(Crossposted to alt.home.repair and sci.engr.civil; followups to alt.home.repair)
Hello all!
Someone I know is proposing to build a privacy fence in his backyard. He's asked me for advice a couple of times - I'm not sure why, as I am most definitely not a civil engineer, but I attempt to be helpful. I have pretty much suggested following the plans in the "do it yourself" type books - in brief, dig a hole every 6 to 8 feet, set a post (4x4 or 6x6 inch (10x10 or 15x15cm) pressure-treated lumber) with about 1/3 of its length in the ground, pour concrete down in the hole, wait for the concrete to set, attach horizontal boards between the posts, then attach vertical boards to the horizontal ones. He keeps coming up with... "unique"... alternative designs that I sometimes attempt to dissuade him from. There are two reasons I care at all: one, I might help him install this fence, and I don't want to be associated with it if it blows over in the spring and lands on somebody's house, and two,
*I* live in the same neighborhood and don't particularly want pieces of this fence landing on _my_ house. My intent in posting here is to see if I'm overly concerned, or if some of these ideas might actually work.Anyway, the basic specs are a mostly-wooden privacy fence that is 6 to 8 feet (2 to 2.5 m) high, 8' being the highest allowed by local code. It is enclosing a rectangular backyard. There will be three long sides; the fourth side is mostly the house but there will be a couple of short sections of fence on either side of the house. I haven't measured it but I would guesstimate it to be around 250 linear feet (75 m) total, plus or minus 50 feet (15 m) or so. The site is a former river flood plain in Oklahoma; because of flood-control projects upstream it hasn't been underwater for at least 50 years or so. I don't know any technical details of the type or composition of the soil, other than I don't think much fill dirt was brought in when the neighborhood was built about 40 years ago; everything is pretty much flat. The highest wind speed recorded in the area has been about 55 mph (90 km/h) from the north- northwest; this is a one-minute average and doesn't include higher gusts associated with thunderstorms, tornadoes, etc.
Idea #1 was to use ready-made fence panels from the lumberyard, 8' high and either 6' or 8' long. But instead of attaching them to 4x4 or 6x6 posts set in concrete, his idea was to attach them to T-posts driven into the ground. T-posts work great for barbed wire fences and temporary crowd- control fences; I've helped repair and build these kinds of fences before. But in both cases, the posts don't have solid panels attached to them, but rather barbed wire, rope, plastic banners, etc. In other words, the wind load on the post is very small - each post might be supporting four strands of wire, each eight feet long, with an average width per strand of maybe 1/8" (3 mm), for a total of something like 0.3 ft^2 (0.03 m^2) per post exposed to the wind. Compare this with 48 or 64 ft^2 (4.5 or
6 m^2) per post with the fence panels. It seems to me that the wind, if perpendicular to the fence, would tend to overpower the T-posts, and either bend them, break them, or push the the fence over, ripping the posts out of the ground in the process.Idea #2 is to use the ready-made panels again, 6' high this time and either 6' or 8' long. For posts, this time, the idea is to use steel pipe of at least a couple of inches diameter, driven an unspecified length into the ground. No concrete or gravel around the pipe - just pound it into the ground. The fence panels would be mounted with their bottom edges 2' (0.6 m) off the ground, so their top edges would be at the legal maximum of 8'. There would be a open area from the ground to
2' above the ground. The posts would extend about halfway up the panels, or to about 5' (1.5 m) from the ground. The steel pipe sounds more substantial to me that the T-posts, but I think that means that it probably wouldn't bend or break in high winds - rather, the whole fence would get pushed over, the pipes would remain straight, and their lower ends would rip up out of the ground.I tried to do some math on idea #2 based on the size of the fence panels, the wind speed, the density of the air, and the diameter and length of the post (assuming 5' above ground and 2.5' (0.75 m) below ground). Basically, I am trying to turn the wind into a point load at the top of the post, and then assuming that all of that load comes out over a relatively small area somewhere else on the post. Based on what happens when you horizontally wiggle a small post or stick that's stuck in the ground, I figured that the bottom end of the post (underground) wouldn't move, so you'd end up with a lever with the fulcrum at the bottom end of the post, the wind pushing on the lever at the top of the post, and the lever pushing on the soil at ground level. When doing this, I come up with an odd unit of "pounds per second" which causes me to doubt my "analysis" somewhat, but the numbers I am coming up with for a horizontal load on the soil are on the order of 200-250 psi, which seems to me like "too much" for the soil to take. (Assuming this line of thinking isn't completely off base, I would be happier with an answer down around 50 psi or even less. Reason: you can drive a pickup truck or a car (30-50 psi) over an unimproved dirt road without beating the road up too much, but much over that starts tearing up the road. I know that this is vertical and not horizontal, but it at least provides some kind of reference.)
Matt Roberds