I just received Metrication Matters 93, and saw the aircraft hanger example
again. As it is snow season in the US, and people need to worry about their
roofs, I have to point out two huge errors in the example as I don't believe
anyone should rely on that example.
80 cm of snow != 8 mm of rain
Even if the 10% rule were true, it would imply 80 cm of snow is 80 mm of rain.
Doing some Googling on snow load and roof designs, I find the density of wet,
heavy snow is more like 32-33% water density, 320 - 330 kg/m³. The Washington
(DC) area is not noted for light, fluffy powder, and light fluffy powder isn't
what collapses roofs. Using the 320 kg/m³ x 0.8 m, the actual roof load was
more like 256 kg/m³ if the drainage system was still working, not the 8 kg/m²
of the worked example.
Flat roofs are a particular problem as snow tends to clog drainage and then you
get slush, a mixture of ice and water. Not surprisingly, the density of slush
lies between 920 kg/m³ (ice) and 1000 kg/m³ (water). The figure above of 320
kg/m³ is for drained (but wet) snow - imagine snow on a screen so any water
melt can drip out.
I can't find the spec for flat roofs, a lot of local codes in the northern US
are 35-40 lb/ft² for sloped roofs. That converts to 170 kg/m². Some extreme
snow areas are higher, and I would expect flat roofs to be higher.
If anybody is worried about their roofs, please use at least 320 kg/m³ for wet
snow on sloped roofs. If you have a flat roof, determine whether the drain is
working, it will be the difference between 320 kg/m³ wet snow and 960 kg/m³
slush.
