This spring (2016) I built a 2 1/4" thick exterior door with three layers of 3/4" flat sawn pine. The wood was kiln dried and then sat in shop for a week on edge to equalize MC. Glued up with TBII , oil primer, latex paint exterior, satin poly interior, and installed this summer. The door is not exposed to weather or sun.

As soon as the heat (direct vent propane) came on this winter the door bowed into the interior. I know, cold wet outside and warm dry inside.

With three ply I thought I was going to safe

WT...?

I would have thought it'd have bowed to the exterior. Hitting top and bottom, and bowed out at the latch (from the inside).

All layers run the same direction? If they did, the number of plies really did nothing to change wood movement. Flat sawn sure didn't help you any, rift or quarter sawn would have been better. I've used flat sawn boards to get rift. You just have to use the cathedral grain section for something else. I thought TBII was the version that was the most flexible.

"...bowed into the interior" is meaningless. Let's use the terms hollow or concave and belly or convex.

Each stile face showed a hollow to the the outside?

Did the entire door develop a hollow/belly top to bottom and/or across the width?

Which face was to the heat?

How much deviation from flat?

Are both stiles showing the same amount of deviation from flat? Kinks?

The latch stile is concave, hollow, bowed away from the stops. The door is in swing so the stile hits at the latch and is out top and bottom. I have not taken the door off the hinges so I don't know if that stile is bowed (oops I meant hollowed/concave) too. The latch stile has about a 1/4" deflection and the door shows no sign of being twisted.

Also, the inside of the building is the heated side.

Geoff

Ok - the interior side is hollow, concave, with the top and bottom of the latch stile bowed into the house about 1/4". The assumption is that the hinge stile would read the same, were it not for the hinges.

This indicates to me (and you) a drying of the interior, and/or an increase in humidity to the exterior. This is normal in the Midwest, but no one likes to see a door like this. Heat comes on, cold and damp outside, even though the RH outside is lower than it is in Summer.

Take good humidity readings in the house - don't ask the owner, don't rely on their info. A thermometer/hygrometer is available from Radio Shack for less than $10.00. Just gathering data. Take a few good measurements on the door to quantify the amount of warp so you will be able to see any change. Make good notes. Get a mirror on a stick and check the bottom of the door and the top of the door to insure there is as much as, or more finish there than the faces of the door - especially the end grain of the stiles. This is the fastest way that door stiles can gain or loose moisture- through the ends. They do not need to 'breathe', they are dead.

If the homeowners or a laborer finished the door, they may have used very thin coats of poly, and it is allowing moisture in the wood to escape to the dry interior. I don't know about water based finishes and what they might do to doors.

Now for the hard part. On the few occasions I have seen this happen, I ask the people to let it sit and acclimate until Spring. If there are no glaring problems to correct (Interior RH less than 35%, unfinished ends, wood stove within 10' of the door, etc) this is all you can do. Often the door will flatten as it acclimates to the house over a period of several months. Once a door is in place, and good, for a year, it will stay that way for decades.

I am not sure what kind of warranty you have and what it says, but it means little to your reputation if you don't remedy the situation. My warranty says 1/4" out of flat is OK on doors up to 95", but no one wants to see that, especially me.

I am one of the last people on the planet to crown a door. We deliberately make the stiles with about 1/8" to 3/16" (8' or taller) crown in them. This is done on the joiner, then they are planed to thickness. The hollow or concave face goes to the exterior on inswing doors. This makes the top and the bottom of the door hit the stops (or damnable q-lon) first, then the latch is caught once the door is given a bit of a push. This makes for a well-fit door, and will lessen the problem you have.

You can fiddle with the strike plate if you have the damnable q-lon and move the latch stile in a bit. You can take the hinge(s) loose except for the top and bottom and push or pull on the door to see if moving any hinges will lessen the problem. They usually can help a bit, but the way they move is counterintuitive.

Last, but not least, good luck.

The three-ply stiles had nothing to do with any of this, as long as you can be sure they were all the same MC.

Did you reverse the grain of the 3 ply's?
I think this is important. We usually use quartered or rift for this type construction. At least quartered for the center lamination. MC variation of different plus can cause this.

Regarding doors and moisture in the environment:
I once ran a large architectural shop that had a pre-hung residential door department that hung about 40 - 50 pine 6 panel doors a day. The guy that pulled them would set aside any that had defects so they could be returned for credit from Morgan Door, the maker. These were good quality doors with solid wood built up stiles with veneered faces.

I noticed that he never pulled the door on top, and asked why - since it was obviously the easiest. He replied that the top door is always warped. I started to climb the racks and look at what he was talking about. On any given day all the doors had about a 3/8" to 1/2" concave or convex bow along the length of the door. All the same direction, and the same amount, only the top door of the stack.

I asked him to flip over the top doors so a concave side that was on top was now on the bottom. Next day, all the top doors had reversed - the convex side was now concave.

I started to measure Relative Humidity in the unheated open warehouse that held the doors. As it increased, the doors bowed so the top face was convex. When humidity dropped, the doors were concave.

I sized some hardboard panels to fit over each stack of doors and had them placed on top. The bowing stopped immediately.

No one that knows wood will tell you that wood expands in length as it gains moisture, but that appears to be what was happening. More like a sponge that dries out as it sits on the counter overnight.

Thanks for the responses.

What I have learned is, if I was to make another 2 1/4" exterior door with a pine interior I would make the stiles stave core with pine faces.

I had thought that three ply would help with stability and give me the thickness but, I was wrong.

This is based on the assumption that I follow fundamental rules of wood working regarding lumber selection, millwork, drying and laminating.

Live and learn. Thanks for the help.

I'm curious to know if the top and bottom of this door was sealed with the same material used to seal the front, back and sides. If not then all kinds of problems will occur.