goes, I agree with my colleagues above, the code is clear on that requirement. What they seem to be silent on, however, is why that's a requirement. And, frankly, I'm pretty skeptical. Other than "that's how it was tested", the only rational explanation for the requirement that I know of is one that I made up myself and am rather unsure of. I speculate that perhaps there is a random/stochastic aspect to which dowels get loaded when and, therefore, the dowels need to be developed for Fy in order to facilitate redistribution and preclude a brittle unzipping failure.

So yeah, I am willing to take some liberties with the Fy development thing in certain situations. And horizontal shear transfer in a two stage flexural element is one of them.

There are certain applications where I roll the other way and actually feel that an app D style, anchorage approach is sensible even though it's not code mandated. Primarily, that is in members where flexural demand would put the shear friction dowels in tension. Shear walls with "stitching" at the slab cold joints are a classic example.

In that situation, I worry that the shear friction dowels will rip out failure cones from the wall as the joint under consideration is opened up in flexure. Were that to happen, the shear friction load path would either be neutered or, at the very least, morph into something substantially messier. I envision the join't devolving to a sort macroscopic "toothiness" which may well have some shear resisting properties. Rationally, it is my feeling that there should be no such stitching and the flexural tension reinforcing in these walls should be the entirety of the shear friction reinforcing. ACI kind of suggests this as well in there soft recommendation that SF bars are most effective in flexural tension zones. Consistent with that philosophy, I also feel that there should be a max limit on SF that is based on the compression zone geometry rather than the overall cross section geometry.



I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[KootK]

Been digging into this a bit. In their 2016 post installed rebar design guide, Hilti covers dowels for shear friction based on a paper by Palieraki (attached). It's a hybrid dowel/shear friction method that seems to yield rather astonishing accuracy with respect to test results. The example that is cover in the design guide uses #5 bars embedded 7.5 in for a cyclic shear example. For monotonic shear, they don't specify a lower limit but imply values of 8db or less. There's no mention of developing for Fy etc. I'm not sure how "street legal" it all is for code compliant use in the the US but this clearly seems to be the way of the future for such applications.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[Ingenuity]

Sorry for the delayed reply. Got bogged-down in reading past E-T posts on shear-friction and post-installed dowels. There a quite a bunch of posts on this subject. Lots of info.

Hilt has a Web program for rebar called profisrebar

Thanks. I am familiar with 'PROFIS Anchor' download program, and am taking the 'PROFIS Rebar' web-based software for a spin.

Hopefully, the "real" shear force can be transferred by providing shear dowel action (right-hand side of Hilti table) rather than by trying to justify the resistance on shear friction. Of course dowel action is much more sensitive to edge distance and spacing than shear friction. Though, I am not certain that it is always reasonable to substitute dowel action for shear friction bars in every type of concrete cold joint.

You bring up a good point. Alternative capacity justified/achieved by dowel-action alone. I have tried to replicate the HILTI capacity numbers on the RHS of Table 19 above, and I get close, but cannot get the same numbers. I calc the tension capacities to the pound. The problem with double-checking the RHS ΦV_n shear numbers is that they refer to their ICC-ERS report, and that gets into the depths of ACI 318-14 Section 17. I have a call into HILTI engineers to explain how they arrive at those concrete capacities - I assume it is pryout failure assuming NO edge effects.

Palieraki paper

I have been digging into this a bunch of papers in the past 24 hours and the Palieraki et. al paper was referenced in a few documents I researched, including the Hilti Post-Installed Reinforcing Bar Guide you cited. The paper I have appears to be a bit different to your paper. The copy I have is entitled "SHEAR TRANSFER ALONG INTERFACES: CONSTITUTIVE LAWS. Dated 2014. What paper is your graphic post from?

Interesting that the researchers split up the capacity into friction and dowel components, rather than 'baking it into the cake' (as you would say!):

Also, the contributions factor for low-levels of slip lean towards dowel action by near 2 times:

More reading/research required on by behalf...back soon.

 

[KootK]

The graphic was from the Hilti guide.

The paper that I referenced was attached to my post. At least I thought it was. I'll try again here.

Can I have your copy? This is pretty interesting.

Given that your EOR must have done one of these things:

1) Based embedment on the latest European research or;

2) Indiscriminately specified the embedment from a place of relative ignorance.

...my money's still definitely on #2. The embedment may well work but I suspect that would be merely a happy accident.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[Ingenuity]

Can I have your copy?

The paper I have/referenced is the same as the paper you attached by Vasiliki PALIERAKI, Elizabeth VINTZILEOU and Konstantinos TREZOS.

2) Indiscriminately specified the embedment from a place of relative ignorance.

Given all the other dowels on the project for other 'non-related' elements are #5 with the same embedment, I agree. A 'copy and paste' note/detail.

 

[Ingenuity]

It's a hybrid dowel/shear friction method that seems to yield rather astonishing accuracy with respect to test results.

...and with a significant database over a significant duration of time:

In total, results from 18 papers, produced between 1960 and 2012 have been collected. Results from almost 580 tests regarding interfaces with different dimensions and geometry, covering a wide range of material properties are included in the evaluated research works.

 

[Ingenuity]

Using the interface shear design method proposed by Palieraki et.al. consisting of the summation of friction and dowel action effects, based upon HILTI's Post-Installed Reinforcing Bar Guide 2016 and assuming UNcracked concrete, and #5 dowels with ONLY 5" (= 8 d_b) embedment, I calc an effective capacity of 13.9 kips, assuming a 6" x 6" grid spacing of dowels:

ACI 318 shear-friction gives me:

ΦVn = 0.75 * 1.0 * 0.31 in² * 60 ksi = 13.9 kips! Co-incidently the same, but I have only embeded the dowels 5" into the substrate!!!​

Also the equation for f_c,vf needs to be limited so that [5 f_bu l_e / d_b] term does NOT exceed f_sy, in my opinion: