I recently attended a JLC workshop. The speaker made the comment that the worst thing you can do is nail porch ceiling boards directly to the framing. I thought, "Oh boy, that's just what I did at my house." I'm looking for input on this because I'm about to put up edge and center bead on the exterior porch of a really expensive house. Should I isolate my materials? If so, with what?
(Architectural Woodworking Forum)
From contributor T:
They've only been doing that for over a hundred years. What was the speaker suggesting, and why?
Second, all of the other framing material in the attic has a high MC, too. As the attic warms up, that moisture dries out through diffusion - air circulation and evaporation - which means the air in the attic becomes hot and moist. I don't have to be a building scientist to know this. I've been in plenty of attics during the summer to know that's true. Without house wrap isolating the soffit beadboard from the attic, that hot moist air migrates into the drier finish material and causes it to swell. End of story.
Except... I've got a picture, too, in case you need a graphic reminder of what a little swelling can do.
I'm also thinking that 1/32" over 3" is well within the range of normal seasonal dimensional changes based on changes in average outdoor relative humidity and (accordingly) equilibrium moisture content, for many species. It seems to me that a sound installation would have to allow for that much movement irregardless of the MC of framing or attic air humidity. Of course paint is likely to moderate the fluctuations in the wood's MC. A little bit of compression failure might come into play too, allowing some degree of expansion before fasteners are pulled loose and a surface buckles, as in Gary's picture.
It's also my understanding that housewraps serve as drainage planes and wind barriers, but not moisture barriers. I'm not clear what Gary intends it to do here.
I'm wondering whether porch ceiling boards shouldn't be acclimated to outdoor conditions much like hardwood flooring is. If it's milled from KD lumber, its initial MC could easily be inappropriately low for a tight outdoor installation.
Yes, attics do make pretty good drying kilns, but where does that initial moisture go? In new construction, as the moisture in framing material dries out, some of it evaporates through ventilation, some of it migrates and is driven into drier material.
I agree with you, acclimating material to the regional seasonal equilibrium moisture content is vital before installation. But the moisture content in new framing is often much higher. Which means, if that ceiling were put up loose in the dry season, allowing for expansion in the wet season, and it swelled from both excessive moisture content in the attic and seasonal humidity, the carpenter doesn't stand much of a chance of stopping it.
Traveling around the country, and seeing frequent failures like this from fresh growth products (and engineered products that are shipped dry), it seems clear that we can't rely on traditional building techniques for installing a lot of new high-tech building materials.
Yes, some housewraps with high perm ratings will allow the penetration of more moisture vapor; low perm rating house wraps provide more of a weather resistive barrier - to water, wind, and moisture vapor. But I don't want to get into that can of worms. My real point is isolation. Though housewraps allow the passage of some moisture vapor, some isolation is better than none.
I think in the scenario that's being discussed, perhaps the most critical item to consider: what is the ideal MC (moisture content) for the ceiling material when it's installed. If the material being installed has too high a MC, shrinkage will be the problem; if the material is too dry, expansion will be the problem.
So... what's the ideal MC for the material? I believe this is determined by making an accurate guess as to what the *in service* MC will be... and I think this is a moving target (pun intended). To me, it would seem the MC of the material will change due to numerous reasons: the season, the geographic location, and job site location (interior or exterior).
From my reading, target MC for interior material is 6-8%, with the low end of the range preferred for less humid regions, and the high end of the range for more humid regions. Note this range is for *interior* use.
In the case being discussed, the ceiling material likely will not be exposed to "dry time" interior woodwork will see during the heating season. So what implications does "no heat" have? Over the years, I have tested the MC on material that has been stored in the unheated area of a building where I store my material, and I would say the average MC of the material hovers around 11%.
So I would think that it's safe to say that in my area, the in-service MC for exterior material that is *not exposed to liquid moisture* (rain, etc.), as would be the case for a porch ceiling, is 11%.
The questioner didn't mention just what the material was, but if the material used for the ceiling was also used (or mostly used) for interior trim work, there's a real possibility that the MC of the material is in the 7% range. Believe it or not, there are good millwork suppliers out there who provide interior material in the lower MC ranges appropriate for interior use.
If the material were 4 inches wide, and the MC went from 7% to 11%, this would result in each board expanding 1/32" (we have a wood shrinkage calculator on the site where you can plug in width, MC swing, species, and see for yourself - link below). If the ceiling boards are installed really tight (no room for expansion), then a 30 foot ceiling will "grow" about 3 inches. And I've seen this happen. I remodeled a kitchen years back, and the cabinets and flooring were both made locally, and the cabinetmaker used a load of overly dry maple. The floor (about 12 feet wide) expanded 1-1/2 inches, and the panels in the raised panel cabinet doors expanded so much that the frames on about half the doors literally blew apart.
My point? Wood moves, based on changes in MC, and there is realistically *nothing* you can do to stop it. And you shouldn't try. The key is to allow for the movement. In the case of an exterior ceiling (and using Gary's photo of beadboard type material as an example), the MC should be checked (more on this), and in the case of 7% MC, the boards should be installed loose to allow for the expansion. Ideally the material should be installed when it is at its in-service MC (11% in this case). But there aren't too many situations where a contractor can let material acclimate until it reaches the in-service MC (I doubt 5 days of sitting on the porch will even come close to moving the MC 4 percentage points).
While the discussions about peripheral issues are worth discussing, what I see over and over again (in the real world, and at WOODWEB) are carpenters who fail to recognize that measuring the MC of the material *at the time of installation* is the most important issue. If you don't know what the MC of the material you're installing is, then no matter what other measures you take, you're gambling.
And the comment about old growth and new growth... Certainly some merit there, and I believe the way the lumber is sawn has much more impact. Flat sawn lumber often moves twice as much as rift sawn lumber, so if you're really interested in reducing movement, avoid flat sawn. You can really see the impact if you substitute rift sawn for flat sawn when using the shrinkage calculator I mentioned earlier.
Take Gary's photo for example. If you're installing a porch ceiling, and measure the MC of the material when it's delivered, then you base what you'll do on knowledge (and if the MC is too far off, refuse the load). A moisture meter really is one of the most powerful tools you can own, and again, it surprises me how many really good carpenters I know who don't use one.
And as for Gary's workshops... I've seen the man in action. If any of you have an opportunity to attend one, do it. You'll come away a better carpenter than when you arrived.
WOODWEB's Wood Shrinkage Calculator
Let's correct some of the information provided above by others. Almost all framing lumber is under 15% MC. If it is much wetter, it will dry out quickly to about 12% MC. In almost all of North America, the outside RH is 65% year round; check the Weather Channel about noon for a good average RH... Do not believe what your body tells you about the outside climate.
Key point: Wood outside (not exposed to rain) will average 12% MC summer or winter.
If the wood is installed at 7% MC (typical MC for much milled wood that has not been stored outside), then it will increase its MC to 12% MC and that will be about 1% swelling. For a 10 ft ceiling, that would be well over an inch.
Sapwood and heartwood shrink the same. Old growth and new shrink the same. All heartwood cells were sapwood cells in earlier years. As stated, sawing patterns have more effect... qsawn material moves half as much.
Incidentally, the framing material does not shrink or swell in length when its MC changes. So, when boards are nailed to framing, the framing essentially is stable. Even if the framing were wet, the amount of water is small and would be quickly evaporated into the air.
House wrap will not fix this problem. The moisture is not coming from the framing but from the outside air. House wrap also provides no moisture vapor barrier at all. If it did, then it would trap moisture between the vapor barrier on the inside (polyethylene) and the wrap... Decay would be rapid.
You need to use wood strips that are higher in MC when installed... close to 12% MC. Do not nail them tightly up to the adjacent piece, but leave a small gap for some expansion on a very humid day.
Housewrap can help protect against that moisture migrating into drier finish material. Housewrap doesn't mean a polyethylene vapor barrier on the inside. I believe Tyvek has a perm rating around 50, Typar much lower, etc. These products allow some transmission of moisture vapor, for the purpose of preventing the rot you mention. But they also protect to some degree against moisture vapor migration, relative to their perm rating. Felt paper, or even building paper, would do the same thing: provide some isolation from the attic air and wetter framing material, so that drier material, with a MC of 12%, wouldn’t absorb excessive moisture.
I'm not sure about wood averaging 12% MC throughout the country. You may be correct, but I'm reading other sources too, like the Forest Products Laboratory, which lists varying degrees of seasonal equilibrium moisture content for locations in America.
I've also read that sapwood cells store more water than heartwood cells; that the wood fibers, being less dense, have more room to absorb bound water. But maybe that source was wrong. Maybe I just assumed this from experience... Take a piece of fresh-growth wood, with few dense growth rings, and drip water on the end grain. That water will be absorbed almost instantly. Drip water on the end grain of old growth wood, and the water will bead up. But then, maybe I'm misreading my own experience. Wouldn’t be the first time!
Your old growth remarks are valid. In the northeast most all exterior trim was made from old growth eastern white pine. It takes a huge amount of abuse and neglect to inflict damage. Usually chronic water damage is the culprit. But wood moves, even the old stuff. The old stuff moves a little less. If you wanted to be perfect on some t&m job, you could strap the framing with dry battens like the bead board itself.
I do believe your Tyvek idea has been debunked by WOODWEB.
I don't believe my idea about housewrap (I never mentioned Tyvek) - which isn't mine, but borrowed from a forensic expert in Canada - has been debunked. Building techniques are changing throughout the industry. Look at a recent issue of JLC and the care that now goes into flashing a recessed window. The time and material expense I'm talking about for protecting soffits from attic humidity is minimal.
I'll be the first to admit when I'm wrong. And maybe I am again. But I've seen too much damage from wood movement in the last few years, wood that was properly acclimated to the job and installed by experienced carpenters. I've seen these problems from one end of the country to the other. I'm just suggesting a way to help resolve the problem. There's more to it than acclimating your material and installing it to accommodate wood movement.
The annual EMC data is in a report from the US Forest Products Lab by William Simpson, and you will find that what I said about 12% MC average, summer and winter for most of the USA, is indeed correct. (I did work at the US FPL for 15 years, incidentally. During part of that time I was project leader for the drying and moisture project.) As just one example, consider Des Moines, IA. In January the EMC is 14.0%, April 12.6%, July 13.1%, Oct 12.7%.
I do believe that acclimating your material and then allowing for some slight movement is indeed the correct procedure. In addition, the design must include features to avoid accumulation of moisture (leaky roof, etc.).
The heartwood and sapwood shrinkage issue is so small a difference that from a practical point of view, it is not important. Please quote a reliable source that shows significant difference in pine or other similar materials. Sapwood cells store more free water, not bound water. You have it backwards. The water drop test you mention is for free water, not bound water. The test also reflects the speed of movement and not the moisture content, per se.
As mentioned, sawing techniques are most important in shrinkage or swelling.
A 2x6 framing material (not treated with CCA or similar) will dry from soaking wet to under 30% MC in 10 days of warm summer weather. The wood dries too fast for it to have substantial residual moisture that will move into an attic space and then will affect the MC of the other wood materials. The material in the South is dried to a maximum of 19% MC, but to achieve this, the average is closer to 15% MC. In Canada and the West, the maximum is typically 15% MC, with a lower average value. (It is possible to buy "S-Grn" material that has no MC guarantee. This is seldom produced.)
Incidentally, you put more moisture into an attic space by venting the fan into the attic or by not using a ceiling vapor barrier on the warm side of a ceiling than the framing itself has in it.
Note that a 2x6x12 would have about 0.2 pounds of water released for each 1% MC change. Even it is installed at a high MC (using the MC 19 grade material and not S-GRN), it will change only a few percent MC and release a very small amount of water. (We certainly are not talking about framing material that was sawn one day, put up and enclosed in an attic the next without any drying, and an attic without any ventilation. This framing would not be grade stamped and so would not be acceptable for building in the USA or Canada.)
It is true that 4x and 6x material is dried differently and often will be treated, which adds a lot of water back to the outer layer of wood.
If you see water squirt out, then the piece is not properly dried (probably not dried at all) and is not at a MC that is in accordance with the standard grades (19% or 15% max.). Most builders will use graded and dried lumber for framing and not try to save a few bucks by buying off grade material. (Exception: Treated material may have lots of water, but that is not typically used for framing except for a sill plate, etc.)
(If it adds credibility to what I say, I have been working with wood moisture relations since 1961. I have been involved in at least a dozen legal cases where dry t&g material was installed in an exterior location and then swelled, leading to swelling related failure, after exposure to the more humid outside. I have been involved in numerous similar cases where there was no legal action.)
Yes, Tyvek does have a very high perm rating. I don't think it's a good product for a situation where you want the housewrap to provide more protection from water vapor. Typar has a perm rating of only 11.7, which probably makes it more useful for protecting wood trim, yet it still allows enough transmission to prevent mold. I'm not saying that the housewrap should stop vapor movement; that would be a mistake. But controlling it is pretty important, from my experience. I guess that's what we're doing by installing it on walls, etc.
The only data I've seen on regional MC is the chart attached to this note. I certainly don't have the experience that you have, or the science background. I have only the bits and pieces of exploratory work, of conversations I've had with other specialists, things I've picked up from reading articles—kind of the same way I've learned carpentry. And yes, I bet I've got a lot of things backwards, like the bound water and free water. But I've crawled into my share of new-construction attics in the summer; it's a lot more humid in there than the air outside. I've visited several jobs and heard of others where beadboard soffits have failed and have seen high MC readings on my meter, 17 and 18%, when the material was installed at 12-13%, and there's no sign of roof leakage. The conclusion that moist air in the attic migrates into the trim wasn't mine.
I recently worked with a Canadian forestry firm on some exterior trim installation instructions. I've learned a lot from that experience. Maybe their suggestions are a bit strong from the US, coming from Canada. But then, it seems like a lot of building techniques, like rainscreen walls, are coming down to us from Canadian building codes. If they are wrong about their science, experience still tugs at me: I can't forget the hot moist attic air I've crawled through and worked in, cutting in skylights, etc. Which is why I'm quick to agree with you: 2x6 framing material will dry from soaking wet to under 30% MC in 10 days of warm summer weather. The problem is, many homes framed in the winter don't dry out until the summer. That moisture can be a problem for trim that's between 10 and 12% MC.
From what you've said, I'm more convinced than ever that there's nothing egregious about adding a cheap layer of protection. I do everything I can to protect my work and the products I use. Most of the material I use today isn't anything like what my father and grandfather used. They didn't even prime the bottoms of their doors - I know because I've re-hung doors on homes they built. The beadboard and v-groove soffits they installed are still in great shape. Maybe I'm jumping to a wrong conclusion by thinking that old growth wood doesn't absorb as much moisture as fresh growth wood. But even if that's true, I'd still want to protect my work from moisture, whatever the source.
Just because the books say framing and finish lumber are supposed to be a certain percentage means nothing. If you want to start losing sleep, buy a good moisture meter and start checking every board you buy. We built a new garage a couple years ago and know what Gary means about getting splashed while cutting 2X6s.
That swelling porch roof could be caused by a lot of things and wet framing lumber is sure a possibility. Here in Colorado that has been blamed for some soffit failures like the picture shows. Even when the T&G was at the right MC and spaced. A lot of homes are framed in the freezing winter months and the framing never dries out till it warms up or the heat goes on in the house. If the roof is on and the house sealed up, the moisture has to go somewhere, and down through the soffit is an easy path.
After spending 12 years reading numerous (i.e. hundreds of) forum threads discussing wood movement, I've come to the conclusion that in 90% of the discussions related to job site movement issues (as opposed to shop environment discussions), the weakest link is that no one tested the MC of the material being installed. When I become involved in these types of discussions now, and find out that the person(s) involved do not agree about the importance of (or practice of) measuring MC before material is installed, and be-laborers other causes as the culprit, I often feel that the discussion leans towards an exercise in futility.
I think contributor O said it all:
"If you want to start loosing sleep, buy a good moisture meter and start checking every board you buy."
You may think this sounds crazy, but he's hit the nail on the head. Finish carpenters really should buy a good moisture meter and start checking every board you buy. Because then, they will start to understand how to recognize serious problems before they start. I'm not advocating that every single board should be checked for every job, I'm just suggesting that finish carpenters who sight-for-trueness every board they pick up should approach the MC of that board the same way. If you reject board(s) because they're crooked, shouldn't you also do your due diligence and reject boards that are not the correct MC?
Wood in-use only changes size or shape for one reason... moisture change. Further, the MC change must be over 2% before any size changes begin to cause problems - rule of thumb. Note that another rule of thumb is that 4% MC change is 1% size change. A 4% MC change is roughly a 20% average RH change. Of course, all MC changes require time, so brief highs or lows are usually not serious.
Moisture changes occur when the moisture in the air (usually expressed as EMC) and the wood moisture are not equal. For outside conditions, the report by Simpson cited a few postings ago in my note (or the short version of his report cited by Gary Katz a few postings above) give EMC values. His longer report covers more US cities and also many non-US cities.
Note that Typar states "Typar’s optimal perm rating (11.7 US perms) allows the wall cavity to breathe. So moisture vapor can escape the wall cavity." The product is intended to prevent liquid water ingress and wind. It is not a moisture vapor barrier. The picture on their site also shows vapor moving through. (I had Typar put on my home, so I am not against the product at all. It is not an effective vapor barrier.) Having any sort of vapor barrier on the cold side of a wall (with or without an interior vapor barrier) will result in mold and other moisture problems within a few months.
Further, their site says that the average home occupants produce 24 to 48 pounds of water per day. Their site talks about how this moisture can easily move through Typar. This water is far more than the water released from partially wet framing lumber. Of course, if the house has an interior vapor barrier, which is commonly done, then the occupant generated moisture is not going to get into a wall.
Also, note that if the framing lumber is installed wet, it will dry quickly to about 12% MC, so wrapping a home with a vapor resistant material, such as polyethylene plastic, to prevent framing lumber moisture from getting into siding or other wood would be effective for only a month or two until the framing lumber is dry (12% MC).
Comment from contributor L:
If you intend to install a tongue and groove wood ceiling on a porch, here is the correct method to help stop the swelling problems associated with moisture intrusion.
1. Select a quality solid wood T&G product, not more than 6" wide.
2. Acclimate the wood in the room it is to be applied (four to seven days). It must be stacked with 1"x2" spacers between rows no more than the width of the ceiling joists apart and placed vertically above each other.
3. Seal all four sides of the wood with a quality penetrating wood sealer - two coats.
4. Install solid wood ceiling leaving 1/2" space around the perimeter and seal all end cuts.
Most problems happen because we are in too much of a rush to see it finished, and do not have the experience to know how to do it right the first time.