Unique punching shear - advice requested with fib MC

[Gallig3r]

Very odd punching shear situation (see attached), with 2 questions below. We do understand what the conservative/ safe routes are, but the question of “can you refine?” is being asked of us.

Looking to assess in a logical manner, I was hopeful to employ mechanical models of punching shear, such as critical shear crack theory (from the fib Model Code), or tangential strain model. ACI’s framework on punching doesn’t appear to be helpful as it is mainly empirically derived. However, fib Model Code is new to us, so appreciate any insights.

1.) If the new concrete for the (fully bonded/ composite) slab enlargement is 2x the concrete strength of the existing slab – what concrete strength should be used for punching capacity? Lets say 4ksi and 8ksi (or C30 and C60). My (perhaps incorrect) understanding of critical shear crack theory, is that for the geometry here, the new concrete is much more influential on ultimate punching capacity (see fig 3 of attached CSCT Mechanics.jpg). But how to quantify when evaluating capacity?

2.) We are obviously checking the 45 degree punching critical sections #1 and #3 in my diagram. However, section #2 seems plausible as well. If one should check it, how does one assess capacity of such a shallow punching cone? I know in theory I could start with Muttoni’s CSCT ground work and rederive the his formulas with a different angle assumption… but not confident in doing it correctly. Was hoping at very least someone might know "is section #2 is qualitatively stronger or weaker than #3?" and start from there.

 

[hardbutmild]

To be honest... I think this is a kind of problem where you MUST be conservative. It's a very specific situation with nothing even close to this ever being tested. Add to that the fact that punching is a dangerous type of failure - go with a conservative solution.

 

[icebloom]

Not familiar with the fib method but I have previously come across this paper by Muttoni when looking for punching shear related info. Best of luck.

https://abcic.org.br/Arquivos/phhmcvll.pdf

 

[Gallig3r]

Not to be combative, but it seems like you (hardbutmild) aren’t familiar with Critical Shear Crack Theory (CSCT)? It is a (semi) mechanical model, so it absolutely does have the ability to extrapolate to some degree. Think of all the weird structures we (designers) can justify using strut and tie models – that we don’t require testing. With a model/capacity based on mechanics, one can account for odd situations.

ACI’s framework on punching however is all based on empirical testing. ACI has no discussion of actual mechanics at play (material failure criterions). Thus, I 100% agree about this situation as “untested” and would avoid any refinements by counting on the higher strength of the new concrete… when using ACI’s framework and equations.

Hence our interest in CSCT. An unfortunate reality is that the CSCT is actually semi-empirical/ semi-mechanical as Muttoni only describes how shear capacity drops with slab rotation opening up a critical crack. The capacity as a function of rotation is empirically derived. He does qualitatively mention that the force transfer is primarily at the base of the cone, but I was hoping others more familiar with the MC2010 or CSCT could weigh in on any nuance.

 

[Gallig3r]

Yesterday I stumbled across a paper that further refines the modeling of punching shear mechanics and offers a fully mechanical model to predict punching capacity.

https://www.sciencedirect.com/science/article/abs/pii/S0141029623006119#f0020

While the approach is impractical to use for typical design, it describes important parameters like 1.)what depth of the slab that is transferring shear via aggregate interlock 2.) what depth of slab is involved in direct bearing, and 3.)what percentage of punching capacity is from top rebar doweling.

In my condition, most of the force flow would be within the depth of the new, high strength concrete. However, the spalling that impacts doweling capacity is a function of the existing, weaker concrete. Since doweling is not a major contributor for my reinforcing ratios, the new, stronger f’c will be primarily influencing the capacity, almost 100% so. Even taking a 50/50 average f’c to use in design equation would still be conservative… according to this model, any way. TBD what we decide to do.

Interesting result from the above: is that if we poured a topping slab (instead of pumping under soffit), the f’c for punching should be based on the existing, weaker concrete f’c. I know of designs that have used weighted averaged f’c in the past for such situations. Good to know for future designs.

 

[hardbutmild]

I am aware of CSCT, you and I just have different design philosophies. STM is significantly different in my opinion, but ok. There’s no hard feelings of course.

 

[Gallig3r]

you and I just have different design philosophies

eh maybe, maybe not.

Either way, the design codes definitely DO have differing philosophies on punching, which is why this thread was created in the first place.

 

[HTURKAK]

My gut says , no! not really.. In this case , the hogging reinf. at weak conc. and if punching failure occurs , it will be initiated at the top of the old concrete . The use of weighted average str , 50/50 seems not reasonable . If the new concrete capacity justify the total demand this could make sense.
But there is a problem; the future trenches which will be cut , seems will also cut the top reinf. where the behavior will be totally different story.

Pls provide more info. ( thickness old and new concrete , trench size , top reinf. a scaled sketch ..) to get better responds.

P.S. My observation for the punching failure is that , the cracks initiate at tension side of the slab and the failure is sudden. If i were , i would resist and simply reject such alterations , cutting such trenches, cutting new holes etc. which will compromise the strength and integrity of the str.

He is like a man building a house, who dug deep and laid the foundation on the rock. And when the flood arose, the stream beat vehemently against that house, and could not shake it, for it was founded on the rock..

Luke 6:48

 

[Gallig3r]

if punching failure occurs , it will be initiated at the top of the old concrete. The use of weighted average str , 50/50 seems not reasonable

It seems you are viewing the problem from ACI's framework, where the capacity is simply a 4sqrt(f'c) that is presumably uniformly distributed across the cross section. I would agree that this simplified framework would frown upon the 50/50.

However, I am attempting to look at it through lens of CSCT (or PSRT model as I recently linked). First, realize the main punching crack initiates at as low as 20% of ultimate. It's there well before failure - its just a matter of how much it opens.

For ultimate capacity, the PSRT is not even assuming any force flow in the top of the existing slab. The existing concrete is barely being utilized in the punching mechanism . CSCT appears to suggest the same, but its not so explicit - hence I was looking for anyone familiar with it. PSRT is seemingly new (and perhaps yet to even be peer reviewed??), whereas CSCT has been thoroughly vetted and incorporated into code (fib MC2010)

 

[Gallig3r]

Pls provide more info. ( thickness old and new concrete , trench size , top reinf. a scaled sketch ..) to get better responds.

In an attempt to keep discussion focused on the 2 questions I posed in original post, I deliberately avoided giving specifics. We have a conservative solution, just looking at possible refinements.

This topic could also influence strategies on enlarging footings so wanted to keep it generalized as possible. Sorry.

EDIT: since top steel is obviously important in punching I will clarify there is a middle mat of rebar due to some trenches (not drawn) that were formed in original concrete. So we do have continual "top" steel, and NOT just relying on existing bottom steel to become the new top steel.