Designing my first FEMA tornado storm shelter 2

[Screwtape]

I'm in the process of designing my first public, FEMA approved tornado/storm shelter. It's in the southeast US and is to be designed using 250 MPH design wind loads. I've run the wind pressures and I nearly fell out of my chair when I saw the loads. I'm currently reading through the FEMA and ICC codes for storm shelters.

The building is approximately 70' long x 30' wide x 14' tall. I was thinking along the lines of 12" CMU reinforced each cell and grouted solid, with closely spaced wide-flange beams at roof level topped with steel deck and concrete. I've attached a working cross section showing the preliminary foundation, wall, and roof framing elements.

The only other issue is the left side of the storm shelter is alongside an existing CMU wall (old Walmart building). Due to the existing footing, the new foundation will be shallow and bear directly on top of the existing foundation. I was thinking about using helical piers for the gravity and uplift loads, as I don't want to overload or undermine the existing footing.

Does all this seem reasonable or should I go in a different direction?

 

[dianium500]

12" CMU walls w/ dbl reinforcing is likely the route I'd use. I've always used a CIP tie beam for my South Florida projects where wind exceeds 160 mph. I would use Helicals as well. I definitely think you are on the right track. I would not be surprised by the loads at 250 mph. I would also check that the footer doesn't need a top layer of steel.

 

[rdubs]

What are the envelope requirements - will the space be conditioned and used for other purposes?

 

[JAE]

Don't forget that per ICC 500 the wind loads based on the 250 mph are considered ultimate (i.e. already factored) loads and you should be using 1.0W for strength design and 0.6W for ASD.

Also use these values (per ICC 500)
Kd = 1.0
Kt = 1.0
I = 1.0
Exposure C
Roof live load = 100 psf

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[Screwtape]

Yes, it will be conditioned.

I'm not sure if it will be used for something other than a shelter. The shelter will be owned by a University (they bought the old Walmart building and converted it for their use). Based on that, I would assume the possibility of it used infrequently for something else. I'm not sure they could put much if any furniture in there, since it would cut down on the max number of users during a storm event.

 

[Screwtape]

Thanks, JAE...I'm at least that far along.

In addition, to the values listed, I'm also using GCpi of 0.55 +/- as recommended by FEMA. Although, I don't really see that, since it's just got 2 small doors (one each end) and a total of 4 - 4'-0" x 4'-0" windows.

 

[mike20793]

I design quite a bit of tornado shelters. The partially enclosed internal coefficients are for atmospheric pressure change. If the mechanical engineer provides a method for allowing atmospheric pressure change, then you can use 0.18. If they don't, which is usually the case since they need to be turned on during a storm, you use 0.55. You will often see precast roofs or steel and composite deck. You want as much dead load as possible to help with uplift. Also, make sure you have an adequate uplift load path, and if you are designing per the new ICC 500, you will need a structural peer review.

Edit: You will also need to make sure the design meets the impact missile criteria. You can find a list of tested assemblies from Texas Tech's website. Also, watch out for the jambs. They get killed when you have side by side openings or if there is a nearby control joint.

 

[JAE]

Screwtape,
A few thoughts on your sketch:

1. The CMU perimeter wall needs to extend up fully to your deck. It appears in your sketch that there is no wall shown within the vertical depth of your roof beams. I presume you had that in mind anyway. Just thought I'd point it out. Usually we'd provide a beam pocket on the inside of the wall for the beam and still have adequate CMU (fully grouted) outside the end of the beam to still provide for projectile protection (i.e. don't allow a projectile to penetrate a think beam pocket outer face shell and get into the space between the beam flanges).

2. The wall needs to be anchored to the deck slab.

3. The deck slab needs to be able to resist C&C tornadic wind uplift pressures without a) the concrete pulling off the deck and b) the deck-slab assembly pulling off the beams. This can be accomplished by using headed studs welded to the beam through the deck and into the concrete.

4. Ensure that the beam is adequately anchored to the wall to resist the full uplift of the beam. (use the 0.6D + 0.6W combo)




Check out Eng-Tips Forum's Policies here:
faq731-376

 

[EZBuilding]

Screwtape,

I expect you will have a very significant uplift load at the ends of your shear wall and I would be surprised if you could account for it based upon self weight of building + typical foundation self weight and overburden. At the very minimum I would expect a thicker and wider foundation than you are showing with top reinforcement. You may need piles even away from the Walmart building wall.

 

[rdubs]

Based on JAEs detailing comments and the possibility that the space might need to be conditioned, I would consider using insulated concrete forms for the walls and the roof. This would provide the strength you need and would eliminate most of the detailing concerns raised by JAE since the concrete joist roof would be monolithic with the walls. Also provides an insulated envelope without extra effort.

 

[JGard1985]

Agree with mike20793

At Nuclear Power Plants, many steel-framed structures have dampers which open to equalize pressure between the sub-atmospheric outdoors and the interior space. This reduces uplift loading on components and cladding (roof, leeward walls) during a tornado.

In regards to missiles, I have constructed reinforced concrete and steel barrier walls and structures to protect openings from wind-borne missiles. Providing missile protection to doors and openings reduces the need for missile qualification of the doors, and louvers themselves.




Jeff
Pipe Stress/Structural Engineer

http://www.xceed-eng.com