Precast Concrete Storm Shelter 1

[archeng59]

I have designed several storm shelters for schools and commercial buildings. Most are either solid grouted CMU or cast-in-place concrete walls with either cast-in-place concrete lids or steel beam & slab-on-metal deck lids. Several years ago, I was asked to be the EOR for a precast concrete storm shelter that would also be used as a gymnasium. After much teeth gnashing on my part about getting connections adequately robust between walls and double tees, I finally convinced myself the local precaster's designs were adequate. The double tees were set in pockets in the walls on that particular project and on a couple more that the same precaster fabricated. I am working on my fourth precast concrete shelter, but this one is smaller than the others. Double tee spans are about 45 ft versus 75-ish feet for the others. I detailed pockets in the walls for double tees, but the precast submittal is showing corbels on the face of the walls to support the double tees. The precaster's calcs show adequate strength, but I am not certain I like the corbel idea for a storm shelter. It seems inherently to me that it is another joint/interface for failure under loads imposed by high winds that a wall pocket would eliminate. Has anyone been the EOR for a precast concrete storm shelter? How were the double tees attached to the walls?

 

[BridgeSmith]

I would be concerned with a few things:

1) Is there adequate connection between the end of the double tee and the wall to ensure the double tees stay fully on the corbel?

Is there enough reinforced corbel width to support the stems of the tees? In dealing with older bridges, approach slabs shifting far enough out on the corbels that the corners outside the reinforcing on the corbels and slabs shear off, letting the approach slabs drop, has been a common issue. Be sure to check the bearing length accounting for the concrete cover, including the bend radius of the corbel stirrups. I would draw a cross section of the corbel and double tee stem in all of its glorious detail, and then draw in the shear failure surfaces of the corbel and the bottom of the stems. See how much bearing area you have left

2)Is there adequate connection between the double tee sections and bracing along the sides parallel to the double tees to ensure stability against overturning (rolling over)? To me, that would seem to be the major difference between the two configurations.

I think your detail with tee sections extended to bear directly on the wall is far superior. Another option is a double tee with an integrated backwall/end diaphragm. That eliminates having to form the pockets in the wall.

 

[TehMightyEngineer]

Usually I'm all over precast concrete questions like this but I fully admit I've never done a precast concrete storm shelter (would love to tackle one, though).

As a precaster I feel that there's nothing inherently wrong with a corbel detail but it definitely does leave room for HotRod's questions. Do you have a detail showing the proposed corbel connection?

Ian Riley, PE, SE
Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

https://www.facebook.com/AmericanConcrete/

 

[dik]

The devil's in the details. Make sure your connections are 'extra strong', with ductility and it should work OK.

Dik

 

[archeng59]

A copy of the detail is attached. I do not believe roll-over is an issue based on the embeds and connections between the dbl-tee and wall shown in the detail. The corbel is added after the wall is cast, so in my opinion the reinforcing is the only thing supporting the dbl-tee loads. At least with the pocket, the dbl-tee is bearing directly onto the wall. The corbel concept for a storm shelter appears to be a collapse waiting to happen if the corbel fails for any reason.

 

[archeng59]

Ian, do you have any knowledge of a precast concrete shelter surviving a tornado? I have several CMU and CIP shelters in place. One of each have been through a tornado and performed well without damage to the shelters. I've asked the precaster for the first 3 precast shelters if they have any info about precast shelters performance during a tornado. They changed the subject and never answered or responded to emails/phone calls about the topic. Makes me feel like a test pilot flying a 747 full of passengers on its maiden flight.

 

[dik]

Can you move the top anchorage up and use a clip angle attachment for a little ductility?

Dik

 

[bones206]

Just took a redvector course by Bill Colbourne on tornado design. He had photos of a collapsed gym with precast walls. He might be a good person to consult with.

 

[archeng59]

This is the detail showing the dbl-tee in a wall pocket.

 

[archeng59]

Bones, was the gym also a storm shelter?

 

[bones206]

I believe he said it was not being used as a storm shelter, but did satisfy all building code requirements. Not sure if it was designed for a tornado or not. Think it was the Joplin, MO middle school gym.

 

[BridgeSmith]

Looking at the detail, I am not as concerned about the bearing issue, since there are steel plates on both the corbel and stem. After seeing the detail, I have a new concern - development of the reinforcing, specifically the sx - 4 bars in the corbel. I hope those are continuous out into the corbel, around the headed stud, and back into the wall. Otherwise, the their tension capacity would be suspect.

Did the precaster include interface shear calcs for the wall - corbel interface?

 

[bones206]

For a tornado shelter, I would detail the connections similar to connections subject to blast loading (i.e. high ductility and capacity). Design non-ductile elements like welds for the nominal capacity of the ductile elements times some factor like 1.25 or 1.5.

This is a good video showing the moment of tornado loading on a gym.
Link

 

[TehMightyEngineer]

Ian, do you have any knowledge of a precast concrete shelter surviving a tornado?

Unfortunately not; though precast parking garage structures use similar corbel end details for double-tee beams and they've definitely gone through a tornado or three. Though they're open structures so will react differently to wind pressures.

Makes me feel like a test pilot flying a 747 full of passengers on its maiden flight.

Well, it's not entirely unknown territory. Perhaps a better analogy is you're the first 747 to land at a new airport.

Ian Riley, PE, SE
Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

https://www.facebook.com/AmericanConcrete/

 

[birneys]

Archeng59, I designed several precast gyms (elementary school size) to the FEMA 361 structural requirements. We used a similar corbel detail as you've shown and went through the FEMA funding/certification process for one of the projects. I don't know how in-depth they reviewed it structurally, but FEMA didn't raise any red flags about the corbel detail (for what it's worth). HotRod10's questions and concerns are all very good things to doublecheck. One other thing I'm wondering though: Is your wall panel solid precast or is it an insulated wall panel? I've run into issues in the past with the architect's desire to have continuous insulation as opposed to having enough solid concrete for the connections.

One other thing about precast shelters: The latest version of the ICC-500 (and FEMA 361) has a requirement about opening protection that is so conservative that your precast wall panel joints (~3/4" wide) are too big . . . Anything wider than 3/8" has to be protected. In order to get FEMA funding, we had to add steel plates to all of the wall panel joints.

As much as I can, I've tried to influence the clients and architects that I work with to provide storm shelters in the classrooms instead of gyms. It's obviously much easier to design the smaller spaces, and, in my opinion, you have a better chance of kids making it to the shelter areas if they're spread out across the school.

 

[archeng59]

Thanks, birneys for the input. The precaster told me the corbel is cast integral with the wall, which is not how it appeared on the detail. I have requested that 2 horiz bars be provided at the hooks shown in the detail adjacent the corbel in addition to the 6x6 wire mesh. The only deformed bars they are providing are verticals at 12" o.c. centered in the wall. I am aware of the joint issue. And, the panels are solid.

 

[BridgeSmith]

Are the bars from the wall into the corbel continuous, or are they hooks in the corbel? If they're just hooks, be sure they checked the hook development length to the critical section at the face of the wall.

 

[birneys]

Archeng59, I only mentioned the joint issue because it took me by surprise. Can I ask how you handle protecting those joints? I'm always interested in hearing how other engineers deal with small details like this. Thanks.