Soil Suction Forces 1

[spats]

I ask this question at the risk of sounding stupid, but here goes. I'm designing metal building foundations for uplift. Unfortunately, I'm dealing with a contractor that insists on thickened slab/haunched footings, or he "will get his other engineer to do it". These types of foundations are always problematic because exterior columns are always well off-center from the haunch, and you wind up with significant overturning moments. I don't want to come back to the contractor showing huge chunks of concrete required, and am looking for other mechanisms to resist the uplift and overturning forces.

It seems to me that there has to be a significant amount of suction developed between the slab and the soil as the slab tries to uplift. I've tried to google the subject, but I'm not having much luck. In the same way as you can place a piece of newspaper over a ruler overhanging the edge of a table, and break the ruler with a downward blow, "soil suction" has to be able to resist a tremendous amount of uplift. A column of air exerts about 14.7 psi of pressure on the top of the slab. It seems that if you had less air pressure exerted below the slab due to the buried/confined soil, you could develop a significant amount of uplift resistance.

Any thoughts or technical information on the subject would be greatly appreciated.

 

[spats]

Sorry! I got testy.

 

[IDS]

I won't comment further because I don't design industrial buildings.

This link is worth a read though:

http://www.structuremag.org/?p=342

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[nonplussed]

I'm not sure about any numbers on this but I know that one of my colleagues, a principle engineer who has been designing thousands of sheds for decades around Australia and had no failures, uses soil suction under shed slabs for wind loading. He says that for gusts of wind there would be considerable suction under the slab, especially for older slabs where the soil has had time to settle and air voids under the slab dissipate. He also says that in all his time designing sheds he has never seen a failure where the slab/footing has "blown across the field like a tumbleweed" as he likes to say.

Other things he does for shed designs is not assume that the maximum wind pressure acts upon all surfaces at the same time. He assumes the wind is very turbulent and, for large areas, does not hit everywhere with maximum force at exactly the same time therefore allowing him to further reduce total uplift (he did his masters project on this and actually tested wind speeds instantaneously across a perpendicular plane 50 metres long).

For cyclones/hurricanes he also reduces the air density designated in the code thus reducing wind pressure and uplift, since cyclones/hurricanes are caused by a low pressure centre and the wind velocity around the centre of the cyclone/hurricane is inversely proportional to the air density. This one can easily be checked by looking at the local meteorological website where they keep pressure and velocity data for cyclones/hurricanes.

Of course, his competitors say that, out of the thousands of sheds he's designed which have been through many cyclones and survived, all of the sheds just didn't undergo "design loading". It's always easier to justify introducing assumptions which make the structure heavier and more costly, than to take risks and introduce assumptions which increase economy and efficiency.

 

[canwesteng]

I've done these types of foundations and used the whole slab to resist uplift. Typically have top and bottom reinforcement in the slab. My preference is usually spread footings and slab on grade but sometimes the client prefers thickened edge mat foundations

 

[spats]

Thanks IDS for the link. I actually have Alexander Newman's book "Foundation and Anchor Design Guide for Metal Building Systems". He covers the slab-with-haunch situation, but I think his approach is overly conservative. He argues that the slab should not be used as part of the foundation. "When the slab is present, some people seem to assume that the slab helps in carrying the load. It what fashion can that possibly take place? The point of transition between the haunch and the slab will most likely develop a crack, as the haunch becomes a point of restraint for thermal expansion and contraction of the slab. Once the crack forms, the continuity disappears and the joint will behave at best as a hinge." In his worked example, he has to count on the resistance of a substantial turned-down edge of slab, a whole bay's worth, to make the foundation work for a modest wind uplift load of 7.7 kips.

Nice story nonplussed! I wish there was some more technical info to go along with that. I like the idea of the lower air pressure in a hurricane, but it doesn't seem like it would make a lot of difference, and I'd have to have some serious technical backup.

Canwesteng - I too am using a chunk of slab around the haunch to increase resistance. I generally go all the way to the slab joints. An example slab area would be 14'-3" wide x 9'-6" into the building. I make the haunch rectangular, with the long direction parallel to the slab edge, to concentrate the foundation weight as close to the slab edge as possible. Regarding top and bottom slab reinforcing: I assume you're talking about just in the area of the haunch and the design slab area around it?

 

[canwesteng]

Well for small buildings (20-30' bay widths) I've just reinforced the whole slab. Is the centroid of your slab area now offset from the centroid of the column by a significant distance?

 

[MikeE55]

Spats, I have calculated this design many times, and it sounds like you probably have too. I count on a certain amount of the slab working with my footing - usually at least a 3' cantilever from the footing, but maybe more, depending on how the slab is reinforced. I also count on a certain amount of the turndown slab on each side of the footing. It doesn't weight a lot, but hey, every little bit counts when you are looking for mass. I will thicken the slab up around my hairpin bar (the bar wrapped around the anchor bolts to resist horizontal load), and I'll count on that also. I don't count on suction, but if I did, I would use about 100 psf on the perimeter of the footing below the gravel base. I don't know if it is suction or friction, but I do know that force is real for the clay type soil we have here. I have tried to use my tractor with a front end loader bucket to lift up embedded concrete. I think the amount of resistance is much greater than 100 psf. But I think the advice you are getting here is good - I would just use an increased mass of concrete, especially in a sandy soil.

 

[SlideRuleEra]

Spats - Here is documentation for Nonplussed's accurate observation about reduced air density, the basic equation for wind pressure:

Pressure in the eye of a major Category 4 hurricane could be, say 27.5 inches of mercury. It would be even lower for a Category 5 storm. Atmospheric pressure in the maximum wind field, close to the eye, may not be as low as the eye, but should be similar. That pressure is only 91.7% of typical 30.0 In Hg sea level atmospheric pressure.

As usual, there is no free lunch, you will get reduced wind pressure... but you will pay for it with increased storm surge. The differential 2.5 In Hg translate to about 3 feet of water. Therefore the storm surge will be about 3 feet higher than wind speed alone would predict.

MikeE55 - I agree with you, if the ground is wet, as in a hurricane, there could be a certain (indeterminate) amount of suction. Simultaneously there would be friction. The uplift load is only a vertical component force trying to lift the slab. At the same time, wind pressure is providing a horizontal component force trying to slide the slab. I would expect these two components to interact, making the slab more resistant to uplift than expected.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[JStructsteel]

I ran into a similar problem with a contractor and footing for a CMU wall and metal building footings. I ended up not doing the work. Best choice ever. I noticed the other day the CMU wall was leaning out by at least 4 inches.

Sometimes, its best to pass on jobs, let someone else pay the attorney down the road.

 

[Joel900]

More than 20 kips of uplift is not an uncommon metal building reaction in my area. There is usually a similarly large thrust to accompany this uplift. If you only have 5 kips of dead load from the roof and frame, a very large volume of concrete is needed to resist the uplift. Do not forget that the weight of the concrete must be factored by the dead load factor.

Using 0.6D + 0.6W and assuming 20 kips uplift and 5 kips dead load, I get a net uplift acting on the foundation equal to 9 kips. Using the reduced weight of concrete (0.6*150 PCF), I calculate 100 cubic feet of concrete is needed or 3.7 cubic yards. Using a hairpin to resist large thrusts has always been questionable in my mind, a grade beam between frame columns with continuous reinforcing across the building is typically how I will resist the thrust and overturning. When countering the uplift, consider portions of the thickened perimeter slabs, grade beam, and triangular area of slab connecting perimeter thickened slab to grade beam as contributing to the total volume of concrete resiting the uplift.

 

[humanengr]

If you would like to use screw/helical piles to resist the uplift. You may wish to review AC-358 (ICC's acceptance criteria for helical piles).

What does the geotech. engineer have to say about soil uplift resistance or slab-soil suction resistance? Let me guess; the contractor said; "if you need a geotech. report, I'll find some other clueless "engineer" to design it."

I would think drilled piers would do the job, but you do need some geotech. data to design properly.

Regarding text books on Metal Building Foundations, the comparison is made with "conventional" buildings. Assuming you have the same required column spacing (column-free spans), what is the difference between a metal building and a "conventional building"?
Presumably a conventional building has hard walls, the weight of which may add to uplift resistance. Does that mean a metal building with CMU or tilt walls is "conventional"?
Presumably a conventional building has a "heavy" roof. Does that mean a metal building with deep joists and a heavy built-up or concrete roof is "conventional"?

In case you're wondering, I'm not a metal building engineer and have no relation to any metal building companies. I am just bothered by the twisted logic which, if taken seriously, can cause the client / owner an increased budget just to make the design easier.
Sure, you could use a braced frame design and roof diaphragm to transfer lateral loads to shear walls at the endwalls. Compare the cost and see if that works out?
By the way, that will eliminate some of the horizontal reactions, but it won't make the uplift forces go away.

 

[spats]

Joel900: you're missing the point about eccentricity and overturning. The centerline of the column anchor rods is typically 12"-14" from the edge of the slab. in order for the resistance to be concentric with the uplift, the "footing" would need to be 24" to 28" wide. At 24" wide, a 10' long footing would need to be 5' thick. This is not practical. When all is said and done, you usually wind up with an eccentricity on the order of 2' or more. I always try to make the footing as thick as reasonably possible to concentrate the resistance closer to the edge, but I don't consider a 2'x10'x5' thick "footing" to be practical. I'm in Florida, so I don't need a grade wall for frost protection.

 

[Joel900]

spats: To counter the overturning due to the eccentricity, I mentioned in my post above a grade beam could be designed that connects the two frame foundations across the building width. This grade beam can be sized and designed to counter the overturning and the thrust.