Solid cabinet doors - wood movement strategy

[Bodgers]

I am completing a set of two wall mounted cabinets that are mostly solid oak.

Wife has 'specified' solid wood on the doors. Doesn't want the frame and panel look as she thinks it will be too kitchen-like for the office room that they will sit in.

So I am faced with the best way to not end up with a wood movement disaster on these things further down the line.

The door itself will be made out of approx 17-18mm thick Oak stock. Strips 50-60mm wide laminated together. Each door (2 per cab) approx 900mm by 350mm. The strips will run vertically, with the grain.

I was thinking of routing a shallow groove on the back of them (across the grain) and creating a contrasting walnut strip (I have other walnut accents on the cab frames) maybe about 12mm thick, and counter sinking screws in across the strip/bar.

I suppose I want to restrict the horizontal expansion and contraction, but not to completely restrict it so that cracks appear.

Crazy?


Sent from my MI 3W using Tapatalk

 

[dzj]

I'd opt for Oak veneered mdf, (or other core) and solid Oak lipping.
You might get away with what you suggest, but you probably won't.

 

[fuster]

Frame and panel isn't a "look"; it's a technique. I have this all the time with customers specifying how they want things with no idea about why things are made the way they are. Fortunately most of them listen to reason when I talk them through it. The worst are always the interior designers who insist on having things made badly because it's always only about the look and they of course know best!

 

[Beau]

"I suppose I want to restrict the horizontal expansion and contraction"

Good luck with that.


A few options.

1. Use solid wood and accommodate seasonal movement.

2. Glue up with epoxy and encapsulate the wood in the resin to completely seal it.

3. If you fitted bread board ends you could veneer the core and have solid wood edges all around. Cut you own structural veneer on a bandsaw to make it all match.

 

[MattRoberts]

The thinner the strips the more of them, and the more chance of movement. If you then try and stop that movement by inlaying a strip across the grain or fixing them to a batten, the movement is more likely to cause something to give.

Containing the strips within breadboard ends that are correctly fitted (fixed in the centre but not the ends) is a great fix, and allows the panels to be flat in rather than the panel look.

 

[Bodgers]

The thinner the strips the more of them, and the more chance of movement. If you then try and stop that movement by inlaying a strip across the grain or fixing them to a batten, the movement is more likely to cause something to give.

Containing the strips within breadboard ends that are correctly fitted (fixed in the centre but not the ends) is a great fix, and allows the panels to be flat in rather than the panel look.

Is that true? I'd have thought the more laminations the less movement - hence the stability of engineered wood.

"I suppose I want to restrict the horizontal expansion and contraction"

Good luck with that.


A few options.

1. Use solid wood and accommodate seasonal movement.

2. Glue up with epoxy and encapsulate the wood in the resin to completely seal it.

3. If you fitted bread board ends you could veneer the core and have solid wood edges all around. Cut you own structural veneer on a bandsaw to make it all match.

I like option 3. Might be able to get away with that...

Sent from my MI 3W using Tapatalk

 

[MattRoberts]

Is that true? I'd have thought the more laminations the less movement - hence the stability of engineered wood.

I think you're thinking of laminated beams where multiple strips are laminated on top of each other.

I believe the Op is talking about edge glueing strips, which have a tendency to cup

 

[Jacob]

...

I suppose I want to restrict the horizontal expansion and contraction, but not to completely restrict it so that cracks appear.

Crazy?..

If you restrict movement you will get cracks. So you need some sort of conventional rail design to which your panels are very loosely fixed - screws in slots, sliding DTs etc.
I saw a clever solution somewhere - they'd used strap hinges but instead of planting them on they'd threaded the straps through the panel via mortices (cut before the panel was joined up obviously). The panel only needed one fixing close to the hing end and could expand contract along the strap. So got a fair face both sides but steel hidden within.

 

[Bodgers]

Is that true? I'd have thought the more laminations the less movement - hence the stability of engineered wood.

I think you're thinking of laminated beams where multiple strips are laminated on top of each other.

I believe the Op is talking about edge glueing strips, which have a tendency to cup

I am the OP, I was more thinking of expansion and contraction.

Sent from my MI 3W using Tapatalk

 

[Chrispy]

Even if you are lucky enough for the panels to stay flat you will still need to accommodate at least 1/4" may be 1/2" of variation due to seasonal movement at the centre join between the two doors.

 

[deema]

Wood moves along the growth rings, so to minimise movement across the panel, select wood that is not flat sawn but quarter sawn. The amount of movement will be negligible.

The other strategy if you are using fiat sawn stuff is to leave it in the room where the cabinet is to be kept for at least 3 months. Movement in a typical central heated home is fairly small

 

[Bodgers]

Even if you are lucky enough for the panels to stay flat you will still need to accommodate at least 1/4" may be 1/2" of variation due to seasonal movement at the centre join between the two doors.

I aren't overly concerned if things get a little 'ripply', it is more the doors getting jammed (or huge gaps appearing) that's my concern.

I had another idea. How about some form of a 2 layer sandwich? 8mm thick panels going vertically (as originally described), and on the back, glued panels horizontally. That's the principal that plywood works on, right?

 

[MattRoberts]

I am the OP, I was more thinking of expansion and contraction.

Sent from my MI 3W using Tapatalk

Ah, sorry. But my point is still valid. If you're edge glueing strips to make a panel, the expansion and contraction is across the width of the strips and can also lead to cupping across the panel.

 

[MattRoberts]

Even if you are lucky enough for the panels to stay flat you will still need to accommodate at least 1/4" may be 1/2" of variation due to seasonal movement at the centre join between the two doors.

I aren't overly concerned if things get a little 'ripply', it is more the doors getting jammed (or huge gaps appearing) that's my concern.

I had another idea. How about some form of a 2 layer sandwich? 8mm thick panels going vertically (as originally described), and on the back, glued panels horizontally. That's the principal that plywood works on, right?

Yes, but ply uses very thin veneers so the internal stresses are minimal.

 

[RobinBHM]

What about a conventional frame and panel, but make the panel flush with the face.

If you then rebate the sides of the panel, it will give you a small recess on each side next to the stiles of the door, which Ive done a few times and looks an attractive feature. It would certainly look nice on office furniture.

Im not sure a slab door in solid timber is actually going to look very attractive, whether it can be done or not.

 

[Bodgers]

What about a conventional frame and panel, but make the panel flush with the face.

If you then rebate the sides of the panel, it will give you a small recess on each side next to the stiles of the door, which Ive done a few times and looks an attractive feature. It would certainly look nice on office furniture.

Im not sure a slab door in solid timber is actually going to look very attractive, whether it can be done or not.

This is the desired look. Ironically, a kitchen...

http://www.nakedkitchens.com/blog/image ... hen-03.jpg

Sent from my MI 3W using Tapatalk

 

[RobinBHM]

I reckon it could be done with a few boards of engineered oak flooring.

Cut the tongues and grooves off and biscuit together.

I realise the inside will be birch ply faced, but the oak face will be a 5mm or so wear layer of oak that is stable as its bonded to a birch pky back.

If you really want to try solid, then it needs to be made with braces like a tgv door. Maybe you could also machine in some stress relief grooves in the back, perhaps in a V shape or other moulding so its a feature. Prob need 1/3rd deep.

 

[ED65]

If someone else hadn't mentioned it I was going to recommend you think about very carefully selecting your wood so that it's all QS to cut movement down to a minimum. It's not negligible, but it might be half what it would be for FS wood from the same tree so quite a big saving. But I think that's made irrelevant by the picture you link to.

This is the desired look. Ironically, a kitchen...

http://www.nakedkitchens.com/blog/image ... hen-03.jpg

Not sure if you've groked but with that orientation the axis of wood movement would be vertical, not horizontal. So the hinges would be put under strain by movement which is obviously a Bad Thing.

With the grain horizontal like this I'm not sure even using QS wood would be viable, so plywood or Robin's suggestion of engineered oak flooring seem like the way to go.

Edit: looking at the photo more closely the door is inset into the frame with a narrow gap all round, so there's essentially zero movement allowable. This is the kind of application that screams for plywood or a veneered manmade board.

 

[Beau]

Is that true? I'd have thought the more laminations the less movement - hence the stability of engineered wood.

I think you're thinking of laminated beams where multiple strips are laminated on top of each other.

I believe the Op is talking about edge glueing strips, which have a tendency to cup

The reasoning for multiple strips is partly to negate movement from any one board. This can defiantly be more stable but does not cut down expansion and shrinkage. As said for least expansion and shrinkage and movement quarter sawn boards are best if available.

An image to show how multiple strips can lessen overall warping. Hope that makes sense

 

[Sgian Dubh]

... the best way to not end up with a wood movement disaster on these things further down the line.

The door itself will be made out of approx 17-18mm thick Oak stock. Strips 50-60mm wide laminated together. Each door (2 per cab) approx 900mm by 350mm. The strips will run vertically, with the grain.

I've made doors with a similar appearance out of solid material, but mostly using wood with smaller expansion and contraction characteristics than the oak you're proposing, e.g., American mahogany. A solution I've used more than once are reinforcing braces fitted across the grain with a sliding dovetail - see below: somewhere I have a photograph of a similar construction, but I can't locate it, so a two minute sketch will have to do. Make the housing in the panel longer than the dovetail on the brace to allow for expansion and contraction, slide the brace in dry, and about 25 - 30 mm before it's pushed completely home, apply a little glue to the still exposed end of the dovetail to lock it in place.

The edge of the panel showing the exposed dovetail becomes the hinged side, and you'll need to allow a gap for expansion and contraction of the door panel. As others have said, quarter (radially) sawn oak has smaller cross grain expansion/ contraction characteristics than tangentially sawn material (5.3% radial and 8.9% tangential). You ought to allow up to 1.5 to 2 mm for seasonal shrinkage and expansion (between roughly 13% MC [summer] and 8% MC [winter]) of each 350 mm wide panel using tangentially sawn wood, and about half that for radially sawn. Slainte.