stile & rail without gaps - help with design

[Helvetica]

I'm making doors for a sideboard - cherry with wenge handles. Trying to allow for seasonal movement but hide the expansion gaps. My idea is to make a stile & rail panel that functions like breadboard ends. Here is what I mean: the handle is deeper than the rest of the door, so the centre panel can expand into the handle. The frame is standard stile & rail, but the right-hand rail is actually the handle. I will post a sketchup file if anyone cares to look, and attach a few images of the sketchup. It's tricky to describe and even harder to draw. Basically wondering is it likely to work or has it been done successfully before.

The door will be 19mm thick (to work with Blum hardware). The handle is also 19mm thick, and 32mm wide where it meets the panel. It has a wider area for a grip.

My house has a thermostat set at 19 degrees all year round in this room. The house is well insulated. The room has large south facing windows and gets lots of light (the cherry has darkened beautifully in no time). Seasonal humidity changes shouldn't be very large. I have made a 5 foot wide mahogany table that hasn't moved more than a millimetre, then again it is mahogany. Thanks for any advice.

 

[Helvetica]

https://www.dropbox.com/s/eyk8fiq93sxe8 ... d.skp?dl=0

link to download the .skp

 

[MikeG.]

Why are you re-inventing the wheel? Woodworkers have been making panels successfully for thousands of years.

 

[Helvetica]

Because panels have gaps, and I don’t want gaps. Is that ok?


Sent from my iPhone using Tapatalk

 

[MikeG.]

Panels don't have gaps. I'm at a loss to understand why you would think they do.

 

[Mrs C]

Because panels have gaps, and I don’t want gaps. Is that ok?


Sent from my iPhone using Tapatalk

You possibly need to look at using tongue and grove.

I have found that storing wood in the same humidity as where the end product will eventually live before you start working on it really helps preventing movement. Mine lives under the stairs for a while!

The other option if your panels are flat is to use veneered mdf, in which case the problem goes away altogether.

 

[MikeG.]

I don't even understand that. Expansion in panels is taken up within the groove the panels sit in. They are loose fitted into channels in the edge of the stiles and rails, and can float about to their hearts content. Whatever style of panel you want (flat, raised, t&g, carved, latticework...whatever) there are no gaps, and no movement issues. Just make up orthodox panels and glue your handles to the outer edge of the stiles. I'm honestly not seeing the issue here.

 

[Mrs C]

I am assuming the problem is that the panels themselves are made of several vertical pieces. If the pieces shrink you get gaps unless you ether glue them together or use tongue and grove. The complete panel itself can then float between the styles.

 

[MikeG.]

Of course you glue them together. The way it has always been done.

 

[Helvetica]

The panel shrinks and expands inside the frame. So where the panel meets the frame you get a gap (or a routed detail to hide the gap). Unless the panel is flat and sits lower than the frame. I want the frame and panel to be flat, with no gaps, and no decoration.

If I just glue it together with no expansion the boards will cup.


Sent from my iPhone using Tapatalk

 

[Helvetica]

You possibly need to look at using tongue and grove.

I have found that storing wood in the same humidity as where the end product will eventually live before you start working on it really helps preventing movement. Mine lives under the stairs for a while!

So the inside panel will be tongue and grooved, with the growth rings alternating to minimise cupping.

The Cherry has been in the room since summer. I thicknessed them to rough size a few weeks ago. When I work on the board in the shed, I bring them back to the house when I’m done. But they will still cup a little.

The dividing panels separating the shelves are tongue and groove panels, and the centre will raise 1mm after being planted dead flat. I’m trying to avoid this warping in the doors.

My understanding of a frame and panel is that the panel is not glued to the frame, or maybe glued to one side to restrict movement in one direction. If it was all glued up would it not break the frame open when it expands?


Sent from my iPhone using Tapatalk

 

[That would work]

What you are saying is that you want the panel to be flush with the front faces of the stiles and rails. ?
If this is the case then you would need something like a bead and butt principal.
So the movement across the grain is disguised in the bead against the stiles. The end grain of the panels is rebated to fit in the grooves in the rails.

 

[MikeG.]

Flush? Then you should be veneering.

 

[That would work]

It's not unusual to have a panel set flush in a frame. Hence the bead and butt type methods.
I think the OP wanted to have a flush panel with no movement provision around the panel edge. (What you refer to as gaps?)
Not really possible with a solid panel.

 

[Helvetica]

It's not unusual to have a panel set flush in a frame. Hence the bead and butt type methods.
I think the OP wanted to have a flush panel with no movement provision around the panel edge. (What you refer to as gaps?)
Not really possible with a solid panel.

This sounds interesting. Is it different to breadboard ends? Is it where you have a backing board with a sliding dovetail or something to keep it straight but allow for expansion?

Yes I am after a flush panel and frame, but I thought the movement could be allowed for in the handle. You would have to look at my attached drawings to see why I think this would work. Cheers


Sent from my iPhone using Tapatalk

 

[deema]

You can do it by allowing the whole panel thickness to slip inside the handle, however the handle would have to be thicker in order for it accommodate the entire thickness of the panel to slide inside. (You could rebate the panel on the back also, to reduce the width of the groove, however you will still need a thicker handle.

 

[That would work]

No not breadboard ends... that's for when the tImber is independent of a frame.
As you know the panel (if solid timber) will move across the grain but practically nothing along.
You won't completely stop this regardless of what second seasoning you employ. Today's 8% centrally heated air can't be relied upon to be consistent for the pieces lifetime.
So if you want the panels face to be flush with the frames face you will need to have a way of disguising the inevitable movement across the grain. So a straightforward tongue (barefaced so like a rebate) goes along the end grain of the panel and into the rails to sit flush on the front.
The vertical with grain edges on the panel also have the same tongue but because the panel WILL move you need a way of disguising it so a bead moulding with the tongue on the edges (that meet the stiles) will do that. A solid panel of around 400mm say can easily gain or loose 3 or 4 mm.

 

[Trevanion]

I didn't really understand what I was looking at yesterday but looking at it again I see where you're coming from, you're going to put a rebate on three edges of the panel so that it's flush with the external face but leave one edge that goes into the handle full size so your expansion and contraction happens there rather than the other end where you'd inevitably get a gap?

I don't see any reason why it wouldn't work so long as the panel is fixed to the gapless side, just be aware that the stiles and rails are also liable to move slightly which may lead to a gap on that side as well as top and bottom so it might be worth having a shadow gap of say 1mm to hide that movement.

 

[Sgian Dubh]

You can do it by allowing the whole panel thickness to slip inside the handle, however the handle would have to be thicker in order for it accommodate the entire thickness of the panel to slide inside.

I think he's incorporated a groove in his proposed handle (see attached image) wide enough to accept the full thickness of the panel. But, looking at the rest of the construction, I'm not altogether sure the simple tongue and groove joint holding the corner of the handle/stile to the top and bottom rail is strong enough. It could theoretically work for a solid wood panel glued to the opposite stile and all the expansion and contraction taking place within the panel wide groove in the handle.

Having said that, I'd probably first explore the veneered panel option (as already suggested by others) to achieve an end result of the panel and framing of the same thickness, similar to veneered table tops, plus lipping. Using that technique, the handle could be simply attached to the door's front face, perhaps using biscuits or similar for alignment, plus adhesive, or even just glue and screw, although this leaves a visible screw head on the door's inside face, which I suppose could be covered with a pellet or similar. Slainte.

 

[That would work]

But... is the movement of the panel going to be taken up only on the handle side? How can that be?