Reverse Deflection in Cantilevered Balconies? 3

[MikeManning]

Hello,
I work for a large general contractor, and we do a lot of high-rise residential concrete construction.
I'm working on a bit of a guide to good balcony design (for waterproofing purposes) in these buildings, and I've got a theory, which might be totally out to lunch (which is why I'm checking).
My theory is that, in buildings with suspended concrete slab construction and cantilevered balconies, deflection of the slabs due to dead load and occupant loading inside the building might actually cause cantilevered balconies to deflect "up" (I'm sure there's a proper word for that).

The reason this would be meaningful is that balconies are often poured with a very slight slope outward, presumably for drainage, but we often find balconies don't drain properly, and I'm wondering if this type of reverse deflection might be part of the reason.

I've done a quick sketch of what I mean.

If this is out to lunch, let me know!
But if it's not out to lunch, then how much extra slope should we put on the balcony slab to allow for this deflection? 2%?

Thanks

 

[structSU10]

It is definitely possible. It is dependent on the ratio of the back span to cantilever how much this can happen. Generally the furnishings are part of the live load so an empty balcony is deflecting up based on dead load and some portion of the interior live load.

If the deflection was kept in code limits, however, I would think this is not a problem but that condition may not be checked rigorously enough.

 

[jayrod12]

You'd be correct and it's generally called hogging of the cantilever.

The way to remedy it, would be for the designer to actually take the potential upward deflection into account when specifying the slope of the balcony slab. They just put a minimum slope, 2% for argument's sake, and the concrete finishers provide that, while the backspan is still shored therefore no upward deflection of the cantilever has happened yet. Maybe in the shored condition the slope needs to be 4% in order to account for it, maybe it needs to be 6%, a lot depends on the cantilever span, and the backspan lengths.

The resistance to this is the architect saying that a 4% slope won't be usable for the occupants, which I can see the argument, but as the structural designer you have to explain that some lifting of the cantilever balcony is expected and therefore we need an exaggerated slope during finishing procedures to ensure the minimum slope is maintained after removal of the shoring.

And honestly, the trickiest bit of all, concrete deflections aren't exactly precise, so you could decide you need 8% slope to maintain 2% after shoring removal and deflections. But then the concrete may not deflect as much as predicted so you've got a significant outward slope on the balcony.

 

[MikeManning]

Thanks for the quick replies. I'll have to look up "Hogging".

But yes this sounds like exactly what I was thinking. Finishers finish in the shored condition, and then shoring is removed and everything deflects. That's why we're always careful, when there are flatness or levelness specs on elevated slabs, to test before the shoring is removed.

 

[BridgeSmith]

It's possible that it's the result of loads applied after construction, but I would think unlikely. Since presumably the floor inside and balcony would be placed at the same time, most of weight of the backspan is in place when the cantilever is finished, it would require a backspan that was severely overspanned and overloaded to get significant deflection of the cantilever after the floor slab has cured.

It's possible, and seems more likely to me, that the metal formwork (and therefore the edge form that defines the top of the slab at the cantilever) deflects up under the weight of the wet concrete when the floor is poured. We see a similar issue with bridge deck construction, if we don't account for deflection of the metal deck forming pans (and/or girders) under the load of the wet concrete, which can be significant.

Edit: I guess with shored construction, as others mentioned, this may not be as much of an issue...unless the shoring is removed prematurely. I would not be surprised if the contractors push to remove the shoring as quickly as possible, and quicker than they are supposed to, if someone isn't watching them, so you may be getting more deflection from creep than what's anticipated in the design. It's similar to what we see in prestressed bridge girder when the strands are released (transferring the forces in the strands to the concrete) too early. The sustained force on the green concrete causes the concrete to creep excessively.

Rod Smith, P.E., The artist formerly known as HotRod10

 

[MikeManning]

Bridgesmith, I should clarify that this is suspended slab construction, with wood formwork and shoring that gets removed after the concrete has cured, so deflection does take place after everything is set.

 

[SirPhobos]

I'm certainly out of my element here, but I'd think that if you saw enough "lift" on the balcony, you'd see a similar or more severe depression on the inside of the balcony. Surely it would be noticeable if the room inside the balcony had a depression between the exterior wall and next interior floor support.

-- SirPhobos

 

[dik]

I generally leave a 1" step at the balcony and taper the slab so that it drains away from the balcony door.

So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[MotorCity]

If the balcony is cast as a true cantilever, the only way it will happen is if torsion causes the supporting edge beam to rotate and causes an upward deflection of the balcony slab. In the case of the balcony slab being poured continuous with the floor acting as a backspan, its theoretically possible but unlikely because:

1. Balconies are typically sloped outward to drain, so even if it rotates slightly upward, the downward slope is probably enough to make it appear as not deflecting upwards

2. The pattern of the loading on successive spans and the magnitude of the loading would need to be just right for this deflection to become apparent. This would likely have a low probability.

3. The slab stiffness will greatly increase in the vicinity of columns. This additional stiffness will also resist upward deflection.

 

[hokie66]

Where are the drains on the balcony? If the drains go into vertical pipes, as in a lot of cases, sloping toward the building is good. You need more than 1" step at the doors for watertightness.

 

[dik]

Most balconies I've encountered, the slope is away to spill off... even a drip edge is often created. Taking to drain pipes in these environs does not work well in winter.

So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[FE_struct1]

Late to the party again, but my 0.02$ here:

It might be good to add some numbers to the discussion. General equations for an overhanging beam (very simplified) can be found here:
[URL unfurl="true"]https://structx.com/Beam_Formulas_024.html[/url]

The overhang will hog when that term inside the bracket goes negative. Doing some very rough and ready checks - this happens for any value of a/L (a = balcony span, L = backspan span) less than 0.43

I'll note that the assessment above is very simplified and doesn't account for rigidity of the supports, nor for variability in loads. Consequently, you may not see upward hogs in practice even if the ratios are less than 0.43

Furthermore, even when the overhang does hog, it may not be an issue if the actual deflection is tiny (this depends on the loads, the stiffness of members, construction sequencing, etc. etc.).

As @MotorCity noted earlier - for slabs supported on beams, the beam itself may be too stiff to allow much hog. However, if we're dealing with flat slabs without drop panels and with columns joints that minimise moment transfer into the columns (e.g. precast column to slab joints), there's a chance this assessment starts to take hold.

Either way, me thinks this a good rule of thumb for when this phenomenon may need to be looked at further with a microscope.

 

[hokie66]

dik,

Point taken. Should have asked the OP where he is located. Where I am, freezing is not a concern, and water running over the edge is frowned on.

 

[dik]

point taken, too... thanks Hokie...

So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[dik]

fe... is that website ever handy...

So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[Brad805]

Very nice link FE. How are their excel sheets?

 

[FE_struct1]

Very nice link FE. How are their excel sheets?

Thanks Brad. I've never used their sheets so can't comment sadly. They do seem to have a good variety but I reckon IDS's sheets will give most of them a run for their money.

fe... is that website ever handy...

Very much so for people like me who don't like flipping through book pages for the same info haha. I sometimes use it for quick and dirty checks when interrogating FEA sub models. Also a good resource to help create personal calc sheets.

 

[hokie66]

Be careful of assuming that concrete behaves like steel, because it doesn't. Even when all the variables are factored in, sometimes disappointment ensues.

 

[dik]

You have room for a book on your desk!

So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[MikeManning]

@FE_struct1

Thanks for this.

The scenarios that we see are columns and flat slabs (consistent 8" thick, no beams or thickened areas in the slab, and no 1" drop at the balcony, which would be nice I admit), most often with spans of 20' between columns in either direction, and balconies that cantilever about 5-7' beyond the outermost column line. So an a/L of about 0.35

The columns and slabs are all cast-in-place, and sequencing is something like this:
- Form and pour the slab on form-tables
- Form and pour the columns
- Drop the slab form tables and install re-shoring
- Repeat on the next story
- Eventually, remove the re-shoring
- Then complete the construction of envelope (usually slab-to-slab window wall, at the outermost column line), interior walls, finishes, furnishings, etc

We always notice deflection once the shoring is removed, and we usually use floor-leveler on inner parts of the building after building walls but before putting down flooring, because there's always depressions in the slab between column lines.

Subsequently, more load is added again during occupancy.

We do our best to waterproof, but we only build to what's designed, which is often either no curb, or maybe a wood curb, at the sill of the exterior wall. There's usually drains in the balconies, with balcony slabs sloped to drain towards them - but once you factor in the standard construction tolerance (1/4" over 10'), and since the entire slab is the same thickness, it's common for there to be some areas that aren't sloped quite as much as they "should" be.

So I'm certainly not saying that Hogging is 100% responsible for the leaks. Simply, I think there are a number of factors, and it sounds like, in some cases, hogging might be one of them.

In a nutshell, here are the recommendations I'm putting together for balcony slabs, which we will then use as "guidelines" to share with our owners/designers:
-

Use curb at the sill of the exterior wall, preferably an integral concrete curb, or else a step down in the surface of the slab at the balcony​

-

Specify an adequate slope in the balcony slab which will allow for minor deviations due to construction tolerances and due to hogging deflection​

-

Apply a waterproofing membrane to the slab, tied into the floor drain​

-

Select a membrane tie-in detail at the base of the wall which will be compatible with the balcony waterproofing and wall waterproofing, and which will allow positive overlaps​

-

EDIT TO ADD - chamfer the slab edge, to allow water to roll off (unchamfered edges sometimes have slightly upturned edges)​