Monday, January 23, 2017

How does a tree slice through a house?

I was feeling privileged this weekend because of my high homebuilding standards. The bad weather that killed 20 people in the southeast this weekend went right over my place. But my built-with-screws tiny house was just fine. Even the jokes about tornadoes being attracted to mobile homes should apply to me since I have a restored bona-fide mobile home a few hundred yards from my house. But it was built in 1951 by an aircraft manufacturer to the same standards as a plane. It laughs at wind. And it isn't going anywhere because of steel straps and 4 augers securing it to a slab.
Frame of my lab with straps installed
New subfloor going in
1/4-20 stainless steel bolts and nuts secure the riveted aluminum body to the frame and subfloor between members. In the foreground note the metal building type screw holding
 the subfloor to the frame
(Yes, I painted the bottom of the plywood before I installed it. Then I painted the top as soon as it was in. I don't like the smell of outgassing plywood and I was trying to seal it. For more on my Spartan project see earlier blog posts, like this one about the finished floor.)

So mobile homes as they were originally conceived back in the 1940s were not doomed to be ripped apart by wind. I'm not sure when it all went wrong for them. But boy, did it go wrong.

Today I got in the car and went looking for signs of damage nearby. A map on weather.com shows 4 tornadoes touched down to the northeast of me less than 20 miles away on Saturday.  I headed that direction. I found the river greatly flooded and a few large trees twisted off but I couldn't find any obliterated mobile homes.

River in flood

Good luck with that
This is the only really twisted tree I saw. Saw several regular broken off ones.
I did see a lot of undamaged mobile homes. They all had skirting around them, which is a county rule that I don't understand. What is it even for? I have researched it and am unable to find any reason for it except varmints. They think it's going to keep them out? I say it's more likely to make it an attractive place for a varmint to explore. Next thing you know you're listening to the incessant chittering of a stressed out raccoon under the house because it went in and then couldn't find a way back out.

What is the purpose of this white mess around the bottom of the mobile home?

It always looks trashy.
I came home and tried to find stories online detailing the damage to homes where people were hurt. I found weather.com had a slide show of images. The mobile homes seem to mostly fail by being physically moved because they are inadequately anchored. Most of the pictures I saw they seemed to be merely resting on dry stacked concrete blocks on a slab. That's basically what I've got for my lab, but I cut holes in my slab and installed 4 augers into the subsoil and then patched the concrete. I ran steel straps over the steel I-beams of the floor and tightened them down to the augur anchors. I got the anchors at Home Depot. I don't see why anybody with a mobile home wouldn't install them.

I also see from the underside of the mobile home below that it is not built like mine at all. The aluminum channel at the bottom of my riveted aluminum strut walls has 1/4-20 bolts through the plywood subfloor every 16". The subfloor is screwed to the steel frame. I don't have any wood floor joists. All welded steel. And I still wouldn't sleep in here. I go in the house if it just rains hard. It's too loud in the all-aluminum lab even with the spray foam insulation.

Photo from weather.com showing underside of a modern mobile home
I feel the science on keeping houses together in the wind was well addressed after Hurricane Andrew. I don't know how many houses that failed in these tornadoes were built according to standards developed after Andrew. I'm guessing none of them.

For a compilation of suggestions and more pictures of failed houses check out Fine Homebuilding's Wind-Resistant Framing Techniques: Cost effective details to help houses weather the storm wind by Bryan Readling. There's an illustration showing something I did wrong on my house. I heard this on the Fine Homebuilding podcast and slapped my head. Hurricane ties go on the outside. Crap.

At least three people this weekend were killed by trees falling on their houses, basically splitting them in two. I'm not sure any building codes have been altered to make houses safe from this particular hazard. But I'm curious about it. At least three of my close relatives have had trees fall on their houses due to hurricanes or tornadoes, and two of them have had it happen twice. The trees didn't slice their house in two though. And they were big trees. What is required for a tree to slice a house in two? Here's some pictures of what I mean from Tallahassee that I saw on Twitter. I don't think anybody was hurt in these incidents.


Are roof trusses made of 2x4s more likely to fail than a house with large dimension rafters? Does a structural ridge matter? Does the tree have to fall in a direction that allows it to go between the rafters or trusses so it only has to slice through sheathing? Did these houses have OSB sheathing or plywood? Which is better? How come the trees were able to plow through the top plate so easily? Makes me feel pretty good about building a rigid frame house. Here's some photos of my roof framing.


I'd forgotten I had to add those blocks on top of the rafters under the purlins to make the metal roofing come out right. They are maple flooring cut offs. Another detail I can see if I zoom in on this photo is the sill seal I stapled to the purlins before I screwed on the metal. It was meant to reduce vibrations conducted through the structure into the house from rain on the roof. It helps.

Here are some photos I took right after the spray foam was installed. Reflects so much light it's easier to see what is going on.

Photo from the front door.
I have 2x6s atop 4x4 corner posts instead of a pair of 2x4s held up by studs like those sliced up houses above. I used Simpson Strong Tie framing brackets to mount the beams on top of the 4x4s.  The rafters are 2x8s screwed in place with the largest Deckmate screws. The 2x6s holding up the rafters are PT and I didn't want to take chances with inferior nail coatings that might corrode. The double 2x6s over the windows are not pressure treated. The outside one is lag bolted to the 4x4 and then another 2x6 is wedged between the posts and nailed to the bolted one. All the beams are spiked together with Simpson Strong Tie hot dipped galvanized nails driven with a palm nailer at the interval specified in my architectural graphic standards book.

Note my hurricane ties are on the wrong side of the beam. I was ignorant and I feel much shame. But I even used hurricane ties on that interior wall. What can I say, I like the palm nailer. I'm not going to sweat it this one mistake. I've got greater than average tie down per roof area.
View of roof framing from bathroom side. Easier to make out that it's a pair of PT 2x6s under the rafters. 
My ridge is a 12' 2x10. It is held up by complicated headers built to accommodate windows I got from Tallahassee Surplus and Salvage. All those big plate glass windows are sliding glass doors I got from my grandparents after a tree fell on their sunroom.

My tiny house plan allowed me to use 8' 2x8s for rafters, a much easier piece of lumber for me to work with alone. I put some 16' 2x6s across the roof on the pressure treated beams to stand on, but no amount of struggling with angles and arranging my center of mass could get the rafters to come out on my penciled line when I shot the nails in with the framing nailer. If the top moved off the line then the birdsmouth on the other end was all out of whack. So I gave up after the first one and started using a 9v battery powered impact driver and Deckmate screws. I could only do about 2 rafters before I had to recharge the battery. But I got the whole roof framed in a weekend. Then I covered it with a tarp and Hurricane Dennis hit Florida as a Category 3 Hurricane July 10, 2004. First chance for a tree to hit my house. Came through it wet but tree-free.

Now back to the present. This house near Albany that was torn all to hell this weekend clearly reveals it was made with trusses and OSB. I don't know if the brick veneer helped or hurt. I don't understand why there is OSB, house wrap, and then bricks, but the bricks don't seem to have been attached to the wall at all. That can't be right. It looks like the whole thing started when a gable end blew out. This seems to be a common failure as described in that Bryan Readling article about. Leaving out the sheetrock on the inside makes the ends blow out if it's sheathed with foam and then the whole thing pressurizes and blows apart.



One of my aunts had a large pine tree fall on her house right across the middle of the longest section of roof of her house, in line with the rafters. It came to a rest on the ridge. One limb punched through the metal roofing. And then later in a tornado an enormous limb from a giant live oak tree fell on another part of the roof. She has a hip roof framed with 2x8 rafters, if memory serves. It has industrial corrugated steel screwed to 2x4 purlins just like my house. No sheathing at all. She and my uncle built that house themselves with boat building precision. Neither tree-on-the-house incidents caused any structural damage. They just got out the chainsaw and removed the trees, unscrewed the sheets of metal with the holes punched in, and screwed on new sheets of metal.

This fall, September 2016, my mother had a huge poplar tree fall on her house in Tallahassee during Hurricane Hermine. She has a plywood roof deck under thinner gauge metal roofing than what my aunt and I have. The tree did not go through the top plate. It did fall over a French door though, so there would have been a big header there.
The photo my mother texted me in the middle of the night when this tree fell

Here's a photo of the poplar tree on the house in the daylight

The tree service came the very next day and cut the tree off the house
Here's the damage to the roofing where the tree fell
A close up of the little overhang where the tree fell shows the plywood sheathing on the roof
This is the second time my mother's house has been damaged by trees in a hurricane. Right after she had it built Hurricane Kate dropped a tree on the main roof. It was right over her bedroom too. But it didn't slice through the house and kill her. It did cost more to repair it than it cost to build the roof in the first place. That was when she had regular composite shingles. She got that metal roof when she built the addition where the poplar tree fell.

I'm curious now about what it takes for a tree to slice a house in two. I found an article about a study Clemson was going to do in 1999 simulating dropping a tree on mockups of walls.
"But if the tree were to fall between the roof structural members and impact the plywood roof sheathing, the roof wouldn't make much of a difference anyway," warned Ed Sutt, a Ph.D student who is helping with the research. "And keep in mind, we're only seeing the effects of a six-inch pine at this point. Larger trees could have far more devastating consequences."
The Clemson research will concentrate on identifying inexpensive techniques that could be used by a construction-savvy homeowner. Research findings will be used to develop technical guidance for contractors and practical application material for homeowners.
Low-tech fixes under study include installation of wood-reinforced wall panels, addition of a layer of metal decking under the wall panel, different combinations of plywood or common insulation foams.
I don't actually agree his assumption. That poplar tree in that photo above is way bigger than 6" and it didn't go through the sheathing. My aunt's house doesn't even have sheathing but the ridge of her hip roof was strong enough to stop a pine tree bigger than 6". The oak tree limb that fell on one of the hip ends was absolutely huge, easily 18" diameter. It didn't break the structure.

A shortleaf pine at least 20" in diameter fell on the gable roof of my grandparent's sunroom and it just spread the walls apart. Didn't even break the plate glass windows. I don't remember it slicing through the roof sheathing at all.

Is the difference in the prediction of that PhD student and my own experience that all the roof/tree incidents I know about involved traditional framed roofs and not trusses?

I feel like the distance of the tree from the house is going to have a lot to do with the damage inflicted too. That poplar tree was so close to my mother's house it hadn't really gotten up much momentum when it rested on the roof. It the house intersected the path when it was most of the way down it would have been much worse, like this giant oak tree that hit that poor air conditioner when it was going full speed.
Another tree that fell during Hurricane Hermine at my mother's house
I may have found the paper with the results of the study teased in the above article in the Journal of Architectural Engineering, but it's $30 to get a PDF. If it doesn't actually answer my questions I'd be annoyed. From the abstract:
The study also considered, on a limited basis, the threat of falling objects such as trees or large branches. Results suggest that a wall designed to resist debris impact will also exhibit satisfactory performance against tree-fall.
Green Building Advisor's article Pros and Cons of Advanced Framing by Martin Holladay is a checklist of ways you can make your house that much easier to slice in two with a tree. But they thought about that and Michael Chandler added a comment with these tree-strike-relevant points.
1- push the headers up to the (double) top plates and cripple down to the window head so the trees hit the header at the ceiling level hopefully stopping at the top plate and keeping most of the water out of the house.
3- solid block w/ 2X stock on the exterior of the upper top plate between the rafters stacked on the sheathing to help distribute the impact of trees to the sheathing and encouraging the rafter tails to break at the plate to help keep water out of the house after a tree strike.
4- block the ridge solid between the trusses, If vented just hold the blocking down on either side as shown in the attached drawing for the ridge vent, if spray foam, block solid and tie both planes of roof sheathing together at the peak.
Click the link above and scroll down to the comments for a drawing.

I would love to make a mockup of my roof and drop trees on it until I get some interesting results. It seems to be an underserved area of study. My online research just now reveals a lot of dubious advice like "cut down all the trees that could fall on your house." I also found a lot of studies advocating the energy savings you get from shade trees. You can't have both. I prefer trees.

Meanwhile whenever there is a tornado warning I will continue to climb down from my loft and get in my cast iron bathtub with my pillows and blankets. I don't have an interior room away from windows, but I feel like the sides of a cast iron tub offer pretty good secondary protection if one of my pine trees decides to fall on my house. It better have precision aim though, because my house presents a pretty small target.

*update*

I asked my father if he wanted to use his backhoe to help me drop trees on a model of a house. He said I should go back to knitting. He thinks the variables involved in trees falling on houses is too vast to be relevant. He thinks the roots of the tree slow the tree down a lot if it is near enough to the house, as in the case of the poplar tree above and the shortleaf pine on my grandparent's house. He doesn't think dropping a cut-off tree on a model would prove anything. I thought it might just be fun to watch it smash. He said if it's so much fun I should get Hollywood to pay me to do it. Anybody want to remake Twister in the Southeast? I'm available for consultation.

My father also told me that the wood is stronger at the bottom of a tree where it has to counteract the bending stress from normal wind loading. Small sawmill owners find this to be the case from vast personal experience with a chainsaw. This explains why I see trees either snapped off about 20 feet up or pushed down from the roots or twisted at the base and not even broken when I survey the woods after a wind event. I rarely see them snap off down by the ground. Only if they are seriously compromised by fire damage. I also rarely see dead trees that have fallen in the woods due to a storm. The insurance company policy of not covering damage from dead trees sounds like a scam to me. I perceive no inherent greater risk to property from a dead tree than a live one. I would like to see their actuarial data that says dead trees do more damage. It's prejudice against snags, which is bad environmental policy. Birds like snags.

So unless I come into a motorized winch for free I guess I won't be making a video of dropping logs on model houses. Sorry.

*** Update Feb 14, 2017 ****

I came across this video describing the Fine Homebuilding 2012 Editors Choice house that is specifically designed to withstand a big tree falling on it. This is pretty much exactly what my father said. If you want to build a house that can't be crushed by a tree you hire a structural engineer and build that house. But most people don't want to pay for that. Well, here's what happens when somebody does pay for it. Big honkin' I beams. Wonder what they do about condensation?

No comments:

Post a Comment