I have to say that I love it when a discussion goes on about something that I post.
A rather lively discussion is going on now about my recent panel experiment over at the Canadian Woodworking Forum. Check out their site! It looks like a lively group of woodworkers.
Being that I'm not Canadian, I thought I would take a little bit of time to respond on my blog, rather than on their forum. This way, y'all (that is, all of you non-Canadians) can benefit from the discussion.
One forum member is thinking of building a coffee table, followed by a dining table, and became concerned when he saw what happened to the panels I glued up.
Here are a few points about my experiment, and why I think what happened, happened.
I glued two panels up using materials and techniques that supposedly gave this project very little chance to avoid failure. The reason I did two, was so I could show the difference in a cross grain batten with glue, and one without.
Don't focus so much on the one that failed, but look at the other. It did not. In fact, it remained plenty flat for a table top while the glued one turned into a banana. That cross grain batten turned out to be plenty strong enough to keep the panel in it's intended position. Using home center lumber, I wasn't sure this panel would not also fail.
The other thing of note is the panel that failed cupped in a direction opposite what it "should" have. Flat sawn boards typically cup towards the bark side of the tree. This panel cupped toward the heart side. Why? Because the batten is trying to keep that side of the panel longer than it wants to be. While the panel dries out, it shrinks in width, but the glued batten is keeping one side of the panel the original lenght. The panel has no other option. I was hoping for it to split and crack, but I think it did not because the batten was attached to the outside of the panel, as opposed to be on the inside like a sliding dovetail.
Another method you could use to keep a panel flat is to use quarter sawn lumber. Most of the movement on a panel glued this way would be in the thickness of the panel, so cupping and stretching are less likely (but not eliminated).
I did not finish these pieces. I completed them in one day. I subscribe to the theory that finish of any kind will not stop wood movement, only slow it down a bit. I did not test this, so one might look at this claim with more scepticism.
At the end of the day, there is a lot of mis-information out there regarding woodworking. One should occasionally test things to find out if they are true.
Me, I'll always glue up panels with the rings going the same direction, as a simple batten is strong enough to prevent warping.