[Beowulf] Re: removing tiles around heavy racks? (David Kewley)
mathog at mendel.bio.caltech.edu
Wed Jan 4 15:44:56 PST 2006
David Kewley wrote:
> Here's how I think about it. This is after talking to a local scientist
> who's also a consulting seismic engineer and a user of my cluster.
> Are the stringers anchored to a load-bearing (concrete) wall? If so,
> pure compression force into the wall should be OK. The stringers will
> transmit the force to the wall, and the wall in turn to the ground,
> which is what the feet are anchored to. So you won't tend to shear the
> floor off the feet.
The caveat being that construction, at least here, tends to put
a sheetrock interior wall in front of the concrete load bearing wall,
separated by the width of a stud. In which case your floor may be
anchored to the studs of the sheetrock wall. These may or may not be
anchored firmly to the concrete on the other side. They absolutely will
be anchored at the bottom and top of each vertical stud. So anchoring
a floor to the intervening wall will be reasonably strong under
compression but tension could be iffy as the studs could be bent
outwards if they are not anchored midway up to the concrete wall.
Also if that happens when the floor snaps back the other way the
forces are going to be extreme when the stud flexes back against the
concrete and stops the motion of the floor in .1 mm. That is
definitely going to bend metal somewhere.
> The first failure mode under such a load would probably be stringer
> buckling, and that'd take a lot of force. (Assuming your stringers are
> something like mine: basically steel C-channels, about 3/4 inch square
> cross-section, 2 feet long). Yes, leaving the tiles in will greatly
> reduce the risk of buckling, but for normal forces, buckling is pretty
> unlikely anyway, I'd think.
I tend to agree but only if the racks are firmly attached to the floor
(and ideally, to another structural frame at the top). There
can be large accelerations both horizontally and vertically and under
the worst case scenario (poor attachment of the equipment) that
might cause racks to hop up and down (force wise) and
pound the floor they are standing on. If the rack is free to bounce
around then the forces can be severe, for instance, if it manages to
get up on one edge and then come down that way, or it walks over a tile
edge. Hmm, the feet on the raised floor are bolted onto the concrete
below, right? Else they can jump around too.
All I could arrange for my lab was a top brace constructed of double
unistrut that runs the length of the room and is bolted to two opposing
walls and to a third via a "T" also composed of doubled unistrut. The
racks are bolted as firmly as possible to that. They just stand on the
concrete floor though and could slip. (A risk somewhat lessened by their
feet having dug holes in the vinyl tiles over that concrete.)
In a huge earthquake the unistrut could punch through or buckle
the two freestanding walls. Of course a quake that severe will
probably also rip loose the AC suspended in the ceiling above, breaking
the chilled water pipes and generally causing total destruction.
> If your stringers are *not* trapped by or attached to walls, then you'll
> be more subject to toppling the raised floor catastrophically, although
> I'd imagine a properly installed floor would be highly resistant to
> human-scale forces.
I've never seen a free standing raised floor like that, they always
go all the way to the walls. I think primarily so that equipment
isn't accidentally rolled off the edge, which would certainly be
catastrophic for anybody unfortunate enough to be underneath it when
> You know, of course, that a toppling full rack can easily
> kill someone? At least, that's the wisdom I've received,
> and it totally makes sense to me.
Ditto for soda vending machines. A couple of people die each year
when they try to shake loose a stuck can and the whole thing topples
over and squishes them.
> Right now, only the tops of our racks are braced, and in a sufficiently
> large earthquake, the bottoms could pendulum out from under the braced
> tops and the racks crash to the ground. We have probably 40-50k pounds
> (mass) of joined racks. In a .5G earthquake, that means 20-25k pounds
> (force) transmitted to the walls.
It's a bit more complicated than that. What you really have is
something resembling a funky two story building, with the raised
floor being its first floor and the racks being its second floor.
Multistory buildings flex at each floor and they have resonant
frequencies that cause them to fail in earthquakes that have ground
motion frequencies that match. The worst mode would be where that
resonant frequency drives the bottoms of the racks across the narrow
dimension in one direction and the tops in the other, which would
maximize the lateral force on the floor. I'm not explaining that well,
but like this:
floor ++++++ --> a lot of force
> Yes, I'm a bit nervous about it. Plus it'd be mighty embarassing for a
> supercomputer owned (in part) by a world-famous seismology department
> to fall over in an earthquake...
They're braced at the top so they probably won't fall over. However
in a really severe earthquake the forces might break the floor enough
so that it could collapse. That's a very big quake though, big enough
that you have to start worrying about the welds which hold the building
If Lucy Jones is giving post quake interviews in front of the
rubble pile that was South Mudd embarassment will be the least of
mathog at caltech.edu
Manager, Sequence Analysis Facility, Biology Division, Caltech
More information about the Beowulf