Snow load evaluation

Anybody doing anything automated in terms of detecting an excessive snow load on a structure's roof? I saw a news report about upstate NY where a guy said "when your doors begin to stick, that's when it's time to clear the snow off the roof. It's good to know one of the warning signs but surely there's got to be a better way!

-- Bobby G.

Reply to
Robert Green
Loading thread data ...

In article , ROBERT snipped-for-privacy@YAH00.COM (Robert Green) writes: | Anybody doing anything automated in terms of detecting an excessive snow | load on a structure's roof? I saw a news report about upstate NY where a | guy said "when your doors begin to stick, that's when it's time to clear the | snow off the roof. It's good to know one of the warning signs but surely | there's got to be a better way!

For something like this (and for the recently mentioned "static" occupancy detection) I would recommend looking into Sure Action's Pulsors with something *other* than their standard fast-change-detecting zone processors.

Dan Lanciani ddl@danlan.*com

Reply to
Dan Lanciani

Yeah, move outta the snow belt.

Reply to
Bill Kearney

On Sun, 11 Feb 2007 11:58:05 -0500, "Robert Green" wrote (with possible editing):

You could put stress gauges on the roof rafters, but you'd still have the problem of calibration. There are just too many variables - rafter size, rafter spacing, decking, shingle material, water content of the snow, ice, etc - to be able to do this in advance.

Reply to
L. M. Rappaport

There shore nuff is. You gotsa looks up. If your response to what you see is: "Jeeehosiphat! That's a whole lotta snow!" Then it's time to get out the shovel, Marvin. :-)

Reply to
Frank Olson

I think it would be difficult to use strain on the rafters as an indicator with a rather evenly distributed load like this. Even wind might have as big an effect as snow.

If you could place load cells to measure the weight on the roof, you should get reliable data that correlates with snow load.

formatting link
snipped-for-privacy@yahoogroups.com

Reply to
Dave Houston

Perhaps Dave means strain _of_ rather than "strain on".

Physics gets in the way of Dave's reasoning. If wind has "as big an effect as snow" on strain of the rafters, it would have a proportionally equal effect on the stress because, Hooke's Law pertaining, strain is directly proportional to stress. So load cells would be subject to the same limitation as "using strain on the rafters as an indicator".

But more importantly, I doubt that this would even approach practicality in a in a typical stick-build US house. Dozens of load cells might be required even if the roof was not also partially supported by the walls.

Much more doable and probably more accurate would be to estimate the mass of water on the roof by obtaining a top-to-bottom sample of the snow pack with a pipe and weighing the amount in the tube. This would be repeated for each range of thickness on the roof (2-3ft, 3-4 ft etc) and multiplied by the estimated square footage of the roof that corresponds to each range of snow thickness. This gives the total load (weight) in pounds.

It also leads to the "small dog corners large cat question": "What's next?" Ok. So you know that the load is x pounds or y pounds/ft^2. So what?

I think you'd end up with Frank's answer/approach.

... Marc Marc_F_Hult

formatting link

Reply to
Marc_F_Hult

I used to do something like this in my last garage.

That garage lacked a BIG A** beam to connect a hoist for lifting motors and other big stuff. As a workaround I placed a 2" steel pipe across 5 roof trusses and connected the chain hoist to the mid point. Using a tape measure I would check the deflection of the middle truss. As long as it was less than .5" (20' span) I didn't worry.

If I were to attempt a similar snow load sensor I would measure the mid span deflection of a rafter in the middle of the largest lowest pitched section of the roof. There are no magic numbers to go by but over time you might be able to get a feel for "well the snow in 07 deflected the rafter .25" and the house didn't collapse" so this snowfall is not that bad (yet)!

Just a thought...

Reply to
Lewis Gardner

Simple solution here!

Step outside and look!

Hey.. I grew up > L. M. Rappaport wrote:

Reply to
AZ Woody

A concentrated load like an engine hanging from your chain hoist is quite different from the more evenly distributed snow load.

As a practical matter there are probably far too many variables such as the type of roof, pitch, age, whether trusses are factory built or assembled on site, whether gussets are metal or plywood, average weight of a cubic foot of the snow, dead weight of the roof, attic insulation, etc. but the purpose of the truss is to uniformly distribute the load with minimal deflection of any of the truss members and to transfer this uniformly to the walls and foundation.

Here's a guide to roof truss design which discusses the various types of loads...

formatting link
formatting link
snipped-for-privacy@yahoogroups.com

Reply to
Dave Houston
[snip]

I meant to include this URL also...

formatting link
formatting link
snipped-for-privacy@yahoogroups.com

Reply to
Dave Houston

Very true.

As a practical matter deflection is a good indicator of the likelihood of failure as in "when your doors begin to stick". Since calculating the actual load and design factors in an existing house are beyond the scope of anything short of inspection by a structural engineer I would tend to stick to simple relative type observations.

If one were to attempt to design such a system, I think a measurement of deflection of the most likely to deflect part of a roof structure ("midspan deflection of a rafter in the middle of the largest lowest pitched section of the roof") would be a simple way to get the most useful data. Over time this data would be helpful in sorting out variables such as "wet" vs "dry" snow. It might be useful to someone. That is the reason I posted.

As a practical matter, if you see a bunch of snow on your roof only an idiot would not get out the snow removal tools! Especially if there have been reports of other roof failures in your area. I would not wait to see how far my roof would deflect before failure.

Reply to
Lewis Gardner

Either I'm having a deja vu, or we've been over this before but with basement joists. I went to their site:

formatting link
and they've got an interesting lineup. I didn't realize how finely they could be calibrated. They offer a good solution to pet problems, although they still can't distinquish between a small child and large dog, I fear. I might have to get some just to play with although they're a little pricey. I was actually hoping one of our resident McGyver's could outline how to make a deflection sensor out of $7 laser pointer and an X-10 Hawkeye. Seem to me the key is in the epoxy they use to hold the sensors down. (-:

If I could shoot a laser at a mirror mounted on a basement joist center from the end of the joist and reflect it back to a sensor with a narrow slit, it should be able to detect very slight deflections. Might even be able to acquire some preliminary data by shining said beam on a piece of paper and photographing the paper using a *very* long exposure and high f-stop. The laser beam should draw a pattern on the paper showing the nature and extent of the deviations caused by motion on the main floor.

If it turns out I can reliably detect motion on the floor above with that test, I might consider looking at roof issues. If snowloading can make doors stick, it should cause a noticeable deflection in a properly aligned laser beam long before that happens. My problem is that getting to the roof rafters is a lot harder than accessing the basement joists.

-- Bobby G.

Reply to
Robert Green

There are some situations that come immediately to mind as to why that's not always an option. One is a summer home in the snow belt that's unoccupied during the winter. Another is an occupant too disabled to get very far out of the house in a severe snow storm. A third would be a situation where you had to shovel 20 cubic yards of snow just to stand far enough away to

*see* the roof. A homeowner might have to decide to unload the roof before clearing the walkways, especially if there's limited snow removal capacity (bad heart, etc.) and a roof full of heavy snow and ice.

I guess the real reason I want some sort of detection capability is that I believe a smart home should be truly smart. YOU would know if someone had put something so heavy on your back it would break your spine but your dwelling wouldn't know if it was about to suffer a similar calamity. Besides, I like to collect data because it often gives advance warnings of trouble. F'rinstance, I use a clock setting utility that also graphs the day-to-day drift of the PC clock. The other day, I noticed it was drifting

*way* out of bounds. That told me the PC CMOS battery was probably failing and I was right. I was able to replace it before it lost the CMOS settings, which can screw some PC's up pretty badly.

-- Bobby G.

Reply to
Robert Green

I recommended them when you wanted something for your floor joists.

formatting link
snipped-for-privacy@yahoogroups.com

Reply to
Dave Houston

This didn't ring true, so I checked the record. Nine years after Sure Action Pulsors were first discussed in this newsgroup by others who have actually used them, Dave participated in discussion of a post titled " Strain Gages " begun Feb 16 2006 by Bobby G. in which Pulsors were discussed by others.

Dave didn't mention Pulsors as best I can determine, much less recommend them. He suggested investigating a different sensor line:

formatting link

Perhaps I need to find the time next week to see what all else Dave has been claiming in this newsgroup these days.

... Marc Marc_F_Hult

formatting link

Reply to
Marc_F_Hult

Less tech than the others but you hang a length of steel pipe inside a slightly larger plastic pipe... (use wire so it doesn't stretch) a quarter of an inch above a pair of contacts... when the roof sinks that quarter of an inch the contacts are closed and what ever you attached them to responds... low tech = light... high tech = sensor that alerts a computer that dials a phone number that alerts the media.. or whom ever. n'est pas?

Beamer Smith Yes *that* Beamer Smith

Reply to
Beamer Smith

Oh and do it in the attic where the wifes visitors won't see it... ;)

Reply to
Beamer Smith

That sounds like it might work but I'd be worried that time would eventually cause enough of a sag to activate the detector without any snow load present.

I've been fooling around with lasers and my video cam trying to see if I can detect joist deflection. I thought whatever worked for the basement and first floor would also work for the attic and roof. I was surprised to discover that the laser beam diverged quite a bit more than I expected. I was also surprised at how little activity on the floor above was detectable.

I mounted the laser on the most unsupported joist I could find. I trained it on a piece of graph paper across the basement and then trained a videocam on the paper. My thought was that if there was deflection of the joist from rolling a heavy power wheelchair over that spot, I should be able to see the laser dot jump around on the video. Nada. About the only thing that made the laser dot jump was jumping up and down on the floor above. I need a better way of amplifying the very slight movements involved. I suppose it's going to have to be the Pulsor sensors and some epoxy!

-- Bobby G.

Reply to
Robert Green

You can "amplify" the laser movement easily enough. Rig a lever with one end attached to the floor joist, the fulcrum attached to a rigid pipe which in turn is attached to the floor. Cheesy ASCII drawing follows

======V====== Joist || ----------------------LASER--->

|| || || Pipe || || ============= Floor

Reply to
Robert L Bass

Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.