Two Way Shear - vs in Thick Elements

[sticksandtriangles]

Recently I was working with some really large loads and trying to deal with punching shear on thick concrete elements.

I worked the ACI equations for punching shear and I needed (32) #5, bars, (8) legs*4 sides, at my critical perimeter, at a spacing of 6"oc.

Traditional examples of punching shear reinforcing always show this thin flat slabs, typically using stud rails that are centered on the column width and then stay within the same column dimensions as they move away from the column.



I aimed to match this traditional layout in my heavily loaded thick foundation.

See option 1 below:





I showed this to a colleague and they asked to splay these out over the critical perimeter, which I initially thought would be fine, but now I have some concerns.

Option 2:





Option 2 has the same amount of reinforcing at the critical perimeter, 8 #5 per face, but the spacing can be greatly reduced, to around 1.25ft in each direction with this setup.

I think the code really is pushing for option 1 as is traditional with shear stud rail layout, but I wanted others opinions.

ACI 318-19 now works in shear stress for vs, but I think the heart of the equation is finding the # of bars that cross the d away failure plane as shown in the red region below:





Option 2 starts to make a come back in raw steel area after d/2 away into the d region, but I don't think it is enough typically. Also the stress concentrations are highest near the column support, so it feels a bit dubious to pick up extra capacity away from the column?

Had a bit too much fun making the images as you can see. Modelling rebar in revit seems pretty slick.

 

[jerseyshore]

We've always done option 1. When SSR's get to heavy or require additional demand then we add drop panels. There's a reason it's called the critical zone. Curious what others who do more concrete design say.

I never want to mess around with punching shear failures. It's one of the few true "sudden collapse" type failures that us building structure engineers analyze.

 

[HTURKAK]

In past , i always preferred option 1, but with closed stirrups. The reason is , the flexural cracks develop firstly following the fictitious column strips.

 

[W460x68]

Option 1 always for me. I have occasionally seen a radial distribution in a similar manner to Option 1.

 

[Luceid]

I work in the US and have only seen Option 1 until I worked with some folks from other countries where for foundations, the initial assumption was option 2.

 

[bookowski]

Note that there's a recommended extra reduction in capacity when using rebar instead of studrails. It's not in 318 so you don't have to follow it but for the cost of a few extra bars its worth considering for heavy situations. It's probably in 421 and I think it's another .75 factor (been a while since i looked it up).

 

[Luceid]

@bookowski Is that for any of the punching shear reinforcement checks in general?

I know that ACI 318 limits the MAX shear stress to be 6sqrt(f'c) for rebar vs 8sqrt(f'c) for shear studs, I'm guessing this is the 0.75 factor you are remembering - let me know if it extends beyond that. I haven't found any other difference between stud rails and rebar calculations other than that, so the main difference is often the premium for stud rails vs ease of construction. Plus in thicker foundations you get into stacked stud territory

 

[bookowski]

Maybe that's what I was remembering, the 8 to 6 reduction.

I never heard of stacking studs.

 

[sticksandtriangles]

Ok, thanks everyone.

Sounds like option 1 is standard in most places.
Clearly option 2 provides some benefit to the punching shear, but quantitively let's try to put a number to it.

I believe that ACI is trying to figure out the total numbers of bars that cross the failure plane as shown in the first picture.

If this is the case i think the formula might look something like this. Do you think if you spaced bars as shown per the calcs, you would get the same performance for punching shear (4 sides of (8) #5's @ 6"oc, verse an area layout of #5 @ 9.5" ea way)?

These give the same number of bars crossing the failure plan.



@bookowski, do you have a reference for the reduction in capacity for rebar vs studrails?

​

 

[Luceid]

@sticksandtriangles Here are the differences I see:

1. 318-19 Table 22.6.6.1: vc is limited to 2sqrt(f'c) for stirrups and 3sqrt(f'c) for stud rails (generally - check the table for other cases). NOTE that almost all ICC-ESR reports for studrails that I have seen will override this value, limiting vc to 1.5sqrt(f'c) instead - so be careful pushing your vc to 3sqrt(f'c) unless you're sure you can get that from the stud rails.

2. 318-19 Table 22.6.6.3: Maximum shear stress limits for stud rails vs stirrups is the 8 vs 6 sqrt(f'c).


@bookowski Dropping this here for reference, studs greater than 10" need to be piggybacked, commentary suggest the head helps to anchor the taller stud

 

[sticksandtriangles]

. 318-19 Table 22.6.6.1: vc is limited to 2sqrt(f'c) for stirrups and 3sqrt(f'c) for stud rails (generally - check the table for other cases). NOTE that almost all ICC-ESR reports for studrails that I have seen will override this value, limiting vc to 1.5sqrt(f'c) instead

Thanks for that, I was aware of this after watching the Bondy video on punching.

I was thinking @bookowski might have something not listed in code

 

[Luceid]

Does option 1 and 2 create the same periphery failure surface? I'm familiar with the octagonish second critical section which seems more applicable to your option 1, which means the two options would have different perimeters - potentially requiring option 1 to extend further from the column to achieve the same punching capacity as option 2. Ballparking it would mean you would need to go 42% further to get the same perimeter area (each corner in this case would be L*sqrt(2) vs 2*L in the option 2 square corner case)

EDIT: I suppose the thicker the element gets, the closer that option 1 will get to becoming more and more squared out. So for a thick foundation element in your case, this is probably less important.

 

[sticksandtriangles]

Does option 1 and 2 create the same periphery failure surface?

For arguments sake, let's say yes.

Technically if you continued the area reinforcing in all directions, option 2 would be even better.

I just want to focus on the first critical perimeter for now.

 

[Luceid]

Got it - I'm with you. I think the logic is sound - since your d is so much larger, your failure cone gets big enough that if you are able to space the punching reinf further apart, I don't see an immediate red flag against it