How to strengthen a 19mm oak panel

[TheTiddles]

Great idea. Routing inside and close to the edge but stopping 20mm short maybe?

I can’t see that working too well. Yes the carbon is good in tension, but its second moment of area is still very small.

Maybe do a scale drawing or computer model and you’ll see how weird it’ll look. You could easily make the top 75mm thick with a large chamfer and it’ll take the load just fine and still look light. As for the hairpin legs, they won’t give any kind of solid feel, so you will have a very wobbly desk indeed.

Aidan

 

[ol_london]

Be careful about cross-grain reinforcement. Your top will expand and contract across the grain by several mm because of humidity changes, and rigid reinforcement means the stress has to go somewhere - usually by the whole thing cupping up or down! So cross-grain reinforcement needs to allow for wood movement.

Thank you. This was one of the things I was concerned about. I thought it could be addressed by making sure the screw hole is wider than the screw shank.

 

[Jacob]

Some over thinking going on here!
I'd agree with Rorton's suggestion.
19mm is a typical thickness for hardwood table tops - often start sawn at 1" and end finished at 3/4"
It follows that a conventional table design is all you need.
4 rails or apron with legs at the corners.
No metal straps - differential movement causes problems.
Top fixed to rails with buttons to allow movement
Details can be out of sight if all set back under, say, 100mm overhang

 

[peter-harrison]

Really, any sag, even under a heavy load is a design failure. Did you try www.sagulator.com to calculate it ?
One idea that springs to mind is to run a few saw kerfs under the table top and glue in slips while clamping the table top to a flat surface. A torsion box gets its strength from the glue lines, not the materials themselves. ! Never tried it so it would need testing on some scrap.

I don't want to be picky but this is not how a torsion box gets its strength. What makes it strong is that it is triangulated in all directions.

 

[Jacob]

I don't want to be picky but this is not how a torsion box gets its strength. What makes it strong is that it is triangulated in all directions.

Actually any box is a "torsion" box, the word doesn't refer to any special box feature. It's just a popular woodworkers' "meme".

 

[ol_london]

What about wooden cross braces that will move as the top moves? A lot of edge joined tops have wooden strips underneath to help hold the staves together, don’t they?

 

[TheTiddles]

Actually any box is a "torsion" box, the word doesn't refer to any special box feature. It's just a popular woodworkers' "meme".

Only if you had something between the walls of said box to inhibit them moving together or apart under load, at which point it’s not a very good box!

 

[Jacob]

Only if you had something between the walls of said box to inhibit them moving together or apart under load, at which point it’s not a very good box!

If you fix the 6 sides of a box together (glue, nails, whatever) then the sides stop the other sides from moving together or apart under load.

 

[Fitzroy]

I made this last year, it's made from an old Victorian wardrobe side (so some kind of hardwood), it is 110cm x 52cm, top is 20mm thick legs are 900mm apart. I made the top and then attached the legs, at this point lengthwise sag/deflection was ok, however back to front was very flexy. Two braces were added which are 20mm thick and 35mm high, they are affixed with slotted holes to manage seasonable movement. The desk is strong enough to perch your bum on but the legs do want to splay a little. On a carpet it works great, on a solid floor the leg splay is noticeable and the desk wants to walk about.

I expect oak could be structurally stronger than this timber but with a 600 x 1500 panel I would fear you are on a loosing streak, you could adopt a similar brace as in my design for back to front stability, however length wise I'd be worried.

Fitz.

 

[TheTiddles]

If you fix the 6 sides of a box together (glue, nails, whatever) then the sides stop the other sides from moving together or apart under load.

That would be a box beam. If you then installed a structure inside to prevent the sides moving towards or apart from each other, then you’d have a torsion box.

 

[ol_london]

I made this last year, it's made from an old Victorian wardrobe side (so some kind of hardwood), it is 110cm x 52cm, top is 20mm thick legs are 900mm apart. I made the top and then attached the legs, at this point lengthwise sag/deflection was ok, however back to front was very flexy. Two braces were added which are 20mm thick and 35mm high, they are affixed with slotted holes to manage seasonable movement. The desk is strong enough to perch your bum on but the legs do want to splay a little. On a carpet it works great, on a solid floor the leg splay is noticeable and the desk wants to walk about.

I expect oak could be structurally stronger than this timber but with a 600 x 1500 panel I would fear you are on a loosing streak, you could adopt a similar brace as in my design for back to front stability, however length wise I'd be worried.

Fitz.
View attachment 103082
View attachment 103081

This is exactly the kind of thing I’m making. My only issue with that is I suspect those braces are visible. Hence a strong flat strap idea with wider slot holes.

 

[TheTiddles]

If you want a really light floaty desk, take your inspiration from bridges they have a thin and stiff deck (that has to accommodate a lot of movement) and little supporting structure. There are lots of examples if you go to Pinterest etc...

The difference between a huge bridge and a desk is only the number of zeros in the calculations, the principles are all the same

Aidan

 

[Fitzroy]

This is exactly the kind of thing I’m making. My only issue with that is I suspect those braces are visible. Hence a strong flat strap idea with wider slot holes.

A flat strap will never be stiff enough. Even though the modulus of elasticity of steel is 17 times that of oak, the moment of inertia (which is used in the calculation of deflection) is proportional to the cube of the thickness. As a result a 30mm deep oak brace and a 13mm deep steel brace would bring the same stiffness increase to the desk.

Fitz.

PS: If you are within 3m of the desk you cannot see the braces, when you are far enough away that you can see the braces you can also see the edge of the plate on the hairpin legs.

 

[ol_london]

A flat strap will never be stiff enough. Even though the modulus of elasticity of steel is 17 times that of oak, the moment of inertia (which is used in the calculation of deflection) is proportional to the cube of the thickness. As a result a 30mm deep oak brace and a 13mm deep steel brace would bring the same stiffness increase to the desk.

Fitz.

PS: If you are within 3m of the desk you cannot see the braces, when you are far enough away that you can see the braces you can also see the edge of the plate on the hairpin legs.

Thanks for explaining.

 

[peter-harrison]

Actually any box is a "torsion" box, the word doesn't refer to any special box feature. It's just a popular woodworkers' "meme".

To me a torsion box is a closed box where all the edges are fixed to each other, and there are some ribs or other internals if it’s quite a lot smaller in one dimension. It’s not a box as a furniture maker would think of one. It makes rigid structures from quite flexible materials by ensuring that one part of the box is always in tension if there’s a load. It is a quite specific thing and not a meme. The commonest torsion box structure is the good old hollow door.

 

[Jacob]

To me a torsion box is a closed box where all the edges are fixed to each other, ..... It makes rigid structures from quite flexible materials by ensuring that one part of the box is always in tension if there’s a load.

Then an empty box where all the edges are fixed to each other is a "torsion box" too.
Sorry, just being pedantic!

 

[TheTiddles]

Then an empty box where all the edges are fixed to each other is a "torsion box" too.
Sorry, just being pedantic!

Pedantic is not the word, it’s wrong, air doesn’t tie adjacent sides together.

 

[Rorton]

could you route out slots in the top for something like an l shaped section of metal like this:

so you have the 'thickness of the metal (13mm in this example) and then the bit at a right angle to it has holes in to secure it to the wood?

 

[Jacob]

Pedantic is not the word, it’s wrong, air doesn’t tie adjacent sides together.

No, but the other sides do. As you add sides a box becomes stiffer. 4 sides floppy, 5 sides stiffer, add the 6th side you have a "torsion box". Anything you put inside it may or may not stiffen it further, but isn't a radically different structural form.

 

[TheTiddles]

No, but the other sides do. As you add sides a box becomes stiffer. 4 sides floppy, 5 sides stiffer, add the 6th side you have a "torsion box". Anything you put inside it may or may not stiffen it further, but isn't a radically different structural form.

Nope, still wrong.

A box beam is reliant on the edges of the panels being joined together to reinforce the structure.

A torsion box relies on the faces being joined together for its strength.

Faces, not edges.