Got handed a project..snow weight

[strokedmaro]

Your scheme will not work.

Yes there will be some load on the sensor as the roof is covered in snow, but you will not get the load per square inch, the roof material has stiffness and will deflect when a load is applied anywhere in the region of your sensor.

To be quite honest, this project sounds like managerial madness, a hare brained scheme dreamed up by a clueless suit, hell bent on having people carry on working in a potentially unsafe building.

I think that you will do just as well to measure the depth of snow at ground level and assume the same depth on the roof.
Knowing the depth of snow, you should be able to calculate the mass of snow on the roof once you know the density of snow.
Your next problem will be to find the density of snow which is far from constant.

JimB

Afterthought:
Why not measure the weight of snow on a 1sq metre board at ground level, and trigger your alarm from that.
(Still a daft idea).

LOL...your not far off. United States Air Force at your service....couldnt possibly stop the mission in favor of safety Seriously though you Idea of measurements on the ground and assuming the same on the roof (makes sense) is not a bad idea.

 

[Willbe]

Post photos.

Borrow a copy of "truss design" or the like for your application.
Put a deflection gauge in a strategic location. You're probably looking for mm deflections. The truss probably acts like a spring with a variable spring constant.
Decide on a warning level of deflection.
Test with a point load centered on truss mid-span and correct for a distributed load, or put a distributed load (cinder blocks or bricks) on the roof and scale up.

I think the density of snow varies by a factor of 10 or so, from something less than 62.4#/ft³

The trouble is that most formulas calculate deflection; you want to go to the elastic limit, to the point of failure, to the point of permanent deformation.

You may want to get books on Structural Failure Analysis for similar buildings to see what you're up against.
Or, ask the MEs or CEs on this site. . .
Eng-Tips Forums

 

[bigkim100]

way too simple an answer

Ok, since finding all the load weights may well be a problem. why dont you just find out what depth of snow you, and your bosses are not comfortable with, and shovel the roof once you have that depth on the ground.
EG: dont really like more than 4 feet of snow on the roof, 5 feet fell last nite...better get shovelling

 

[Rolf]

KISS....... (Keep It Simple Stupid)

{snip}
What sensors to you guys think would work best for this application? Any ideas welcome! THANKS!!!

Strokedmaro

Since the building is fairly new the architect should be able to supply the required information on the roof designs snow load capability.

To obtain the current snow load use a heated (remote if needed) rain gauge.
Then calculate the amount of rain water that is equivalent to the max safe load of snow.

When it snows have some idiot look at the posted numbers and compaire it with that on the rain guage and take appropriate action when the numbers get close.

Cost less than $100.

 

[Boncuk]

You might want to look at the Elektor article posted in January 2009.

The project is called "Schubkraftmesser", translated about "thrust measuring device".

It uses a normal digital kitchen balance which are fitted with strain gauges and an ATMEL ATMEGA8-P µC.

I would not recommend to measure the strain on the roof for structural deformation. Having a deformation already there will be no need to call for men shuffling the snow.

I'd rather recommend to measure the snow weight directly at four points of the roof and go for the max weight to alarm the shufflers. Snow sludge might be very thin, but it has almost the same weight as water. The same applies to hail. (Make sure the hail won't jump off your pan )

Fitting a heater and a drain hole to the pan you won't have to climb the roof when snow fall has seized and the heater is activated.

Regards

Boncuk