ornadic shelter design loads

huangyhg

tornadic shelter design loads
i am currently researching design loads to withstand a tornadic event for an upcoming project.  i have found the fema 361 document that contains a bulk of the information i have been trying to locate.
however, one issue i am having difficulty in finding information deals with impact load due to a collapsing structure.  my structure is a two story building, but the first level will only be designed as the shelter in an effort to save money.  this assumption would indicate that the upper level is not designed to withstand the extreme tornadic loads, thus giving the potential for collapse of the second level onto the hardened structure below.
does anyone have any recommendations or references on how to account for the impact loading created by the collapsing structure above?  i am sure this involves impulse-momentum theory to a certain extent, but was hoping to find something simple so i do not have to relearn kinetics.
my guess is that if the first floor can survive a tornado - then a relatively light second floor falling on it won't do much damage.
design the second floor to blow away and you won't have to worry!!  probably what will happen anyway....
toto - we aren't in kansas anymore!!
my initial thought was to design a light second floor, but the owner requests cmu or concrete walls with a brick veneer.  while the roof is light (bar joists and metal deck), i'm concerned with the possibility of the cmu/concrete wall collapsing onto the floor.  this would be ~100 psf plus whatever effect for impact.  would an impact factor of 2 be reasonable?
why not design the second floor cmu to survive a 210 mph wind force?  simple span between the floor and the joists/roof with pressure at 100 psf (+/-).
one objective is the cost...the idea behind hardening the first level only was to lighten the roof structure and prevent the use of ballistic resistant windows at the second level.  while hardening the entire structure would lighten the second level structure (reduce floor live load from the tornadic prescribed 100 psf to office live loading), it would require a much heavier roof structure to combat the tornadic prescribed live load of 100 psf in addition to the roof wind pressures of (+100 psf/-200 psf).  for missile resistance, the roof also needs to be a 6" minimum thick slab.
let the roof fail, just design the walls/joists for horizontal loads from tornado winds.  this negates the walls becoming impact loads on the first level ceiling.
if the roof fails, then i would lose the diaphragm action that supports the cmu walls laterally (out-of-plane).  then the walls would effectively become 16' high cantilevered parapet walls that would have to resist ~175 psf wind pressure.  this is something to consider, but would add significant cost to the wall reinforcing.
since you are costructing the home in cmu, i would think just designing the entire structure as partially enclosed for the wind load, and move the harded concrete roof to the top of the 2nd floor (or steel trusses and roofing) would not be much difference in cost.
design a light roof system with the minimum possible connections to cmu wall.  reinforce the second floor masonry wall with horizontal and vertical reinforcing (not every cell necessarily).  increase the vertical reinforcing as you move from the roof to second floor so that when it fails it will just be a mangled mess but not actually laying on the second floor.  not sure how to put numbers to this just an idea.
the real problem sounds like the owner/architect has placed too many constraints on you to be able to arrive at a reasonable solution.  if the owner justs wants to see cmu on the interior maybe you could use light framing with a thin adhered interior that resembles cmu.
good luck
j  
why not just follow the fema standards for safe areas. you are concerned with a tornado and not just a hurricane. look to fema 320. your client may feel this is a good option as opposed to total protection from the very, very rarely hit of an f5.
the impact of the collapsing upper floor and still being in the same county during a storm is a very nebulous situation.
is this a residence or a commercial structure? if you are attempting to design the first floor and everything that can happen to the second floor, complicated by the usual overly conservative, simplying design assumptions, you can easily create a monster. - a tornado covers a very narrow path and f3 to f5 storms are not that common on a probability scale for a single location.
mass and reinforced conrete masonry/concrete are obviously your best options.
the fema suggestions (life safety - not property) can easily be adapted to below ground (basement) or to first floor situations. frequently, a large walk-in closet (8x10) can serve the purposes and an adjoining bath can easily/economically be an added safe area.
trying to design for an f5, especially with the simplifying assumptions is just a folly, since the economics and constructability pose huge problem.
i saw a home near long beach/biloxi, mississipi that withstood the 29' direct hit of storm surge from katrina (lesser winds, but storm surge). it looked just fine 4 months later. it was on over 1200' ($10,000/ft) of shorefront, so that owner could afford to pay for the design and constructon. - i hope you have that type of client.
the project is for a utility company.  the floor to be hardenend contains their control room and adjacent electrical/hvac rooms to keep the control room running.  even without the hardened structure request, they are requesting this to be classified as an essential facility, which would dictate an importance factor of 1.15 under normal wind loading (90 mph - 3 second gust).  the upper/"disposable" level contains management offices, conference rooms, etc.