Shear stress in a slab 3

[WARose]

I’m modeling a elevated reinforced concrete slab that will pick up a pretty heavy point load. To make a long story short: the (one-way) shear stresses look ok....except for a strip that represents about a third of the slab’s smallest possible failure plane. (I.e. the width of the bearing area + d.) It’s somewhat overstressed at the peaks of this zone. The average looks ok....but I’m not sure here. There certainly is lots of heavy re-bar in the slab to distribute this force.....and the total force (over the failure plane) isn't in excess of the allowable.....but having such a large percentage over worries me a bit.

As a sanity check I modeled some point loads via about the only other code I know that treats them: AASHTO. And modeling some worst case scenarios via that code.....I think the approach is valid......but some second opinions are welcome.

 

[KootK]

If punching shear is good and you've got lots of flexural distribution steel, I'd say that you're fine. It's a very pedantic thing but I actually have a theory that, to a degree, shear stresses actually re-distribute laterally to go where the flexural capacity is +/-. That said, I typically like for my peaks to not exceed my averages by more than about 50%.

 

[Yao1989]

Do you have any shear studs in the one-way shear check in the column width + d zone? How about interface shear in the same zone?

 

[WARose]

[blue](Kootk)[/blue]

If punching shear is good and you've got lots of flexural distribution steel, I'd say that you're fine. It's a very pedantic thing but I actually have a theory that, to a degree, shear stresses actually re-distribute laterally to go where the flexural capacity is +/-. That said, I typically like for my peaks to not exceed my averages by more than about 50%.

Thanks for the advice Kootk. It just unnerves me a bit because I've been around a lot of FEA guys who get nervous when more than about 10% of a plane is in excess. But they were not working with reinforced concrete.....so the question has always loomed large with me: how much is too much?

[blue](Yao1989)[/blue]

Do you have any shear studs in the one-way shear check in the column width + d zone? How about interface shear in the same zone?

No.

 

[Shu Jiang PEng SE]

Check punch strength, conservatively treat the slab as one way slab and check bending and shear strength respectively. There would no worries if all the three items are acceptable.

—————————————————————
Shu Jiang, SE (Nevada). PE(Michigan, South Dakota), PEng (Ontario)
J&J Structural Consulting Inc.
Structural design, analysis, inspection, drawing review and stamping, and connection design

 

[WARose]

Check punch strength, conservatively treat the slab as one way slab and check bending and shear strength respectively. There would no worries if all the three items are acceptable.

The one-way didn't work (in a strip of the failure plane).....that was the whole point of the thread.

 

[DaveAtkins]

I have been wondering the same thing. Very often, when I model a wall or mat footing in RISA-3D, there will be "hot spots" where the shear is quite high. I generally average them out over a limited distance, but I still think this is too conservative. By definition, isn't one way shear across the entire width of the structure you are checking (wall, or mat, or...)? Put another way, you can't get a localized one way shear failure, can you?

DaveAtkins

 

[jayrod12]

I would say it may be possible to have a zipper effect type failure. Crack starts and then propagates. Personally I feel it would depend on the location of the hot spot.

 

[bones206]

One perspective, for what it's worth:

3.8—Finite element analysis (FEA)-based slab design resultants
To complete slab design, an engineer should determine the quantity of reinforcement required for each design cross section. Although design cross section locations and lengths are often guided by code rules, in general, sections are needed at peak stress locations and their lengths based on the extent of the slab that can be assumed to act as a unit in resisting internal forces. The width of each section should be chosen so that the moment distribution along the section is reasonably uniform, does not change sign, and can be resisted by uniformly distributed reinforcements.

I got dinged once by a reviewer who said my one-way shear averaging was unconservative because my design strip was too wide. I think it was like 3 or 4 feet wide on a 2 foot thick slab. The reviewer was one of the head honchos on the ACI 349 committee so I couldn't argue much, but he didn't provide any hard and fast reason for his concern beyond his gut feeling.

Did you model this load as an actual point load? It may help to refine the mesh locally and model the load as a pressure over the actual bearing area.

 

[TehMightyEngineer]

I have been wondering the same thing. Very often, when I model a wall or mat footing in RISA-3D, there will be "hot spots" where the shear is quite high. I generally average them out over a limited distance, but I still think this is too conservative. By definition, isn't one way shear across the entire width of the structure you are checking (wall, or mat, or...)? Put another way, you can't get a localized one way shear failure, can you?

I've had this exact same question. I generally try to make sure such hot spots are limit and not too far over the allowable but have always had the question in my mind of how much is too much. My general out was also thinking that you can't get a localized one-way shear failure in a typical concrete slab or wall.

Edit: Nice reference Bones! I almost forgot about that section which I believe is also in 318-14. That's probably our answer there.

Ian Riley, PE, SE
Professional Engineer (ME, NH, VT, CT, MA, FL) Structural Engineer (IL, HI)

 

[WARose]

Did you model this load as an actual point load? It may help to refine the mesh locally and model the load as a pressure over the actual bearing area.

No it's put in as pressure.

Thanks for the reference.

 

[GC_Hopi]

Here is the AASHTO for slab type bridges because it pretty generous on the equivalent strip width.

 

[WARose]

Yeah, AASHTO is kind of how I am justifying it in my mind. A I alluded to in my OP: you can model a lot of loads on slabs with the thicknesses they prescribe and not get acceptable shear stresses until you have eaten well more than half of the failure plane.

But by that approach, you need distribution steel perpendicular to the main steel.

 

[GC_Hopi]

Agreed, you should have distribution steel. AASHTO gives specific guidelines but ACI seems to hint at that when it says "based on the extent of the slab that can be assumed to act as a unit in resisting internal forces"

 

[TehMightyEngineer]

The only reason I wouldn't want to use AASHTO as more than just a rough check is slab bridges per AASHTO have specific distribution reinforcement requirements that are much higher than the equivalent in typical building slabs IIRC.

Ian Riley, PE, SE
Professional Engineer (ME, NH, VT, CT, MA, FL) Structural Engineer (IL, HI)

 

[Settingsun]

Did you apply the load to the area it hits the real slab's surface (actual bearing area) or spread it at 45 degrees down to the mid-depth of the slab? I'm assuming you've done a plate model at mid-depth rather than bricks in which case you can adopt this additional load spread.

This document might also assist - give you a reference if nothing else.

http://publications.lib.chalmers.se/records/fulltext/176734/local_176734.pdf

 

[WARose]

Did you apply the load to the area it hits the real slab's surface (actual bearing area) or spread it at 45 degrees down to the mid-depth of the slab? I'm assuming you've done a plate model at mid-depth rather than bricks in which case you can adopt this additional load spread.

That's a good point Steve. I just put it on the surface. (And yes, it is a plate model.)

As thick as my slab is, that helps.