Why does my bending fail?

[Wade Lippman]

I want to bend some 5/8"x7/8" walnut strip into a 8" radius quarter circle.
I got some fresh walnut and milled it to size, as I read that kiln dried won't bend.
I made a steaming box out of 1.5" pvc pipe, with dowels to keep the wood out of the condensate on the bottom.
I used a wall paper stripper for steam; it worked pretty well as steam came out the condensate drain on the far side of the steam box.

I steamed my first piece for an hour. As soon I bent it, it cracked. i figured I was too slow, so I practiced so I could do it instantly; and steamed for 90 minutes.
Second piece cracked immediately also.
It was much easier to crack than unsteamed wood would have been, but that doesn't help much.

The grain isn't perfect, but it is pretty straight.
Any ideas what I could be doing wrong?

I should mention that the wood came from a branch the electric company trimmed, rather than from a trunk; but can't see why that would matter.

[John TenEyck]

That's a pretty tight radius for something that thick. I prefer to use glued laminations rather than trying to steam bend a thick piece, but it's possible. You need really straight grain to start with. AD is better than KD. Walnut doesn't bend too well though, from what I've read. Anyway, you need to use a metal strap on the outside of your wood to keep it from cracking around a radius that small. The strap needs to have two blocks attached to it on either end of your piece of wood, too, so that the wood is completely captured. It's also best to put your piece of wood into nearly boiling water with a little Downy fabric softener for any hour or so, then into the steam bath for another hour. That has always worked better for me than just steam.

Soften in hot water, steam, remove and insert into the metal strap, bend quickly and clamp to your form.

John

[Wade Lippman]

Maybe I am expecting too much; I thought 5/8" was fairly thin and 8" was okay. I can't do much about the 8", but I can plane the wood down to 3/8".

I am not sure what a metal strap is. A piece of aluminum flashing maybe?

[Al Weber]

See this link. http://www.leevalley.com/us/wood/page.aspx?p=42173&cat=1,45866

[mreza Salav]

That's a lot of bend for a relatively thick material. Try bent lamination; works much better and less chance of spring back.
Walnut will change color when gets wet/damp (probably you have noticed already) and makes a mess that doesn't go away with sanding.
I've done quite a bit of bent lamination of walnut and if you use darker glues you can't tell the glue lines.

[Brian Henderson]

Yeah, you're trying to do too much. Too much material at too small a radius. As everyone else has suggested, go for thinner material laminated together and you should have better luck.

[Steve Demuth]

I disagree that 8" radius is too aggressive for 5/8" stock. I've bent Black Walnut to half that radius successfully. But you absolutely need a bending strap to succeed at that tight of radius. Also you need straight grain, with as little runout of the grain in the long dimension as possible.

If you really want steam bent, rather than using bent lamination as others have suggested, you also need to be prepared for quite a bit of relaxation of the bend. Since you don't have the experience to have a feel for your bend, getting a precise 8" radius is going to be hard. You'll have to over bend and probably experiment a couple of times before you get the right degree of bend to match relaxation to your target radius. Rather than that, to get a precise curvature from steam bending I generally use stock 1/8" to 1/4" thicker than my target and then after over bending to an approximate match, template rout down to the final dimensions. If it's long stock with longer radii, this isn't necessary, because you can generally pull those into their final radii with other parts of a structure.

So here's how I would do this:

1. Create a bending form with a 7.5" radius over 100 - 110 degrees of curvature and 6" of straight at either end of the curve.

2. Cut straight grained walnut 7/8" X 7/8" X 24" long. This should cover the entire length of the form.

3. Steam for about an hour (once you have full steam)

4. Bend with the bending strap (making certain that the strap is tight on the outside of your curve), with slow even pressure - you should spend at least 15 seconds on the actual curve - which is longer than you think.

5. Clamp for an hour.

6. Remove from the form, put a clamp across the ends to prevent further relaxation, and let it set for a day.

7. Joint flat and dimension in the width.

8. Rout the final curve in the depth dimension.

9. Cut at the ends of your quarter circle.

There is plenty of margin of error in what I suggested above that you should come out fine.

[Wade Lippman]

I assumed I could use a series of clamps to gradually make the bend, but that isn't going to happen in 15 seconds.
$80 for a strap is making me seriously reconsider the project.
Maybe I will just saw them out of a board. The end grain shouldn't be a problem.

[William Shelley]

You could probably make your own strap clamp with some metal pipe hanging strap from home depot. It's probably not the right kind of steel compared to the Veritas strap clamp, but i doubt that matters if you're using it for one job.

Make a loop in each end and use a bolt or something through two of the holes in the strap, so you have something to put your clamp in.

Here's a 10ft roll for $3.24: http://www.homedepot.com/p/Oatey-3-4-in-x-10-ft-Galvanized-Steel-Hanger-Strap-339232/100167964

[Lee Schierer]

Bending works best with riven pieces of wood and green lumber. Lumber that has been cut with a saw has grain ends exiting the side and those places usually fail. Try splitting off a piece of stock from a wider board and then use a plane to smooth the sides and reducing the thickness. It should bend better.

[Wade Lippman]

Bending works best with riven pieces of wood and green lumber. Lumber that has been cut with a saw has grain ends exiting the side and those places usually fail. Try splitting off a piece of stock from a wider board and then use a plane to smooth the sides and reducing the thickness. It should bend better.

Wouldn't it warp badly while drying?

I've got green wood and a splitting maul...

[dennis davidson]

You don't need to spend money on a bending strap. I've made several from the steel banding straps that loads of dimension lumber used to be banded together with, they use plastic now, but steel coils are still banded with steel and the bands are thrown out. Find a steel warehouse near you or perhaps a small manufacturer of steel products.

[Steve Demuth]

Wade,

First, you can't just use clamps. But you can easily make the bending strap for a few bucks. For what I wrote above, you just need a strip of reasonably heavy gauge sheet metal about 3' long and 2" wide. 12 or 16 gauge should work. Bolt a 4" X 2" X 1" block of wood on one end, and an 18" X 2" X 1" handle on the other, with 24" between the blocks. I drew a picture.

The key is after you've made this jig to cut the wood to be bent so it fits snugly in the gap between the blocks. When it comes out of the steam, bang it between the blocks. If necessary, put a wedge at the end so that it fits so snugly that the metal strap is tight along the length of the wood to be bent when the strap is straight. Clamp the 4" block to your form, and use the handle to bend the wood around with the strap to the outside. Then clamp strap and all all the way 'round to the handle end. As you wrap the wood and strap around the form, the metal strap will apply a steady heavy pressure to the outside of the bend, preventing the wood from separating. Go fairly slowly, and you'll be find.

354730

[Phil Mueller]

Here's some general guidelines from online regarding the thickness of each ply:

• 2″ to 4″ radius — 3/32″ thick
• 4″ to 8″ radius — 1/8″ thick
• 8″ to 12″ radius — 3/16″ thick
• 12″ radius or larger — 1/4″ thick

The article goes on to say, wood type, grain, etc., could require some variance.

[Robert Engel]

Certainly no expert on steam bending, but I know cracking like that is probably from grain fibers breaking out because they are not continuous fibers, but rather they'e been cut obliquely through the milling process. Normally you don't care, but when steam bending it can matter. I suspect this is probably species dependent to some degree. You also need a board with straight grain.

Check out Curtis Buchanan on YouTube he makes Windsor chairs. Lots of bending info there. He uses mainly white oak, but I think the principle would apply to any straight grained piece of wood.

One thing I noticed, he rives all his bending wood out of green logs rather than using dimensional lumber. This produces a board with all the long grain fibers running continuous and parallel from one end to the other. Then there is a much better chance of bending without the convex surface fracturing.

[russell lusthaus]

IMO, that is a tight bend for walnut that thick. Laminating it will be easier, for sure. Getting straight grain from end to end is the best place to start troubleshooting. I also agree a bending strap is a good idea, but when I do it, I use a thin strip of hickory as a bending strap. Say like, as wide as my work piece, longer than necessary, and about 1/8 inch thick. I always have Hickory laying around and it is a good tough wood and bends well on its own. Not sure about other woods, but easy enough to try out what you have on hand. Dont use oak or the steam and water may leave stains. I bet ash or maple would work great too. Good Luck

Russ

[Jeff Duncan]

I would question whether you have a lot of stress in that wood already being that it came from a branch and not the trunk? A brand would be under constant tension/compression which may cause problems when trying to bend. For instance if your bending against the way the woods natural tendency is, I would think that would be significantly more difficult to bend than just a straight piece of wood with no pre-existing stress?

Just a though but may be worth trying the same bend with a piece of trunk wood and see if there's any difference?

good luck,
JeffD

[Chris Fournier]

You can easily make the bend that you are proposing but you will need a back strap to prevent the wood from splitting. This being said by the time that you get set up properly to do this you'd be further ahead to use a bent lamination and spring back would be controlled easier as well. Plastic resin glue. Take your time to resaw from a billet and make sure that you put the pieces together in proper sequence from the saw and you will have an almost invisible glue up. Steam bending can do incredible things but you really have to fixture up for success.

[Brian Holcombe]

If you have green lumber available, use that. You can actually use sawn lumber, but you need to align the grain and it's quite wasteful. Probably just as wasteful as riving, but with the price tag of KD sawn lumber on the waste.

I start with rift cut material and cut along the grain lines on both faces, the results are free of grain runout. I do this for most parts (legs, door frames, etc) because the resulting board moves more predictably and I hate the look of boards with runout.

[Frank Drew]

As others have also recommended, I prefer bent laminations for the predictability of the results and they're also easier to do unless you do lots of steam bending and already have your setup.

[Curt Harms]

Certainly no expert on steam bending, but I know cracking like that is probably from grain fibers breaking out because they are not continuous fibers, but rather they'e been cut obliquely through the milling process. Normally you don't care, but when steam bending it can matter. I suspect this is probably species dependent to some degree. You also need a board with straight grain.

Check out Curtis Buchanan on YouTube he makes Windsor chairs. Lots of bending info there. He uses mainly white oak, but I think the principle would apply to any straight grained piece of wood.

One thing I noticed, he rives all his bending wood out of green logs rather than using dimensional lumber. This produces a board with all the long grain fibers running continuous and parallel from one end to the other. Then there is a much better chance of bending without the convex surface fracturing.

I recall a New Yankee Workshop episode where they went to Mike Dunbar's Windsor Chair Institute. Dunbar used all riven stock for steam bending and for the reason stated.