Plywood laminate table - design advice

[hertspretendchippy]

Hello all,

I'm looking for some advice on my dining room table project. The aim is to produce something that looks like this but also extendable....

As it's extendable - I'm making the table tops separately instead of making one big slab and cutting it up. I've mocked up the design of the table

Underside - Non-extended

Underside - Extended

Top

The top oak layer will be glued down onto the plywood surface using Gorilla wood glue and likewise for the oak trim around the edges. I decided on adding an apron around the table as this would hide the metal extending runner. The table will have alignment pins on the edge of the aprons and I'll add a toggle latch to prevent the table from sliding open.

Here's the thing...I jumped the gun and I've already applied the oak parquet planks to the top of the plywood. But after coming downstairs the very next morning there was some slight bowing (concave side being the exposed plywood face) in the leaf and the larger pieces (table top) seem ok but not perfect. The deflection was about 5mm over a 900mm span. This scared me and so I placed all the planks on the kitchen floor and added weight to straighten them out. I've just checked today and they have straightened out well enough to be acceptable. I've read this is due to an in balance of moisture in the faces - guessing that came from the glue!

I like to keep them flat as possible now and for the future. I've gone down a rabbit hole trying to figure out what the best approach is and there are a few:

1. Additional plywood layer - maybe 9mm?!
2. Torsion box - I've only seen the torsion box applied to the entire underside of a plywood faced table. I would need to make the top thicker and try to source oak strips (£££) to accommodate the extra thickness so I would like to avoid this one.
3. Partial torsion box - building the torsion box within the apron and making the apron taller to hide the extending runners.
4. Mild Steel C channels - Adding 40mm x 20mm x 5mm channel parallel with the legs. I would then have to add filler pieces so the extension runners can be fixed.
5. Plywood stretchers - Like above but instead of steel, cutting strips 15cm strips out of 18mm plywood and gluing two pieces together.

After looking everywhere for answers I realise now that I should have either used an old table top or started off with a torsion box. I bought some mild steel 20mm angle and tested it out on another piece of bowed 18mm plywood to see if it would straighten it out. The results were much better but again, not perfect. The steel bar was 700mm in length and the plywood was 900mm (steel bar placed in the centre and secured with 5 evenly spaced 13mm wood screws).

Any help or advice would be greatly appreciated.

Tom

 

[hertspretendchippy]

I've sketched up a drawing of option 4 and after a lot of pondering the one I will most likely will chose. With this design I can add a layer of thin plywood on top, hiding everything and have a cleaner appearance too.

I'll be using 70mm x 20mm Meranti for the Apron/Torsion box.

 

[hertspretendchippy]

Have I posted this in the wrong forum?

 

[profchris]

This puzzled me until I realised your second post was Option 3, not 4!

I build musical instruments, not tables, but that means I have quite a bit of experience with humidity movement.

You have two issues here:

1. Cupping or bowing after the initial glue up. This is likely to be because the top laminate of the plywood expanded across its width as it absorbed moisture from the glue. I'd expect it to return to pretty much flat as it dried and lost the moisture, and this is what you say has largely happened. I'd suggest leaving it for a week or so in its final environment and then, if it's close enough, plane any slight ripples out of the top. One way to avoid the problem is to use epoxy for the glue up, as this doesn't introduce moisture - some guitar makers do this when attaching fretboards.

2. Movement in use, as the oak absorbs moisture from the air when humidity in the room changes. You can't avoid this unless you encase the whole thing in resin or plastic, but if you are using the herringbone pattern then it looks to me as if the expansion will be roughly equal in all directions (proportionally the same, wider/longer panels move further in total). There is a useful table here to help calculate that: https://www.popularwoodworking.com/wp-content/uploads/2010/10/WoodMovement.pdf

Taking a guess that your kitchen humidity will change around 3 percentage points over the year (ignoring short term changes such as steam from cooking), and that your oak has a widthwise expansion coefficient of .002, that suggests expansion of 900 x 3 x .002 mm across the width, or 5.4mm, and 2.58 mm across the length of each herringbone. Then halve that for the whole panel because the expansion should be equal in both dimensions, so 2.7 and 1.27. The plywood shouldn't move, so that force has to go somewhere, and the whole panel will either dome or dip - doming if humidity was low when you glued up, dipping if high. If the panel were as thin as a guitar top then, at the most humid and without bracing, the centre of the panel might rise 11mm or so. But a guitar top's bracing resists that to around 1/3 of the amount. Your 18mm plywood provides a lot of bending resistance, but the oak is thick and so exerts a lot of force. We're guessing here, but I wouldn't expect the centre of these panels to dome more than 4mm. In practice you might not notice that.

Aprons and/or a torsion box will resist movement further, maybe halving it to 2mm.

You can't stop wood moving, but you could reduce it by reducing the thickness of the oak top - again a guess, but halving the oak's thickness might mean it only moves 1 or 2 mm vertically.

Hope this helps!

 

[Hornbeam]

The general rule to maintain stability is that whatever you glue on one side, you should also do to the other side. You dont say how thick the oak herringbone pattern layer is but I would not go over 4mm max as you will get differential movement between the plywood and the oak
All the bracing is just fighting the tension/compression caused by only covering 1 face

 

[hertspretendchippy]

@Hornbeam Just an update - I went ahead with the build after reading your feedback. Here is the build!

It's not 100% complete - still need to finish middle leaf but overall very happy with the outcome.

Thanks for your help!

 

[Hornbeam]

Glad you are happy with it
Wood movement and construction details is often a balance of past experience, judgement and skills
I am just making a veneered cabinet with a starburst pattern on the top and doors. I have done the starburst pattern on both sides as I know that the wood grains/tensions will be equalised and although it is a bit more work it also gives me teh option of a B side if I cock something up