Wood movement in segmented turnings

[Richard Dooling]

I’ve been wondering about segmented vessels - how they are glued up and stability issues.

I think that some consideration must be made for grain orientation and species compatibility. I know I’ve felt unevenness at segment edges of many bowls I’ve seen at shows and in galleries.

How do you account for differences in wood movement between segments that have different properties?

Reading recommendations?

Thanks,
Richard

[John Beaver]

Richard, it should be taken into account. It isn't always.
Most important is the grain direction. A segmented turning will hold up with varying species if the grain direction is correct. It will hold up better if you use compatible species.

[Glenn C Roberts]

John, Any chance you could expand on that a little? Like, what is the correct grain direction and how do you pick compatible wood?
\Thankx,

[John Beaver]

John, Any chance you could expand on that a little? Like, what is the correct grain direction and how do you pick compatible wood?
\Thankx,

When you make segmented rings you are basically gluing pieces of wood into a circle. Generally you have the long grain running around the circle - pie shaped pieces glued end grain to end grain so the face of the ring has the long grain. Any element you add to that, be it a small spacer or fancy design element should continue with that pattern. If you insert a piece with the grain running 90 degrees to another piece you will end up with problems as the expansion and contraction of the wood in different directions will fight the glue joint. There isn't necessarily a correct way as long as all the pieces are the same. You cant have a piece with the grain running horizontal and then one running vertical.

If you look at wood density charts, you want to use wood that is similar density - also referred to as Specific Gravity. Google "wood specific gravity" or "wood density" and you will find many charts. Something like Poplar has a gravity around .4 while Purpleheart is around .8. These would probably cause problems as the Purpleheart is essentially twice as dense as the Poplar and will therefore move (expand and contract with humidity changes) at a very different rate putting stress on the joint, causing the uneveness Richard asked about. Most N. American hardwoods fall in the same neighborhood while exotics tend to be more dense. Depending on the size of the segments you can push the difference farther, but the closer you can be the more stable your piece will be.

There are probably people here that can explain this better then I can, but hopefully you can understand the concept.

[Glenn C Roberts]

John, Excellent explanation. This will definitely help with the learning curve. Thankx.

[Richard Dooling]

Thinks John,

That's a nice concise explanation.

[Doug Rasmussen]

"pie shaped pieces glued end grain to end grain so the face of the ring has the long grain. "

Nice explanation John.

What about gluing long grain to long grain so the face of the ring has end grain? I'm not visualizing why it would make any difference which way it's done as long as it's consistent on all rings. Either way it seems the bowl will expand and contract with humidity.

Turning might be easier with end grain to end grain gluing, huh?

[John Beaver]

What about gluing long grain to long grain so the face of the ring has end grain? I'm not visualizing why it would make any difference which way it's done as long as it's consistent on all rings. Either way it seems the bowl will expand and contract with humidity.

Turning might be easier with end grain to end grain gluing, huh?

Doug, side grain gluing is possible, but then you would end up with a piece showing entirely end grain (All the exterior walls would be end grain blocks). I've often thought about it but never tried it. With standard gluing the piece breathes across all diameters with roughly 1/4" walls. With side grain gluing the piece would expand and contract around the circumference and the full width of each block. That is something to ponder.

With end grain to end grain gluing the entire piece is side grain so turning is super easy. The end grain is never exposed

[Brian Brown]

One other thing to consider about the expansion and contraction of the woods is the gluing itself. When wood dries, it shrinks. Segmented turning is usually done with dry wood, because the wood movement is less. When you glue the pieces together with a glue like Titebond, the water in the glue causes the wood to expand slightly. If you were to drill a 1/4 inch hole in a piece of wood, and create a same species 1/4 plug to fill the hole, they may dry fit superbly. However, if you apply glue to both pieces, and try to put the plug back into the hole, both pieces have expanded slightly around the glue area. The plug fit is now usually very tight. Sometimes so much that the plug will not go into the hole. If you do this with different species, the problem can be worse depending on several factors.

In a segmented piece, after the glue dries the wood will usually still contains moisture, and a bit of expansion. If we do a full glue up of the piece, and finish turning and sanding before that residual moisture has gone, the wood will later shrink slightly around the joints. This can cause a slight depression in the areas around the joint. Visually we don't usually see them, but your fingers can detect the change. It is especially problematic when using thin veneer pieces for accents between the segments. Most turners are not very conscious of the grain orientation of the veneer as it is so thin. If you orient the veneer with the end grain to the outside of the piece, you will feel the expansion contraction even more because the veneer will shrink less than the side grain. Wood shrinks less longitudinally with the grain. Now, your veneer will poke out more. Solution; keep the veneer grain in the same alignment as the segments, and give the wood more drying time after a glue-up. As for how long to let the wood dry after gluing, I don't know. If you let it sit for two weeks after final turning, and then sand, it would probably be enough but that may also be over kill. I hope this makes sense. I have no science to back this up, it is wholly anecdotal based on my own experience, so YMMV.

[John K Jordan]

I think anyone who works with wood would find this book valuable, "Understanding Wood" by R. Bruce Hoadley, subtitled "A Craftsmans Guide to Wood Technology":

https://www.amazon.com/Understanding.../dp/1561583588

It explains what wood will do, and why. It is technical in places but also highly practical. Hoadley is an expert on wood AND a craftsman. For woodworkers and turners the knowledge here is almost like the difference between working in the dark and with the lights on.

JKJ

[Ted Reischl]

Expanding and contracting due to humidity changes certainly occurs. BUT.....

I looked up changes for walnut and cherry. On a 12 inch panel, flat sawn, the difference in expansion for a 9% change in humidity is about .03 inch. That is a lot. BUT.....

Bowls are not 12 inches thick. Most I have seen are about 1/4 to 3/8 thick. So, that .03 has to be divided by 12, and then divided again to obtain the movement difference in a 3/8 thick piece. My number came to about .0009. Not much.

BTW, a 9% change is HUGE. Most bowls will never see that unless they were turned in the desert and then shipped to a jungle and stored outside.

[Brian Brown]

One idea if didn't mention in my previous post, is that when the damp wood shrinks back to it's original size, the glue lines remain the same size, and sometimes the raised ridges you feel are the thin glue lines that poke out from the wood. Also corrected by waiting for the wood to dry before final sanding.