Eric Viglotti
06-27-2019, 10:08 AM
Hello,
I am working on an engraved serving tray idea and to do this, it looks best if the wood used is one solid, wide boards rather than a glue-up of smaller boards. I've done a few of these and have used 4/4 maple one time and cherry another time. The boards I have been using have been 9 to 10" wide and about 18" long all in one piece. Since they are wider boards, they are definitely flat sawn. I end up coating them with 2 coats of urethane on the front, back and all sides. One of them has been completed for about a month and doesn't seem to have cupped or warped, but what are the chances I can still get away with this concept? The wood has all been properly dried and is from a reputable hardwood dealer. Am I still doomed to using a board this wide for this function or do you think I might be able to get away with it? Granted it doesn't need to stay absolutely perfectly flat forever, but as a serving tray, the person using it ideally shouldn't notice anything warping, rounding or cupping.
And Dave, great post on this topic, super informative which is what led me to quote it here :)
Thanks!
Sean,
I apologize if this is all a review for you, but your questions made me think that maybe you aren't thinking about wood movement and why bench glue-ups are always done quarter sawn.
One of the most critical requirements for a good bench is that it is flat. This is why torsion box benches (like your door) are good assembly tables. If you're working on assembly on a warped bench, you'll have problems.
Flat sawn boards edge joined don't stay flat. Therefore, benches are not usually built that way--because it turns out to be a waste of time and wood. There's a reason bench tops and counter tops all look the same, and it's not just looks.
Remember that as wood changes moisture, the rings expand and contract. If you slice a board tangent to the rings, flat-sawn, then it will expand and contract across it's width the most. Boards almost don't change length along the grain at all. And if you saw a board where it's "width" goes from the bark in towards the middle pith of the tree (quartersawn), it will expand and contract across it's width the least. The THICKNESS of a QS board is the part that is tangent to the growth rings, so the thickness will change.
Now, if you have flatsawn boards (which 14" wide pine is likely to be), and you look at the ends, you'll see that there are rings included, it's a slice of across the surface of a log. As the board changes moisture, it will "cup" and un-cup, and they way it does it is this: The "curve" of the rings will flatten out. So if you start with a flat board, with the rings facing up like "U", those rings flattening out will make the actual board center raise up and the edges turn down.
If you edge join flat sawn boards, you should alternate the rings, up..down...up...down so as the whole thing flexes, it will on AVERAGE stay a wrinkled flat.
But that's not really what you want right? Therefore, the way to keep the top surface from cupping/flexing, is to cut boards into 2" strips, turn the strips on their side... and viola, you have them in a quarter sawn configuration now! You glue a bunch together, all the end grains should look like this: ||||||||. You alternate any grain curve like this: ()()()()
NOW when the wood moves, the top and bottom surface will remain flat, and the only movement will be that the thickness of the top increases and decreases.
If you could get a giant 2" thick slab of hardwood from a tree with 27" radius, that was cut from the bark to the center of tree (quarter sawn) then it would make a perfect bench top. Since it's expensive as hell and not readily available since 1900, we instead just buy a bunch of cheaper flatsawn lumber, cut it in strips and fabricate that giant QS slab.
When you do your glue ups, its all long-grain to long-grain, and will be massively strong without any other reinforcement or splines. The only function of biscuits, splines, etc. would be to keep it all aligned as you glued up a bunch of strips. Not otherwise needed. Many folks just glue on 2 boards at a time and build up their slab over a week or so. Or you could assemble on a flat surface... oh wait ;) ... OK, the door would work with some paper on it, then use a router sled to flatten it. Or a planer, build two and join.
Here's a great link of more reading on the subject that I've found useful:
Basic Wood Anatomy and Behavior (http://www.organicjewelry.com/woodanatomy.html)
Hope that helps!
I am working on an engraved serving tray idea and to do this, it looks best if the wood used is one solid, wide boards rather than a glue-up of smaller boards. I've done a few of these and have used 4/4 maple one time and cherry another time. The boards I have been using have been 9 to 10" wide and about 18" long all in one piece. Since they are wider boards, they are definitely flat sawn. I end up coating them with 2 coats of urethane on the front, back and all sides. One of them has been completed for about a month and doesn't seem to have cupped or warped, but what are the chances I can still get away with this concept? The wood has all been properly dried and is from a reputable hardwood dealer. Am I still doomed to using a board this wide for this function or do you think I might be able to get away with it? Granted it doesn't need to stay absolutely perfectly flat forever, but as a serving tray, the person using it ideally shouldn't notice anything warping, rounding or cupping.
And Dave, great post on this topic, super informative which is what led me to quote it here :)
Thanks!
Sean,
I apologize if this is all a review for you, but your questions made me think that maybe you aren't thinking about wood movement and why bench glue-ups are always done quarter sawn.
One of the most critical requirements for a good bench is that it is flat. This is why torsion box benches (like your door) are good assembly tables. If you're working on assembly on a warped bench, you'll have problems.
Flat sawn boards edge joined don't stay flat. Therefore, benches are not usually built that way--because it turns out to be a waste of time and wood. There's a reason bench tops and counter tops all look the same, and it's not just looks.
Remember that as wood changes moisture, the rings expand and contract. If you slice a board tangent to the rings, flat-sawn, then it will expand and contract across it's width the most. Boards almost don't change length along the grain at all. And if you saw a board where it's "width" goes from the bark in towards the middle pith of the tree (quartersawn), it will expand and contract across it's width the least. The THICKNESS of a QS board is the part that is tangent to the growth rings, so the thickness will change.
Now, if you have flatsawn boards (which 14" wide pine is likely to be), and you look at the ends, you'll see that there are rings included, it's a slice of across the surface of a log. As the board changes moisture, it will "cup" and un-cup, and they way it does it is this: The "curve" of the rings will flatten out. So if you start with a flat board, with the rings facing up like "U", those rings flattening out will make the actual board center raise up and the edges turn down.
If you edge join flat sawn boards, you should alternate the rings, up..down...up...down so as the whole thing flexes, it will on AVERAGE stay a wrinkled flat.
But that's not really what you want right? Therefore, the way to keep the top surface from cupping/flexing, is to cut boards into 2" strips, turn the strips on their side... and viola, you have them in a quarter sawn configuration now! You glue a bunch together, all the end grains should look like this: ||||||||. You alternate any grain curve like this: ()()()()
NOW when the wood moves, the top and bottom surface will remain flat, and the only movement will be that the thickness of the top increases and decreases.
If you could get a giant 2" thick slab of hardwood from a tree with 27" radius, that was cut from the bark to the center of tree (quarter sawn) then it would make a perfect bench top. Since it's expensive as hell and not readily available since 1900, we instead just buy a bunch of cheaper flatsawn lumber, cut it in strips and fabricate that giant QS slab.
When you do your glue ups, its all long-grain to long-grain, and will be massively strong without any other reinforcement or splines. The only function of biscuits, splines, etc. would be to keep it all aligned as you glued up a bunch of strips. Not otherwise needed. Many folks just glue on 2 boards at a time and build up their slab over a week or so. Or you could assemble on a flat surface... oh wait ;) ... OK, the door would work with some paper on it, then use a router sled to flatten it. Or a planer, build two and join.
Here's a great link of more reading on the subject that I've found useful:
Basic Wood Anatomy and Behavior (http://www.organicjewelry.com/woodanatomy.html)
Hope that helps!