PEMB net uplift and thickened slab footings

[UcfSE]

Another thread about pre-engineered metal buildings (pemb) brings to mind some questions I've had about designing footings for PEMB for high net uplift. The footings to which I refer are the thickened-slab type where the edge of the building - and column - is at the edge of the footing. Since the net uplift force does not coincide with the center of area of the footing mass, there should be a net moment on the footing, but where do you assume the point of rotation is located? It seems to me that you could assume the point of rotation, for purposes of making sure moment equilibrium is satisfied, is at the lower corner of the thickened slab but that puts a lot of moment into the floor slab as well. Others in my office do not consider this, they only sum forces in the vertical but do not account for the moment arm between the uplift and resisting forces. That to me seems like we would be leaving something out. How would you properly handle this situation?

 

[haynewp]

Are you tying in the turndowns at each side to the thickened slab footing? Or are you just using the area of the footing itself in your calculation?

 

[DaveAtkins]

UcfSE,

You are exactly right, and your colleagues are ignoring a very real moment. I design the floor slab for the bending moment induced by the eccentricity between the uplift and the weight of the thickened slab. It is critical where the floor slab "thins down" to its typical interior thickness. Or I don't make localized footings at each column, but design a continuous thickened slab along the edge of the building (you still have to take the moment due to eccentricity into the interior slab, but you can spread it across the entire bay length).

DaveAtkins

 

[haynewp]

I think you should also be able to add in the weight of the turndown in resisting the moment and uplift if it is adequately tied in. I have done this before.

 

[UcfSE]

I do tie in the turn downs on each side and will typically take a length of turn down of about 2B (B of the smaller width of perimeter footing) on each side to help with the uplift. I'm just not comfortable designing the footing with eccentric load as I would one that is concentric. I haven't found literature on this and the rest of my office does not agree with me.

So what I'm looking at is designing the footing for that net moment but I am not sure where to define the point of rotation. Bascially it comes out to the footing having to be, say, twice as large as a concentric footing because the uplift force has more lever arm than the resisting weight. I want to go with the inside edge of the footing but that would put moment into the floor slab also.

|---typical perimeter ftg
v
--------------------------------------
+ column ^
----- -------- | -------
| | | lever arm
| | |
| + ftg cg| v
| |
| |
---------------* point or line of rotation

 

[haynewp]

So you have the weight of the footing with its lever arm, the weight of the turndown with its lever arm, and the weight of the slab adjacent to each side and connected to the footing with its lever arm (not shown in your diagram above).

I would sum moments about the interior edge of the footing with all that resistance (where you have your point indicated), and try not to use the interior slab where the thckened footing begins to resist any moment (zero moment at this point). Be aware of control/construction joints when including the slab in the resistance. This all should give you a pretty good bit of resistance, and then increase the footing as required. I assume you don't have tie beams since you haven't mentioned them.

This is just the way I would look at doing it. If all this still gives you a very large footing, I would then look at using typical concentrically loaded footings.

 

[SacreBleu]

UCFse,
To minimize the footing volume required, make it rectangular "long ways" to reduce the lever arm.