Wednesday, January 6, 2010

The Other Failure

The other morning was too lovely to pass up.

It was about 12 degrees out as I made my way to the shop, which isn't unusual this time of year. I try to bank the fire before bed so that the shop isn't too cold, but it's usually around 35 or 40 degrees when I get to work. I don't mind too much, it keeps me moving and within an hour the layers start peeling off as I get it to around 50 degrees.

As you can see, I went well beyond my usual comfortable working temp in order to do a glue up. Using hot hide glue in winter can be tough because I try not to use too much retarder in my mix. Any additive to the glue weakens it (less glue in the mix), so I heat the shop ahead of time when I know that I have a tricky glue up.

The "Other Failure" that the title refers to is compression failure when steam bending. Anyone who has bent wood or even imagined bending wood has probably run into the more common failure, which is on the expansion side of the bend. It can come as a few fibers separating and lifting (easily fixed with little or no loss of strength) or a catastrophic shearing. As I sawed the wood for my walnut chair, I thought that my greatest issue would be on the expansion side of the equation because the parts don't strictly follow the fibers the way that shaved parts do.

I am happy to report that the air dried walnut bends beautifully. It seems to melt into the form. Of course, the chair doesn't have the extreme bend of a sack back etc... but I was surprised at the ease of the bends. I steamed each piece for about an hour, which is longer than my normal 30-40 minutes for green wood. I did have a couple of breaks on the bent stretchers, but the radius was quite tight and the shape of the parts (turned) are the main cause of the failure.

Where I've had trouble is on the compression side. The walnut collapses under the pressure when the bend is too tight. Here is a photo of a crest rail that I bent. It was smooth before the bend, what you are seeing are the waves of collapsed cells. Big bummer.

If this were a painted chair (never mind that the oak that I normally use doesn't fail so intensely in this way), I could smooth the surface and paint away happily, but the walnut has some ideas of it's own. Unless I was to shave away the material beyond the failures in my smoothing efforts, which are at least 1/16th of an inch deep, the faint line will refract light differently and be visible!

Perhaps the some combination of steaming time, thickness of material and radius of bend could minimize this issue. I've tinkered a bit (I'm on my third crest) and haven't got any real conclusions. It's been one more learning experience with the walnut, luckily I'm filled with curiosity about the limits of the new material and not beating myself up over the failures.

No comments: