contemporary horizontal boarded solid oak front door

[woodcarver]

Been ask to make a contemporary horizontal boarded solid oak front door.

I have concerns over expansion . Has any one done these before in solid?

Example pic

https://uk.pinterest.com/pin/121949102385711486/

 

[ColeyS1]

Providing you leave gaps between each board it will be fine. I'd put a finish on the back of the boards just to help, perhaps sandwich in a piece of celotex to help with insulation.

Coley

Sent from my SM-G900F using Tapatalk

 

[RobinBHM]

When I get enquiries for these, I refuse to make them -conventional construction is to me a no no. The commercially made ones are built with constructional veneer

A joiner on this forum did show some WIP images of making such a door (I cant recall who, may have been Nathan at Murdoch joinery)

 

[woodcarver]

The customer wants solid conventional construction with no gaps between the boards.

 

[RobinBHM]

The customer wants solid conventional construction with no gaps between the boards.

Thats simply a non starter, doomed to fail. I wouldnt entertain it

 

[woodbrains]

Hello,

They don't exist as solid, meaning single timber boards throughout. The ones you can buy are engineered wood. Why don't they buy one of those? Trying to emulate what is an industrial process for a one off is going to be expensive, trying to get the same result without emulating the industrial process will end in disaster. Avoid!

Mike.

 

[Phill joiner]

I made one for my last house. Made it like a frame ledge and brace door but the head was only tenon thickness to allow enough to slide the T&G panels down from the top each side in grooves. Then glued front and back faces to the top rail after it was assembled. Still left 2mm expansion gaps between the boards though. This allowed me to spray stain it inside and out. Lasted 15 years and still going strong .

 

[transatlantic]

sorry, could someone explain why this is a bad idea? is it because if the boards were solid, then they would all expand making the door taller/shorter? why is that any different to a normal door where the boards are vertical and it expands wider/narrower?

 

[Fitzroy]

On a horizontally boarded door the main expansion/contraction is in the same plane as the door is fixed by the hinges, not so on a vertically boarded door.

F.

 

[transatlantic]

On a horizontally boarded door the main expansion/contraction is in the same plane as the door is fixed by the hinges, not so on a vertically boarded door.

F.

Why is that more of an issue though?

 

[Fitzroy]

On a horizontally boarded door the main expansion/contraction is in the same plane as the door is fixed by the hinges, not so on a vertically boarded door.

F.

Why is that more of an issue though?

- on a vertically boarded door the door will grow wider or narrower, so could stick or get drafts, but there is nothing trying to constrain its movement in this direction
- on a horizontally boarded door the door will grow taller or shorter, but the hinges are fixed to the door and frame, they will attempt to fix the same point on the door at the same height. Either the hinge will be destroyed or the door will bend to accommodate the extra length or pull itself apart if it wants to be shorter than the hinges will allow.

F.

 

[RobinBHM]

There are a number of reasons:

-the customer wants conventional construction with no gaps.

-traditional framed L&B doors have tongued board ends so the top can be a full thickness rail. Usually the bottom has a thinner rail so the boards run through to the bottom and any rain water running down the V has an escape route.

-there is greater chance of water retention on the horizontal boards. Water could run along to the ends and soak in to any unsealed end grain.

-the horizontal boards have end grain where they meet the stiles, so potential for water ingress.

-timbers largest movement is tangentially, so 1980mm of timber has an overall movement of perhaps 20mm at a 1% movement limit. Possible if gaps are included in design and boards are fixed to expand and contract within their own space.

 

[transatlantic]

But in the picture posted above, wouldn't the side rails just be like two bread board ends?

 

[woodbrains]

But in the picture posted above, wouldn't the side rails just be like two bread board ends?

Hello,

You could look at it like that, but you should understand, breadboard ends help keep the board it crosses flat. It absolutely will not prevent seasonal movement and therefore that must be taken into account when fitting BB ends to a wide board. In other words, it cannot be glued for most of its length to allow the timber through seasonal movement to extend or recede across the board as it wishes.

See the post above about a 20mm change in length for only a 1% change in moisture content. I would expect a door to have swings in moisture content a lot more than 1% and an exterior door, if that is what it is, considerably more again. Imagine a door becoming a 20-50 mm taller or shorter in its frame!

Mike.

 

[Sgian Dubh]

See the post above about a 20mm change in length for only a 1% change in moisture content. Mike.

Mike, I don't think Robin was suggesting that. I read it as 1% change in dimension (width), 20 mm being approximately 1% of 1980 mm. Just for fun I quickly ran a few numbers on the potential dimensional change a piece of European oak of that width might experience across a range of moisture contents. I came up with an MC range of about 6%, e.g., from say about 12% MC to 18% MC, perhaps not too far off reality in well finished external door sheltered from extremes of weather.

Actually, the shrinkage factors I use, which are pretty much industry standard for various wood species, to estimate likely wood dimensional change tend to overestimate. There are a number of reasons for this such as a form of hysteresis in dried wood as it regains moisture, the effectiveness of the moisture barrier created by various wood finishes, etc. But perhaps one of the biggest causes of inaccuracies in the formulae used is that wood contraction and expansion is non-linear as wood moves from fibre saturation point to oven dry and vice-versa. In general, the formulae, including the one I devised, ignore this, along with ignoring other factors such as those I mentioned above. Slainte.

 

[ColeyS1]

Imagine boards stuck both sides of a sheet of ply with 2-3 mm gaps between each board.
Replace the ply for a solid wood frame- so about 25mm thick, top, middle, bottom rails and stiles. Fit a diagonal brace and cut celotex in the triangular voids.
The only time you'll see its not 100% solid is when the doors open or at the letter box cutout.
Keep your tongues on the boards facing upwards and secret nail through the tongues working your way to the top.
That's how I'd do the doors that look to have boards running the full width of the doors.
You could nearly allow a solid edging on the sides- like they do with the thin cheap hardboard type doors. It'd certainly help make the edges of the door look a little prettier.

The doors that have stiles visible are just normal framed ledge and brace. The only difference being they've rotated the boards 90 degrees.

Coley


Sent from my SM-G900F using Tapatalk

 

[woodbrains]

See the post above about a 20mm change in length for only a 1% change in moisture content. Mike.

Mike, I don't think Robin was suggesting that. I read it as 1% change in dimension (width), 20 mm being approximately 1% of 1980 mm. Just for fun I quickly ran a few numbers on the potential dimensional change a piece of European oak of that width might experience across a range of moisture contents. I came up with an MC range of about 6%, e.g., from say about 12% MC to 18% MC, perhaps not too far off reality in well finished external door sheltered from extremes of weather.

Actually, the shrinkage factors I use, which are pretty much industry standard for various wood species, to estimate likely wood dimensional change tend to overestimate. There are a number of reasons for this such as a form of hysteresis in dried wood as it regains moisture, the effectiveness of the moisture barrier created by various wood finishes, etc. But perhaps one of the biggest causes of inaccuracies in the formulae used is that wood contraction and expansion is non-linear as wood moves from fibre saturation point to oven dry and vice-versa. In general, the formulae, including the one I devised, ignore this, along with ignoring other factors such as those I mentioned above. Slainte.

Hello

Oh, I see, misread the percentage as a rise in moisture content, not a change in width.

However, the point is still that all those boards will have an enormous swing of seasonal expansion and contraction, and should not be entertained a a way of making a door.

TBH, I seldom work out the seasonal changes with sums, even that is a bit theoretical.

Once we did an experiment with a 12in wide board of acacia. It was crosscut in 2 and both halves were put in different locations, one indoors and one sheltered by an overhang, but essentially outside. It was summer, admittedly in California, (Northern, coastal) but there was no rain. After one week, we reunited the boards and the difference in width was 3/8in! Imagine a door 6 ft 6 in tall made from boards arranged horizontally. Would it jack up the house?

Mike.

 

[Sgian Dubh]

Once we did an experiment with a 12in wide board of acacia. It was crosscut in 2 and both halves were put in different locations, one indoors and one sheltered by an overhang, but essentially outside. It was summer, admittedly in California, (Northern, coastal) but there was no rain. After one week, we reunited the boards and the difference in width was 3/8in! Imagine a door 6 ft 6 in tall made from boards arranged horizontally. Would it jack up the house? Mike.

Mike, I spend only a minute or so calculating expected wood expansion and contraction - I simply feed four bits of data into a spreadsheet I created, and the answer spits out at the end. I then simply note the result, and mentally knock off about 20 - 25% from the number given to allow for those factors I mentioned in my earlier post.

I agree that cross grain dimensional change can be spectacularly large, particularly in some species with European and American white oaks being a couple of wood types displaying some of the largest movement factors, i.e. pretty close to 9% dimensional change tangentially from FSP to oven dry. In service, of course, most wood species only experience a proportion of the potential full range of movement. For example, most furniture for habitable buildings will seasonally change in moisture content by perhaps 5 - 7 or 8%, e.g., from about 8% MC to perhaps 13 or 14%.

I agree with you and others that building such a door out of solid wood (oak) with no allowance for expansion and contraction in the horizontal boarding has real failure potential. It's do-able in my opinion, but there would need to be provision for every horizontal board to expand and contract individually, e.g., a generous tongue and groove arrangement with each board fixed at each end at the centre, but that sounds like it's not what the customer wants. Otherwise, I suspect the only other suitable form of manufacture would be an engineered product where the horizontal boards are pretty thin and bonded to a stable substrate. Even so, I'm not fully convinced even that form of structure is something that will really last, but those that engineer such products (doors, floors, cross laminated timber, glulam beams, etc) seem fairly (very?) certain they've managed to defeat wood's propensity to expand, contract and destroy adhesive bonds as it changes MC.

Just for fun, here's a snap of a piece of European oak sequentially cut into three short sections from one board that I then subjected to three different states of moisture content, i.e., from greater than FSP (~30% MC for oak) to oven dry, and the third to 15% MC. You probably already know this, but wood only grows in size up to FSP. Above that there is no further increase either across the width and thickness (tangential/ radial) or lengthways. Slainte.