ICC 500 - Host Building Collapse Load?

[etimm]

Anyone have experience in designing storm shelters for host building collapse? I have an ICC 500 storm shelter that is below 2 stories of its host building. Curious to see what kind of impact loading others have used or would recommend for host building collapse. Any advice is welcome.

Some more information about the host building above the shelter...The floor above weighs around 200 psf (including partitions) and is 14ft above the shelter. The roof above weighs around 25psf and is 40ft above the shelter. Assuming the host building collapses in a free fall directly onto the shelter and a 1/4sec impact duration...I am coming up with a 900psf impact load. Does this seem reasonable?

 

[phamENG]

Do you have access to Blodgett's Design of Welded Structures? Chapter 2 has a really good section on analysis for impact. The method proposed looks at the potential energy of the object and the stiffness of the loaded structure (rather than trying to approximate impact time which is a bit of a crap shoot). Even if your shelter is concrete, the basic principles would still hold up.

 

[JLNJ]

I would think the punching shear from a big girder would be the controlling factor, not simply a “psf” number. I’m not sure that the ICC really covers this.

 

[SlideRuleEra]

...I am coming up with a 900psf impact load. Does this seem reasonable?

Using the given weights, distances, and impact duration I get impact loading of 746 psf to 839 psf (depending on what happens to the roof when the floor impacts the shelter). Those values assume constant deceleration during the 1/4 second impact duration. IMHO, that is unrealistic, I would double those values to allow for an initial deceleration "spike" followed by ramped downward deceleration for 1/4 second. This makes the range for impact loading roughly 1500 psf to 1700 psf.

I assume the host building floor/roof "shatter" when they impact the shelter making this calculation very tricky and highly dependent on assumptions used.

http://www.SlideRuleEra.net

 

[etimm]

phamENG, I agree approx the impact time is a crap shoot. I have spent some time looking through section 2.8 in Blodgett's Design of Welded Structures. It looks like a great reference, thank you. In looking at the stiffness of the shelter I think need to look at the shelter elements separately (ie girder vs column). The shelter is ~40ft x ~70ft, structural steel framing, exterior load brg walls, and an interior girder line with columns at 20ft spacing. With the proposed approach the impact at the middle of a girder will be less than impact above a column or wall, that makes sense. A blanket load wouldnt capture that. What do you think?

 

[phamENG]

Sounds about right. JLNJ brings up a good point. Rather than looking at the 200 psf weight above, pick the biggest piece of debris. Whether it's a column segment, a girder, or something like that. Then assume it lands like a spear on one of your columns, just off of a column on a beam (maximize shear in the connection), and at mid-span. Determine the impact loading using Blodgett's formulas, and apply an additional uniform static load to each of the cases. Unless there is a plausible failure mechanism in the building above that would result in a sudden and essentially uniform pancake type failure (I haven't come across one in a steel frame before, but I suppose it could happen), it's probably a little unreasonable to use the entire mass above in your dynamic considerations. It's more likely to come down piecemeal. They'll be hitting quickly, but from the little I've been able to explore in dynamics the the seemingly insignificant time shifts can really change the results.

 

[bones206]

I did this recently where my host building had precast double tee roof members. I took the approach that the double tee is a free falling missile and made sure the shelter roof could take the punching shear. Then I checked the bearing walls for a load corresponding to the punching shear capacity of the roof. There are so many assumptions in this kind of analysis, I just went with what my gut said would make a survivable shelter, then ran it through that test scenario as a verification of sorts.

I think it's the best you can do, unless you have access to software like this:

https://www.extremeloading.com/els-applications/progressive-collapse-analysis-software/