Steam bending

I've always wanted to try steam bending, and the upcoming work of installing the carlins (under deck supports) and gunwales is going to provide a good opportunity to try it out.  The plans call for several narrower piece of wood to be bent and laminated in place without steam, but I was not looking forward to handling and securing long skinny, slippery, epoxy covered pieces, so I thought I would gear up for steam bending.

Traditional steam bending involves placing the wood to be bent in a box and applying steam until the wood gets flexible.  Then you remove the wood from the box and quickly put it in place before it cools off and stiffens up again.

Recently I saw a video on the net showing a variation of this, where the wood is steamed in place on the boat. Using this method there's no frantic dash to beat the cooling wood, as you can keep the steam on until everything is in place.  The key is to do away with the traditional steam box, and substitute a plastic tube.

So, excited about trying something new, I placed an order for an Earlex steam generator and soon it showed up at the door.  A quick visit to eBay and 200 feet of 6-inch wide 4 mil lay-flat tubing was also on the way. 

The steam generator comes with the tank, a hose, and a electrical cord.  The hose is nice and long, about 10 feet, and has a flare fitting on one end and a threaded fitting on the other - very nice.  The steam generator is pretty simple, and consists of a heating coil in a plastic box, but it has a pressure relief valve and over-temperature shut off (if you run it dry).  It's a nice setup.

For my test bend I planed some pine to 20mm x 30mm.  Nothing special about this stick, it has knots and all - kind of a worst case scenario.  I cut off a section of plastic the same length as the wood.

And installed the hose in the middle.  The threaded end and jam nut is inside the tubing in this photo.

Then I put the wood inside the tube, filled the steam generator with water and plugged it in.

After about 15 or 20 minutes steam began to flow.  Some flows through the tube and out both ends, and some condenses in the tube and eventually runs out the ends.  I put a bucket under each end to catch the drips.

After about 45 minutes I put on gloves and easily pulled the wood into place.  This bend is in two dimensions - curving both upward and inward as it approaches the ends of the boat. 

Pulling the plug on the steam generator, I let the wood cool.

And then slipped it out of the plastic.  There was very little spring-back - this experiment was a success! 

The wood took on a nice smooth curve for the most part.  If you look in the background you can see there is a straight section there as the piece ended and there was nothing pushing it into a curve.  With a full-length piece this shouldn't be a problem.  Or if I wanted this shorter length to curve in that area I could have pushed it out a little with a brace - the bending was easy and did not require much force.

Since the final, full-length, part will have one or more scarf joints joining shorter sections of wood, I plan to do another experiment like this, but with a scarf joint in the test piece to make sure the steam heat does not cause the joint to fail.  If it does I'll have to come up with a plan to work around that.