Monday, March 11, 2013

Let's Blow Up a Panel - Part II: 7 Days Progress

Still waiting for the good stuff.
It's only been seven days since I built the two panels for my experiment.  You can read the original post here.

To summarize, I made two nearly identical panels out of home-center construction lumber for demonstration purposes.  They were about the same size, and used cross battens screwed with three screws each for stability.

The only difference was on one of the panels I also applied glue to the cross-grain battens.

OK, there were a couple other differences, too.  I got lazy after planing all of the rough stock for the first panel by hand.  The second panel didn't wind up quite as flat due to my lack of diligence.  I glued the first panel, the good one.  The cross battens actually flattened the second panel a bit.  I do not expect this to have much of an effect on the results.

So, here is some photographic evidence after only seven days:

The glued and screwed panel.  Notice the daylight under either end.
Glued panel on the left.
The non-glued panel still looks pretty straight.
While the panel without glue seems pretty stable for the moment, the other panel is starting to go crazy.  If I were using this panel on a project, I would have to throw it away and start over.  It is already starting to get a fairly pronounced cup across the width of the panel.  This cup is visible without holding a straightedge up to it.

Interestingly enough, the cup goes the opposite direction it should:  i.e. it is cupping toward the center of the tree.  Normally, lumber left on it's own will cup the other way.  This panel is showing the stunning strength of the long-grain battens.  They are staying the same length, while the panel is shrinking and drying out along the width.  Because it is glued on, the panel can not cup the normal way.  But, it also can not shrink in width where it is in contact with the battens.  Therefore, the battens are forcing the panel to cup to the other side.

Notice the non-glued panel.  This batten with only three screws is plenty strong to keep everything where it should be.  While the panel shrinks, the screws give a little due to the elongated pilot holes that were drilled in the battens, preventing the kind of stresses seen on the other panel.  Also of note, the battens are made of the same 7/8" spruce that the panels are.  Not known for being the world's strongest wood, it nevertheless is plenty strong to keep this panel mostly flat.

Amazing, considering that this is crappy construction lumber from the home center.

I'll keep posting the results of this test until something spectacular happens.  I would think over the course of a month or two that I should get some interesting splits or other wood breakage.  I can't wait.

I would expect the non-glued panel to remain looking exactly as it does now.  In fact, I have some examples from previous builds that should prove this point.

First, is the underside of the panel on a coffee table I built about five years ago.  This panel was constructed exactly in the same manner as the panels in this experiment.
Maple and glass coffee table.

The second shows the lid of a blanket chest that is more than ten years old.

Walnut blanket chest.
Both of these examples from my earlier work have remained as flat as the day I put them together.  They both also are constructed with only three screws in each cross-batten and no glue.

The moral of the story?  Don't be afraid of wide panels for table tops or the like.  Keeping them flat is easy.  There are many ways to join battens to the panel, such as with sliding dovetails, breadboard ends,  etc.  If this panel can stay flat with only three screws, any other method should also work.

As long as it isn't glued.


  1. This is a great experiment, Brian. Please keep us up to date.

    1. Thanks, Dyami. I would have done this experiment with a few more panels to see some other variables, but processing all that by hand just for it to fail wasn't too fun.

  2. Good stuff. Woodmovement should not be underestimated (personal experience)

    1. When I built the walnut blanket chest pictured, I glued and screwed the battens. Before I could put the hinge on, there was a pretty spectacular failure. The lid you see is the result of learning that lesson.

  3. If you want to accelerate the explosion, just leave the panels outside when it's raining... but it may defeat your controlled conditions to simulate furniture inside.