About 6 weeks ago I built a storm door for a client. I used 1-1/8 x 4" poplar rails and stiles, with an MDO lower panel. Both sides primed and painted. After about two weeks it had expanded so much that the door would no longer open and close properly. I planed it down and after another two weeks, same thing. Planed it again and now after about two weeks, it is again sticking. I think I've already planed close to a half inch off (before the third fix). It is a double door setup and I've only planed one door, so the actual expansion is probably half that amount.
The joints are still tight, so the panel seems okay. In fact, there is just a very slight paint line where the panel goes into the stile, so if anything, the rails have expanded along the grain just slightly more than the panel has grown.
I just can't believe that I'm getting that much expansion across the 4" stiles and/or along the rails. Is poplar okay, or would another wood have been better? I can keep planing till it eventually stops, but I'm afraid if it someday contracts, it will end up smaller than the opening. I'm in Wisconsin, and we have had a pretty damp spell of weather.
What did I do wrong, and what can I do to correct it? (I normally build furniture and cabinets, so this is a little out of my normal realm.)
(Architectural Woodworking Forum)
From contributor J:
Poplar is not a wood to be used on exterior projects or any project that might be exposed to moisture. As I'm sure you know, poplar is very stable in interior conditions. It is very prone to expansion and rots easily when exposed to temp swings and moisture. There is a good chance you will always have problems with this door. I'm sure as the wood moves with the seasons, the gap will be huge and since it is a storm door, exposed to moisture, it has a greater chance of rotting.
The door will not rot unless it is exposed to direct wetting. Is this door protected from direct rain? If so, do not worry about rotting, as there is no such thing as dry rot (that is, wood will not rot when it is dry). In other words, y-p is a fine choice when protected from rain.
I assume that you are referring to the door expanding in width, which means that either the rails are growing lengthwise (and the MDO is growing in the same direction) as they gain moisture, or the stiles (I believe you have 4 stiles total and each is 4") are growing. The stiles would gain a maximum of 0.00158" per inch of width for a 1% MC increase. So, for your 16" total, that is 0.025" per 1% MC gain. If you built the door with 7% MC wood and now the RH is 80% RH average (16% MC), you would have a 9% MC change or swelling under 0.25". So, your removal of wood must therefore be because the rails and MDO are also swelling. Lengthwise movement of y-p is not unknown, but is not common enough that all the rails would be swelling that much. What is the frame of the door made of? I would suspect that you might have some frame movement instead of all door movement. Is the exterior door okay? Also, it is disturbing because the air inside a storm door is very hot and therefore dry, compared to the outside environment.
Did you make the doors a specific and precise size? If so, you can measure them now and see how much they have grown. Further, is the door opening a specific accurately known size? If so, you should be able to measure the opening size now to see if it has changed.
My choice for a wood is sapele, but it is fairly heavy compared to pine.
The specifications are not in dispute, but in practice, poplar moves more and faster than any other wood I have ever encountered. It responds to changes in RH darn near as fast as my barometer! Current cuttings of poplar (here in Indiana), have wide spaces in the annual rings, and the wood is light, less dense, and moves like a wet noodle. We are seeing less and less stability in interior use as well.
Even if it is kept out of the wet (I also have seen poplar grow mushrooms in 3 years or less), it still moves way too much. This movement also does nothing to help the doors stay flat and true. Since we all mostly work with KD wood at 6-8%, as the doors move to 12% or so, they likely will not stay flat, even if enough gap is provided at original fit and hang.
One oversight on paired doors is to at least double the normal gap, since there is twice the wood to move across the opening. In my experience, the frame movement is negligible. 3/4" to 1" thick jambs, if existing, have reached their equilibrium moisture content and will move only a hair or two through the seasons.
Regarding other species, rift white oak (not quartered, however) is equal to y-p in the amount of movement. All softwood species do indeed move less.
In my previous calculations, I misread the number for poplarů It should be 0.00289. Again, however, it is equal to or more stable than most other hardwoods.
Regarding the movement of y-p in an interior environment, it will not move at all if the MC does not change. If the MC changes 2%, the movement is extremely small. If high movement is encountered, it is because the wood was subjected to high MC change. Almost all hardwoods will change as much or more than y-p when there is a high RH change. One problem with y-p is that because it dries so easily, it is often over-dried and then the MC change between manufacturing and use can be large. This is not a problem with the species, but with the manufacturer.
Although you may feel free to challenge the historical values of how much y-p moves with MC changes, appreciate that the US Forest Products Lab has spent many years and done careful measurements of these values using a wide range of samples. They are not incorrect and do accurately represent the species. High movement represents high MC changes.
I have been involved in historic renovations and seen (and used) old growth poplar that was a dream - dense, dark green heartwood as straight and stable as anything else I have ever worked with. This wood had been outside, under cover or paint, for 160 years and looked great. What we produced from this wood went back outside and has weathered very well in the few years it has been outside.
The poplar I get from various suppliers is the same species, but does not have the same characteristics that the old growth does. It moves - as does any wood - with changes in RH, but is just more squirrelly when it does move. That is, it does not stay flat, will curve and bow and cup more than any other wood when the inevitable happens - the RH changes.
It also appears to attract rot. I have replaced doors and shutters and all sorts of exterior millwork under warranty (all at my cost) because it rotted in four years or sprouted mushrooms in three years. We even tried making beveled siding as a premium upgrade. While this was a bit thicker than the Western examples of bevel siding, it still curled up on (off) the walls and was totally worthless in less than one year. Yes, it was properly nailed, back primed and vented. I recommended the wood, as did many shops in this area, based upon observation of historic structures, the advice of a major State agricultural University, and the FPL and others.
This recommendation nearly took out the shop I worked for - the owners just stopped honoring warranty requests as a means of damage control, all the while looking at me quite angrily. Another local shop that produced very well made, properly built and expensive custom casement and double hung windows, with 8 employees, based their entire production upon the recommendations of the State Agricultural University that poplar was an ideal wood for exterior use. When they closed their doors after seven years, they left a trail of rotted windows, unhappy customers, open warranty claims, and a few lawsuits.
My point is twofold: That the numbers may well be accurate in a lab setting and analysis, but real world experience counters the conclusions some would draw from said numbers, using the standard product available today. And two, that there is absolutely no reason to skirt so close to disaster by using poplar in any sort of an exterior situation, especially when there are so many other choices readily available.
Comment from contributor B:
From the amount of swelling that you describe, it looks to me like your door is absorbing moisture, not so much from the atmosphere as from direct water absorption. Our experience with screen and storm doors is that the clearance between the stiles and the frame must be large enough so as not to hoard water when it rains. You cannot use nice 1/16 inch clearances for exposed doors: it should be more like 1/8 inch.
When we install one of our doors we make it 5/16 narrower than the opening since kiln dried wood will first swell, through atmospheric absorption, to an annual average size then shrink and swell about that point. We find the door is largest in late November (Ontario, Canada) and at that point it should still have about 1/8 inch clearance to allow water to drain out of the frame rabbet. After several callbacks on our first doors we came to realize that planning to have the door just clear the frame was not enough, it had to be planed to 1/8 inch clearance or upon the next rain the water would just sit in the rabbet again and start the cycle all over again. Better still was to make the door the right size in the first place. For the bottom we give the door a 1/4 inch clearance and close the gap with a sweep- another hard learned lesson.
We have doors on display year round outside our shop with only paint to protect them and I have seen any species of wood swell amazing amounts once the paint deteriorates and direct water gets into the parts. Direct water throws all the swelling equations out the window. We once put the moisture meter on the bottom rail of one of these doors after observing it swell 5/8 inch across the grain and found it was off the scale, fully saturated. Needless to say it was also rotting.