Cover Slab Design 6

[Joe Grant]

Looking for examples/help for designing a concrete cover slab. Attached is an example picture of a cover slab. My thoughts would be to design for punching shear around the missing area where the manhole cover is to be place however I am unsure how to make up for the missing concrete.

Any tips/advice is welcome

Thanks

 

[KootK]

I see where you're coming from with the punching shear check but, here, I think that one way shear checks around the perimeter would be more appropriate. For flexure, probably just treat it like four, overlapping simple span beams around the perimeter of the opening.

One of our members, TheMightyEngineer, is a precast specialist and has designed similar items in the past. Hopefully he shows up here to offer his 2c.

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.

 

[TehMightyEngineer]

[POOF!] I have been summoned! Thanks for the props KootK.

KootK's got the best approach, dividing this up into beams will give you a lower-bound approach that will give you acceptable results. See the attached PDF sketch for what we mean.

I'm assuming this cover will be designed for H20 wheel loads which will be a 16 kip wheel load (21.28 kip if you need to include highway impact loading) distributed to a 10 in x 20 in area. Your major load cases are going to be a wheel on the edge of the opening and a wheel at the corner of the opening. If you have a metal cover over the opening then also check a wheel load in the middle of the cover (equally distributed to all sides of the opening) but this doesn't usually control for symmetric covers. Similar approaches can be done if this is just distributed live loads.

For flexure a wheel load on the middle of the "beams" shown in the sketch will be the controlling flexural design for these "beams". Don't consider the entire width to the perpendicular edge, the high stress locations will be near the opening. Here's the flexural results from a FEA I did for a similar thin cover with a large opening:

Circumferential flexure
Transverse flexure
Mxy flexure

You can see that only about 2/3 of the perpendicular dimension from opening to free edge is being utilized. From the transverse flexure you can see that there is transverse flexure spanning from the edge support of the cover to the "beam". On this thin cover I had to have hoops to get properly developed reinforcement for this transverse flexure, as shown here.

So, that's flexure. For shear you have a number of checks. First, if you have a metal frame and cover you need to make sure that the two-way shear of breaking the opening out doesn't control. It rarely does but often covers are set into the slab and thus reduce your thickness available for two-way shear. You can also have two-way punching shear at the wheel load itself (the 10"x20" tire contact area). This almost never controls but must be checked. Lastly, you have the one-way shear of the "beams" where they're supported at the edge of walls or whatnot that's supporting the cover. This is often what controls the thickness require for the cover. It's almost entirely impractical to design shear reinforcement for these covers; increase the thickness until the concrete shear strength is sufficient (this counts as a slab for the ACI 318 provisions in my opinion). If these covers have a "lip" where there's a bond out in the edge of the cover for whatever this sits on, you'll have to design the shear for the thinner dimension for the one-way shear.

With weird or offset cover geometry you can also get some interesting shear stresses or stress risers. Also, square or odd-shaped covers can get negative flexure in certain situations even though they're simple span. For a round cover with the opening shown though these shouldn't be an issue.

Finally, unless these are mass-produced items, you'll help out everyone to overdesign it a bit. Avoiding bending rebar in weird shapes (the shop guys hated my hoops in that sketch above) and try to stick to just a bunch of straight bars. Don't forget those diagonal "beams", you will likely need bars in the diagonal areas as well.

Professional and Structural Engineer (ME, NH, MA)
American Concrete Industries
www.americanconcrete.com

 

[TehMightyEngineer]

Oh, and definitely don't forget the development length requirements for embedment at a simple support. This can often cause you to need to hook your flexural bars around the corners of the opening to get them developed properly. Not a bad idea for crack control at those corners as well (even with the haunches I suspect you'll want to be sure there's no cracking issues at the corners).

Professional and Structural Engineer (ME, NH, MA)
American Concrete Industries

http://www.americanconcrete.com

 

[KootK]

If I could, I would give myself a star for summoning TME. Clearly, that's worked out well.

First, if you have a metal frame and cover you need to make sure that the two-way shear of breaking the opening out doesn't control. It rarely does but often covers are set into the slab and thus reduce your thickness available for two-way shear.

This I'm curious about. I wouldn't consider a two way shear check here to be appropriate because I don't feel that the frame/lid assembly would possess the kind of internal stiffness required to produce a proper punching shear failure frustum like you'd get with a column. This would be exacerbated, in my mind, by the lack of concrete flexural capacity across the opening and the absence of the struts that usually emanate up/down from the support/load.

On the other hand, if one treats this as one way shear as I've suggested above, it tends not to work out so well. I've struggled with this when designing slab lift out panels. As the wheel load approaches the edge of the lid, your effective shear area for the adjacent concrete gets smaller and smaller and rather difficult to define. For plaza slabs, I've concluded that this phenomenon requires a minimum slab thickness of about 9" which is quite a bit when considered in a historical context.

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.

 

[TehMightyEngineer]

Hmmmm, interesting. So you essentially consider one-way shear for the load on the frame/lid? Assuming the load is evenly distributed to the circumference do you just consider the shear width (b_w to use ACI 318 terminology) to be the circumference of the frame?

I've typically assumed two-way shear failure as I've personally seen failures that I ran the numbers on and they roughly matched the punching shear calculated strength when a overload point load was applied to a thin slab. Granted, this was just one data point and was a heavy load on a pad rather than a heavy load on a frame and cover at an opening. Similar enough that I've gotten in my head that two-way shear was the way to go. Wish I could show pictures; not sure if my NDA still applies as that client went out of business but their parent company still exists. Plus, examples for pile caps and column to footer shear design seemed to match this approach (though, again, not an opening).

I've also struggled with effective shear width as wheel loads approaching the edge of the lid. I've taken a number of pages out of AASHTO for this. Their equation for effective width of slab bridges doesn't decrease for shear. In addition, they note that shear will likely not control for slab bridges. Obviously, a single-span slab bridge is quite a different animal than a thin cover or slab which likely can't even fit both wheel loads on it. In the past I've usually made an FEA when I thought shear from a wheel load near an opening could control. Plus, as I said above, I almost always try to design the cover so that I was comfortable that shear wouldn't control.

Professional and Structural Engineer (ME, NH, MA)
American Concrete Industries

http://www.americanconcrete.com

 

[TehMightyEngineer]

It's been a while since I went through this and I have to get back to my work but perhaps something can be gained from this: [URL unfurl="true"]http://www.slideshare.net/elantsoght/shear-in-reinforced-concrete-slabs-under-concentrated-loads-close-to-supports[/url]

Professional and Structural Engineer (ME, NH, MA)
American Concrete Industries

http://www.americanconcrete.com

 

[KootK]

So you essentially consider one-way shear for the load on the frame/lid?

Yes.

Assuming the load is evenly distributed to the circumference do you just consider the shear width (bw to use ACI 318 terminology) to be the circumference of the frame?

No. I'll estimate the plf based on where I've put the load and design for that value locally.

Granted, this was just one data point and was a heavy load on a pad rather than a heavy load on a frame and cover at an opening.

If this was a solid slab without the frame and opening, I'd consider it apples and oranges.

I've taken a number of pages out of AASHTO for this.

I'll look into that.

Their equation for effective width of slab bridges doesn't decrease for shear. In addition, they note that shear will likely not control for slab bridges. Obviously, a single-span slab bridge is quite a different animal than a thin cover or slab which likely can't even fit both wheel loads on it.

Interesting. Like I said, once your 9" or so you can pretty much handle a fire truck outrigger. How thin might a bridge slab be in the wild?

Thanks for turning me on to the TU Delft stuff. Just what the doctor ordered.

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.

 

[TehMightyEngineer]

Oh, are you talking about a point load on the metal frame and cover in the opening of the slab? Or are you talking about a load right next to the opening?

Yes, somewhat apples to oranges. Without going prestressed I'd suspect most slab bridges of normal to short span are going to be at least 9 inches thick so perhaps that's why AASHTO can get away with that. I also agree that it's a little apples to oranges but if I recall the TU Delft slides there imply that you do get quite a wide effective shear width when near to supports.

Professional and Structural Engineer (ME, NH, MA)
American Concrete Industries

http://www.americanconcrete.com

 

[KootK]

Oh, are you talking about a point load on the metal frame and cover in the opening of the slab?

This one exactly. Mine was a 12' X 12' lift out panel for heavy data center equipment. Over those kinds of spans, it was hard to justify using more than a few edge angle studs at a time to resist a load applied at the edge of the lift out panel. I ended up building a 7" channel into the perimeter of the lift out panel itself to help with the load distribution. Load at the corner was the trickiest for me to get cozy with. All of your distribution things are essentially cantilevering from the interior then.

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.

 

[TehMightyEngineer]

Oh! We're probably thinking entirely different scales. I'm thinking 24"x24" opening (or 600 mm x 600 mm as shown in the OP post). For small openings like manhole covers I still think two-way shear and even distribution of the load to the edge is reasonable. On a larger opening cover I agree completely that you have to consider an effective width.

Professional and Structural Engineer (ME, NH, MA)
American Concrete Industries

http://www.americanconcrete.com

 

[KootK]

No, I understood the scale from the get go.

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.

 

[TehMightyEngineer]

Hmm, so you think a 10"x20" tire on a 36"Ø manhole cover wont distribute evenly to the frame?

Professional and Structural Engineer (ME, NH, MA)
American Concrete Industries

http://www.americanconcrete.com

 

[KootK]

I see it like this:

1) I agree that the lid likely will distribute the load rigidly. Obviously, for non-centered loads, there's an M/S component in addition to P/A.

2) If the lid will not be fastened down positively, then M/S + P/A needs to be adjusted to reflect the potential for uplift and separation where you're outside the kern.

3) As you know, you get much better values for two way shear stress than you do for one way shear stress. But why is that? I'm sure there are several reasons of which I am unaware but I'm fairly certain that a big one is that you get struts emanating from the point of load application. And those struts are tied locally by the continuous slab top steel. Here, you won't have the top steel, you won't have the struts and, in my opinion, you won't have the spiffy two way shear stress values to play with. Remember a while back when everybody lost their *()^! because those U of M studies showing that punching shear stud rail values were un-conservative when the top steel was deficient? Just imagine what would happen if there were no top steel at all.

Obviously, #3 is the one that concerns me the most. I will concede, of course, that there are tons of these things out in the wild being successfully load tested to a far greater degree than your average building.



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.

 

[TehMightyEngineer]

I've stared that post because it's a great summary and I see what you're getting at, but also because that's a lot I hadn't considered, despite this being something I'm designing all the time.

1 & 2) Interesting thoughts; I'll have to consider this more. Maybe I need to have one of our delivery trucks drop an outrigger on frame and cover and observe the results as they load box culverts. I would think that the weight would keep it from actually lifting but I must concede that you're right that it could have a "kern" area where outside that results in a non-uniform support reaction around the cover. The only counter I have to this short of running a 3D non-linear FEA analysis on this is that when they load test manhole cover castings they use a central load. Obviously this is the max-flexure point but I would assume if what you say is true then the frame would need to be load tested for a non-centered load. More thought on this is definitely needed.

3) This.... I actually will freely admit I wasn't aware of this. KootK, you ruin all simple things. First you take away my retaining wall joints and now you ruin my two-way shear equations.

Wow, that's quite a thing. I agree 100% that I thought two-way shear got extra strength because of compression struts in the top of the slab that is being sheared. But I assumed these compression struts could form around the opening without much consideration for opening or rebar. So you're saying that lack of reinforcement prevents these struts from developing as was to be expected for two-way shear in a typical slab? But, if that's the case, why isn't this lack of rebar considered in anything I've seen regarding footings or pile caps with reinforcement in one face only? What about for plain concrete? What about very thin slabs? Too many questions...

Well, I now see what you're getting at; one-way shear with an effective width that takes into accounts the "lifting" of the unloaded side of the frame/cover. Hmmmm... beard stroking must commence. I need to sleep on this one.

Professional and Structural Engineer (ME, NH, MA)
American Concrete Industries

http://www.americanconcrete.com

 

[KootK]

KootK, you ruin all simple things.

Yeah, that's my thing. KootK. Complicating the simple since the early 90's. Probably explains my astounding wealth and jamb packed social calendar.

So you're saying that lack of reinforcement prevents these struts from developing as was to be expected for two-way shear in a typical slab?

Yes, that's the crux of my concern.

But, if that's the case, why isn't this lack of rebar considered in anything I've seen regarding footings or pile caps with reinforcement in one face only? What about for plain concrete? What about very thin slabs?

Excellent questions. Easy first.

1) Thin slabs. Not sure that I understand the concern. Short struts?

2) Footings. With isolated, single column footings you've got steel on the bottom, opposite the load application point, right where it should be for strut formation. Check. I have always been surprised that we don't concentrate our rebar a bit under the load however.

Now for some brown belt level, unsubstantiated theory before dealing with caps and plain concrete:

START THEORY

Originally, I thought that you always needed rebar opposite the loaded face in order to provide lateral restraint to those wonderful struts. I've come to harbor a more nuanced understanding. I now believe that all that is needed is for there to be a competent flexural mechanism in place. Any competent flexural mechanism. On the face opposite the load, that can entail:

A) Rebar resisting flexural tension (the normal case)

B) Concrete resisting flexural tension without rebar (modulus of rupture).

C) Concrete in compression if that reflects the moment at the location being considered.

Three's the surprising one in my opinion. It jives though. You can have the struts so long as you can restrain them laterally at the unloaded face. And there's no reason that restraint can't be obtained by the diagonal struts butting up against the Whitney stress block flexural compression struts rather than being tied by rebar. In these cases, the rebar would be in compression anyhow. It's actually quite a common condition. Most transferred column are near one another and, as such, deliver their load where underside of the slab is actually in compression.

END THEORY

Back to business:

3) Plain concrete. See point B above.

4) Pile caps. See point C above. All of the piles push upwards on the cap similar to a simple footing. As such, the unloaded face of the cap opposite the piles is in compression -- at least theoretically -- and it is that compression that restrains the shear struts.

Let me know if a sketch is required to clarify any of this.

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.

 

[TehMightyEngineer]

Let me know if a sketch is required to clarify any of this.

No, I follow you. After some beard stroking and morning coffee I arrived at similar, but less thought-out, concepts.

1) Thin slabs. Not sure that I understand the concern. Short struts?

I was more wondering if you would consider the positive rebar in a thin slab as being "close enough" to the upper surface to provide restraint to the struts. In short, do we need two discrete layers of reinforcement?

With isolated, single column footings you've got steel on the bottom, opposite the load application point, right where it should be for strut formation.

I was wondering if this was the case. Glad you came to the same conclusion. But if this is the case don't we have reinforcement on the side opposite the load when we look at our frame and cover? Obviously this reinforcement is distributed around the opening and could be "insufficient" for two-way shear, but it at least it's on the "correct" side, right?

See attached sketch (a little rough, was in a hurry).

I have always been surprised that we don't concentrate our rebar a bit under the load however.

I've done this on occasion; plus the provisions in ACI 318 for rectangular footings prescribe this to some degree (though not for two-way shear reasons).

3) Plain concrete. See point B above.

Would you say that it would be appropriate to use the plain concrete two-way shear equations for our cover shear design but with the phi factor for reinforced concrete in shear? I would say that this makes the most sense to me.

Professional and Structural Engineer (ME, NH, MA)
American Concrete Industries
www.americanconcrete.com

 

[KootK]

I was more wondering if you would consider the positive rebar in a thin slab as being "close enough" to the upper surface to provide restraint to the struts. In short, do we need two discrete layers of reinforcement?

Ahh. I'd say that one layer is fine so long as your punching shear "d" is measured to that layer.

Obviously this reinforcement is distributed around the opening and could be "insufficient" for two-way shear, but it at least it's on the "correct" side, right?

I agree, it's on the right side. It is the fact that it goes around the load rather than under it that bothers me. ajk1 had a problem a while back where he was replacing a slab around an existing column. He wanted to dowel in new bars through the column to make it coshure for punching. It seemed onerous and possibly destructive. I held back on mentioning it but, at the time, I wondered if bars very close to the sides of the column would suffice in place of bars through the column. It is technically a code violation in that instance but, at some point, close has to be close enough.

Would you say that it would be appropriate to use the plain concrete two-way shear equations for our cover shear design but with the phi factor for reinforced concrete in shear?

I wouldn't say that. While I can see some log in it, ans surely there is some truth to it, I think that the litmus test is to ask yourself this: "if I'm going to use enhanced, two way shear stress values, what makes this situation substantially different from a one way situation?". The answer, in my opinion, would be "not much".

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.

 

[TehMightyEngineer]

So, as coincidence has it I have a project I need to finish which includes a precast 6 ft x 6 ft span cover with a 36"Ø manhole to be placed over a cast-in-place valve pit structure. I've used the time to rough up a FEA model in RISA for a 2" thick steel cover supported on the edge with 30 k/in compression only springs. I loaded it eccentrically with a H20 wheel load to try to capture the shear effects you expect.

Here's the 40x exaggerated deflection which shows the supports being loaded on one quadrant only:

Short answer, you appear to be correct; this does appear to cause a non-uniform shear in the frame of the metal cover.

Here's the various plate stresses from the model:

Circumferential plate shear (most telling IMO)
Von Mises stress
Tau shear stress

Now the question remains what to do with the shear calculation in the slab.

Professional and Structural Engineer (ME, NH, MA)
American Concrete Industries
www.americanconcrete.com

 

[KootK]

Awesome. I love me some FEM exploration in threads.

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.