Building and Using Your Own Steam-Bending Rig

Turkey fryers and pressure cookers, PVC pipes and plywood boxes woodworkers show their inventive side in this thread about steam-bending apparati. January 27, 2008

This is my first time bending wood. I am building a steam tube of PVC pipe. Caps on both ends, steam vent on far end. Going to use a metal can and turkey fryer gas cooking element, run a tube in the end with a removable cap. The research that I have done says to soak the lumber in water for a day and then steam for one hour per inch of thickness. I am going to do some practice runs with poplar, but am going to use walnut for final curve. Am I on the right track?

Forum Responses
(Furniture Making Forum)
From contributor K:
Don't waste your time on the poplar - I don't think it is a good one for steam bending. I would also forget the soak cycle. Use air-less-than-dried lumber if you can get it. Most important is to use compression straps. Lee Valley has a free publication with good info.

From contributor J:
I highly recommend the booklet that contributor K mentions. I was too cheap to buy Lee Valley's hardware, but the principles described there are right on, and you can get the same effect less expensively with steel strapping available at big-box stores.

From contributor R:
Wherever did you read about soaking the wood? Bad info. It's the heat that softens the wood. The steam is the delivery system. Contributor K is on the money. You need to over-bend for springback. Varies with thickness and species. It takes practice to learn this stuff. Grain orientation affects bending, of course. Wood Bending Handbook (Paperback)
by Wendelle C. Stevens (Author), N. Turner (Author) is an excellent book. Was put out by Woodcraft for years. Tells what woods bend best and how much of a radius you can successfully bend to.

From the original questioner:
Thanks! As far as the soaking, I found a personal site from a Google search showing pictures and everything of soaking for 12 hours or so and then steaming. The Veritas booklet is great.

From contributor T:
PVC is thermoplastic (may be more than one kind, like grades of stainless steel) and the 4" tube we used for steam bending sagged severely when hot. We had to reinforce it with an aluminum spar that we pop-riveted to the tube.

Riven air-dried walnut will bend the best. For sawn lumber you really should have a jig that utilizes a strap on the outside of the bend and ideally traps the ends of the bending stock. Soak - no! Don't be afraid to steam longer than minimum times, and practice with walnut.

From contributor J:
I don't know much about the thermal properties of various plastics, but I made my setup from black ABS DWV pipe and had no sagging problems.

From contributor M:
Don't pre-soak the wood. I used PVC pipe but I ran blocks every 10" along the box bottom to keep the wood up out of the water and drilled holes through the blocks to run the pipe through. There isn't any sagging in it. If you can get your hands on walnut that is air-dried to around 20-25 percent, that would help a lot. Kiln-dried wood is harder to get good results. As mentioned before, don't be afraid to let it cook longer than required. Soaking the wood doesn't necessarily raise the moisture content throughout the wood, just gets it really wet (there is a difference). Practice some first, that's the best way to get used to it. Once you steam bend successfully, you'll find yourself doing it more often than you think you will.

From contributor A:
I don't know anything about steam bending (although I am thinking about trying it), but I read an article in the March '07 issue of Wood Magazine that seemed fairly good.

From contributor K:
One last tip. Since it is all about getting BTU's into the wood, if the outside of your steam-box conducts enough heat to be too hot to touch, then you are losing heat through it, which could be substantial in cold weather. I use a fiberglass pipe for one of my chambers, and some exterior grade plywood for others that I cut little windows through for only the part that will bend to pass in one side, and out the other.

I insulate both by wrapping them with several wraps of some 1/8" packing foam. It is easy to do, and anything that I can do to improve the chance for success is worthwhile. I drilled a little hole in the far end of my tube chamber for a thermometer. I have forgotten how much rise in temperature I got from insulating, but it sure helped. Five or ten degrees off of the top is enough to cause failure.

From contributor C:
All good responses. I have used a PVC pipe, 4 inch I think, hooked up to a teapot on a burner. There was a difference when I put a moving blanket on the top to keep the heat in. I did have to strap it to a solid piece of wood to keep it from sagging. Have fun.

From contributor W:
Instead of using schedule 40 (white) PVC, try using the heavier schedule 80 (dark grey) to minimize heat bending.

From contributor L:
Heavy wall PVC pipe will work. Drill a 1/4" hole at the lowest point to allow the escape of condensate and air. Be sure to have sufficient steam available. Forget the recommended steam times; the requirement will vary depending on your material and ability to control stretch.

From contributor M:
One more thing to add: when you are going to take the wood out of the box, don't put your face in front of the door and look at the wood. The steam coming out is hot!

From the original questioner:

Thanks a lot for all the information. I built a steam tube with 6" PVC pipe and did a trial run yesterday with a piece of poplar. Everything went well until the propane tank was empty after only 30 minutes in the box. Anyway, I clamped it up anyway and just had some spring-back today when I unclamped it. So now for the trial and error with the walnut... and the joinery of the curve!

From contributor K:
While you are at this, for future reference, you might want to put a mark on your water container, and measure how much you use over a given period of time, so you will know how long before you boil dry or need to add water. I use a 5 gallon pressure cooker for mine. I have it written on the lid for reference that I use 2" per hr., and the empty pot weighs 17 lbs empty. If I think I am running close, I can pick it up by the top handle with a spring scale, and weigh it at any point. Water is about 8 # per gal. This may not matter now, but in the future, you may have a longer run of parts that you will need this information for.

From contributor Z:
I agree with contributor K. Only apply heat where it's needed. Also, with a wooden steam box, your work piece can be set in the middle of the chamber. That's how the shipwrights do it. With pipe, such as insulated stove pipe, the work piece will be sitting in a puddle of condensate, and thereby damage said work piece. Further, a cautiously judicious amount of fabric softener in the steam pot can work wonders, provided that the part is non-structural.

From contributor O:
What's the fabric softener do?

From contributor V:
Made this portable steamer, with PVC and a wallpaper steamer from the local home center. Works like a champ. Cost less than $50.

From contributor F:
I built a box about 4"x6"x60" out of regular old cabinet grade interior plywood and much to my surprise, it's still going strong. It has delaminated a little bit but not enough to effect performance. It actually looks a lot like the aforementioned steam box in "Wood," but I swear I built mine first.

I use a 1300 watt electric hot plate to heat an old ~1.5 gallon pot with a sheet metal lid welded on. I then cut a hole in the sheet metal lid and welded a 1.25" pipe nipple to it. I drilled a hole in the bottom of the plywood box and screwed a 1.25" pipe flange with a 2" long nipple over the hole. If you don't want to weld, it would work just as well with a tea kettle, you would just have less water capacity. I had some flexible PVC (for hot tubs) that I used between the two nipples, but radiator hose would be a better choice. My access lid just fits snug so excess pressure can escape and the steam can move through the box. I put 5/16" dowels across the bottom about every 8" so that the work piece doesn't sit on the bottom of the box, which would prevent that edge from being steamed.
I set the box on a table so that it slants down toward the steam inlet with the flange hanging over the end of the table. With the burner on the ground, the re-condensed water runs back into the pot and is recycled. Theoretically, if no steam escaped the box, you would be able to run the closed system indefinitely without running out of water. Obviously steam does escape and the plywood absorbs some water, but even still, it will provide plenty of steam for definitely more than three hours (I have no idea how long it runs, but I know I've run it for 3 hours straight) with a lot less energy than heating 5 gallons would require. When I have to refill the water I usually just dump it down the open end of the box and let it run into the pot. When I'm done I leave the lid open so the plywood can dry out. Sometimes some mold will form on the inside but it turns out that steam is a great way to kill mold.

I purchased the hot plate from Meijers for $9.99. It has a spiral burner just like an electric stove. It still amazes me that it's possible just to manufacture them for ten bucks, never mind shipping it to the states and generating any profit. It takes probably 15-20 minutes this time of year before the water gets to a full boil. I'm sure that's longer than a propane burner would take, but I think that am few extra minutes of pre-heating is definitely worth the convenience of electric power.

I have very little money in my setup, it works well, and sometimes it's fun to build something that you can just recklessly nail together.

The moisture is definitely a critical element in the bending. If you think you only need heat, trying using an oven and let me know how much success you have. Heck, if it does work, please do let me know so I don't waste any more effort on steam boxes.

I've also heard before that you should soak kiln dried wood in a water/fabric softener solution before bending. I've never tried that but always figured I could pull it out of the bag of tricks as a last resort if my plans called for bending more than my lumber is limber enough to handle. I suppose it makes sense that the fabric softener would leave the cellulose in wood more pliable the same way it softens cellulose in cotton. I've also heard that fabric softener is a good lubricant but I'm a little surprised that it makes a difference when added to the steaming because I'm pretty sure that the active ingredients are rather hydrophobic.

My biggest secret to success has been building good forms. I try to build the form with an outside radius which assures the radius stays uniform throughout the piece while you're applying the bend. If you bend the piece into an inside radius form, the piece being bent tends to fold in one place and that's an easy way to crack it.

Obviously when you bend the wood, the outside face becomes longer than the inside face. The point of the straps is to keep the outside edge from stretching so that the inside edge is forced to compress. Wood is better at compressing than stretching, so if you keep the outside edge from stretching and tearing, you should be able to apply more bend without cracking.

I use the 1 hour/1" thickness rule-of-thumb for a starting point, but I usually forget what time I put the wood in anyway. I've heard you can overcook and make the wood more brittle, but I've never experienced that problem.

Another important thing to remember is that it's almost impossible to bend the last couple inches on either end of the board. If you need curve throughout the length of the board, cut it long, bend it, and then cut it to length. Also remember that if you're not laminating multiple pieces together, you need to account for pretty substantial springback.

I went into my first wood bending project thinking it would be way harder than it was. The actual bending of the wood isn't very difficult. However, cutting, joining, sanding, routing and everything else gets much more complicated now that your work piece is curved. As such, radius projects are a great way to challenge yourself and improve your skills. However, if you want to make money, wacky wood and veneer is the way to go.