Precast/Tilt-Up lifting anchors - Combined Shear/Tension 1

[530_3D]

I'll keep it short...half the hardware vendors and consulting engineers say combined shear and tension should be checked for lifting hardware, other half say it doesn't apply. Sometimes it controls.

Assuming its origin is in ACI Ch. 17, there is a disclaimer in the first paragraph of the chapter: "Safety levels specified are intended for in-service conditions rather than for short-term handling and construction conditions." There is also the 4:1 safety factor specified in OSHA 1926.704, which is further straying from ACI Ch. 17.

I feel comfortable not checking it when such a large safety factor is used for tension and shear capacities, and a recent check of an anchor using Ch. 17 calculations showed it passed when maxed out in tension and shear, but the catalog-listed capacities are more conservative.

Any precast or tilt-up folks able to weigh in?

 

[KootK]

I agree with your perception that it is uncommon to see folks checking combined tension and shear in practice in this space.

I can, however, think of no valid technical reason to justify that practice. Be they large or small, I don't believe that it is the intent of safety factors that they should compensate for undercooked design approaches.

I kind of feel as though the catalog purveyors lean into this a bit. "Here are some simplified design values that you are welcome to combine in more complex and penalizing ways if you wish". Wink wink.

 

[530_3D]

So when checking it, would you not use the larger safety factor for allowable, or is that “undercooking”? Some catalog purveyors consider the check with the 4:1 for allowable “overcooking”

 

[KootK]

So when checking it, would you not use the larger safety factor for allowable, or is that “undercooking”?

I would use the larger factor of safety and would not consider that undercooking. By "undercooking", I meant using a design method that is not adequately rigorous. In my mind, safety factors are always about load and material uncertainty and never about compensating for inaccurate design methodologies.

If it is the case that the 4:1 is indeed meant to compensate for not considering combined shear in tension, then I would at least like to see that in writing someplace. Otherwise, how do you know what design cheats are to be covered by it and which aren't?

 

[530_3D]

Unfortunately there is not a single available example of how the capacities are determined, or even a failure mode for more specialized anchors.

The 4:1 safety factor is from OSHA, which doesn't mention combined interaction or ACI Ch. 17. Adopting the most conservative aspects of different codes while disregarding their context and disclaimers could be considered exceeding the standard of adequacy. This is the take of one of our more prudent consultants.

 

[KootK]

This is the take of one of our more prudent consultants.

It's also been my approach at times in the past. That said, when I take that approach, I'm am at least honest with myself about why I've taken that approach. I've done it:

1) Because it was expedient and;

2) Not because I thought that it was technically correct in some nebulous fashion.

The 4:1 safety factor is from OSHA, which doesn't mention combined interaction or ACI Ch. 17

Of course it doesn't mention interaction. Once that can of worms be opened, they would then have to speak to every other possible, combined situation. I feel that the 4:1 is concerned with load and material uncertainty and nothing else.

Adopting the most conservative aspects of different codes while disregarding their context and disclaimers could be considered exceeding the standard of adequacy.

Considering the interaction is not just the most conservative approach, it's the most accurate approach. And it's that latter bit that matters to us. Just because less rigorous approaches may form an "average" that is less than the most accurate one is no rational justification for using something less than the most accurate approach. Shall we include not checking the capacity at all in the "average"? Obviously not.

Unfortunately there is not a single available example of how the capacities are determined, or even a failure mode for more specialized anchors.

You could:

1) Ask the suppliers for back up information be it calculation or testing.

2) Not use products that don't have the back up information.

3) Maybe do your own , DIY interaction based on a some of the components raised to the 5/3 power or something.

4) Ignore interaction and follow what you believe to be the industry standard of care.

I'm trying very hard to be real with you about the situation here. What the industry does, what I've done, whether or not ignoring the interaction has technical merit... That said, if you wish to convince me of the technical merit of ignoring the interaction, expect that to be a protracted slog. And that make sense because this thread exists precisely because you are also struggling to convince yourself of that same thing.

Search your heart Padawan, you know this to be true.

 

[530_3D]

And that make sense because this thread exists precisely because you are also struggling to convince yourself of that same thing.

I am struggling to convince others. If it was my call I would not require interaction to be checked.

I got some calcs from the vendor that use ASD from ACI318-02, and another vendor advised that it shouldn't be used for ASD with 4:1 SF.
KootK can you tell me if interaction is part of that particular version of Appendix D, or was the interaction chart and equation not included until anchorage switched to strength design? Serious question, I don't have a copy.

 

[KootK]

I am struggling to convince others. If it was my call I would not require interaction to be checked.

Fair enough.

KootK can you tell me if interaction is part of that particular version of Appendix D

I've only been paying attention to Appendix D since the 2002 ACI. My recollection is as follows:

a) There's been strength design available since at least 318-02. If there was an ASD method available at that time, I was not / am not familiar with it.

b) There has been some accounting of interaction effects at least as far back as 318-02.

For historical context, the clip below is from the 1977 Nelson Stud Design Manual. That's going back a ways obviously. Like, "Jesus and the apostles" back a ways. And acknowledgement of interaction effects even then.

 

[530_3D]

I found this in a Chapter 17 clarification webinar document from the Director of Engineering for ACI. Since the capacities are without a doubt derived with the methodology in Chapter 17, I like this as clarification for any lifting anchor that relies on reinforcement. I like it a lot. I'm sure its not enough to sway you KootK, so I will cut the protracted slog short here. God speed.

 

[KootK]

I'm sure its not enough to sway you KootK, so I will cut the protracted slog short here.

I agree with that approach if, and only if, you have anchor reinforcement. Just like the blurb says. In my mind, that means u-bars, hairpins, etc that deal with one or both of the breakout failure modes. If you've got that, great. It's simply not something that you mentioned previously.

In the absence of hairpins, u-bars etc, I do not take that PPT blurb to mean that interaction can be disregarded.

As a purely theoretical matter, I take it as patently ridiculous that one would disregard interaction when there is reinforcement. If anything, concrete that has to crack to engage the reinforcement is surely less capable to resist the other failure mode(s) for which it is not reinforced. I feel as though you can hear that in the voice of the author of the slide. The code "implies". As in, this is obviously BS but it's what the code seems to say so let's run with it.

 

[530_3D]

Yes we completely agree on all points!!! Not a protracted slog!!!
...yes this is brand new info that I found today, and i promise it wasn't a bait and switch

 

[KootK]

.yes this is brand new info that I found today, and i promise it wasn't a bait and switch

Can you point us to the source document?

 

[530_3D]

Can you point us to the source document?

https://www.concrete.org/portals/0/files/pdf/webinars/ws_F23_Matthew Senecal.pdf

 

[Brad805]

For dowels I have plotted the Vf/Vr, and Tf/Tr values like you see below. PCI and CPCI contain this clause. I am not sure why you are not specifying a produce like Dayton Superior where they are tested. Some on the market nowadays seem to be copying others, but they should have data to satisfy you.

 

[530_3D]

PCI and CPCI contain this clause.

PCI specifies the interaction equation for connections (Chapter 6), which I don't disagree with, especially if they don't use reinforcement (ductile) to achieve capacities.

Meadow Burke only specifies to check interaction for items that remain in service after initial erection (i.e. corbel brackets and slotted inserts), which also happen to not use reinforcement and have a SF less than 4:1.

 

[Brad805]

We have put in a ton of lifters, and my crew has never balked at some rebar. This is one of the most important parts of precast and in many cases this can govern your design. Depending on the plant setup you can handle the parts after casting, loading, and site erection. At each stage you have humans near the parts. If they are stressed panels, the extra rebar is not always appreicated, but it is only a few bars that typically fall off when you cut pieces for fixed dimensions. Keep in mind precasters do this everyday. Broken parts, or OSHA investigations are bad for business so most have their own safe working values they like that they have devised over endless years.

 

[530_3D]

As a purely theoretical matter, I take it as patently ridiculous that one would disregard interaction when there is reinforcement. If anything, concrete that has to crack to engage the reinforcement is surely less capable to resist the other failure mode(s) for which it is not reinforced.

Picture a slender wall with edge lifters being lifted by a sling at 45 degree angle, with equal shear (acting on wall out of plane) and in-plane tension. If somehow the tension breakout cone occurs due to long developed legs of a vee hairpin bar acting similarly to a headed stud (theory from KootK's take in another post), then shear friction would apply across that plane if the cone's apex (or apexes if acting as a group) does not occur at the ends of the long vee hairpin and the legs cross the failure plane. In that case the shear friction at the large cone failure plane could still resist enough out-of-plane shear to force a typical smaller cone failure at the embedded plate on the lifting anchor itself. That shear cone would be within and not intersecting the larger cone. That's a scenario where the crack that engages the anchor reinforcement in tension would not render the anchor less capable in shear, the mode for which it isn't reinforced.

 

[icebloom]

It is interesting that the ACI presentation notes that interaction should not be checked when designing reinforcement. The Eurocode explicitly states that reinforcement utilisation in tension, shear, or both, should be inserted into the interaction equations. They even kill you with the index on the equations, if you have reinforcement for shear or tension ONLY your index becomes 2/3, which kills your capacity. For reinforcement for both shear and tension, the index becomes 1.5. This is all based on the empirical data for tests of anchors loaded under combined shear and tension.

On the other hand, when you do traditional strut and tie design you never have some kind of interaction equation you check. Seeing as anchor reinforcement is designed via simplified strut and tie analogy this may be the reason for that slide, however it seems to go against the guidance found in Europe which is where the majority of research on the CCD method is from.